Talk:Journal of Morphology 22 (1911)
THE ORIGIN OF THE SEX-CELLS OF AMIA AND
LEPIDOSTEUS
BENNET M. ALLEN
From the Department of Anatomy, University of Wisconsin
TWENTY-SEVEN FIGURES
INTRODUCTION
There has been an increasing amount of attention given in the last few years to a study of the origin and migration of the sexcells of the vertebrates. The number of forms in which this subject has been studied is being constantly extended. While much conclusive work has been done, upon the history of these cells in the elasmobranchs, and an equal amount in tracing them in the teleosts, up to this time, they have never been studied in the ganoids. This work was begun over two years ago, and was reported at the 1908 meeting of the Association of American Anatomists in New York. (Allen, '09.)
The material for the present work is obtainable in great plenty within less than half a mile of the grounds of the University of Wisconsin. Since the breeding habits of these two fishes have been thoroughly treated by other writers, it is not necessary to redescribe them. Telleyesniczky's bichromate-acetic fluid and Zenker's fluid have been used as fixing agents and have proved in every way satisfactory. One secret of obtaining good sections is to secure most thorough infiltration by placing the material in a solution of paraffine in turpentine. While turpentine has a bad reputation, no deleterious effects were noted in the course of the work. Paraffine sections were made without difficulty 7m and 10/^ thick, and were stained, for the most part, in haemalum and orange G. Heidenhain's iron haematoxylin stain was sometimes used for the later stages, but showed no superiority over the haem-alum. In the earher stages of development it could not be used at all, owing to the deep stain that it gives the yolk material.
With the abundance of material and the amount of time given to the work, it was possible to make a careful study of a large series of stages, much larger than it has been found necessary to use in the preparation of this paper.
It is not necessary to enter into a detailed account of the earlier work upon the origin of the sex-cells, because that has already been done in earlier writings. Since the author's articles on the sex-cells of Chrysemys and of Rana, some important papers have made their appearance, which, with one exception, bear out in a most gratifying manner the conclusions expressed by the writer in the two papers mentioned above and in somewhat less confident manner in the earlier writings of Wheeler, Woods and Beard.
These papers will be discussed in the light of the facts set forth in this paper in the last part of this article, since they are to be considered in a more or less controversial manner.
OBSERVATIONS UPON LEPIDOSTEUS OSSEUS
Lepidosteus, 4 '^nm. total length. Cells which appear to be sexcells lie in the ventral portion of the single layered gut entoderm. They can be but dimly distinguished from the other cells of the entoderm among which they lie. They have a more spherical shape than the other entoderm cells, never being flattened as the neighboring cells frequently are. Another difference lies in the fact that the sex-cells contain more and decidedly larger yolk spherules than do the adjoining entoderm cells. Unfortunately these differences are masked by the large quantities of yolk found in the entoderm at this stage. This is true to so great an extent than one can not be certain at this stage as to the identity of the cells in question. At this period the hind gut has a much greater diameter than it has at later stages. At a point onequarter the distance from its cranial to its caudal end it has a dorso-ventral dimension of .24 mm. and a transverse diameter of .20 mm.
Lepidosteus 6.8 mm. total length. In a similar part of the hind gut of a specimen 6.8 mm. in length the dorso-ventral dimension of the gut endoderm is .084 mm. while the transverse dimension is .056 mm. The total length of the hind gut in this later stage is 1.70 mm. as compared with .76 in the 4 mm. embryo. It is seen that there has been a very decided diminution in the diameter of the gut, and, furthermore, that this is out of proportion to the increase in length. It is correlated with a thickening of the gut wall, due to the drawing together of the component cells.
In the 4 mm. stage the gut entoderm was composed of a single layer of cells, while in the 6.8 mm. stage under consideration its lateral and ventral portions are made up of two, and in some places three irregularly arranged layers of cells, while the dorsal wall is made up of a single layer as in the 4 mm. stage. Two series chosen from several of this stage may be taken as showing typical differences. Both show an advance over the preceding stages in the greater ease with which the sex-cells may be distinguished. This is due to the contrast between the ordinary entoderm cells in which a considerable amount of the yolk material has been absorbed and the more rounded yolk-filled sex-cells. In neither embryo has the process of sex-cell migration commenced. This is clearly evident in one, while in the other there are a very few scattered yolk-filled cells of rather problematical character in the loose mesenchyme above and at the sides of the gut entoderm. One striking individual difference between • these two specimens is found in the fact that while in one the sex-cells have retained their primitive position in the ventral and lateral portions of the gut entoderm, in the other they have migrated up into its dorsal portions. Although it is somewhat difficult to establish with absolute certainty this migration from the ventral to the dorsal portions of the gut entoderm owing to the difficulty of distinguishing the sex-cells in the preceding stages, the individual differences in this regard observed in this stage, together with the fairly reliable observations upon earlier stages seem to point to an actual migration of this character.
BENNET M. ALLEN
Lepidosteus 7.3 mm. total length. In the single series of 7.3 mm. embryos the sex-cells in general still occupy the ventral and lateral portions of the gut entoderm, having come to lie in the dorsal wall at only a few points, especially toward the cranial end of the hind gut. A very few sex-cells are found to have migrated into the loose mesenchyme between the gut entoderm and lateral plates of mesoderm, occupying positions in it lateral and dorsal to the latter. These migrant cells are merely the precursors of a general migration which does not become conspicuous until the embryo has reached a length of 8.5 mm.
TABLE 1 Number of sex-cells in Lepidosteus
NUMBER
Entoderm
Mesoderm
mm.
6.8 8.6
No count
15
73
9.2
"
136
9.3A
«
41
9.3B
((
311
9.3 C
«
425
10.7
133
125
674
S07
Int.
Root
R.
L.
12.0
104
163
180
179
751
14.1
128
222
37
197
153
737
17.0
No count
262
235
18.0
"
171
173
24,0
'(
147
154
Int. — Intestine.
Root — Root of mesentery.
R. — Right sex-gland. L. — Left sex-gland.
Lepidosteus, 8.6 vim. total length. Passing over several intermediate stages studied, the conditions found in a specimen 8.6 mm. long may be described. At this stage the lateral plates of mesoderm are just beginning to split and to form the coelomic cavities (fig. 7). The interval between the plates is filled with loose mesenchyme, of which that portion lying between the gut entoderm and the aorta will later be condensed by the apposition of the lateral plates of mesoderm in formation of the mesentery.
SEX-CELLS OF AMIA AND LEPIDOSTEUS 5
It will be seen by comparing the figures drawn to scale that the mesentery is at this stage not only relatively but actually much thicker than it is during later stages. The accompanying table serves to show the number of sex-cells and their distribution in certain stages.
In this specimen 73 sex-cells have already migrated out of the gut entoderm into the surrounding mesoderm tissues. While .most of them have migrated upward into the loose mesenchyme and splanchnopleuric mesoderm of the anlage of the mesentery, a few have passed laterally into the portions of the splanchnopleuric mesoderm that enter into the formation of the intestinal wall. While the migration of the sex-cells is seen to be well under way at this stage, the great majority of them still remain in the dorsal portion of the gut entoderm. Very few, indeed, are to be found in the ventral half at this stage. The sex-cells of the gut entoderm are easily distinguishable from the other entoderm cells; the latter have lost very nearly all of their yolk material and have become cylindrical in shape. These features stand out in sharp contrast to the large yolk content and spherical form of the sex-cells.
The migration of sex-cells from the gut entoderm into the mesenchyme dorsal or lateral to it may be clearly seen at a few points, as illustrated in figs. 7 and 8. They retain for the most part their spherical form, but cases like that shown in fig. 7 can be readily found. The shape of this sex-cell clearly indicates the mode of progression. They, undoubtedly, pass through the loose network of mesenchyme by an amoeboid movement, however slow or intermittent it may be.
In this stage sex-cells are found in the hind gut from its cranial end to within .2 mm. of the cloaca, a distance of 2.6 mm.
Lepidosteus 9.2 mm. long. The number of sex-cells that have migrated out of the gut entoderm is 136 in this specimen. The number of these is still increasing but solely by migration from the entoderm, since there is no evidence of division of the sexcells during these stages of sex-cell migration.
In this stage the coelomic cavities have appeared in the dorsal portion of each lateral mesodermal plate and the mesentery is
6 BENNET M. ALLEN
consequently much more clearly defined. In three specimens 9.3 mm. long the following counts of sex-cells outside of the gut entoderm were made:
A = 41 B = 311 C = 425.
A and C are extreme cases, indicating that the process of migration is an irregular one in point of time. The mesentery in specimen A is .46 mm. wide.
In specimen C those sex-cells destined to migrate out of the entoderm have for the most part already done so, while in A, an embryo of the same stage, the process is just beginning. The coelome is least developed in A and furthest advanced in C. This would indicate that the extent to which the migration of sex-cells has been carried on is correlated with the degree of development of the mesentery, resulting from the enlargement of the coelomic cavities. WTiile these three specimens belong, no doubt, to slightly different stages of development, they were very carefullj^ matched as to length, and are most certainly of the same age.
Lepidosteus 10.7 mm. total length. At this period of development, the mesentery is well formed, being much thinner (.18 mm.) than in the 9.3 mm. stage. This results in its possessing a denser texture (fig. 9). The great majority of the sex-cells are scattered through the mesentery, showing no definite arrangement; but lying for the most part in the mesenchyme enclosed between the somewhat denser splanchnic laj'ers of mesoderm. A few are found in the mesodermal layers of the intestine, while a fairly considerable number have remained in the gut entoderm. At this time such sex-cells as are found in any but the dorsal wall of the intestine, at its junction with the mesentery, are most probably destined to remain in their present positions. A few of the sex-cells have migrated to places immediately dorsal and lateral to the root of the mesentery. The latter may be considered to have reached the sex-gland anlagen, although their position relative to the root of the mesentery will be shifted, as we shall see, in the later stages, probably by a general shifting of the tissues in which they lie.
SEX-CELLS OF AMIA AND LEPIDOSTEUS 7
The number of sex-cells which have migrated from the entoderm is found to be 674. It can be fairly taken as the number destined to undergo migration from the entoderm in this particular individual. Those still remaining in the entoderm number 133.
A few scattered sex-cells are found as far forward as the cranial end of the hind gut. The latter is 3.41 mm. in length. Opposite to the cranial portion of the hind gut the sex-cells are rather sparse, increasing in number as one follows the series caudally. The}'- become most numerous a short distance caudal to a point two-thirds the distance from the cranial to the caudal end of the hind gut. The last one in the mesoderm is found at a distance of .46 mm. from the cloaca, and the last one in the entoderm lies at a point .27 mm. from the cloaca.
Lepidosleus 12 mm. total length. At this period migration of the sex-cells has progressed to the point where most of them have reached their final positions. They are still to be seen in the entoderm. This number (125) is quite close to that (133) of the similarly situated sex-cells of the preceding stage. The density of the mesodermal tissues surrounding the gut entoderm makes it seem quite unlikely that any more sex-cells could migrate into them from this source.
The distribution of the sex-cells is as follows:
392 outside of sex-gland anlagen.
Gut entoderm 125
Mesoderm of intestinal wall and Mesentery 104
Root of mesentery between sexgland anlagen 163
Right sex-gland 180l
Left sex-gland 179/ ^59 ^ sex-gland anlagen.
Total 751
The table may be allowed to speak for itself. The sex-gland anlagen grade into one another by an intermediate region at the root of the mesentery. More or less arbitrary limits had to be assumed to distinguish between these three regions. In later stages, illustrated by the 17 mm. stage, fig. 11, we shall see
8 BENNET M. ALLEN
that the sex-cells undergo lateral migration, either apparent or real, so as eventually to lie at some distance on each side of the median point.
The narrowest portion of the mesentery is at about one-quarter the distance from its origin to its insertion. Its minimum width, as measured here amounts to .028 mm., thus showing a great reduction as compared with the 10.7 mm. stage. This reduction in width is shared by the entire mesenterj', certain regions remaining broad only on account of the enclosed blood vessels. No doubt the migration of the sex-cells out of the mesentery is in large part responsible for this, but a considerable share of it must be ascribed to the fact that there has been a tendency for the tissues to become more compact.
The total length of the hind gut at this stage has reached 4.03 mm. Sex-cells are found in the entoderm at its cranial end, and from there extend to within 0.62 mm. of the cloaca. The distribution of sex-cells within the sex-gland anlagen is somewhat more restricted, since they extend from a point 0.31 mm. caudal to the beginning of the hind gut, to a point 1.00 mm. cranial to the cloaca. The}' are rather sparse at these two extremes.
As in the preceding stages, there is no clear evidence of division of the sex-ceUs, although one can not be absolutely certain upon this point. While at this time many are free from yolk material, others show but little diminution of it. It is true that the sexcells are often found arranged in clusters, but there is no evidence to show that these are due to repeated division of a parent cell rather than to a tendency for them to congregate through mutual attraction. ^Yhsit the nature of this attraction might be, we do not know; but it might well be akin to that influence which causes the sex-ceUs to migrate toward the sex-gland anlagen from their source. Similar clusters of sex-cells were found in earty stages in Chrysemys.
Lepidosteus 14-1 inm. total length. Little radical change is to be seen in this stage. The sex-cells were counted and gave the following results:
SEX-CELLS OF AMIA AND LEPIDOSTEUS
Gut entoderm 128
Mesoderm of intestinal wall and
mesentery 222
Root of mesentery between S. G.
anlagen 37
Right sex-gland 197 \
Left sex-gland 153 /
Total 737
387 outside of sex-gland anlagen.
- 350 in sex-gland anlagen.
There is a strikingly close correspondence between the results of the count hi this specimen and those in the preceding one. Attention may be called to the fact that in this specimen a materially greater number of sex-ceUs is found in the right sexgland than in the left. At the same time there is a very close correspondence in the total number of sex-cells that have reached the sex-gland anlagen as compared with the total number in the 12.0 mm. stage (359).
Lepidosteus 17 mm. total length. In this specimen those sexcells destined to occupy the sex-glands are seen to have migrated some distance to each side of the root of the mesentery, fig. 11. Their position relative to the root of the mesentery and to the Wolffian duct varies at different points along the sex-gland anlage. In the most cranial portion of the latter they lie just medial to the Wolffian duct. As one follows the sex-glands caudally, the sex-cells are found to lie closer and closer to the mesentery, being situated midway between the latter and the Wolffian duct in the middle region of the sex-gland anlage. The most caudally situated sex-cells lie close to the root of the mesentery.
In this and the succeeding stages the intestine had become so voluminous as to make the counting of the sex-cells in its walls very difficult and inaccurate. It is in fact not easy to distinguish them from the cells of the gut entoderm because of their rather small size and their entire lack of yolk material at this stage.
The total number of sex-cells in the sex-gland anlagen of this specimen is rather high, there being 235 in the left sex-gland and 262 in the right. The total number is 497.
10 BENNET M. ALLEN
The slightly greater number of sex-cells in the sex-glands of this specimen as compared with that in the previous ones is of little significance. It most certainly does not indicate that there has been any extensive division of them. In a previous work upon Chrysemys, (Allen '07), it was shown that there was an extreme amount of individual variation in the number of sex-cells. This variation in Lepidosteus is relatively slight compared with that observed in Chrysemys. In a specimen slightly older than this stage (18 mm.) there were 171 sex-cells in the right sex-gland, and 173 in the left one, the total number, 344, being not far from the average.
In these two stages, 17 and 18 mm., the sex-cells usually occur singly, although in places they are aggregated into clumps so thick as sometimes to show as many as five or six in a section of one of the sex-glands. Whether the sex-cells occur singly or in clumps, they are surrounded by peritoneal cells which contribute materially to the formation of the ridge-like anlage of the sex-gland.
Lepidosteus 21+ mm. total length. In a specimen of this length, fig. 12, there is no essential advance in the development of the sex-gland. There were 147 sex-cells in the right sex-gland, and 154 in the left one. The total number, 301, is distinctly below the average.
Comparison with other forms leaves no room for doubt as to the identity of these sex-cells. Since the aim of this paper is merely to trace out their origin, we will not follow them through later stages in their history, but will describe the conditions found in a specimen 110 mm., in length, fig. 13. A complete series of sections through the sex-gland region of this specimen was not made, so it is impossible to give a full account as to the number of sex-cells and general condition of the sex-gland at this time. In running through the series one is struck with the sparseness of the sex-cells. Never are more than two or three to be found in a single section, and often none at all. This would lead one to infer that there has been little or no multiplication of the sexcells even at this late stage of development.
SEX-CELLS OF AMIA AND LEPIDOSTEUS
11
A glance at table 2 shows that there is a general tendency to a reduction in the average size of the cell body in the later stages. This may be due to the absorption of the contained yolk material. There is no marked change in the size of the nucleus.
TABLE 2
Dimensions of sex-cells of Lepidosteus
CELL BODY
Stage
Nucleus
LARGEST
SMALLEST
AVERAGE
mm.
8.6
6.04
15,10
12.08
13.74
9.3
6.04
18.12
12.08
14.95
10.7
6.04
15.10
11.32
13.59 .
14.0
5.81
12.08
9.06
10.27
17.0
6.04
13.59
9.06
11.63
24.0
5.81
9.06
7.55
8.65
110.0
6.53
14.50
9.22
12.40
AMIA CALVA
Amia 4- fnm., total length. In the text figure A is shown a transverse section of an Amia larva of this stage. It will serve as a starting point from which we shall proceed to consider still earlier stages in tracing out the earliest phases in the origin and migration of the sex-cells. The section shown is taken "just anterior to the hind gut, the gut entoderm being clearly marked by its greater thickness and dorsal curvature. The cavity of the intestine at this point opens into the large sub-germinal cavity. The extra embryonic portions of the entoderm, i.e., those which do not form part of the anlagen of the alimentary tract and its appendages can logically be divided into four different regions : (1) The roof of the sub-germinal cavity which is distinguishable from the gut entoderm, as indicated; (2) The layer forming the floor of the sub-germinal cavity; (3) The peripheral layer of entoderm lateral to the sub-germinal cavity (peripheral entoderm) ; (4) The central yolk mass, or vitellus (vitelline entoderm). In the first three of these regions the cells are arranged in a single layer. They are characterized by the fact that the yolk spherules of the component cells are distinctly smaller than are those of
12
BENNET M. ALLEN
the vitellus, their diameter being from one-quarter to one-half of that of the typical spherules to the vitellus. In the latter cells are scattered a few of these smaller yolk spherules; but the distinction between the first three divisions of the entoderm and the vitellus is a very sharp one.
Perjph. C
Text figure A
In connection with this distinction it is interesting to note that the yolk spherules along the cleavage planes that cut through the vitellus are found to belong to this small type. It is easy to see that if the vitellus were cut up into cells as small as those comprising portions 1, 2, and 3 of the entoderm, the thickness of the layers of small spherules which form merely a border to the large cells would be so great as to comprise the entire body of the more finely divided ones. This difference in the size of the
SEX-CELLS OF AMIA AND LEPIDOSTEUS 13
yolk spherules is then probably associated with the difference in the size of the cells. The peripheral entodermal layer which we have designated as division three is interrupted lateroventrally by blood vessels lying in the mesoderm.
The lateral plates of mesoderm have long since broken away from the mesoblastic somites. Their inner margins lie at some distance to each side of the median line. While there is the slightest tendency in places for the splanchnic and somatic layers of mesoderm to split apart along the medial margins of the lateral plates, the remainder of the lateral plates show no indication of a splitting, even in the arrangement of the nuclei. It is, however, quite probable that such a plane of cleavage is already laid down. This is shown by the sex-cells (text fig. A) being imbedded in the lateral plates. When the somatopleure and splanchnopleure separate later, these will be found to lie in the coelomic cavity, being for a time merely adherent to the coelomic surface of the medial portion of the somatic mesoderm. One can fairly assume that during the period of migration, represented by fig. 5, the sex-cells push their way between the two layers of mesoderm following the potential cleft that separates them.
Text fig. A is very suggestive, as it shows sex-cells situated at intervals from a point just beyond that at which the roof and floor entoderm join the peripheral entoderm. The path of their migration is thus clearly marked out. In this figure it should be noted that the most laterally situated sex-cell lies in the entoderm, while all of the others are clearly imbedded in the lateral plates of mesoderm as already indicated. In but one or two of the many specimens examined were there any sex-cells found in the roof or gut entoderm. They arise in the peripheral entoderm from which they migrate into the lateral plates of mesoderm and through them to their medial borders, whence, as I shall later show, they pass into the sex-gland anlagen after the formation of the coelomic cavity.
The total number of sex-cells found in the mesoderm of the specimen of this stage was 87. Of these 40 were found on the right side and 47 on the left. Text fig. A will indicate their distribution.
14
BENNET M. ALLEN
Table 3 serves to show for purposes of comparison the numbers of sex-cells found in different specimens of Amia.
TABLE 3 Number of sex-cells in Amia calva
A.GE
SPECIMEN
NUMBER
3P SEX-CELLS IN MESODERM
ST
R.
L. Total
Hours.
mm.
132
A
None
None
132
B
None
None 1
132
c
None
None
132
D
None
None
137
A
7
4 1 11
137
B
9
7
16
137
C
21
8
29
147
A
15
7
22
147
B
14
17 31
147
C
22
11 33
147
D
48
18 66
147
E
39
28 67
147
F
50
26 76
155
3.0
15
34 49
3.4
62
41 i 103
3.5
A
39
53
92
3.5
B
59
48
107
3.7
A
42
30
72
3.7
B
45
47
92
4.0
40 ■
47
87
5.0
23
20
43
6.6
42
56
98
6.0
28
34
62
7.0
38
36
74
7.6
33
42
75
9.1
36
40
76
11.4
28
54
82
15.0
A
28
49
77
15.0
B
38
45
83
16.0
A
19
14
33
16.0
B
22
17
39
16.0
C
99
23.7
47
55
102
This stage is a convenient starting point from which to proceed in the study of earlier stages.
SEX-CELLS OF AMIA AND LEPIDOSTEUS 15
A7?iia 3.7 mm. total length. The conditions are, in the main, quite similar to those found in the 4 mm. stage. In one of the two specimens (B) in which the sex-cells were counted there were 92 sexcells in the mesoderm and 10 in the entoderm. Although this total number of 102 is greater than the number found in the 4 mm. stage (87), yet, as shown in table 2, no significance is to be attached to this on account of the great individual variation in the number of sex-cells observed, not only in Amia, but also shown by the author to be so obvious in the turtle, Chrysemys. In A of this stage, 72 sex-cells were found, 42 on the right and 30 on the left side.
Amia 3.5 mm. total length. Two larvae of this stage were studied. It was rather difficult to measure the specimens accurately, owing to the fact that the caudal portion of the body free from the yolk has a strong ventral bend. It can be straightened out only in later stages. The two specimens of this length were taken from the same nest and both are distinctly younger than the preceding, yet they showed decided differences from one another in the positions occupied by the sex-cells, probably owing to the fact that this, in all likelihood, is the period of their most active migration. In specimen A the sex-cells are quite numerous in the portion of the lateral plate of mesoderm, which lies immediately above the border of the subgerminal cavity. They occur in fair numbers in the mesoderm between this region and a point one-half the distance from this point to the median edge of the lateral plate of mesoderm. Only three were found nearer the median line than this. Of these, one had scarcely passed the midway point, one was still some distance from the median edge of the lateral plate, while one had actually reached that point.
In specimen B of this stage a large proportion of the sex-cells have reached the median edge of the lateral plate of mesoderm of each side. This is especially noticeable on the right side. The conditions in this specimen approach those described for the 4 mm. stage but do not show quite such an advanced condition, owing to the fact that a larger proportion of sex-cells are scattered along the outer portions of what we may call the sex-cell path. There
16
BENNET M. ALLEN
were noted two or three instances in which the sex-cells were migrating from the peripheral entoderm into the mesoderm.
Amia, 3 mm., total length; 155 hours. In a specimen of 3.0 mm. total length, the free caudal portion has but recently separated from the vitelline mass, and has attained a total length of .56 mm. By comparison with a number of embryos of 132, 137, and 147 hours old, the age of this embyro was estimated to be very close to 155 hours. This estimate was made by counting the number of sections passing through the posterior part of the embryo free from tJie yolk mass. Sufficient numbers of embryos were used to give a fairly accurate determination, there being seven specimens of the 147-hour, three of the 137-hour, and two of the 132-hour stages studied.
TABLE 4
The numbers of sex-cells in each were as follows:
RIGHT SIDE
LEFT SIDE
TOTAL
In A
39
59
53
48
92
In B
107
There were 49 sex-cells counted in the 3 mm., 155-hour embryo. This, it will be seen, is decidedly below the average and yet the number is greater than that found in the 5 mm. stage and in the much later 16 mm. specimens.
Only two of the sex-cells have migrated a very short distance along the lateral plate of mesoderm, beyond a point overlying the lateral boundary of the subgerminal cavity; the remainder of them all lie lateral to it. It will thus be seen that they show a much earlier phase of migration than that observed in the 3.5 mm. embryo, not only as regards the number that have migrated into the mesoderm, but likewise in the distance through which they have travelled in their journey in that layer toward the sexgland anlagen.
Amia, 11^7 hr. stage. That there is a great amount of individual variation in the rapidity with which this migration from the peripheral entoderm to the lateral plates of mesoderm is accomplished may be readily seen by referring to the numbers counted
SEX-CELLS OF AMIA AND LEPIDOSTEUS 17
in the mesoderm of seven specimens of the 147 hour stage. These specimens were all taken from the same nest and kept in the same dish, so there can be but very slight difference in their ages, due, if it exists, to the small difference in the time at which the eggs were laid. It will be seen that the total number of sexcells in the entoderm in these specimens varies from 22 to 76. The latter number is not only greater than that observed in the 3 mm., 155 hour stage, but almost equals that counted in many specimens of older stages after migration has been completed, as, for instance, the 11.4 mm. and 15 mm. stages (see table). In this stage clearly defined sex-cells can be seen in the peripheral entoderm just below the lateral plates of mesoderm, figs. 16 and 17. These cells are distinguishable from the other entoderm cells among which they lie, by the greater size of their contained yolk granules as contrasted with the small size of the yolk granules in the other cells that make up this layer. The difference is further marked by the more rounded form of the sex-cells. Comparison of these sex-cells in the peripheral entoderm shows them to be identical with other more clearly defined sex-cells in the mesoderm. Of this identity there can be no question, and it is equally clear, from a study of later stages, that these cells, having once migrated into the lateral plates of mesoderm, pass unaltered along the latter to come finally to rest in the sex-gland anlagen. There can be no doubt about the origin of the sex-cells from the entoderm. A number of cases were observed in which the sexcells were actually in process of passing from the peripheral entoderm into the lateral plates of mesoderm.
At this stage, sex-cells have a wide distribution in the peripheral entoderm, being scattered through a region extending from a point opposite to the region where the blood cells originated to the junction of the peripheral, sub-germinal and roof entoderm. In three specimens of the 137 hour stage, conditions are quite similar to the foregoing. In these embryos the number of sexcells ranged from 11 to 29. It will be seen that the maximum number of sex-cells counted in this stage is greater than the minimum number of the 147 hour stage, although in all three of these 137 hour embryos, the caudal end of the embryo, that part
JOURNA.L OF MORPHOLOGY, VOL. 22, NO. 1
18 BENNET M. ALLEN
that has been Hfted off the yolk, is decidedly shorter than in any of the 147 hour specimens.
Amia, 132 hr. stage. In four specimens of the 132 hour stage, the caudal end of the embryo was just ready to undergo separation from the yolk. Only in one of them had this really commenced, the separated portion having reached a length of but 20)U. Not one of these four specimens showed a single sex-cell in the mesoderm. There can be no question upon this point because they could be very readily detected if present. In the 137 and 147 hour stages those that migrated into the mesoderm stand out most clearly and sharply from the surrounding mesodermal cells. The points of difference between the two kinds of cells are very striking and unmistakable. The sex-cells on the one hand are large, spherical, have sharply defined boundaries, and are filled with large oval yolk grains; while the mesodermal cells are small, flattened, syncytial, and contain a very few minute yolk granules.
It is very much more difficult to trace the earlier history of the sex-cells in the peripheral entoderm, owing to the slight differences that may be taken as criteria in distinguishing them from the neighboring entoderm cells. Numbers of cells with all the characteristics of sex-cells are found just beneath the anlagen of the blood masses. This stage is just before the development of blood vessels within the embryo, and the blood-forming cells occur in the form of two sharply limited bands, one on each side of the embryo and at some distance lateral to it. Here and there, sex-cells are found in the peripheral entoderm, medial to these areas; but clearly defined cases of this sort are rather rare as compared with the large number seen in this region a little later in the 147 hour stage. It is quite likely that many of these sex-cells are overlooked at this stage owing to the fact that the neighboring entodermal cells contain rather large yolk grains at this time, while those seen in these cells in the 147 hour stage are much smaller than at this stage.
It is quite possible that the sex-cells may migrate medially in the entoderm from an entodermal source beneath the blood anlagen to various points between this region and the edge of the sub-germinal cavity. It is possible that a large proportion
SEX-CELLS OF AMIA AND LEPIDOSTEUS 19
of them may have developed in the peripheral entoderm throughout this entire extent. On the other hand, it is also possible that sex-cells may migrate up into this region from the central entoderm beneath.
We have traced the history of the sex-cells from the 4 mm. stage where they are readily identified by any one who has had any experience in observing these cells, back to the earliest stage at which they are distinguishable in the entoderm. We shall now follow them up to the period when they are enclosed in the definitely formed sex-glands and finally to the stage at which they are found to have begun to increase in number.
Amia 5 mm., total length. Passing from the 4 mm. stage to the next represented in our series, 5 mm., we find that the sexcells have made but little progress in their migration toward the median edge of the lateral mesodermal plates. The total number of sex-cells counted in this stage was surprisingly small, being 43 as compared with 87 in the 4 mm. stage. This difference in number is probably due to individual variation. The hind gut has materially lengthened, being 1.3 mm. in length, compared with .88 mm. in the 4 mm. stage. There has been a corresponding increase in the length of the region over which the sexcells are distributed. In the 4 mm. stage they extend from a point 0.06 mm., in front of the beginning of the hind gut, caudally to a distance of 0.35 mm. In the 5 mm. stage that we are considering, this region begins at the same point relative to the hind-gut and extends caudally for 0.50 mm., one isolated sexcell being found at a distance of 0.57 mm. behind the cranial limit of their distribution.
In the more caudal portion of this region the splanchnic and somatic layers of mesoderm have begun to separate to form the coelome. This separation does not at first lead to the formation of a continuous cavity but rather to a series of isolated, somewhat rounded cavities. Further caudad, the coelome becomes more and more completely developed, appearing as a large cavity on each side.
Amia 6 mm., total length. At this time the first sex-cells appear in the splanchnopleure just at the entrance of the hind gut.
20 BENNET M. ALLEN
The first sex-cells in the somatopleure are found in the sex-gland anlagen a short distance (0.04 mm.) behind this point. The sexcells are distributed somewhat irregularly from the cranial end of the hind-gut to a point 0.90 mm., caudad to this point and there are a few scattering sex-cells still further caudad than this.
The coelome is apparent as a continuous cleft on either side of the hind-gut along the entire extent of the region occupied by the sex-cells. The majority of the sex-cells are to be found in the dorso-medial extremity of the coelome, i.e., near the root of the mesentery. A few lie lateral and ventral to the intestine. The coelomic cleft has not as yet become wider than the diameter of the average sex-cell and we consequently see them usually bridging across it, fig. 18. In no case have they penetrated into the somatic mesoderm as we find them doing later. One sex-cell was found in the gut-entoderm, whither it may have migrated from the mesoderm. It is, on the other hand, quite possible for it to have migrated in the entoderm in the manner of sex-cell migration in the turtle. This is a point of minor significance and an occurrence which is at best very infrequent.
Amia, 7 mm., total length. Up to this time, the mesentery has been only potentially present, the two lateral plates of mesoderm being in contact above the gut-entoderm. Now, however, we find that it has begun to elongate and become thin. This is naturally correlated with the increase in the extent of the coelome, fig. 19. Two well defined sex-cells are found in the gut-entoderm, 0.06 mm., cranial to the opening of the hind-gut. These are to be interpreted in the same way as the cell in the entoderm mentioned above. The first sex-cell occurring in the mesoderm is found 0.08 mm. caudad to the beginning of the hind-gut. The sex-cells are distributed through a region extending from a point immediately back of the opening of the hind-gut to a point 1.05 mm. behind it, with a few scattering ones behind these. The total number of sex-cells is 74.
Amia, 9.1 stage. Sex-cells first appear .18 mm. cranial to the opening of the hind-gut. They extend from this point to a point 1.59 mm. caudad to this, giving a total extent of 1.67 mm. The total number of sex-cells counted at this stage amounted to
SEX-CELLS OF AMIA AND LEPIDOSTEUS 21
76. Of these all were in the sex-gland anlagen except three; one of which occurred in the gut-entoderm and two in the parietal peritoneum. I am inclined to consider it unlikely for these misplaced sex-cells to reach the sex-glands. One is struck, however, with the great difference in the relative number of misplaced sexcells in Amia as compared with Lepidosteus. This may be apparent rather than real, owing to the possibility that in Amia large numbers of them may have failed to migrate from the entoderm into the mesoderm during early stages. Owing to the difficulty of certainly distinguishing sex-cells in the entoderm from ordinary entoderm cells, it was quite impossible to make any count of those left behind in migration. All but a very few, however, that reach the mesoderm succeed, as we have seen, in reaching the sexgland anlagen. A considerable number of cells seen in the entoderm in later stages contain small yolk spherules and show other points of resemblance to sex-cells. In this stage the mesentery has become quite lengthened and the coelome very large. The sex-cells have penetrated into the root of the mesentery, fig. 20.
The sex-cells, with rare exceptions, still contain large quantities of yolk material. In these exceptional cases a finely granular appearance gives at least the suggestion of small unstained yolk spherules. The yolk appears in the shape of particles varying in size from small granules up to large lemon-shaped pieces quite as large as those with which the cells of the yolk entoderm are so completely filled.
Amia 11. j^ mm., total length. The sex-cells are fairly numerous over a region 1.85 mm. in length, beginning at a .point 0.06 mm. back of the yolk stalk and ending at a point 0.85 nam. cranial to the cloaca. Two isolated sex-cells are found caudad to the point named, one of them occurring very close to the cloaca. Their total number in this embryo is eighty- two. The sex-cells have much the same characteristics as in the previous stage.
This stage is marked by a decided increase in the length of the mesentery and by a decrease in the size of the yolk-sac, which is now but 0.7 mm. in diameter and is greatly hollowed out to form a portion of the intestinal wall.
22 BENNET M. ALLEN
While the sex-cells of the 9.1 mm. stage are imbedded in the mesoderm at the root of the mesentery and always close to the median line, they are found in the 11.4 mm. stage to occupy a position a short distance on each side of this point. Not only have they moved laterally, but they have also protruded into the body cavity, accompanied by a few mesoderm cells which are intercalated between them, fig. 21, and surround them with a thin peritoneal investment as well.
Amia 15 mm., total length. In this stage the sex-cells extend over a distance of 2.70 mm. in the caudad 0.50 mm. of which they are very sparse. The sex-glands protrude further into the body cavity than in the preceding stage, and the ligament of attachment becomes narrower. The genital ridge is very much lower in the gaps between sex-cells than it is in the sex-cell regions. In spite of the fact that it may be very low for quite a distance, it is continuous throughout. The genital ridges diverge quite widely at their cranial ends, approaching the median line at a point .4 mm. caudad to their point of commencement.
The sex-cells have almost uniformly used up their contained yolk material, although a few scattered ones are still closely packed full of them. The sex-cells in specimen A, numbered 28 on the right side and 49 on the left, the total number being 77. The number of sex-cells in specimen B was 38 on the right side and 45 on the left, the total being 83.
Amia 16 mm. long. In two 16 mm. larvae, conditions very similar to those of the 15 mm. stage were found. None of the sex-cells contained yolk material in a sufficiently large amount to be clearly recognizable. The striking thing about these two specimens is the very small number of sex-cells present, 33 in one case and 39 in another. There is no indication of degeneration or of a failure to migrate to the proper positions.- The case seems to be similar to one cited in Chrysemys, both being due to individual variation.
These two specimens were taken from the same brood and no doubt had the same parentage. Another 16 mm. specimen taken from a different brood showed 09 sex-cells, a number not very far below the maximum. From this fact, and from the
SEX-CELLS OF AMIA AND LEPIDOSTEUS
23
total absence of any indication of degeneration of sex-cells in these or earlier stages, I feel convinced that this small number does not indicate any tendency to degeneration of sex-cells.
Amia, 23.7 mm. total length. In the next stage studied, 23.7 mm., the sex-cells numbered 102. Here again there is no evidence of a change in the number of sex-cells originally present. The number, although somewhat high, is exceeded by some of the specimens of very much earlier stages. There is no evidence of sex-cell division nor of any degeneration.
Amia, 1^0 mm. total length. At this stage the sex-gland is elongated oval in transverse section. It has become bent over in such a way that the proximal edge is medial and the free edge
TABLE 5
Dimensions of sex-cells of Amia
STAGE
NUCLEUS
CELL BODY
Hours
137
7.10
18.03
147
6.71
18.70
mm.
3.7
6.45
21.88
5.0
6.51
17.80
9.1
8.00
14.96
11.4
7.48
11.59
15.0
7.48
12.64
16.0
7.74
14.06
23.0
7.22
14.20
lateral in position. The mesodermal cells have increased greatly in number. The peripheral cells have become arranged into a somewhat poorly defined layer, while the sex-cells lie in the interior of the sex-gland. No attempt was made to determine the time at which the sex-cells begin to divide, or to study the further, development of the sex-glands.
Measurements of the nuclei and cell bodies of the sex-cells gave the following averages, two diameters being measured in each of five sex-cells chosen at random in each stage.
Although the number of cells measured in each stage is hardly sufficient to justify one in considering these average dimensions to have any high degree of accuracy, I feel that we are quite justified in concluding from these figures that: (1) there is a fair decrease in the size of the cell-body as development proceeds, and (2) that there is a slight increase in the size of the nucleus. The decrease in the size of the cell-body is probably due to the absorption of the yolk material with which the sex-cells are so richly filled during the earlier stages. No good explanation to account for the slight apparent increase in size of the nucleus presents itself.
DISCUSSION OF RESULTS
We can not consider this work as completed without making a comparison between the sex-cells and the other cells of the embryo. This subject will first be taken up in Amia where we have traced the sex-cells back to earlier stages than in Lepidosteus. It has already been pointed out that the sex-cells, as first seen in the peripheral entoderm, are to be distinguished only by the size and arrangement of the yolk spherules. The nuclei bear a close resemblance to those of surrounding cells of the same size, while the larger nuclei of larger cells show many points of similarity to them. In all except the earliest stages studied, these nuclei are quite rounded. The chromatin appears in the form of slender strands that take a peripheral position in the nucleus. There is invariably a plasmosome present and rarely two of them. In the 147 hour stage the nuclei of the sex-cells bear a resemblance not only to those of the neighboring cells but also to those of the gut entoderm. In fact, many nuclei of the mesoderm show similar characteristics.
After development has gone a little further, as in the 3.4 mm. and 4 mm. stages, the mesodermal nuclei and those of the gut entoderm are found to have become smaller and are more deeply stained than those of the sex-cells and peripheral entoderm. In all of these later stages, which include 5 mm., 6 mm., 9. 1 mm., 11.4 mm. and 16 mm. larvae, these differences are found to increase. Although the sex-cells undergo a migration from the peripheral entoderm into the lateral plates of mesoderm and through the latter to the sex-gland anlagen, they still bear a close resemblance
SEX-CELLS OF AMIA AND LEPIDOSTEUS 25
to certain cells of the peripheral entoderm. This not only involves a similarity of the nuclei but of the dimensions of the cell bodies. This is true even after the sex-cells and the corresponding cells of the peripheral entoderm have lost their yolk through absorption.
In the stage of 11.4 mm., the yolk mass has been greatly reduced (figs. 25 and 26). Only here and there about its periphery are cells to be found with well defined outlines. The great mass is syncytial, with large nuclei of varying size scattered here and there. While these nuclei of the vitelline mass are much larger than the sex-cell nuclei, they bear a close resemblance to the latter. The nuclei of the well defined peripheral cells are practically identical in size and appearance with those of the sex-cells.
While the similarity between sex-cells and between these two classes of cells is not so marked in Lepidosteus as in Amia, yet it appears to be equally true. In the 17 mm. stage (figs. 14 and 15) the yolk mass is still of fair size. There is a layer of peripheral entoderm that is largely made up of cells with clear boundaries, whose nuclei are similar to those of the sex-cells in respect to the presence and character of the plasmosome and in the form and distribution of the chromatin material. In many cases these nuclei are larger than those of the sex-cells; but many are found which are quite as small. These grade into the very large nuclei of the syncytial vitelline entoderm.
At this stage the tissues of the body have taken on their distinctive characters and their component cells have undergone in many cases a high degree of specialization. This emphasizes strongly the similarity between the sex-cells and the cells of the peripheral entoderm.
As we pass back to earlier stages, such as those of 9.3 mm., 5.9 mm., etc., we still find this similarity between these types of cells, although the nuclei of all the body cells tend to show greater and greater similarity to one another in the earlier stages. For instance, it becomes quite difficult to distinguish the nuclei of the gut entoderm cells from those of the sex-cells. Even the nuclei of the Wolffian ducts show quite a close resemblance to the sexcell nuclei during the early stages of development.
26 BENNET M. ALLEN
There are two ways of viewing the similarity that the sexcells of Amia and Lepidosteus bear to these cells of the peripheral entoderm. The well defined cells of the peripheral entoderm might be interpreted as sex-cells that have failed to migrate into the lateral plates of mesoderm. It would then remain to give an explanation of the resemblance that the nuclei of these cells bear to the nuclei of the vitelline entoderm and to account for the intermediate types of nuclei by which they grade into one another.
The other view of this problem is to consider sex-cells, peripheral entoderm cells, and vitelline entoderm cells as slightly differentiated blastomeres, dating from an early stage of development, and to consider the similarity that they bear to the cells of the peripheral entoderm as due to the fact that they too have remained in a relatively slightly differentiated condition. This view seems the more probable of the two. It is by no means a new one, having been advanced by Nussbaum in 1880.
It would be rash in the extreme to claim that the sex-cells might not differ in some essential chromosomal characters from the cells of the peripheral entoderm which they so closely resemble, and yet careful study has failed as yet to show any real differences. While such differences may exist, these cells all have much in common with one another.
In a recent paper by A. P. Dustin ('07), this author gives a new view of the origin and movements of the sex-cells of Triton alpestris, Rana fusca and Bufo vulgaris. Since his view is so greatly at variance with my own, it will be necessary to review this work in some detail. He begins with an account of the sex-cells of Triton, and stress is laid upon this form, the author showing a strong tendency to bring his studies upon Rana and Bufo into line with his work upon Triton.
He first recognizes the anlage of the sex-cells in the medial portions of the lateral plates of mesoderm in the 3 mm. larva of Triton. They occur only in the caudal half of the body and involve only those parts of the lateral plates of mesoderm lying medial to the Wolffian ducts. In the early stages these cells are filled with large yolk spherules and do not greatly differ from the mesodermal cells that surround them. At a later period the sex
SEX-CELLS OF AMIA AND LEPIDOSTEUS 27
cell anlagen are pushed together in the median line, between the aorta and the roof of the archenteron. They fuse into a median longitudinal rod of cells lying just above the dorsal root of the mesentery. By this time the sex-cells have lost their yolk material and have, to a large extent, assumed their definitive character. During these stages the number of the sex-cells has increased from one hundred to one hundred and fifty, occasional mitoses being observed. Soon after this stage of the median anlage (9 mm.) has been reached, the sex-cells migrate laterally to their final positions on each side of the root of the mesentery. At the stage of 14 mm., a large number of them degenerate, leaving only 60. A second generation of sex-cells soon begins to form from a source entirely different from the first, namely, from a transformation of ordinary peritoneal cells. Dustin is, in this regard, quite in accord with Bouin who expressed similar views regarding Rana. Dustin considers somewhat more briefly the corresponding stages in Rana and Bufo. Here he finds what he considers to be a substantially similar source of origin of the sexcells, namely the medial borders of the lateral plates of mesoderm. An incredible feature of his account is the statement that the lateral sex-gland anlagen contain no sex-cell at all comparable in size to those of the yolk-filled entoderm, at the period immediately prior to their union in the median line. Dustin would have us believe, nevertheless, that these selfsame sex-cells show a close resemblance to the entoderm cells immediately after this union of the lateral anlagen, and this in spite of the fact that both of these stages of development are so close together that the embryos upon which he made these observations were all of the same length. His own statement is as follows :
" Au moment ou les ebauches paires separees par une sorte de clivage des lames laterales du mesoblaste se sont rapprochees de la ligne mediane, les cellules sexuelles futures passent par une serie de transformations cytologiques a la suite desquelles elles auront presque les caracteres des cellules de I'hypoblaste vitellin. Les dimensions des corps cellulaires augmentent dans de fortes proportions; les grains vitellins deviennent beaucoup plus nombreux et plus volumineux; ils se colorent mieux par I'orange G. Par le fait de I'augmentation du nombre des plaquettes
28 BENNET M. ALLEN
vitellines, le noyau, souvent refoule a la peripherie de la cellule, presente a sa surface une serie d'encoches lui donnant un aspect herisse (p. 476).
He finds the number of sex-cells in Rana to increase gradually, from 75 in the 8 mm. stage to 90 in the 15 mm. stage, at which time sex-cells begin to be formed by the transformation of ordinary peritoneal cells. Simultaneously there is a degeneration of sex-cells which is overbalanced by this process of transformation.
In criticism of the above views I wish, first of all, to admit the possibility that Dustin may be perfectly correct in his account of the origin of the first line of sex-cells from the lateral plates of mesoderm in Triton. His account of this feature is circumstantial and rather convincing. His account of a transformation of peritoneal cells into sex-cells during later stages is by no means so easy of acceptation. His figures to demonstrate this are not convincing.
His counts of sex-cells are not given in any circumstantial detail and there is no indication as to whether the number of sexcells recorded for any given stage is the result of a count of the sex-cells in one specimen or in several. One can not be blamed for being skeptical of the value of such counts if made upon but one specimen of each stage, when so few stages are chosen to demonstrate general processes of degeneration and new formation. Such a process can only be established by a count of the sexcells of numerous specimens.
I wish to express my complete disbelief in the first appearance of the sex-cells in the lateral plates of mesoderm of Rana and Bufo in the manner described by Dustin. In my paper upon An Important Period in the History of the Sex-Cells of Rana pipiens" ('07) I showed that the sex-cells migrate upward from the median dorsal portion of the gut entoderm at the time when the two lateral plates are pushing together to the median line in the process of forming the mesentery. Attention was called to the resemblance that this process bears to an actual pinching off of the mass of sex-cells by the inner margins of the plates of
SEX-CELLS OF AMIA AND LEPIDOSTEUS 29
mesoderm. As pointed out in my article, the lateral plates of mesoderm, examined immediately before their approximation in the median line, show no cells which, as regards size or yolk content, in the least compare with the sex-cells.
It is especially gratifying to me to find support for my views in two recent papers. In one of these Kuschakewitsch ('08), referring to my paper of a few months before, stated: Der Verfasser hat die Abschniiring von Dotterzellen langs der dorsalene Sagittallinie des Dottersackes im hinteren Telle des Rumpfes beobachtet und die Theilname dieser Dotterzellen am Aufbau einer kompakten Mesenterial-anlage festgestellt, die Bouin (1900)) als ebauche genitale primordiale" aufgefasst hatte. Wie aus meiner Schilderung der entsprechenden Vorgange in der Normalreihe von Rana esculenta zu ersehen ist, kann ich die Angaben von Allen voUstandig bestatigen."
Another paper, appearing the same year (King, '08), gives an account of the origin of the sex-cells in Bufo lentiginosus which is in complete accord with the above, and states : "Allen's recent account of the origin of the sex-cells in Rana pipiens agrees essentially with what I have found in Bufo." Miss King finds no evidence in the course of development of any transformation of peritoneal cells into sex-cells as asserted by several writers among whom may be mentioned Bouin and Dustin. This is quite in accord with my observations upon Chrysemys ('06) in which the sex-cells were traced to the period of sexual maturity without finding any evidence of such transformation.
Miss May Jarvis ('08) in a paper upon The Segregation of the Germ-Cells of Phrynosoma cornutum" (preliminary note) finds the sex-cells to take their origin in the entoderm of the vascular area on all sides of the embryo, even cranial to it, and notes a few in the region of the brain. Her results are in their main features confirmatory of my own work upon Chrysemys. The following quotation from her paper is self-explanatory: Through the courtesy of Dr. Allen, I have been enabled to examine the more important stages in the migration of the germ-cells of Chrysemys; they are similar to my own material, as my conclu
30 BENNET M. ALLEN
sions, although differing from Dr. Allen's in details of early distribution and periods of migration, uphold his."
Rubaschkin ('08 and '09) in a couple of recent papers, has shown that the sex-cells of the rabbit and guinea-pig are first to be found in the entoderm at some distance on each side of the hind-gut and that they follow a path almost identical with that followed by the sex-cells of Chrysemys. These references to the coincidence of the views of other recent writers with my own are made to show that I do not stand alone in placing emphasis upon the entodermal origin of the sex-cells in the vertebrates. At the same time I wish, however, to disclaim any intention of making at this time a sweeping claim that the sex-cells of all vetebrates arise in the entoderm. Wheeler's work on Petromyzon ('99) shows that they may be included in the mesoderm at the time when that layer is split off from the entoderm. He has, however, pointed out their similarity to the entoderm cells and their dissimilarity to the mesodermal cells among which they lie.
I do not seek to discredit the work of Dustin upon the sexcells of Triton; although his statements about the origin of the sex-cells in Rana and Bufo strike me as being very far from the mark, because they are so radically at variance with not only my own observations, but with those of King and Kuschakewitsch as well. Dustin, in his attitude toward the work of others, seems to consider that there must be a strict uniformity in all forms in both the place of origin and in the movements of the sexcells. He has apparently studied this problem first in Triton and at some length. His results, probably correct for that form, he has attempted to apply to Rana and Bufo as well, undeterred by the difficulties to which attention was called above. Dustin is quite ready flippantly to dismiss my work upon Chrysemys, because the results there expressed did not coincide with the views that he had formed regarding the origin of the sex-cells in Triton, Rana, and Bufo.^ The process of migration through the entoderm is so clear in Chrysemys, that it is unmistakable. The sex-cells are not only characterized by their larger size,
1 See postscript.
SEX-CELLS OF AMI A AND LEPIDOSTEUS 31
definite, rounded outlines and fine chromatin network, but by their large yolk content and the fact that they do not divide during the stages in dispute.
The sex-cells are migratory to a high degree. The path and time of their migration may vary greatly within a given group of animals, as illustrated by the case of Amia and Lepidosteus. While in the forms that I have studied they are first to be observed in the entoderm, I am quite open to conviction that in other forms they may migrate from this layer into the potential mesoderm before the two layers are separated, as shown by Wheeler in Petromyzon. It is even conceivable that they may lie, from the very beginning of development, in material destined to form mesoderm — that they may never have existed among cells actually or potentially entodermal. The more recent development of our work along these lines, however, most certainly tends to show that it is usual among the vertebrates for the sex-cells to first appear in the entoderm.
SUMMARY AND CONCLUSIONS
1. The sex-cells of both Amia and Lepidosteus have their origin in the entoderm. In Amia they are first distinguishable in the peripheral entoderm from the lateral angle of the subgerminal cavity to the anlage of the blood cells.
In Lepidosteus they are first seen in the ventral and lateral portions of the gut-entoderm, although analogy with Chrysemys leads us to assume that they may have migrated through the entoderm to these regions from more lateral anlagen, similar to those from which the sex-cells of Amia arise. In both forms, the sex-cells arise only in the region of the hind-gut. None were found at any considerable distance in front of it.
2. The path of sex-cell migration in Amia carries them out of the peripheral entoderm directly into the overlying lateral plates of mesoderm, along which they travel, to come to rest near the medial edges of the latter. These portions are destined to join above the intestine to form the mesentery. As the splanchnic and somatic layers of the lateral plates of mesoderm
32 BENNET M. ALLEN
split to form the coelome, the sex-cells adhere to the somatic layer at a point near the root of the developing mesentery — the sex-gland anlage. They later sink into the peritoneum of this region, which afterwards proliferates to form a long ridge — the sex-gland. Very few sex-cells fall by the wayside in this migration, practically all reaching the sex-glands.
3. In Lepidosteus the sex-cells, first seen in the ventral and lateral portions of the gut-entoderm, migrate to occupy a position in the dorsal portion of it, from which they pass dorsally into the loose mesenchyme that forms the substance of the developing mesentery. As the mesentery becomes more narrow and compact, owing to the increase in size of the body cavity, the sexcells migrate to its dorsal portion and laterally to the sex-gland anlagen. Roughly speaking, one-half of the total number of sex-cells reach the sex-gland anlagen, the remainder being distributed between the intestinal entoderm, the mesodermal layers of the intestine, the mesentery and the tissues at and dorsal to the root of the intestine.
4. The number of the sex-cells in Amia and Lepidosteus is a matter of individual variation for those periods of development during which they do not undergo division. The average number in Amia, after the period when the migration from the entoderm to the mesoderm has been completed, up to the latest stage in which counts were made, was found to be 75. In Lepidosteus it was 765, an average of 636 of these occurring in the mesoderm.
5. There is a close resemblance between the nuclei of the sexcells and of the yolk cells. This is especially true of certain cells of the peripheral entoderm, although these grade by gradual transition forms into the large nuclei of the vitelline entoderm. This is probably due to the fact that both types of cells have undergone but little differentiation in the course of development.
POSTSCRIPT
A few days before proof of this article came to hand, I received, through the courtesy of the author, a reprint of an article by A. P. Dustin, entitled, "L'Origine et I'Evolution des Gonocytes chez
SEX-CELLS OF AMIA AND LEPIDOSTEUS 33
les Reptiles," (Archives, de Biologie, 1910). This article deals with the origin of the sex -cells in Chrysemys marginata, the form which served as a subject for my own work of 1906. As noted above, Dustin in his paper "Recherches sur Torigine des gonocytes chez les Amphibiens" 1907, exhibited scant respect for my work on the sex-cells of Chrysemys. It was, no doubt, in large part, this feeling that prompted him to repeat my work. While he, no doubt, expected to find in this form a confirmation of his previously expressed views, he is led to substantiate completely my statements regarding the entodermal origin of the sex-cells. He traces them along the same migration path that I demonstrated four years before. For all this he now gives me full credit and support; but takes issue with my statements regarding the distribution of the sex-cells prior to their migration into the embryo, and, furthermore, claims to have -evidence to show that there is a new formation of sex-cells, due to a transformation of ordinary peritoneal cells. These points of controversy and certain other minor ones can not be considered here, but I promise a full discussion of them in another place. I may say that I am fully prepared to maintain my views upon all of the points at issue.
On my part, the work that I have carried on upon Necturus since this paper was written, has given me results quite similar to those at which Dustin arrived in his work upon Triton. I may say that preliminary studies have convinced me that the sex-cells arise in an essentially similar manner in Amblystoma. We then see that, in all three of these urodeles, the sex-cells arise from the inner edges of the lateral plates of mesoderm. I owe it to myself to call attention to the fact that I have at no time disputed the accuracy of Dustin's work upon Triton. While the evidence seems to me quite clear that this is the usual, if not the universal, mode of origin of the sex-cells among the urodele amphibians, I am ready to maintain with equal vigor the entodermal origin of the sex-cells in the aruran amphibians, at the same time admitting the possibility that exceptions to this apparent rule may be discovered. I do not feel however, that Dustin has proved his case in Rana fusca and Bufo vul
34 BENNET M. ALLEN
garis. The discussion of his work above gives the reasons for my position in this matter.
Not only does it seem probable that the sex-cells arise during early stages in the mesoderm of the urodeles, but this seems to be the case in the teleosts as well. The most recent and satisfactory support of this view is contained in the excellent paper of Dr. Gideon S. Dodds upon the Segregation of the Germ-Cells of the Teleost, Lophius, " in the Journal of Morphology, 1910. Here again, we must urge caution in forming a sweeping generalization from the facts thus far at hand. There is certainly a wide field for work in the study of the origin of the sex-cells of the vertebrates. It is a subject which should be approached in a spirit of broad toleration for the views of others. The sexcells are cells that retain their early embryonic character after the somatic cells have undergone specialization. It seems, from a number of observations made by different authors, that in most forms the sex-cells first make their appearance in the entoderm — the germ layer whose cells appear to maintain their primitive embryonic characters longer than do those of the other germ layers. At the same time, unimpeachable evidence shows that this apparently logical process is not universal, and I have at no time claimed that it is. The sex-cells, as show^n by Nussbaum, Eigenmann, Beard and others, do not belong to any one germ layer, but are, in a sense at least, independent of the somatic tissues. They are free to follow their own path in their travels from the place of origin to the sex-gland anlagen, where they finally come to rest. While this path is no doubt identical or similar in closely allied species and in more general divisions of the vertebrates, I do not feel that we are justified in attributing a high degree of phylogenetic importance to the different steps in the migration paths through which they travel.
I wish to express my indebtedness for the work of our departmental artists. Misses Hedge and Battey. I am indebted to Miss Hedge for the execution of diagrams 1-6 and for figs. 9, 10, 14, 15, 21, 22, 25 and 26; and to Miss Battey for figs. 11, 12, 13, 23, and 24. The remaining drawings are my own.
SEX-CELLS OF AMIA AND LEPIDOSTEUS 35
BIBLIOGRAPHY
Allen, Bennet M. 1906 Origin of the sex-cells of Chrysemys. Anat. Anz. Bd. 29.
1907a A statistical study of the sex-cells of Chrysemys marginata, Anat. Anz. Bd. 30.
19076 An important period in the history of the sex cells of Rana pipiens. Anat. Anz., Bd. 31.
1909 The origin of the sex- cells of Amia and Lepidosteus. Anat. Rec, Vol. 3.
DusTiN, A. p. 1907 Recherches sur I'origine des gonocytes chez les Amphibiens. Arch, de Biologie, tome 23.
Jarvis, Mat. 1908 The segregation of the sex-cells of Phrynosoma. Biol. Bui., Vol. 15.
King, Helen Dean. 1908 The oogenesis of Bufo lentiginosus. Jour. Morph., Vol. 19.
KuscHAKEwiTSCH, S. 1908 Ueber den Ursprung der Urgeschlechtszellen bei Rana pipiens. Stzber. math. phys. Klasse, k. bayer. Akad. Wiss., Bd. 38.
RuBASCHKiN, W. 1907 Zur Frage von der Entstehung der Keimzellen bei Saugetierembryonen. Anat. Anz., Bd. 31.
1909 Ueber die Urgeschlechtszellen bei Saugetieren. Anat. Hefte, Bd. 39.
Wheeler, W. M. 1899 The development of the urogenital organs of the lamprey. Zool. Jahrbuch., Anat. Abth., Bd. 13.
ABBREVIATIONS FOR ALL FIGURES
Arch., Archenteron
Coel., Coelomic cavity
Ect., Ectoderm
Gut End., Gut entoderm
Int., Intestine
Lat. Mes., Lateral plate of mesoderm
Mes., Mesentery
Meson., Mesonephros
Myo., Myotome
Nolo., Notochord
P. Card., Post cardinal vein
Periph. End., Peripheral entoderm Roof End., Roof entoderm S. €., Sex-cells S. GL, Sex-gland
Sub-Germ. Cav., Sub-germinal cavity Sub-Germ. End., Sub-germinal entoderm Siv. Bl., Swim bladder Vit. End., Vitelline entoderm
Wolff. D., \ W. D., J
Wolffian duel
(36)
PLATE 1
EXPLANATION OF FIGURES
1 Diagram to show the migration path of the sex-cells in Chrysemys marginata.
2 Diagram to show the migration path of the sex-cells in Rana pipiens.
PLATE 2
EXFLANAIION OF FIGURES
3 Diagram to show the migration path of the sex-cells in Lepidosteus osseus.
4 Diagram to show the last phase of the migration of the sex-cells in Lepidosteus osseus.
5 Diagram to show the migration path of the sex-cells of Amia calva.
6 Diagram to show the last phase of the migration of the sex-cells in Amia calva.
JOURNAL OF MORPHOLOGT, VOL. 22, NO. 1
PLATE 1
PLATE 3
EXPLANATION OF FIGURES
7 Transverse section through the hind-gut of an 8.6 mm. larva of Lepidosteus osseus. X 300.
8 Transverse section through the hind-gut of a 9.3 mm. larva of Lepidosteus osseus. X 300.
9 Transverse section through the hind-gut of a 10.7 mm. larva of Lepidosteus osseus. X 300.
10 Transverse section through the hind-gut of a 14.1 mm. larva of Lepidosteus osseus. X 300.
11 Transverse section through the hind-gut of a 17 mm. larva of Lepidosteus osseus. X 300.
SEX CELLS OF AMLV AND LEPIDOSTEUS EENNK.T M. A1.M:X
ERRATA
Gelatin plates 1, 2 and 3 should have been numbered 3, 4 and 5
JOURNAL OF MORPHOLOGY, VOL. 22. NO. 1
PLATE 3
EXPLANATION OF FIGURES
SEX CELLS OF AINHA AND LEPIDOSTEUS F.ENNET M. ALLEN
JOURNAL OF MORPHOLOGY, VOL. 22. NO. 1
PLATE 4
EXPLANATION OF FIGURES
12 Transverse section of the rudimentary sex-glands of a 24 mm. larva of Lepidosteus osseus. X 300.
13 Transverse section of a sex-gland of a 110 mm. specimen of Lepidosteus osseus. X 300.
14 Part of a transverse section of a 17 mm. larva of Lepidosteus osseus, showing the reduced vitelline mass.
15 Detail drawing of a portion of the vitelline mass of the above section. X 300.
16 Transverse section through the region immediately lateral to the posterior portion of the sub-germinal cavity of a 147 hr. embryo of Amia calva. X 300. This shows the place of origin of the sex-cells.
17 Section passing similarly through another specimen of the same stage of Amia calva. X 300.
One sex-cell shown as it is pushing up into the mesoderm.
18 Transverse section through the hind-gut of a 6 mm. larva of Amia calva. X 300.
19 Transverse section through the hind-gut of a 7 mm. larva of Amia calva. X 300.
SEX-CELLS OF AML\ AND LEPIDOSTEUS
KENNF.T M. AI.I.EX
Periph. EncL.jg.^
My
«ir.-J?>Vsw B,, ^f:'^ ^1^ r^x
JOURNAL OF MORPHOLOGY, VOL.23, NO. 1
PLATE 5
EXPLANATION OF FIGURES
20 Transverse section through the hind-gut of a 9.1 mm. larva of Amia calva. X 300.
21 Transverse section through the hind-gut and sex-gland anlage of an 11.4 mm. larva of Amia calva. X 300.
22 Transverse section through the young sex-glands of a 16 mm. larva of Amia calva. X 300.
23 Sketch to show the orientation of the sex-glands in q 40 mm. specimen of Amia calva as seen in transverse section.
24 Detail drawing of the sex-gland as seen in above sketch. X 300.
25 Drawing to show the orientation of the much reduced vitelline mass of an 11.4 mm. larva of Amia calva.
26 Detail drawing of a portion of the vitelline mass indicated above. This shows the resemblance that certain cells of the peripheral entoderm show to sexcells of this stage. X 300.
SEX-CELLS OF AMLV AND LEPIDOSTEUS BENNET M. ALLEN
PLATE 3
JOURNAL OF MORPHOLOGY, VOL. 32, NO. 1
THE CYCLIC CHANGES IN THE OVARY OF THE GUINEA PIG
LEO LOEB
From the Laboratory of Experimental Pathology of the University of Pennsylvania, and from the Pathological Laboratory of the Barnard Skin and Cancer Hospital, St. Loxds, Mo.
In the course of an experimental investigation into the causes of the cychc changes taking place in the uterine mucosa and into the factors underlying the formation of the maternal placenta in mammals, we observed that cyclic changes in the structure of the ovary correspond to the uterine cycle. It has of course been known that at certain times ovulation takes place in the mammalian ovary, and furthermore, changes have been described as occurring in the ovarian follicles of certain mammals in connection with copulation and during pregnancy; but the cyclic changes taking place in the ovary 'quite independently of copulation and of pregnancy and merely dependent upon ovulation have, as far as we are aware, not yet been recognized. While we know of no publication dealing with the cyclic changes in the ovaries in general, a valuable study of the changes taking place during pregnancy in two species of Insectivores and in one species of Lemurid has been made by C. H. Stratz.i This author comes to the conclusion that in the period following copulation all the ovarian follicles become atretic; that during pregnancy small follicles are formed but also become atretic before they can develop; that only towards the end of pregnancy the follicles begin to grow to a considerable size, and that they reach the stage of maturity during the puerperium.
Stratz was not in a position to determine in an exact manner the time elapsed since the last copulation of the animals the ova ^ C. H. Stratz: der geschlechtsreife Saugethiereierstock. Haag. 1898.
37
38 LEO LOEB
ries of which he examined. He also seems to have examined a relatively very limited number of ovaries of animals during the different stages of pregnancy, and furthermore he studied only certain parts of each ovary. A methodical study of ovaries of non-pregnant animals was not undertaken. While his observation that after copulation all follicles become atretic is approximately, but not altogether correct, as far as its general validity is concerned, in the guinea pig the processes taking place in the ovaries during the subsequent stages differ from the conditions described by Stratz in the case of Tupaja, Sorex and Tarsius.
Furthermore Stratz does not recognize the essential factor upon which the cyclic changes in the ovaries depend. The conclusions in the last chapter of his publication show this clearly.
He summarizes as follows: If we find all follicles atretic, the animal has been pregnant. If at the same time a new corpus luteum is present, we have to deal with an early stage of pregnancy. If we detect some normal follicles, besides numerous atretic follicles and a new corpus luteum, we have to consider a puerperal condition of the animal. A large number of atretic besides a few normal follicles also* suggests a puerperal state.
These general conclusions are not justified; the changes of the follicles do not, as Stratz assumes, depend upon pregnancy, and if we should attempt to use the criteria given by Stratz in the case of guinea pigs and mammals in general we would be liable frequently to make mistaken diagnoses. Notwithstanding, these necessary criticisms, the work of Stratz is very valuable and it advanced to a considerable extent our knowledge of the ovaries.
Since his publication no more detailed investigation into the processes taking place in the ovaries under various conditions has appeared, as far as we are aware. Within recent years, however, the question has been raised whether a new ovulation can take place during pregnancy.
We limited our investigations to the study of the ovary of the guinea pig. We examined several hundred pairs of ovaries of animals in which the period of the sexual cycle at which the ovaries
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 39
were obtained had been ascertained. In each case the entire ovary was cut into serial sections.
During the progress of our work new problems arose and an accident made it impossible for us to re-examine all our material in order to answer several questions which were raised at a later stage of our investigation. We especially regret our inability to determine the existence of follicles which were ready to rupture, in certain cases in which these data would have been of considerable interest. Our work is therefore incomplete in some respects. We expect, however, very soon to be able to supplement our present work, wherever necessary.
OVARIES OF GUINEA PIGS IN THE LAST STAGE OF PREGNANCY
The condition of the ovaries of a guinea pig in the last days of pregnancy is as follows: there are small, medium sized and large follicles without degeneration of granulosa cells. In other large follicles various stages of granulosa degeneration are present. Many follicles show further advanced stages of atresia, in which connective tissue grows into the follicular cavity. Especially numerous are the last stages of atresia in which the zona pellucida is directly surrounded by very cellular connective tissue. Mitoses are seen in the granulosa cells of the well preserved follicles. We also find here a few mature follicles which are characterized by an increase in cytoplasm of the granulosa cells. These follicles are large; their cavity is very wide. The nuclei of the granulosa cells are not as densely packed in these follicles as in the ordinary large follicles, this peculiarity being due to the marked development of the cytoplasm. They can be easily recognized in sections stained by haemotoxylin and eosin, inasmuch as they appear stained more reddish, in contradistinction to the ordinary large follicles in which the blue color of the nuclei predominates, while in the mature follicles the red stain of the cytoplasm is a distinguishing feature. In these mature follicles the number of mitoses is very much smaller than in the ordinary large follicles. With the increase in the quantity of cytoplasm and the relative decrease in the nuclear material,
40 LEO LOEB
the cell proliferation is diminished. The number of mitoses is usually very small, or mitoses may be absent in such follicles. Another characteristic feature is the relative lack of degeneration of the granulosa in these follicles. While the ordinary large follicles degenerate in the large majority of cases, the granulosa cells becoming karyorrhectic, as soon as the follicle attains a certain size; the mature follicles are very much more resistant. The changes in the granulosa cells described above and which lead to the transformation of an ordinary large follicle into a mature red-staining follicle, and simultaneously to a decrease in cell proliferation of the granulosa and to a diminished karyorrhexis of the granulosa cells, probably produces a decrease in cell metabolism, and this decrease in cell metabolism stands perhaps in a causal relation to the decrease in cell multiplication and to the greater resistance of the granulosa cell. A slight degree of degeneration of the granulosa may even occur in the mature red-staining follicles; a few of the central granulosa cells may degenerate; and in one case we observed even a fargoing degeneration of the granulosa in a mature follicle. It becomes therefore probable that these mature follicles also degenerate, if ovulation does not take place. This transformation of an ordinary large follicle into a mature follicle takes place only to a limited extent; the large majority of the follicles degenerate before they have reached the stage of full maturity. This holds good even in the case of guinea pigs before delivery, in which a rupture of follicles will soon take place.
The corpora lutea of pregnancy which, at the time at which we examined the ovaries, were approximately fifty-six to sixty-four days old and which had formed soon after copulation, show already some retrogressive changes in the lutein cells. A considerable number of the vessels entering the corpora lutea have a very thick wall consisting of several rows of cells. A large number of the vessels, however, have merely an endothelial lining. In many of the vessels no lumen is visible, the circulation through the corpus luteum being evidently not very active; some of the capillary vessels have, however, a widely open lumen. The quantity of the connective tissue in the centre of the corpus luteum is small.
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 41
on account of the previous proliferation of lutein cells which encroached more and more upon the space originally filled by the connective tissue. The corpora lutea are large. The lutein cells show signs of degeneration ; they are finely vacuolar and may have a foamy appearance ; a certain number of cells take less eosin and appear therefore pale. Many cells have a sharply defined, redstaining outline. The nuclei also show changes; they are frequentquently deformed, indentated ; or they are round, vesicular, but stain less with haematoxylin ; they appear somewhat karyolytic. Mitoses could not be seen in the lutein cells. The degree of retrogressive changes may vary in different corpora lutea even in the same ovary.
We see therefore that even before delivery and before a new ovulation has taken place, degenerative changes set in in the corpora lutea, and it accords with these retrogressive changes that mitoses are absent or at least very rare in such corpora lutea.
Besides these corpora lutea of pregnancy we may find in such ovaries 'yellow bodies' representing the last stage of retrogression of corpora lutea. In the corpora lutea which were transformed into such yellow bodies, degeneration must have set in approximately sixty to sixty-five days ago. These 'yellow bodies' have the following structure: In their centre and periphery we find hyaline connective tissue; between these two zones of hyaline connective tissue a relatively small number of degenerated large lutein cells is enclosed, in which, during the process of retrogression, a large amount of yellow pigment was produced.
OVARIES OF GUINEA PIGS WITHIN TWO DAYS AFTER DELIVERY
In the period directly following delivery the condition of the ovaries, as far as follicles and corpora lutea are concerned, is approximately the same as in the period preceding it. The growth and degeneration of the follicles still continue to take place, and in follicles in which the granulosa has completely or almost completely degenerated an ingrowth of connective tissue and complete atresia of the follicles occur. The retrogressive changes in the corpora lutea also progress, but at a slow rate, and on the whole the
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 1
42 LEO LOEB
corpora lutea are not very different from those found in the preceding period. This description holds good for instance for ovaries of a guinea pig extirpated ten minutes after complete delivery.
Soon after delivery (usually within a few hours) the guinea pig is ready for a new copulation and ovulation, and after ovulation changes take place in the follicles which will be described later.
The corpora lutea of the preceding pregnancy undergo no very marked changes within the next two days after delivery, although vacuolization of the lutein cells and degenerative changes in the nuclei show probably a slight advance ; the lutein cells do not stain as well with eosin and appear pale. If copulation take place soon after delivery, a rupture of the mature follicles occurs within the succeeding six or ten hours; but if copulation be prevented by isolating the female, ovulation frequently occurs, but does not need to take place within thirty-six hours after delivery. In several cases in which an actual copulation was prevented, in which however the male was in contact with the female for a short time after delivery, the rupture of the follicles and the formation of new corpora lutea took place in the usual way. The changes in the new corpora lutea within the first two days after delivery are the same as those described in a previous paper.^
In three cases the lower part of the uterus or the vagina of guinea pigs were tied completely or incompletely towards the end of pregnancy. This procedure led to the death of the fetuses, followed by expulsion of the dead fetuses in a case in which the occlusion had been incomplete. In another case the animal was killed by chloroform six days after the application of the ligature, and the fetuses were found dead; furthermore autolysis of the placenta had set in. In these cases especially the periphery of the corpora lutea of the preceding pregnancy showed vacuolization of the lute tein cells. The nuclei were shrunken or somewhat chromatolytic. Notwithstanding the degenerative changes visible in the corpora lutea, no new ovulation had taken place. From these and other observations it follows that delivery as such does not lead to far 2 The formation of the corpus luteum in the guinea pig. Journal American Medical Association, February 10, 1906.
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 43
going changes in the ovaries; that merely a slow progress takes place in changes which had set in before delivery. We furthermore see that without copulation a spontaneous ovulation does not need to take place after delivery, notwithstanding the degenerative changes in the corpora lutea; that ovulation can, however, occur without copulation, and this seems to be the rule, if the male had been in contact with the female for some time after delivery, a copulation having been made impossible during this period of contact.
OVARIES OF NON-PREGNANT GUINEA PIGS IN THE PERIOD DIRECTLY PRECEDING OVULATION
This description applies to ovaries of guinea pigs which had copulated a few hours previously, in which an ovulation had however not yet taken place — ovulation usually taking place approximately six to ten hours after copulation. In another case we examined the ovaries of a guinea pig that was ready for copulation ('in heat') in which, however, an actual copulation had been prevented by occluding the vagina by means of a strip of plaster.
The condition of the follicles in these ovaries was similar to the condition found in ovaries preceding and immediately following delivery; we find good follicles of small, medium and large size; mitoses are present in the granulosa of such follicles. The majority of the large follicles however show more or less degeneration of the granulosa, with the exception of the few large follicles which progressed to complete maturity; they showed the cytoplasmic changes described above. In these as well as in some other well preserved large follicles the theca interna appears somewhat hyperemic. We also find the various stages of connective tissue ingrowth and of the subsequent diminution in the size of the follicles ('connective tissue atresia') which we described in the case of the other ovaries. In this case we do not find corpora lutea of a preceding pregnancy, but corpora lutea of an ordinary ovarian period, not accompanied by pregnancy. These corpora lutea are much smaller than those of pregnancy ; their lutein cells show vacuolization, indicating the beginning of retrogressive changes. Notwithstanding these retrogressive changes an occasional mitosis
44 LEO LOEB
can still be found in lutein cells. The corpora lutea of the second last ovulation have in the meantime been transformed into yellow bodies. Processes of degeneration have therefore set in in the corpora lutea of non pregnant as well as of pregnant guinea pigs before ovulation. These beginning degenerative changes do however not prevent the occurrence of a few mitoses in the corpora lutea of previously not copulated animals, while in the degenerating corpora lutea of pregnancy we have so far not been able to detect the presence of mitoses in lutein cells.
OVARIES OF GUINEA PIGS WITHIN THREE AND ONE HALF DAYS AFTER OVULATION
In connection with ovulation certain far reaching changes take place in the ovaries. All follicles, with exception of very small ones, degenerate. These changes set in with ovulation, or they may perhaps start somewhat earlier, namely, simultaneously with those processes that bring about ovulation. As we have pointed out above, the general degeneration of the follicular granulosa which we find directly after ovulation cannot yet be observed before ovulation. This sudden degenerative process is quite independent of copulation; we found that it can be produced through ovulation without a preceding copulation. We discovered experimental means through which we can produce a spontaneous ovulation without a preceding copulation. Such an ovulation is followed or accompanied by the same degeneration of the granulosa. Moreover, if we keep a number of female guinea pigs separated from the males and if we examine their ovaries after various periods of isolation, we find occasionally ovaries in which the rupture of follicles had taken place a few days before. In this case also the typical follicular degeneration takes place independently of a preceding copulation.
Six and a half hours after a preceding copulation the ovaries showed, besides the presence of newly ruptured follicles, the following changes in the follicles : All, with the exception of very small follicles, show granulosa degeneration; in the large majority of the follicles almost the whole granulosa is found in a process of degeneration. We also find folHcles in the process of connective tissue
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 45
atresia. Similar conditions are found in other ovaries at the same period.
Twenty-two hours after copulation some granulosa cells are found degenerated even in small follicles, (follicles having a small cavity) ; these degenerated granulosa cells are dissolved.
Similar changes take place in ovaries of guinea pigs in which ovulation followed delivery. In a guinea pig in which copulation took place two hours after delivery and in which the ovaries were examined seventeen hours after copulation, only a few quite small follicles without granulosa degeneration were found ; in the large and also in the medium sized follicles much granulosa degeneration had taken place, the central granulosa cells degenerating first. Almost no entirely good follicles were left. As soon as the interna becomes exposed, phagocytic cells (rounded off interna cells) peneirate into the follicular cavity and these cells take up debris of the granulosa. The degeneration of the granulosa cells is as usual followed by ingrowth of connective tissue.
In other ovaries the granulosa may be degenerated to a great extent, but some remnants may still be left. Especially the granulosa cells of the discus proligerus survive usually the rest of the granulosa. We find of course various stages of connective tissue atresia besides the degeneration of the granulosa. From these observations it follows that the onset of degeneration of the granulosa must be extremely rapid.
If we extirpate the corpora lutea, from two to eight days after copulation a new spontaneous rupture of follicles takes place in most cases approximately from thirteen to fifteen days after the previous copulation, even if the female had been kept entirely isolated during the whole period following the extirpation of the corpora lutea. This early spontaneous ovulation is accompanied by the same follicular degeneration which we described above.
It is an interesting problem, whether an artificially produced rupture of a follicle, with the subsequent development of a corpus luteum, is accompanied by the same acute follicular degeneration. Several years ago we made experiments in which we pricked or cut the surface of ovaries of guinea pigs which were either 'in heat,' without however having copulated, or which copulated a few hours previously, or which had in some cases copulated from three to six
46 LEO LOEB
days previously. In only one case did we find a young corpus luteum the origin of which could reasonably be attributed to the cutting of the ovary and to the artificial ru^pture of a follicle. In this case an animal had been used which showed the first symptoms characteristic for the period of heat. Three days after the cuts had been made the ovaries were examined. One young corpus luteum was found in the cortex of the ovary. Blood and connective tissue were found in the center of the corpus luteum; connective tissue and vessels grew into the corpus luteum, which was very small. In this ovary we found good follicles of small medium and large size; we also found large follicles with beginning and wdth further advanced granulosa degeneration, and with beginning ingrowth of connective tissue. In as much as in no case of spontaneous rupture the follicles were found in a similar condition at that period after the rupture, it is very probable that we have in this case to deal with an artificial rupture of follicles and that such an artificial rupture of follicles is not accompanied by the rapid degeneration of the follicular granulosa.
On the basis of our previous results we can easily understand, why in all probability we succeeded in one case only in causing an artificial rupture of a follicle. Such an experiment does not promise to be successful, unless we have the chance of opening a mature follicle, and such an opportunity exists only at periods of very short duration.
In these ovaries we find usually two or three generations of corpora lutea; namely:
1. The young corpora lutea, developing in the recently ruptured follicles. These corpora lutea we have described elsewhere in their development up to the sixth day.
2. Corpora lutea that had formed at the time of the preceding ovulation, which had not been followed by pregnancy in female guinea pigs which had been kept separated from males. These corpora lutea are therefore in all probability approximately nineteen to twenty-eight days old. They show signs of beginning retrogression. Their lutein cells are more or less vacuolar, especially in the periphery, where the vacuolization usually begins ; gradually the vacuolization progresses to the central part. In the center of
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 47
the corpus luteum we find a relatively small amount of fibrous tissue. We not only find capillary vessels but also vessels the wall of which consists of two coats penetrating the corpus luteum.
These corpora lutea begin to shrink very soon, and three days after the new rupture they are usually smaller than immediately after the ovulation. Notwithstanding the degenerative processes which are apparent in these corpora lutea, it is not uncommon to to find still mitoses in the lutein cells of such corpora lutea within the first twenty hours after the new rupture of follicles has taken place. At a later period mitoses were not seen in this series. The mitoses appear in the relatively well preserved, but they may be present even in somewhat vacuolar lutein cells. It is possible that occasionally mitoses occur also in endothelial cells of the capillaries.
3. The third generation is represented by yellow hyaline bodies. They are the remnants of corpora lutea that formed forty or more days ago.
If we examine ovaries of young guinea pigs, two and a half to three months old, we may find only the first, or the first and second generations of corpora lutea, but yellow bodies may be lacking.
We see therefore that preceding and following the rupture of new follicles in non-pregnant animals, processes of degeneration have begun in the corpora lutea of the preceding ovulation, and that notwithstanding such processes of degeneration, mitoses may occur in such corpora lutea for a short period following the new ovulation. These corpora lutea which are not accompanied by pregnancy are much smaller than the corpora lutea of pregnancy and they shrink more rapidly. The absolute diminution in size is more rapid than in the retrogressing corpora lutea of a preceding pregnancy. Concerning the relative rapidity of retrogression (the percentage decrease in size, the full size of the corpora lutea being taken as the standard), we cannot make any definite statement, not having carried out any measurements.
The mode of retrogression is the same in both ordinary corpora lutea and in those of pregnancy. The vacuolization begins in the periphery, where it becomes most marked, and from here it proceeds into the interior of the corpus luteum.
48 LEO LOEB
OVARIES OF A GUINEA PIG APPROXIMATELY THREE TO FOUR DAYS AFTER ABORTION
In one case the ovaries of a guinea pig were examined which on examination had previously been found to be in a well developed stage of pregnancy, but which had aborted about three to four days previously. The four corpora lutea showed signs of degeneration. The lutein cells were vacuolar in the periphery, in the center the cells stained pale red with eosin, the vesicular nuclei showed a diminution in the amount of chromatin. The cell outlines were very sharp, staining red with eosin. In the center there was dense connective tissue and many blood vessels had very thick walls.
Follicles of small, medium and large size, with well preserved granulosa, were present. A few mature, red staining follicles without mitoses or degeneration in the granulosa were also found. Many other large immature follicles showed various stages of granulosa degeneration. There were of course also present various stages of connective tissue atresia.
We see therefore that abortion is not followed by or associated with marked changes in the follicles. Whether the mature follicles which we found in these ovaries matured as a result of abortion, or whether the mature follicles were present before the onset of abortion, we cannot state with certaintj^, although it is more probable that maturation of the follicles followed abortion. We also note the beginning retrogressive changes in the corpora lutea. But in this case also we cannot be sure that the degenerative processes had not set in before the abortion had commenced.
OVARIES OF GUINEA PIGS FOUR TO SEVEN AND ONE HALF DAYS AFTER OVULATION
Six days after an ovulation we find in the ovaries on the whole the following condition of the follicles: There are well preserved follicles of small and medium size, with mitoses in the granulosa cells. A limited amount of granulosa degeneration is found only in rare instances. In such follicles mitoses are absent or their
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 49
number is decreased. Follicles in an advanced state of connective tissue atresia are frequent.
The character of the follicles at this period -of the sexual cycle is the same in cases in which the last ovulation was preceded by delivery, in which, therefore, in the previous period of the sexual cycle a pregnancy was present, and in other cases in which the previous period of the sexual cycle had not been comphcated by pregnancy. We see therefore that within six days quite small follicles, possessing only a very small follicular cavity, grow and reach medium size. During this period the granulosa of medium sized follicles did not degenerate, and no large follicles had as yet developed. We find therefore principally, besides the follicles with preserved granulosa, follicles in an advanced state of connective ,^issue atresia.
Six days after ovulation we find the corpora lutea of the last generation (corpora lutea six days old, as follows : The center of the corpus luteum is filled by a more or less loose connective tissue. Mitoses are present in the lutein cells as well as in the endothelial cells of the capillaries. Almost all the vessels have a capillary character. They penetrate into the central connective tissue. At that period vessels with two coats (intima and muscle coat of the media) can be observed for the first time, although they become more frequent at a somewhat later period.
In guinea pigs in which a pregnancy and delivery preceded the last ovulation, the corpus luteum of the preceding pregnancy shows marked signs of degeneration. Especially the peripheral cells are frequently coarsely, while the more centrally situated cells are more finely vacuolar; but even in the latter the protoplasm stains less with eosin and the nuclei are slightly chromatolytic ; the cells appear distinctly pale. The vessels are very thick and at certain places in the periphery the connective tissue of the neighborhood seems to begin to grow into the peripheral parts of the corpus luteum.
The ordinary corpora lutea of the second generation (not accompanied by pregnancy) show marked vacuolization; they diminish in size and in one case yellow pigment developed in a few of the vacuolar cells. Therefore in the course of five to eight days
50 LEO LOEB
since the beginning of degeneration the retrogressive changes have much advanced. The retrogressing corpora lutea of pregnancy of the corresponding generation are much larger at this period than the ordinary corpora lutea.
In a certain number of ovaries we also find a further (third) generation of retrogressing corpora lutea, represented by yellow bodies.
One corpus luteum deserves especial mention. In an ovary of a guinea pig which had ovulated approximately four and a half days before, five corpora lutea were found, four of which showing the typical structure. In the fifth of these corpora lutea, however, the lutein cells were arranged in the shape of glandular ducts. This condition has perhaps been produced through a dissolution of the central cells. Otherwise the corpora lutea in this ovary were normal.
The same typical changes in the folhcles noticed in ovaries of this period after a preceding copulation and ovulation are also found in ovaries in which a spontaneous ovulation took place independently of a preceding copulation. As we stated above, such a spontaneous ovulation can be produced through an early excision of the corpora lutea. The same follicular changes take place also in pregnant animals in which, through an excision of the corpora lutea about six to eight days after copulation, a spontaneous ovulation is produced approximately thirteen to fifteen days after the beginning of pregnancy, without the pregnancy being interrupted.
We see therefore that these cyclical changes in the ovaries are essentially independent of copulation and of pregnancy and are directly connected only with ovulation.
OVARIES OF GUINEA PIGS SEVEN AND ONE HALF TO EIGHT AND ONE HALF DAYS AFTER OVULATION
At this stage of the sexual cycle we find good follicles of small, medium and large size with no, or only very little, granulosa degeneration. We also find follicles in connective tissue atresia. We see therefore that in approximately eight days follicles origin ally
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 51
very small have reached a large size. The new (eight days old) corpora lutea grow actively during this period and show frequent mitoses in lutein cells. The corpora lutea of the preceding ovulation (second generation) continues to shrink and show marked vacuolization of the lutein cells. If the second last ovulation were accompanied by pregnancy, the retrogressing corpora lutea were still larger.
The third generation of corpora lutea was represented by atretic yellow bodies the age of which varied approximately between forty-eight and ninety-five days.
OVARIES OF GUINEA PIGS TEN TO ELEVEN DAYS AFTER OVULATION
We fmd good follicles without granulosa degeneration of small, medium and large size, besides various stages of granulosa degeneration and of connective tissue atresia, early stages with beginning ingrowth of connective tissue included. In the granulosa of well preserved follicles mitoses are present as usual.
At this stage — ten days after ovulation— the ovary presents again its normal aspect. The follicles have grown to a large size and undergo the ordinary retrogressive changes. The ten to eleven days old corpora lutea are well developed; in the centre a relatively small amount of connective tissue is present. Mitoses in the lutein cells are usually frequent ; they occur perhaps also in endothelial cells of capillaries. The large majority of the vessels have a capillary character, but occasionally a vessel is seen with a double coat of cells. Marked signs of degeneration are absent, but a few slightly vacuolar lutein cells may occasionally be seen.
The second generation of corpora lutea, originating in the second last ovulation, are small vacuolar bodies with much connective tissue and thick vessels. If, however, this second last ovulation had been followed by pregnancy, the retrogressing corpora lutea of the previous pregnancy are as yet much larger; the lutein cells have become very vacuolar; many thick vessels are present. In some of the vacuolar lutein cells yellow pigment appears.
A third generation of corpora lutea is represented by yellow bodies. They are, however, not found in all ovaries.
52 LEO LOEB
In this series of animals pregnancy had been prevented after a preceding copulation, either by ligaturing the tubes within the first two days after copulation, or by making long incisions into the uterus approximately four to six days after copulation.
The ovaries were also examined in a certain number of other guinea pigs of this period in which pregnancy existed. The accompanying pregnancy does not produce any marked change in the ovaries and the preceding description applies on the whole equally well to these ovaries.
OVARIES OF GUINEA PIGS THIRTEEN TO FIFTEEN DAYS AFTER OVULATION
In this series of animals pregnancy was prevented in the same manner as in the series of animals examined ten to eleven days after ovulation. The follicles have approximately the same character as in the previous period. We see the same varieties of follicles. Small foUicles grow and become large and, after having reached this stage, or even at a slightly earlier stage, granulosa degeneration sets in with consecutive connective tissue atresia. In the granulosa of well preserved follicles numerous mitoses are present, and mitoses may even be found, if a slight amount of granulosa degeneration has taken place. The corpora lutea of the last ovulation (I generation) show more generally the beginning of vacuolization, especially in the periphery of the corpus luteum; but on the whole the corpus luteum is still well preserved and usually mitoses are found in some of the lutein and occasionally in cells belonging to blood vessels.
In the center we find connective tissue with thin spindle-shaped nuclei, and a number of vessels with walls consisting of several rows of cells penetrate into the central connective tissue. In some of the lutein cells the protoplasm appears dense and stains deeply with eosin. It appears probable that in such cells the nucleus had started to divide by mitosis, but degenerative processes seem to have set in and interrupted the process of the mitotic division. We are however not certain that this interpretation,
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 53
which would perhaps agree with an opinion expressed by Regaud and Dubreuil,^ is correct.
The second generation of corpora lutea is represented by small vacuolar bodies with relatively much connective tissue and thick vessels. These atretic corpora lutea originated at the time of the second last ovulation and are therefore approximately thirty-three to forty days old. If this second last ovulation had been followed by pregnancy, the corpora lutea of this period are still much larger than the corpora lutea of the corresponding generation without an accompanying pregnancy; but a considerable shrinking of these corpora lutea has also taken place. The vessels are to a great extent collapsed. The lutein cells are finely or coarsely vacuolar, take less stain, still possess nuclei and a distinct cell wall, staining with eosin. The third generation of corpora lutea is again represented by yellow bodies. They are not present in all ovaries, but are found especially in the ovaries of older guinea pigs. Occasionally the degenerating corpora lutea of the second generation may also be absent.
In guinea pigs in which the last ovulation was followed by pregnancy, the condition of the follicles is very similar. The corpora lutea of the first generation, however, are large and show frequent mitoses in lutein cells, occasionally also in lutein cells the periphery of which is vacuolar. There are possibly also mitoses present in the endothelial cells. The retrogressing corpora lutea of the second and third generations are in pregnant animals of a similar character as those described in the ovaries of guinea pigs of the same period without an accompanying pregnancy.
OVARIES OF GUINEA PIGS FIFTEEN TO NINETEEN DAYS AFTER OVULATION
Pregnancy had in most cases been prevented by the same means which were used in the preceding stages. In a few instances in which pregnancy had occurred an early abortion followed. The follicles exhibit on the whole the same character as in the preceding stage ; we find good foUicles of small, medium and large size,
3 C. R. Soc. Biol., 54. 1908.
54 LEO LOEB
and follicles in various stages of granulosa degeneration and of connective tissues atresia. We may also find large mature follicles. In how many cases these latter are present, will still have to be determined. In such animals a rupture of follicles is imminent.
In three guinea pigs a spontaneous ovulation had taken place at this period, notwithstanding the absence of male guinea pigs. In such cases young corpora lutea were found and, accordingly, a condition of the follicles characteristic of a period directly following ovulation. In the large majority of cases however a spontaneous ovulation did not take place in ovaries at this period of the sexual cycle. In such cases the folUcles showed the character described above.
The corpora lutea of the first generation, which originated as a result of the last ovulation, show more or less signs of beginning retrogressive changes as indicated by fine or coarse vacuolization of the lutein cells. The intensity of this degenerative change varies is different ovaries. On the whole the retrogressive changes seem to be more marked in the nineteen days than in the sixteen days old corpora lutea; but variations seem to occur, even in corpora lutea of the same age. The vacuolization is usually most marked in the periphery and progresses toward the center. Other lutein cells are still more solid and mitoses in lutein cells can be seen in the majority of the corpora lutea of this period. In cases in which mature follicles are present and a spontaneous rupture of follicles is therefore soon to be expected, the corpora lutea show much vacuolization; but here also mitoses are still present in lutein cells.
In some cases the retrogressive changes are still further advanced and a connective tissue capsule may appear in the periphery of the corpus luteum. The marked vacuolization of peripheral lutein cells may be accompanied by a diminution in the lumen of blood vessels. Vessels with coats consisting of several rows of cells are seen regularly in these corpora lutea. The connective tissue in the center of the corpora lutea is usually dense and relatively small in amount.
In those cases in which a new spontaneous ovulation had taken place the vacuolization of the corpora lutea had still further pro
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 55
gressed and under such circumstances mitoses were no longer present in them.
The corpora lutea of the preceding (II) generation, originating in an ovulation that took place at least thirty-seven days ago, are sometimes represented by small bodies which are surrounded by a thick connective tissue capsule; much fibrous tissue is found in the center and the lutein cells between these two zones show very large vacuoles. The vessels remaining in such structures have very thick cellular walls. In other cases some yellow pigment appears in such vacuolar cells and in still other cases we see only yellow, atretic bodies. It is probable that the latter structures are found in cases in which a still longer time has elapsed since the preceding (second last) ovulation. There may of course have occurred a longer interval than twenty days between the last and second last ovulation.
When the second generation was represented by a corpus luteum of pregnancy, the retrogressive changes were also marked, shrinking of the corpus luteum and vacuolization of the lutein cells are pronounced, but such corpora lutea are still considerably larger sixteen to nineteen days after the completion of pregnancy than ordinary corpora lutea of the corresponding generation. Some of the vacuolar cells may show a yellow pigmentation. In such ovaries we may find a still older generation of retrogressing corpora lutea present, represented by yellow atretic bodies which owe their origin to an ovulation that took place more than a hundred days ago; and if the last named (third last) ovulation were followed by a pregnancy, this ovulation may have taken place approximately one-hundred and fifty days ago. Not in all animals are so many generations of corpora lutea found ; especially in young animals (two to three months old only one generation may be present.
If the last ovulation that took place fifteen and one half to nineteen days ago were followed by pregnancy, the follicles in the ovaries of pregnant animals of this period do not show any marked difference from the follicles of non-pregnant animals at the corresponding period after ovulation. In both cases we find good follicles of various sizes and the different stages of retrogression of
56 LEO LOEB
follicles which we mentioned above. In the ovaries of pregnant animals of this period we may also find mature follicles, the granulosa cells of which have more cytoplasm that stains red with eosin. Such follicles show less granulosa degeneration and a decrease in the number of mitoses is visible in the granulosa cells. Some degeneration of granulosa cells may however occur in these follicles and their further fate will still have to be determined.
The corpora lutea of pregnancy (first generation) are well preserved. Fine vacuoles may however be present, especially in the peripheral lutein cells. Mitoses are also present. They do not show such pronounced signs of retrogression, as occur in corpora lutea of non-pregnant animals of this period.
OVARIES OF GUINEA PIGS TWENTY TO TWENTY-SEVEN DAYS AFTER OVULATION
At this period the proportion of animals in which a spontaneous ovulation had taken place, notwithstanding the separation of females and males, is much greater than in the preceding period. Among twenty-two guinea pigs a spontaneous ovulation had taken place in eight, while in the fourteen other females no rupture of follicles had as yet occurred. In at least one and possibly in more of these fourteen guinea pigs a rupture was however imminent, as indicated by the presence of mature, red-staining follicles. In those animals in which ovulation had taken place within the last few days the follicles were in the condition corresponding to that stage after ovulation. The corpora lutea that originated as a result of the ovulation twenty to twenty-six days previously showed marked degeneration; the cells were vacuolar; in one case the lutein cells formed a hyaline material in which the vesicular nuclei were imbedded. Mitoses were present in only one case, in which the rupture had taken place apparently within the last twenty-four hours, but even vacuolar cells may divide mitotically. Many blood vessels have thick cellular coats and the blood vessels in general do not seem to be patent.
In all the other guinea pigs in which a new rupture of follicles had not yet taken place the follicles behave approximately in the same manner as in the previous stage; we see follicles of various
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 57
sizes without granulosa degeneration/ and follicles of large and also of medium size in various stages of granulosa and connective tissue atresia.
In the ovaries of the guinea pig in which a spontaneous rupture of follicles was imminent, the twenty-two days old corpora lutea also showed the signs of early degeneration; some of the cells were still good, but the majority were vacuolar.
In the guinea pigs, in which a spontaneous ovulation had not yet taken place, the corpora lutea of the last ovulation were also in a process of degeneration, which was especially marked during the later stages, twenty-four to twenty-six days after ovulation: here the vacuolization was very pronounced, and occasionally connective tissue began to grow into the periphery of the corpus luteum. The vessels of these corpora lutea were very thick. In some other ovaries, especially in those examined twenty and twenty-one days after ovulation, the number of relatively well preserved cells was still greater. On the whole the number of mitoses found in lutein cells at this period is distinctly diminished.
The older generations of corpora lutea are represented by atretic yellow bodies, which are however not present in all animals. In one case a corpus luteum was present that originated as a result of an ovulation that took place approximately ninety-three days before and was accompanied by pregnancy. In this cases twentyseven days after delivery very little of the lutein tissue was left, the blood vessels had very thick coats, and the fibrous tissue of the remnant of the corpus luteum was very prominent.
If the ovulation which took place twenty to twenty-seven days before were followed by a pregnancy, no new spontaneous ovulation took place, 'the conditions of the follicles was the same as in those guinea pigs in which the last ovulation was not followed by pregnancy and in which no new spontaneous ovulation had as yet taken place. The corpora lutea of pregnancy of this period showed much less vacuolization, although a slight amount of it may have been present, especially in the periphery of the corpus luteum. Mitoses were more common in these corpora lutea of pregnancy than in the ordinary corpora lutea of the same period. Their size was also greater.
JOURNAI^ OF MORPHOLOGY, VOL. 22, NO. 1
58 LEO LOEB
In regard to the ordinary corpora lutea and the corpora lutea of pregnancy of previous generations, the same retrogressive changes which were described above in the ovaries of non-pregnant guinea pigs of this period, were found in pregnant animals.
We see therefore that the condition of the corpora lutea indicates the condition of the follicles, and conversely the condition of the follicles indicates the history of the corpora lutea. At a certain time (approximately ten days) after the ovulation a certain equilibrium is reached between the growth and the degeneration of the follicles. Whether a quantitatively exact equilibrium is reached, cannot yet be stated. In proportion to the length of time which elapsed since the last ovulation, the probability of a new spontaneous rupture, with the subsequent changes in the follicles, becomes greater. At this and the preceding period signs of degeneration are present in the ordinary corpora lutea, which become the more marked the older the corpus luteum; the number of mitoses in lutein cells decreases with advancing age ; they may however still be present in corpora lutea immediately following a new ovulation; the latter however is soon followed by further progressing degeneration of the corpus luteum of the preceding ovulation. If the ovulation that took place twenty to twenty-six days previously was accompanied by pregnancy, no new spontaneous rupture of follicles took place, the proliferation of the lutein cells continued, and degenerative processes in the corpora lutea were retarded.
Approximately twenty-five days after the completion of pregnancy the corpora lutea of pregnancy (second generation) have become small vacuolar bodies with thick vessels and fibrous tissue, while corresponding ordinary corpora lutea have at this time apparently been transformed into yellow bodies.
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 59
OVARIES OF GUINEA PIGS TWENTY-SIX TO FORTY DAYS AFTER OVULATION
In five animals in which, in order to prevent pregnancy, both (and in one case one) of the Fallopian tubes had been ligated within twenty-six hours after copulation, and in which at a later operation incisions had been made into the uterus, no new ovulation had taken place at the time of the examination, twentysix to thirty-four days after copulation. The corpora lutea (twenty-six to thirty-four days old) showed very marked retrogression; they were very vacuolar; their size was always diminished especially after thirty-two to thirty-four days, but differed somewhat in individual cases. Some corpora lutea formed small bodies containing very dense fibrous tissue in the center and enclosing in the periphery a relatively small number of very vacuolar cells. Other corpora lutea were still somewhat larger and contained a. few better preserved cells.
Besides the retrogressing vacuolar corpora lutea some atretic yellow bodies could be found in some cases ; they were remnants of corpora lutea at least fifty days old. In two other ovaries a spontaneous ovulation had taken place recently and the condition of the follicles was in accordance with the age of the new corpora lutea. Here also the thirty to thirty-two days old corpora lutea of the preceding ovulation were very vacuolar and contained blood vessels with a thick coat and much dense fibrous tissue.
OVARIES OF A PREGNANT GUINEA PIG APPROXIMATELY THIRTYFIVE TO FORTY DAYS AFTER COPULATION
In these ovaries we found good follicles of small, medium and large size without granulosa degeneration and with mitoses in granulosa cells; other follicles showed various stages of granulosa degeneration and of connective tissue atresia. Mitoses were absent or diminished in number in follicles in which granulosa degeneration existed.
In addition to the ordinary large follicles mature or almost mature follicles were seen in which the cytoplasm of the cells was well developed, and in which the granulosa contained only very
60 LEO LOEB
few mitoses which were found especially in the discus proligerus. Some of the nuclei of the granulosa cells appeared somewhat contracted in these follicles, but no marked degeneration of the granulosa cells was found.
The corpora lutea of pregnancy were large, the cytoplasm of the lutein cells stained red yellow with eosin; the cell outlines were quite distinct. The large majority of the lutein cells were compact and did not show vacuoles; the nuclei were vesicular. A few mitoses were found in lutein cells. Only very little connective tissue was present in the center of the corpora lutea. Some of the vessels had thick walls, while other vessels were of a capillary character and had either a wide or narrow lumen. We see therefore that also at later stages of pregnancy the follicles continue to grow and to degenerate, and that even at this period of pregnancy follicles may mature. The lutein cells of the corpora lutea of pregnancy continue to show mitotic nuclear figures and well preserved cytoplasm at a time when, in the ordinary corpora lutea, retrogression is very far advanced.
OVARIES OF GUINEA PIGS IN WHICH COPULATION HAD BEEN PREVENTED
A large number of ovaries were examined of female guinea pigs which had been kept separated from males for various lengths of time.
One set of guinea pigs was separated from males before sexual maturity had been reached. The ovaries were examined, when the animals were six and twelve months old. In every instance ovulation had taken place repeatedly and we usually found the three generations of corpora lutea which we described in the case of guinea pigs which had copulated, namely relatively young corpora lutea, retrogressing vacuolar corpora lutea and atretic yellow bodies.
In another series guinea pigs were guarded against contact with males after delivery, and were kept separated from males for various periods of time. In this case a spontaneous ovulation took place after delivery, at least in the majority of cases, even
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 61
without contact with males, and subsequently further ovulations occurred. Under such conditions the successive ovulations do however not occur in the same intervals in all animals; in some cases a delay in ovulation may take place : this accords well with our previous observations. The conditions of the follicles correspond to the time elapsed since the last ovulation, as indicated by the state of the corpora lutea.
Not in every case however does a spontaneous ovulation take place without contact with male. In several cases neither new nor retrogressing corpora lutea could be found in the ovaries of guinea pigs which, according to their age, ought to have ovulated, but in which no sign of heat had been noticed during an observation extending over a certain period of time. In other guinea pigs which had been in heat recently, but in which copulation had been prevented, no new ovulation corresponding to the period of heat had taken place at the time of examination.
SOME OBSERVATIONS ON THE POSTFETAL DEVELOPMENT OF THE OVARY OF THE GUINEA PIG
In connection with the cyclic changes in the adult ovary of the guinea pig, just described, we thought it of interest to determine the time at which these cyclic changes set in. For this purpose we studied a series of ovaries at differents stages of the growing guinea pig 1. In the ovaries of a fetus near the time of birth many follicles are present in the cortex. These follicles have not yet a cavity and the largest follicles have a granulosa consisting of three, or four rows of granulosa cells ; in the latter some mitoses can be seen. No distinct differentiation appears in the connective tissue of the different parts of the ovary.
2. In the ovaries of guinea pigs four, five and seven days old we find a cavity in a certain number of the follicles ; no atretic processes have as yet taken place. The theca interna cells are distinguished from the surrounding connective tissue through the
62 LEO LOEB
increase in the size of their nuclei. The connective tissue around the medullary canals is relatively dense. In the granulosa, theca interna and in the ordinary connective tissue stroma mitoses are frequent.
3. The ovaries of guinea pigs eighteen days old are larger; the follicles also have increased in size. Small and medium sized and in proportion to the as yet small size of the ovaries, relatively large follicles are present. In some of the follicles degenerative processes appear at this time, but the extent to which such changes have taken place differs in the ovaries of different animals. In the ovaries of some guinea pigs no degeneration of the granulosa has as yet taken place. In the ovaries of another guinea pig a few follicles showed a trace of granulosa degeneration, while in another follicle the granulosa degeneration was pronounced.
In the follicles of some ovaries we find even a beginning ingrowth of connective tissue into the follicular cavity, and in one case a cavity of a follicle was filled with loose connective tissue. The majority of the follicles are in a good condition; their cavity is larger than at the preceding stage and the interna is better developed and consists of more rows of cells. Mitoses are present in the theca interna and in the granulosa. The connective tissue between the follicles is a little more fibrous, and around certain blood and lymph vessels it is somewhat edematous and rarefied.
4. In the ovaries of guinea pigs twenty-eight days old the majority of follicles are in good condition and non-atretic; they are of small and medium, but not yet very large size. In some ovaries hardly any degeneration of follicles is visible ; in others we see some follicles which have not yet attained their full size (corresponding to the as yet small size of the ovaries) , presenting various stages of granulosa degeneration. In some follicles the granulosa has been entirely destroyed and connective tissue begins to grow into the cavity.' In some cases we find quite atretic connective tissue follicles. In some small and medium sized follicles the ova may undergo (probably amitotic) nuclear division and a corresponding segmentation of the cytoplasm, the granulosa being still intact.
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 63
In other cases, however, such ova are surrounded by connective tissue.
The connective tissue of the ovaries shows more differentiation at this period and is somewhat more fibrous.
5. In ovaries of guinea pigs one to two months old the size of some of the folhcles, in correspondence with the growth of the ovaries, enlarges. We see various stages of granulosa degeneration and of connective tissue atresia. Granulosa degeneration may take place in medium sized and in large follicles. In some ovaries the large majority of follicles may show either granulosa degeneration or connective tissue atresia. Corpora lutea are not yet visible.
6. Ovaries of guinea pigs three months old: Approximately at this period the ovaries have become mature. We find various stages of developing follicles and occasionally mature follicles ready to rupture. We find the various stages of granulosa degeneration and of connective tissue atresia. We notice a greater differentiation in the structure of the stroma in different parts of the ovary.
Corpora lutea, which occasionally are already in the beginning of degeneration, are present in some ovaries; in other animals ovulation has not yet taken place.
It follows from these observations that degenerative processes in follicles set in approximately fourteen to eighteen days after birth, and ovulation and formation of corpoa lutea appear in guinea pigs two to three and a half months old. The ovaries and follicles must have reached a certain size, before ovulation sets in. The time required for the development of small into large follicles, with subsequent beginning of degenerative processes, is somewhat longer in the young growing animal than in the mature guinea pig, but in both the periods of time are of a similar order (approximately nine and fourteen days respectively).
64 LEO LOEB
SUMMARY
The principal result of our investigations we can state as follows : In the ovary of the guinea pig (and probably of mammals generally) cyclic changes take place independently of copulation and of pregnancy.
A sexual period (the period between two ovulations) is accompanied by a series of changes in the follicles. As a result of the conditions leading to or accompanying ovulation the granulosa of all large and medium follicles undergoes a very rapid degeneration, which is very marked within an hour or two after ovulation, or perhaps even sooner. In the follicles in which the cavity is as yet very small, the degenerative processes are very slight or absent. These follicles do not seem to perish. These degenerative changes affect equally both ovaries of one animal, even if a rupture of follicles should have taken place in only one of the two ovaries. The local effect of the rupture of the follicle can therefore not be the cause of the follicular degeneration. Within the next few days the small follicles grow and gradually attain a large size. Eight days after ovulation large follicles are again noticeable. As soon as good sized and medium sized follicles have been formed they begin to undergo degenerative processes, the granulosa degenerating and becoming dissolved and connective tissue growing into the follicular cavity. This process ends in an almost complete disappearance of these follicles. In the meantime other follicles grow and, having reached a large size, they also degenerate. Thus after a first stage of general growth, comprising approximately ten days after ovulation, a certain equilibrium is reached in which new follicles are growing to a certain size, and in which other follicles of large or medium size degenerate. Whether certain quantitative differences in the proportion of the number of growing and degenerating follicles exist at different periods of this second part of the sexual cycle, will still have to be determined. This second period of equilibrium begins approximately ten days after the last ovulation, and it lasts until a new ovulation occurs. Gradually a few large follicles undergo still further changes, the cytoplasm of their granulosa cells enlarges, the number of mitoses in these
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 65
cells decreases and they become more resistant to those processes which lead to degeneration in other follicles. The follicles in which such changes have taken place are mature and ready to rupture. In the meantime the follicles that ruptured during the preceding ovulation developed into corpora lutea. The latter represent principally the hypertrophic granulosa cells of the ruptured folUcles, which proliferate mitotically. After a certain stage of development has been reached, degenerative processes set in in the corpus luteum, which start in its periphery and proceed to the center. These degenerative processes set in very early, are noticeable eighteen to twenty days and are usually marked twenty to twenty-four days after the preceding ovulation. Throughout this period of beginning degeneration, however, some mitoses are still visible in certain lutein cells. At this period usually a new ovulation takes place. The exact time at which the new ovulation occurs varies however somewhat in different animals, ovulation occurring earlier in some animals than in others. In some cases it can be hastened through certain external factors, especially copulation, but in the large majority of cases it occurs sooner or later even without a preceding copulation.^ After the new ovulation has taken place, the degenerative processes progress in the corpus luteum, although within the first twenty hours after ovulation mitoses may still be found in certain lutein cells. In the following period a considerable shrinking of the corpus luteum takes place ; the connective tissue in the cortex and in the periphery becomes hyaline and forms a relatively prominent part enclosing a small number of very vacuolar cells. Gradually yellow pigment is deposited in these vacuolar cells and thus the corpora lutea become transformed into the atretic yellow bodies. The new ovulation was of course again followed by the typical changes in the follicles.
If the ovulation be followed by pregnancy, the principal changes taking place in the ovaries are on the whole the same. The only
■* Whether or not in the guinea pig ovulation can take place independently of a preceding copulation has been a subject of controversy. Concerning the literature of this question see William H. Kirkham, Biological Bulletin, vol. 18, no. 5, April, 1910.
66 LEO LOEB
difference consists in a prolongation of the sexual cycle, which lasts as long as the pregnancy continues. The changes in the follicles are identical with those found in the ordinary sexual period not accompanied by pregnancy.
After copulation the period of growth following the sudden degeneration of the follicles is the same as in the ordinary sexual period, but the period of follicular equilibrium is much prolonged.
During this period of follicular equilibrium certain follicles can not only grow to a considerable size, but may even undergo the additional changes which indicate the maturation of the follicle. A rupture of follicles does not however take place during pregnancy under ordinary circumstances.
The corpus luteum of pregnancy differs from the ordinary corpus luteum mainly in its prolonged duration of growth and of life. At a time when, in the ordinary corpus luteum not accompanied by pregnancy, mitoses have ceased to be present and the retrogressive changes are very marked, mitoses are still seen in the corpus luteum of pregnancy. In the corpus luteum of pregnancy degenerative changes set in before the end of pregnancy has been reached, and they continue after delivery. A short time after delivery a new ovulation usually occurs, even if no copulation had taken place after delivery. The retrogression of the corpora lutea of pregnancy continues, but it requires much more time than the retrogression of an ordinary corpus luteum.
The mechanism that governs the sexual cycle in the ovary can be recognized only incompletely by observation and it has been the subject of an experimental investigation, the results of which we shall report in more detail elsewhere. We may however state that our experiments have shown that through extirpation of the corpora lutea the sexual cycle is shortened. The presence of well functioning corpora lutea inhibits a new ovulation. Pregnancy as such does not prevent ovulation. Ovulation can be made to take place even in pregnancy, if the corpora lutea be extirpated at an early period after copulation. And under such conditions the typical follicular changes follow the ovulation during pregnancy. As soon therefore as degenerative processes have set
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 67
in in the corpora lutea, either during pregnancy or outside of pregnancy, a new ovulation can take place. How far the presence of ,the corpora lutea influences the transformation of ordinary large follicles into mature follicles and how far its action merely concerns the rupture of the mature follicles, remains still to be determined.
It follows from our observations that the time of ovulation depends upon at least three different factors: (1) Changes taking place in the ovaries. It is necessary that mature follicles have been produced, before rupture can take place. Our experiments indicate that cuts into an ovary causing an opening of a follicle may possibly lead to the formation of a corpus luteum only at a time when mature follicles are present. A certain time must therefore have elapsed after ovulation before another ovulation can take place. During this period small follicles reach their full size. Thus a minimal time which must elapse between two ovulations is required. (2) The time at which the influence of the corpus luteum preventing ovulation ceases to be exerted. Our observations make it very probable that the retrogressive changes observed in the corpora lutea before ovulation indicate the necessary cessation of functional activity. It is however noteworthy that, notwithstanding such a cessation of activity, mitoses can still be observed in the lutein cells at this period. Whether the corpus luteum acts principally upon the last stage in the development of follicles (maturation) or merely upon the rupture of follicles will still have to be determined with certainly. We recall however the fact that we observed the occurrence of mature follicles during various stages of pregnancy, notwithstanding the existence of corpora lutea. (3) Certain more or less accidental conditions, as for instance copulation. It is probable that other circumstances also may accelerate or retard the rupture of the follicles. Such factors act probably indirectly by causing changes in the circulation in the ovaries. In the guinea pig these are not indispensable, but their place can be taken by other factors; or even the total absence of corpora lutea may in the guinea pig be sufficient to allow a new ovulation.
68 LEO LOEB
In the guinea pig ovulation occurs in the large majority of cases without any previous copulation. In many cases however copulation is not without significance even in the guinea pig; it accelerates ovulation. While, after delivery, a spontaneous rupture may take place without copulation, in other cases it does not occur without ovulation. Also in the ordinary period of heat ovulation does not need to take place without copulation. Copulation is therefore not without importance; but in almost all of these cases ovulation is only deferred and sooner or later it will take place without the male. So far as the literature has been accessible to us it appears that the role copulation plays had not been fully appreciated by former investigators. Certain observations which we made indicate that other factors besides a preceding copulation may influence ovulation, and we intend to continue our investigation in this direction.
Our observations enable us to give some data concerning the time relations in the growth of various ovarian structures.
a Follicles. In about six days after ovulation small follicles reach medium size. In approximately eight days large follicles have developed and now degenerative processes set in. Mitotic cell division is most pronounced in the granulosa before degenerative processes have commenced; but mitoses may still be seen, if a slight degree of degeneration exist.
h Ordinary corpora lutea. The development of corpora lutea within the first six days after ovulation has been described in a previous paper. At six days we see for the first time, besides the capillary vessels, blood vessels with walls consisting of two rows of cells penetrating into the corpus luteum; they become somewhat more frequent from the tenth day on. In the meantime mitotic division of lutein cells continues and the increase in these cells causes the central connective tissue to become smaller in amount.
In corpora lutea ten to eleven days old a few vacuolar cells are present in the periphery of the corpus luteum. From ten to fifteen days after ovulation vacuolization is still very slight in peripheral luetein cells. From fifteen to eighteen days more fine or coarse vacuolization may appear. Other lutein cells are still
CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 69
more solid and mitoses are still present. If no new ovulation have taken place, degeneration becomes more marked after twenty days; twenty-four days after ovulation we noticed a small amount of connective tissue growing into the periphery. At this period the number of mitoses is already diminished. In cases in which, between the eighteenth and twenty-sixth day after ovulation, a new rupture of follicles sets in, the degenerative processes are still more marked; mitoses may still be seen in the course of the first day after rupture of the follicles, but they disappear afterwards and the degenerative processes progress. The vacuolization of the lutein cells increases, the corpora lutea shrink, the connective tissue becomes gradually hyaline and is relatively preponderating in quantity over the lutein cells. About six days after the new ovulation (in approximately twenty-six days old corpora lutea) yellow pigment may be seen for the first time in the vacuolar lutein cells. Eight days after the new ovulation the corpus luteum is much shrunken, and ten to eleven days after the new ovulation corpora lutea approximately thirty-one to thirty-two days old have been reduced to small vacuolar bodies, around which a strong connective tissue capsule may appear. Corpora lutea thirty-three to forty days old (twelve to nineteen days after new ovulation) still represent vacuolar bodies; but now gradually the transformation into a yellow body sets in. Corpora lutea about forty-five days old have the appearance of yellow bodies and they may probably persist as such for a long time, after the third ovulation has taken place. Thus three generations of corpora lutea may be present side by side in the same ovary, c. Corpus luteum of pregnancy. The corpus luteum of pregnancy differs from the ordinary corpus luteum in the longer duration of mitotic division, and the delay in retrogressive changes. Although slight vacuolization may be noticeable at relatively early stages, the corpora lutea of pregnancy are still in a good condition thirtyfive to forty days after ovulation and they may still show mitoses at this period. Towards the latter part of pregnancy however degenerative processes set in, vacuolization and loss in staining power of the nuclei, and other changes, are noticeable. Mitoses could not be seen at this stage, and they appeared to be absent
70 LEO LOEB
after delivery had taken place. From ten to twelve days after delivery yellow pigment was seen in a few of the lutein cells in the corpus luteum of the previous pregnancy.
Thirteen to twenty days after delivery the corpus luteum is still much larger than an ordinary corpus luteum at the same period after ovulation, but considerable shrinking has already taken place. Twenty-seven days after delivery the corpus luteum is very small and vacuolar, with much hyaline connective tissue, but has not yet been transformed into a yellow body; but at a later stage, approximately sixty days after delivery (or possibly somewhat earlier) the corpus luteum appears as a yellow body, and as such it may persist for some time.
d. In the developing ovaries degeneration of the granulosa and connective tissue atresia of follicles are found as soon as the follicles have reached a relatively large size; these retrogressive changes first appear in guinea pigs approximately fourteen to eighteen days old, while the first ovulation appears much later, namely two to three and a half months after birth.
STUDIES ON CHROMOSOMES
VII. A REVIEW OF THE CHROMOSOMES OF NEZARA; WITH SOME MORE GENERAL CONSIDERATIONS
EDMUND B. WILSON
From the Zoological Department, Columbia University
NINE FIGURES AND ONE PLATE
CONTENTS
Introduction 71
Descriptive 73
1 The second spermatocyte-division in Nezara 73
a The idiochromosomes 73
b The double chromosome 77
2 The first spermatocyte-division 78
3 The growth-period and spermatocyte-prophases 80
4 The diploid chromosome-groups 83
General 84
5 The idiochromosomes 84
a Composition and origin cf the XY-pair 85
b Modifications of the X-element 88
c Sex-limited heredity 94
d Secondary sexual characters 99
6 Modes in which the chromosome-number may change 99
Conclusion 105
INTRODUCTION
In the first of these 'Studies' ('05a) I described the idiochromosomes (X and Y-chromosomes) of Nezara hilaris as being of equal size in the male, and reached the conclusion that in this species no visible dimorphism appears in the spermatid-nuclei. In my third 'Study' ('06), after examination of the female diploid groups, this species was assigned a unique position as the single then
71
72 EDMUND B. WILSON
known representative of a type in which a pair of idiochromosomes can be identified in both sexes, but are of equal size in both, and in which, accordingly, no visible sexual differences appear in the diploid nuclei. These conclusions, as is now apparent, were based upon a wrong identification of the idiochromosome-pair, which is not the smallest pair, as I then believed, ^ but one of the largest. When this fact was recognized, the true conditions soon became evident.
I was led to re-examine Nezara hilaris by the fact (very surprising to me) that in Nezara viridula, a southern species closely similar to N. hilaris, the idiochromosomes of the male are extremely unequal in size, and the dimorphism of the spermatidnuclei is correspondingly marked. Upon returning to the study of N. hilaris it soon became manifest that the dimorphism is present in this species also, though in far less conspicuous form. The size-difference between the X- and Y-chromosomes is here often so slight that I did not at first distinguish it from an inconstant fluctuation of size, such as is sometimes seen between the members of the other chromosome-pairs. When, however, the identity of the XY-pair was correctly recognized, its constancy of position and of size in the second division enabled me to make an accurate comparison between it and the other bivalents; and this fully established the constant inequality of its members, which is constantly greater than that now and then seen in other pairs. Both species also exhibit some other very interesting features that I overlooked in my former studies.
Nezara can therefore no longer stand as a representative of the third of the types distinguished in my third 'Study,' but belongs with Euschistus, Lygaeus, etc., in the second type
1 This was in part because in most of the other forms known at the time the idiochromosomes are in fact the smallest, or one of the smallest, pairs. In part, also, I followed Montgomery ('01) who described in this species two small " chromatin nucleoli" in the spermatogonial groups, and believed them to be identical with the chromatic nucleolus of the growth-period. In a later paper ('06) Montgomery states these "chromatin nucleoli" to be "apparently not quite equal in volume," and asserts that I was in error in describing them as equal. In my material they are certainly equal in the great majority of cases. However, this is not the idiochromosome-pair.
STUDIES ON CHROMOSOMES 73
DESCRIPTIVE
Since the two species agree very closely save in respect to the idiochromosomes they may conveniently be considered together. Before describing the divisions, attention may be called to a striking difference between the two species in respect to the size of the cells and karyokinetic figures. As a comparison of the figures will show, the spermatocytes and maturation divisionfigures of N. hilaris are much larger than those of N. viridula. In the spermatogonia this difference is also apparent, though less marked. In the ovaries, strange to say, it cannot certainly be detected, either in the dividing cells or in the nuclei of the folliclecells or of the tip-cells at the upper end of the ovary. It would be interesting to make a more accurate study of these relations; but I will here only state that the differences between the two species seem to arise mainly through greater growth of the spermatocytes in N. hilaris. With this is correlated a greater size of the testis as a whole; but the size of the entire body in this species is but little larger, as far as I have observed, than in N. viridula.
As regards the general features of the divisions, the diploid groups of both sexes uniformly contain fourteen chromosomes, the first spermatocyte-division eight and the second seven, the idiochromosomes being, as is the rule in Hemiptera, separate and univalent in the first division.
1. The second spermatocyte-division
a. The idiochromosomes. Polar views of the second division always show 7 chromosomes which are usually grouped in an irregular ring of six with the seventh near its center (fig. 3 j-m, figs. 14, 15). In both species one chromosome of the outer ring (s) can usually be distinguished as the smallest, though this is not always evident owing to the apparent variations produced by different degrees of elongation. This is the chromosome that I formerly supposed to be the idiochromosome-bivalent, despite its peripheral position, and despite the fact, which I had myself described, that a similar small chromosome, also peripheral in posi JOUBNAL OF MORPHOLOGY, VOL. 22, NO. 1
EXPLANATION OF TEXT FIGURES
Figures 1 to 9 are from camera drawings, and are not schematized except that in a few instances the chromosomes have been artificially spread out in a series in order to facilitate comparison. Figs. 2 k-l are somewhat more enlarged than the others. In all the figures d denotes the double chromosome or 'd-chromosome,' s the small chromosome, X the large idiochromosome and Y the small.
Fig. 1 The second spermatocyte-division in Nezara viridula. a-d, metaphases
in side view; e-gr, anaphases; h, i, polar views of two sister-groups, middle anaphase, from the same spindle and in the same section.
tion, appears in several other pentatomids (e.g., in Euschistus, Coenus and Mineiis). But Nezara forms no exception to the rule that the central chromosome is the idiochromosome-bivalent. In N. viridula this is immediately apparent in side views (often also in polar views) where the central chromosome is seen to consist of two very unequal components, the smaller being not more than one fourth or one fifth the size of the larger (fig. 1 a-c). In the anaphases these separate and pass to opposite poles, while all the others divide equally (fig. 1 e-g). Polar views of middle or rather late anaphases, when both daughter-groups can be seen superposed in the same section, clearly show the marked difference of the two groups in respect to the idiochromosomes (fig. 1 h-i). All the facts are here so nearly similar to those seen in Euschistus or Lygaeus as to require no further description.
STUDIES ON CHROMOSOMES 75
In N. hilaris the conditions differ only in tiiat the two components of the central chromosome are but sUghtly miequal; but in the examination of at least two hundred of these divisions I have never failed to detect the inequaUty. A series of side views are shown in fig. 2 a-i, figs. 16-21, two of which show all the chromosomes. These figures illustrate practically all the variations that have been seen in the idiochromosomes. The most characteristic condition is that seen in 2 a, b, d, in which both idiochromosomes (X and Y) are more or less elongated and united end to end. Less often one of them assumes a more spheroidal form (fig. 2 e, h, i, fig. 17). The size-difference, though always evident, seems to vary slightly (perhaps because one or the other component may be more or less compressed laterally), but is always distinctly greater than that now and then seen in other bivalents.
Fig. 2 j shows a mid-anaphase^ (cf. figs. 21-23) in which the inequality would hardly be noticed without close study and the comparison of other cases. Figs. 2 k and I are similar stages showing all the chromosomes spread out in a series for the sake of comparison. In both, the two idiochromosomes are easily distinguishable,^ and the larger is seen to be 07ie of the three largest chromosomes. Figs. 2 m-n, o-p, q-r and s-t are pairs of sistergroups, in each case from the same spindle in anaphase. All of these are selected from cases in which a distinct size-difference appears between X and Y, but there are also many cases in which this cannot be seen. Such a case was figured in fig. 4 e-f of my first ' Study' the correctness of which is confirmed by re-examination of the original section. This condition is due simply to the fact that the large idiochromosome is more elongated than the small, so that the size-difference cannot be seen in polar view; and for the same reason it is often not evident in polar views of the metaphase.
2 This and the two following figures are a little more enlarged than the others .
' Fig. 2 I is the same group figured in fig. 4 d of my first 'Study,' carefully redrawn and corrected. A comparison of the two drawings will show that in the latter a distinct size-difference between X and Y is actually shown but is minimized by the fact that the former is represented a trifle too small, the latter a little too large. It is now also evident that they are connected by two connecting fibres instead of by one.
Fig. 2 The second spermatocyte-division in Nezara hilaris. a-i, metaphase figures in side view, a and e showing all the chromosomes; j~l, mid-anaphases; in A; and I all the chromosomes are shown artificially spread out in series; m-n, o-p, q-r, s-t, four pairs of sister-groups from late anaphases, in polar view, in each case from the same spindle.
b. The double chromosome. A seco;iid interesting feature of the
second division that I formerly overlooked is the presence of a
remarkable double chromosome which in the metaphase has exactly the appearance of a butterfly with widespread wings. This
chromosome (which may be called the d-chromosome) is shown in
profile view in 2 b~e and 1 a-d, 16, 17, 20, 24, 25. This is the only
chromosome in the second division that shows any approach to a
quadripartite form, audits characters are so marked as to constitute
the most striking single feature of the division. As the figures
show, it is one of the largest of all the chromosomes. It always
has an asymmetrical tetrad shape, giving exactly the appearance
of a smaller and a larger dyad in close union; and it always lies
in the outer ring, so placed as to undergo an equal division, and
with the larger wings of the butterfly turned towards the axis of
the spindle. In polar view (3 j-m) the duality is far less apparent
and sometimes invisible, even upon careful focussing. In N.
viridula the duality is always apparent in side view, but the butterfly shape is usually less evident than in N. hilaris.
In the initial anaphases the (i-chromosome divides symmetrically, drawing apart into two bipartite chromosomes (2 j, k, I g); but this is seldom evident save in profile view. Viewed from the pole the duality does not now ordinarily appear, though it may still sometimes be seen upon careful focussing. In the later anaphases the two components tend to fuse, and often can no longer be distinguished. Not seldom, however, the duality is visible even in the final anaphases; and sometimes this is so conspicuous that the spermatid-group seems at first sight to comprise eight instead of seven separate chromosomes (n, r, s, t).
Since the duality of this chromosome does not certainly appear in the spermatogonial groups or in the first spermatocyte-division, its peculiar form in the second division might be supposed to result from some special mechanical relation to the spindle-fibers in that division. This is, however, excluded by examination of the interkinesis, in which the chromosomes are irregularly scattered. • In these stages, even when the spindle is still very small and the chromosomes lie in a quite irregular group, the butterfly shape is already perfectly evident; and it shows no constancy of
78 EDMUND B. WILSON
relation to the spindle-axis^ often lying at right angles to the latter. Apparently therefore its duality arises quite independently of the spindle or astral rays, and its constant position in the fully formed spindle is the result of a later adjustment. In this species, as in many others, each chromosome is connected with the pole by a bundle of delicate fibers. In case of the d-chromosome this bundle is very broad, but I cannot be sure that it is double.
At first sight any observer would, I think, take the c?-chromosome to be merely a result of the accidental superposition or close adhesion of two separate dyads of unequal size ; but such an interpretation is inadmissible. When all the chromosomes can be unmistakably seen, the d-chromosome is found to constitute one of the seven separate elements invariably present in this division; and since the diploid number is 14 in both sexes this chromosome must represent one chromosome, not two, of the original spermatogonial groups. It is certain, therefore, that the double appearance does not result from close apposition of two separate chromosomes; it is therefore not a tetrad" in the ordinary sense of the word — i.e., not one that results from the synapsis of two chromosomes that are originally separate in the diploid groups.
2. The first spermatocyte-divisiori
This division requires only brief mention. As stated, it shows eight separate chromosomes, of which the only one that can be positively identified is the Y-chromosome of N. viridula. This chromosome, always immediately recognizable in this species by its small size (3 c, d, f, g, i), figs. 12, 13), is usually central in position, like the m-chromosome of the Coreidae, but this is not invariable. Since it divides equally, and without association with any other chromosome (3 g) it is evident that the two idiochromosomes must be separate and univalent in this division. In N. hilaris (3 a, b, figs. 10, 11) the eight chromosomes usually form an irregular ring, there is no central chromosome, and neither idiochromosome can be certainly recognized. It nevertheless seems a safe inference from what is seen in N. viridula that the two idiochromosomes are here also separate and univalent.
Fig. 3 First and second spermatocyte-divisions in the two species of Nezara.
a, 6, first division, hilaris, polar views: c, d, corresponding views of viridula; first
division, hilaris, side view showing five of the chromosomes in position and the
other three to the right above;/, corresponding view of viridula; g, middle anaphase, viridula, showing division of Y; h, first division metaphase, hilaris, all the
chromosomes artificially spread out in series; i, corresponding view of viridula;
2, k, second division metaphase, hilaris, polar views; I, m, corresponding views
of viridula.
80 EDMUND B. WILSON
In this division the d-chromosome can not be identified in either species. Figs. S e, f, h, i, show all the chromosomes of the two species, in each case from a single spindle in side view. Most of them have a simple bipartite form, but in each species two or three of them often appear more or less distinctly quadripartite as is, of course, often the case with the bivalents in this division. In N. hilaris one of the largest chromosomes is usually more elongated than the others, and each half shows a slight transverse constriction. I suspect that this may be the d-chromosome, but cannot establish the identification.
3. The groivth-period and spermatocyte-prophases
These stages fully bear out the conclusions based upon the divisions and establish the identity of the idiochromosome-pair with the chromatic nucleolus of the growth-period. Throughout the growth-period each nucleus contains a single intensely staining spheroidal chromatic nucleolus and in addition a very large, ■clearly defined pale plasmosome, which is sometimes double. Series of drawings of these two bodies (in each case from the same nucleus, and in their relative position) are given in figs. 4 i-l and m-p, from cells of the middle growth-period. They are also shown in figs. 26-29. In these stages no sign of duahty is to be seen in the chromatic nucleolus, even after long extraction or in saffranin preparations. In later stages, as the chromosomes begin to condense, this nucleolus becomes less regular in outline, and gradually assumes a tetrad form, which becomes very clear as the chromosomes assume their final shape. This transformation may be traced without a break, successive stages being often seen within the same cyst. Just before the nuclear wall breaks down this tetrad is still clearly distinguishable from the others by its asymmetrical quadripartite form, as seen in 4 y, z, which show all the chromosomes (in each case from two successive sections). Figs. 4 q-t show four views of this tetrad at this period in N. hilaris, while u-x are corresponding views of N. viridula. These figures (which might be indefinitely multiplied) show the marked differences between the two species in respect to this tetrad, obviously corresponding to that seen between the idiochromosome
Fig. 4 The diploid groups, nucleoli of the growth-period, and late prophasefigures of the two species of Nezara. a, b, spermatogonial groups, hilaris ; c, d, the
same, viridula; e,f, ovarian groups, hilaris; g, h, the same, viridula; i-l, chromatic
nucleolus and plasmosome from the growth-period, in each case from the same
nucleolus in their relative position; m-p, corresponding views, viridula; q-t, the
idiochromosome-tetrad (chromatic nucleolus) from prophase nucleoli, hilaris;
u-x, corresponding views, viridula; y, late prophase nucleus, showing all the
chromosomes, hilaris (combination figure from two sections) ; z, corresponding
viridula, three of the chromosomes from adjoining section at the right.
82 EDMUND B. WILSON
bivalents of the two in the second division. ^ The two species may in fact readily be distinguished by mere inspection of the chromatic nucleolus at this period. Already at this time the two components are here and there seen to be separating, but as a rule they do not finally move apart until the nuclear wall has dissolved. From this time forward they cannot be individually identified with exception of the small idiochromosome of N. viridula, which is obvious at every period.
As far as my material shows, the earlier stages of the idiochromosomes can not be so readily traced in Nezara as in some other species, and the chromatic nucleolus can not actually be followed backward to the spermatogonial telophases — as can be done in such forms as Lygaeus or Oncopeltus, of which a detailed account will be given in a later publication. The prophase -figures, however, decisively establish its identity with an unequal pair of chromosomes that divide separately in the first spermatocytedivision; and in N. viridula, one of these is certainly the small idiochromosome. It may therefore confidently be concluded that the chromatic nucleolus is identical with the idiochromosome-pair, as in so many other cases. Comparison of the divisionfigures proves that this pair can not be identical with the small pair that I formerly supposed to be the idiochromosome-pair; and this small pair is moreover usually recognizable in the prophase groups (s, in 5 y, z) in addition to the unequal pair.
The foregoing facts make it clear that in Nezara the idiochromosomes undergo a process of synapsis at the same time with the other chromosome-pairs, and that their separation before the first division is a secendary process, to be followed by a second conjugation after this division is completed. A similar process often takes place in many other Hemiptera. There are, however, some forms, like Oncopeltus, in which the idiochromosomes are always separate, from the last spermatogonial division through all the succeeding stages up to the end of the first division. In this case, which I shall describe more fully hereafter, there can be no doubt that the conjugation which follows the first division is a primary synapsis, to be immediately followed by a disjunction.
^ CJ. the earlier figures of the corresponding tetrad in Brochymena in my first 'Studv,' fis. 7.
STUDIES ON CHROMOSOMES 83
4. The diploid chromosome-groups
In these groups the interest centers again in the identity of the idiochromosomes and the d-chromosome. Of the 14 separate chrosomomes present in the diploid nuclei of both sexes, none shows any constant indication of duality (figs. 4 a-h). The dchromosome can not, therefore, be identified in these stages. Secondly, in both species the diploid groups of the two sexes show the same relation as in other Hemiptera of this type, though this is, of course, more readily seen in N. viridula than in hilaris, owing to the small size of the Y-chromosome. In the spermatogonia! groups of this species (4 c, d) this chromosome is at once recognizable while in the female groups {g, h) it is lacking, its place being taken by one of larger size. In both sexes the small pair (s, s) is also recognizable. In this species, accordingly, the Ychromosome is confined to the male line, and the Y-class of spermatozoa must be male-producing, as in other forms.
In N. hilaris the Y-chromosome can not be identified (4 a, h), but the relation of the spermatozoa to sex-production is shown in another way, though less unmistakably than in N. viridula. As already described, the large idiochromosome or X-chromosome is one of the largest three chromosomes seen in the second division. We should therefore expect to see five largest chromosomes in the male diploid groups. This is clearly apparent in figs. 4 a, h, and is also shown in the corresponding figures of N. viridula (c, d) though not quite so clearly. One of these five in the male should be the X-chromosome; and if the usual relation of the spermatozoa to sex hold true, there should be six largest chromosomes in the diploid groups of the female. This relation actually appears in nearly all cases, and is shown in figs. 4 e, f, g, h, in each of which, again, the small pair (s, s) may be recognized. Though this evidence is in itself less convincing than that afforded by N. viridula (since the relation can not always be made out with certainty) it is fully in harmony with the latter, and sustains the same conclusion.^
^ This relation is shown in my original figures of N. hilaris, though not quite as clearly as in the groups here figured. In my first 'Study' ('05) the five largest chromosomes are very clearly shown in fig. 4 h, and are also evident in 4 q. In the third 'Study' the relation is hardly evident in the male but fairly so in the female (figs. 5 I, m).
84 EDMUND B. WILSON
GENERAL
5. The idiochromosomes
The case of Nezara shows how readily a morphological dimorphism of the spermatid-nuclei may be overlooked when the X- and Y-chromosomes do not differ markedly in size ; and it emphasizes the necessity for the closest scrutiny of these chromosomes in the study of this general question. In my fourth 'Study' I placed with Nezara hilaris, as a second example of my original 'third type/ the lygaeid species Oncopeltus fasciatus (Dall.), on the strength of Montgomery's account of the conditions in the male ('01, '06) and my own unpublished observations on both sexes. While I have carefully re-examined this case also, I am not yet prepared to express an unqualified opinion in regard to it. Certainly, in very many of the cells of this species, at every period of the spermatogenesis, the idiochromosomes (which are always separate up to the second division) seem to be perfectly equal. A slight inequality may indeed be seen in some cases; but as far as I can yet determine this seems to fall within the range of the sizevariation in other chromosomes and gives no positive ground for the recognition of a morphological dimorphism in the spermatozoa. A similar condition has been described in several other insects, notably in some of the Lepidoptera (Stevens, '06; Dederer, '08; Cook, '10), in the earwig Anisolaba (Randolph, '08) and apparently also in the beetle Hydrophilus according to Arnold ('08). I see no reason to question these observations; but the interpretation to be placed on them is by no means clear at present. The experimental evidence on the Lepidoptera seems to demonstrate that in at least one case^that of Abraxas according to Doncaster and Raynor, — it is the eggs and not the spermatozoa that are sexually dimorphic ; that is, in the terms that I have recently suggested ('10a), in this case it is the female that is sexually 'digametic' whUe the male is 'homogametic' Baltzer's careful work on the seaurchins ('09) clearly demonstrates a cytological sexual dimorphism in the mature eggs of these animals, and shows that the spermnuclei are all alike. In cases, therefore, where no visible dimorphism of the spermatid-nuclei is demonstrable, two possibilities
STUDIES ON CHROMOSOMES 85
are to be considered, namely, (1) that it may be the female which (as in sea-urchins) is the digametic sex, and (2) that one sex or the other may still be physiologically digametic even though this condition is not visibly expressed in the chromosomes. The first of these possibilities may readily be tested by cytological examination of the female groups. The second can only be examined by means of experiment, and especially by experiments on sex-limited heredity. It is interesting that the work of Doncaster and Raynor, cited above, and the more recent one of Morgan on Drosophila ('10) have given exactly converse results, the former demonstrating a sexual dimorphism of the eggs, the latter of the spermatozoa. This agrees with the cytological data, as far as they have been worked out. The researches of Stevens ('08, 10), on the Diptera establish the cytological dimorphism of the spermatozoa in these animals, while all observers of the Lepidoptera have thus far failed to find such dimorphism in this group. It thus becomes a very interesting question whether a cytological dimorphism of the mature eggs may be demonstrable in the Lepidoptera; though a failure to find it would in no wise lessen the force of the experimental data. Physiological differences between the chromosomes are of course not necessarily accompanied by corresponding morphological ones — indeed such a correlation is probably exceptional.
(1) (a) Composition and origin of the XY~pair. The facts seen in Nezara again force upon our attention the puzzle of the Y-chromosome or 'small idiochromosome.' It is remarkable that two species so nearly akin as N. hilaris and N. viridula should differ so widely in respect to this chromosome; though this is hardly so surprising as the fact that in Metapodius this chromosome, as I have shown ('09, '10) may actually either be present or absent in different individuals of the same species. These facts show, as I have urged, that although the Y-chromosome shows a constant relation to sex when it is present, it can not be an essential factor in sex-production. As the case now stands this might be taken as a direct piece of evidence against the view that the idiochromosomes are concerned with sex-heredity. Further, as I have pointed out ('10) in Metapodius the introduction of super
86 EDMUND B. "WILSON
numerary Y-chromosomes into the female has no visible effect upon any of the characters of the animal, sexual or otherwise; and this might be urged against the whole conception of qualitative differences among the chromosomes and of their determinative action in development. It is especially in view of these and certain other general questions that I wish to indicate some of the many possibilities that must be taken into account in the consideration of this problem. My discussion is throughout based upon the assumption that the chromosomes do in fact play some definite role in determination, and that they exhibit qualitative differences in this respect. I do not hold that they are the exclusive factors of determination; though it is often convenient, for the sake of brevity, to speak of them as if they were such.
(2) Cytologically considered, the morphological dimorphism of the spermatozoa seems to have arisen by the transformation of what was originally a single pair of chromosomes comparable to the other synaptic pairs. We have at present no information as to whether the members of this pair were equal or unequal in size; but in either case there are grounds for the assumption that its two members differed in some definite way in respect to the quality of the chromatin of which they were composed. This pair, which may be called the priixiitive XY-pair, has undergone many modifications in different species, but without altering its essential relation to sex. In the insects (Hemiptera, Coleoptera, Diptera) its most frequent condition is that of an unequal pair, consisting of a 'large idiochromosome' or 'X-chromosome,' and a small idiochromosome" or ' Y-chromosome,' the latter being confined to the male line, while the former appears in both sexes — single in the male and paired in the female. That all gradations exist between cases where X and Y are very unequal (as in many Coleoptera and Diptera and in some Hemiptera) and those in which they are nearly or quite equal (Mineus, Nezara, Oncopeltus) gives some ground for the conclusion that in the original type the XY-pair was but slightly if at all unequal.
By disappearance of the free Y-member of this pair has arisen the unpaired odd or 'accessory' chromosome, which accordingly
STUDIES ON CHROMOSOMES 87
has no synaptic mate. This condition seems to have arisen in more than one way. It is almost certain that in many cases the Y-chromosome has disappeared by a process of gradual and progressive reduction (as indicated by the graded series observed in the Hemiptera (Wilson, '056, '06). In some cases (of which Metapodius is an example) the same result may have been produced suddenly by a failure of the idiochromosomes to separate in the second spermatocyte-division (Wilson, '096). A third possibility, first suggested by Stevens ('06), is that the X-element may have separated from a YY-pair with which it was originally united. This possibility seems to be supported by recent observations on Ascaris megalocephala, where the X-chromosome is sometimes fused with one of the other pairs, sometimes free (Edwards, '10).
(3) We have as yet no positive knowledge as to how the Xniember of the XY-pair originally differed, or now differs, from the Y, or as to how this difference arose— a definite answer to these questions would probably give the solution of the essential problem of sex. There are, however, pretty definite grounds for the hypothesis that the X-member contains a specific ' X-chromatin' that is not present in the Y-member, and that the XY-pair is heterozygous in this respect. If this be so, the primary sexual differentiation is therefore traceable to a condition of plus or minus in this pair, accompanied by a corresponding difference between the nuclear constitution of the two sexes. (Cf. Wilson, '10a.) Further, there is also reason for regarding the heterozygous condition of this pair as due to the presence of the X-chromatin in one member of a pair which is (or originally was) homozygous in respect to its other constituents. The latter may be called collectively the 'Y-chromatin'; and we may, accordingly, think of the XY-pair as being essentially a YY-pair with one member of which the X-chromatin is associated.^ Both the X ^ This suggestion is in principle the same as one earlier made by Stevens ('06, p. 54) that the Y-chromosome represents "some character or characters which are correlated with the sex-character in some species but not in others," with one member of which the X-chromosome is fused; and that "a pair of small chromosomes might be subtracted from the unequal pair, leaving an odd chromosome."
55 EDMUND B. WILSON
chromatin and the Y may themselves be composite, thus giving the possibility of many secondary modifications. The point of view thus afforded opens many possibilities for an understanding of sex-limited heredity, as indicated beyond.
(6) Modifications of the X-element. This view of the XY-pair is based upon two series of facts, of which the first includes the various modifications of the X-member of the pair seen in different species. It is, perhaps, most directly suggested by a study of the pentatomid species Thyanta custator. ' In this common and widely distributed species I have found two races, which thus far can not be distinguished by competent systematists, ^ but which differ in a remarkable way in respect to both the total number of chromosomes and the XY-pair. In one of these races (which I will call the 'A form'), widely distributed throughout the south and west, the total number in both sexes is 16, and the XY-pair of the male is a typical unequal pair of idiochromosomes, exactly like that seen in many other pentatomids {e.g., Euschistus, Coenus or Banasa). These are shown in fig. 5 a, b, their mode of distribution being the usual one. The second race (the 'B form') is thus far known from only a single locality in northern New Jersey. It differs so remarkably from the A fonn that I could not believe the observations to be trustworthy until repeated study of material, collected in four successive years, established the perfect constancy of the cytological conditions and the apparent external identity of the two forms. In this race the XY-pair is represented by three small chromosomes of equal size, which are always separate in the diploid groups and in the first spermatocyte-di vision (fig. 5i), but in the second division are united to form a linear triad series (5 c, d) . This group so divides that one component passes to one pole and two to the other {oe, h), the
^ I am indebted to Mr. E. P. Van Duzee for a careful study of my whole series of specimens of both races. He could find no constant differential between them. Additional studies of this material are now being made by Mr. H. G. Barber.
Addendum. Since this paper was sent to press Mr. Barber, after prolonged study, has reported his conclusion that the 'A form' is Thyanta custator of Fabricius, while the 'B form' is probably Thyanta calceata of Say, which has long been regarded as a synonym of former species.
Fig. 5 Comparison of the XY-group in various Hemiptera. (a-i are original; the others from Payne.) a, b, Thyanta custator, 'A form,' second division
in side view; c, d, corresponding views of the 'B form'; e-h, anaphases of same;
i, polar view of first division of same; j, k, metaphase chromosomes, second division, Diplocodus exsanguis; i, similar view of Rocconot^ anrtailicornis; m, similar
view of Conorhinus sanguisugus; n, Sinea diadema; o, Prionidus cristatus; p,
Gelastocoris oculatus; q, anaphase chromosomes of the same species; r, the XYgroup, from the second division, AchoUamultispinosa; s, diagram, slightly modified
from Payne, to show the distribution of the XY-components in the second division of the same species.
latter being usually in close contact and in later anaphases sometimes hardly separable (5^), though now and then all three components are for a time strung separately along the spindle in the early anaphases, so that no doubt of their distinctness can exist (5/). Comparison of the diploid groups of the two sexes shows that those of the male contain but three of these small chromosomes and those of the female four, the total respective numbers being 27 and 28 (instead of 16 in both sexes, as in the A form).
These facts make it perfectly clear that one of the small chromosomes in the male passes to the male-producing pole, and therefore corresponds to the Y-chromosome ; while the other two, taken together, represent the large idiochromosome, or X-chromosome, of the A form — precisely as in the reduvioids the single X-chromosome of Diplocodus is represented by a double element in Fitchia, Rocconota or Conorhinus (Payne). Had we no other evidence on this point we might assume simply that the original X-chromosome has divided into two equivalent X-chromosomes. But there are other facts that give reason for the conclusion that the breaking up of a single X-chromosome into separate components means something more than this. In the B form, as in' Fitchia or Rocconota (fig. 5 I), the X-element consists of two equal components, but in Conorhinus the two components are always of unequal size (5 m). In Prionidus and in Sinea there are three equal components (5 n, o), in Gelastocoris four equal ones (5 p, and in A choUa multispinosa five, of which two are relatively large and equal and three very small (5 r, s). In every case these components, though quite separate in the diploid groups (and usually also in the first spermatocyte-division) act as a unit in the second division, though not fused, and pass together to the female-producing pole (Payne, '09, '10).
In the foregoing examples the X-element is accompanied by a synaptic mate or Y-chromosome. The following are examples of a similar breaking up of the X-element into separate components when such a synaptic mate is missing. In Phylloxera (Morgan) the X-element consists of two unequal components, sometimes separate, sometimes fused together. In Syromastes (Gross,
Wilson) it consists of two unequal components, always separate, in the diploid groups but closely in contact (not fused) in both spermatocyte-divisions. The recent work of Guyer ('10) indicates a similar condition in the X-element of man. In Agalena (Wallace) there are two equal components, always separate. Finally, in Ascaris lumbricoides (Edwards, '10) there are five components, separate, and scattered in the diploid groups but closely associated in the spermatocyte-divisions.
In all these cases the significant fact is that not only the number but also the size-relations of these components are constant ; and in many of these forms this fact may be seen in such multitudes of cells, and with such schematic clearness, as to leave no manner of doubt. It seems impossible to understand this series of phenomena unless we assume that the single X-chromosome is essentially a compound body — i.e., one that consists of different constituents that tend to segregate out into separate chromosomes. We are led to suspect, further, that the composition of the Xelement, even when it is a single chromosome, may differ widely in different species because of its great variations of size as between different species. For instance, in the family of Coreidae it is in some cases very large (Protenor), in others of middle size (Chelinidea, Narnia, Anasa), in others one of the smallest of the chromosomes (Alydus). Similar examples might be given from other groups.
In the case of Thyanta, therefore, it seems a fair assumption that the double X-element of the B form likewise represents at least a partial segregation of the X-chromatin from other constituents ; and the latter may plausibly be regarded as representing the 'Y-chromatin' of the original X-member of the pair. In other words, we may think of the triad element as a YY-pair, one member of which is accompanied by a separate X-chromosome. In accordance with this its formula should be X.Y.Y, while that of the A form is XY.Y; and this may also be extended to other forms of similar type. If this be admissible, the male formula, as regards essential chromatin-content, becomes in general XY.Y and the female XY.XY, both sexes being homozygous for the Y-constituents, while in respect to X the male is heterozygous, the female homozygous. The puzzle of the Y-chromosome would thus be solved; for although a separate Y-chromosome, when present, is confined to the male line, its disappearance only reduces the male from a homozygote to a heterozygote in respect to the Y-chromatin, and the introduction of supernumerary Y-chromosomes into the female (as in Metapodius) brings in no new element.
Fig. 6 Compound groups formed by union of the X-chromosome with other
chromosomes in the Orthoptera. (a and b, from Sinety, the others from McClung. )
a, triad group; first division of Leptynia, metaphase; b, division of similar triad in
Dixippus; c, triad group formed by union of the l^-chromosome with one of the
bivalents, first spermatocyte-prophase, Hesperotettix; d, the same element from
a metaphase group; e, the same element in the ensuing interkinesis; /, the compound element of Mermiria, from a first spermatocyte prophase; g, the same element in the metaphase (now, according to McClung, united to a second bivalent
to form a pentad) ; h, the same element after its division, in the ensuing telophase.
The same general view as that outlined above is suggested by the constant relation known to exist in some cases between the Xchromosome and a particular pair of the 'ordinary chromosomes.' The first observed case of this was recorded by Sinety ('01) in the phasmid genera Leptynia and Dixippus (fig. 6a,h), where the X-chromosome is always attached to one of the bivalents in the
STUDIES ON CHROMOSOMES 93
first spermatocyte-division, and passes with one half of the bivalent to one pole. Since the spermatogonial number in Leptynia (36) is an even one and twice that of the separate chromosomes present in the first spennatocyte-division, it may be inferred that the X-element is already united with one of the ordinary chromosomes in the spermatogonia, though Sinety does not state this. Somewhat later McClung ('05) discovered essentially similar relations in the grasshoppers Hesperotettix and Anabrus (fig. 6, c-e) , and in case of the first named form was able to establish the important fact that it is always the same particular bivalent with which the X-chromosome is thus associated. In respect to the intimacy of this association, a progressive series seems to exist, since in Leptynia it seems to take place in the spermatogonia, in Hesperotettix only in the prophases of the first spermatocytedivision, while in Thyanta the union is only effected after the first division is completed.
Finally, the recent observations of Boring ('09), Boveri ('09) and Edwards ('10) seem to establish the fact that in Ascaris megalocephala the X-element, whether in the diploid groups or in the maturation-divisions, may either appear as a separate chromosome (which has the usual behavior of an accessory chromosome) or may be indistinguishably fused with one of the ordinary chromosomes.
These relations may, of course, be the result of a secondary coupling; and I m^yself formerly so interpreted them ('09c). But in view of what is seen in Thyanta or the reduvioids we may well keep in mind the possibility that they are expressions or remnants of a more primitive association, like that which I have assumed for an original XY-pair. Whatever be their origin, the effect is the same — a definite linking of the X-chromatin with that of one of the other pairs.
Fig. 7 shows, in purely schematic form, the general conception of these relations that has been suggested above, the X-chromatin being everywhere represented in black. A is the primitive XYpair from which all the other types may have been derived. By simple reduction of such a pair arises the ordinary or typical idiochromosome-pair (B) ; and from either A or B may be derived
the other types (C-G)/ or the more compHcated ones shown in fig. 5. I represents the possible mode of separation of the X-element from a YY-pair, as suggested by Stevens; and this may be reaUzed in Ascaris megalocephala (H). J and K are, schemes of the relations seen in Hesperotettix, Anabrus and Mermiria (cf. fig. 6). These may be direct derivatives of a primitive XY-pair, as the diagram suggests, or may be a result
Fig. 7 Diagram illustrating the possible relation of the various types of idiochromosomes to a primitive XY-pair. Explanation in text.
of secondary coupling of X with other elements. In either case X may itself have such a composition as is indicated in F (Protenor).
(c) Sex-limited heredity. (1) The foregoing considerations have an important bearing on the problem of sex-limited heredity, for they give us a very definite view of how such heredity may be effected. It is not my intention to consider this subject in ex
- These figures are not intended to indicate the precise mode of segregation of
the X- and Y-chromatins of the X-element, but only illustrate possible modes.
STUDIES ON CHROMOSOMES 95
tenso; but I wish to indicate some of the possibiUties that have been opened by the cytological results, even at the risk of offering what may be regarded as too speculative a treatment of the matter. It is obvious that atiy recessive mutation should exhibit sex-limited heredity when crossed with the normal or dominant form, if it be due to a factor contained in {or omitted from) the X-element. For instance, in the remarkable Drosophila mutants discovered by Morgan ('10) the experimental data establish the fact that white eye-color (which seems to follow the same type of heredity as color-blindness in man) is linked with a sex-determining factor in such a way that when the white-eyed male is crossed with the normal red-eyed female, the former character is never transmitted from father to son, but through the daughters to some of the grandsons (theoretically to 50 per cent), though the daughters are not themselves white-eyed ; that is, after such an initial cross, white eyes fail to appear in the Fi generation in either sex and in the F2 generation appear only in some of the males. As Morgan points out, this follows as a matter of course if the factor for white eye be identical with, or linked with, a sex-determining factor in respect to which the male is heterozygous or simplex, the female homozygous or duplex. The X-element exactly corresponds in mode of distribution to such a sex-determining factor; for this chromosome, too, is simplex in the male, duplex in the female and its introduction into the egg by the spermatozoon produces the female condition, its absence the male. This chromosome therefore, as I have shown ('06), is never transmitted from father to son, but always from father to daughter. Conversely, the male zygote always receives this chromosome from the mother. So precise is the correspondence of all this with the course of sexlimited heredity of this type that it is difficult to resist the conclusion that we have before us the actual mechanism of such heredity — in other words, that some factor essential for sex is associated in the X-element with one that is responsible for the sex-limited character.
This will be made clearer by the accompanying diagram (fig. 8) where the X-element assumed to be responsible for a recessive sex-limited character is underscored (X) . This character may
Fig. 8 Diagram of the distribution of the X- and Y-elements in successive generations, illustrating sex-limited heredity. The underscored X-element (X) is assumed to bear a factor for a recessive character {e.g., white eye-color), while X represents the normal or dominant character {e.g., red eye-color). Y (being the absence of X) likewise represents the recessive character.
Upon pairing the affected male (XY) with the normal female (XX) there are in the Fj generation but two possible combinations, XX and XY. The affected X-chromosome here passes into the female, and the male is normal; but the female of course likewise shows only the normal (dominant) character. In the following F2 generation (5) there are four possible combinations XX, XX, XY and XY, two of each sex. Though X is present in half of each sex, the character appears only in the males, XY, again because of its recessive nature. By crossing together males of the composition XY and females of composition XX, some of the resulting females will have the composition XX , and the sexlimited character is thus made to appear in the female.
When the female is the heterozygous or digametic sex — as in sea-urchins, in Abraxas, the Plymouth Rock fowls, etc. — exactly the converse assumption has to be made. Here, as Spillman ('08) and Castle ('09) have pointed out, the observed results follow if the sex-limited character {e.g., lacticolor color-pattern in Abraxas) be allelomorphic to, or the synaptic mate of, a sexdetermining factor, X, that is present as a single element in the female but absent in the male. The formulas now become'^ (as Spillman has indicated) XG (9 grossulariata), GG (d^ gross.) XG (9 lacticolor) and GG (cf lact). XG X GG then gives in Fi XG and GG (gross. 9 and d^), G having passed from the female to the male. The following cross, XG X GG gives in F2 the four types XG, XG, GG and GG, — i.e., grossulariata appearing in both sexes but lacticolor only in the female. By crossing XG with GG some of the progeny will have the composition GG (cT lacticolor). The other combinations follow as a matter of course.
This interpretation is in all respects the exact converse of that made in the case of Drosophila, which is also the case with
^ These formulas are in substance the same as those stated by Mr. SpiHman in a private letter to the writer, and are a simplified form of those suggested by Castle ('09). The interpretation thus given seems both the simplest and the most satisfactory from the cytological point of view of all those that have been offered. It obviates the cytological difficulties that I urged ('09) against Castle's formulas, and renders unnecessary the secondary couplings that I suggested. All these ways of formulating the matter conform, of course, to the same principle and differ only in details of statement. Whether the synaptic mate of X is directly comparable to the Y-chromosome of other insects (in which case the female formula becomes XY and the male YY) is an open question.
the experimental results, as Morgan has pointed out. It seems probable that all the observed phenomena may be reduced in principle to one or the other of these schemes, though many modifications or complexities of detail probably exist. A possible basis for many such modifications seems to be provided by the cytological facts already known.
(2) We might assume that in cases of the first type {e.g., Drosophila) both sex and the characters associated with it are determined by the same chromatin; and such a possibility should certainly be borne in mind. But in view of the widely different nature of the characters already known to exhibit sex-limited heredity it seems improbable that we can regard them as all alike due to the same chromatin. In the light of the conclusions that have been indicated in regard to the composition of the X-element, it seems more probable that such characters may be determined by the various other forms of chromatin (' Y-chromatin') associated with the X-chromatin. If these constituents be identical with those contained in the free Y-chromosome (the synaptic mate of X) sex-limited heredity would of course not appear, since the two members of the pair would be homozygous in this respect. It should make its appearance as a result of the dropping out, or other modification, of certain Y-constituents of the X-element, and such a mutation might arise in either sex.
We may perceive here the possibility of understanding many different kinds of sex-limited heredity, perhaps of complex types that have not yet been made known. Such a possibility is suggested, for example, by the remarkable relation discovered by McClung ('05) in Mermiria (fig. 6/-/i, fig. 7 in diagram), where the X-chromosome is in the first spermatocyte-division attached at one end to a linear chain of four other elements to form a pentad complex, to which may be given the formula XA . ABB. This so divides as to separate XA from ABB. The interpretation to be placed upon this is a puzzling question under any view, and apparently must await more extended studies on both sexes, perhaps also on other forms, before it can be fully cleared up. Even here the possibility exists, I think, that the entire complex may have arisen by the differentiation of a single original XY-pair; but this question is clearly not yet ready for discussion. However such associations have arisen, the result is equally applicable to the explanation of sex-limited heredity.
(d) Secondary sexual characters. Castle ('09) has offered the interesting suggestion that the free Y-chromosome may be responsible for the determination of secondary sexual characters in the male. Though I have criticized this view ('09c) I now believe it may be true for certain cases. It is obviously excluded when the Y-chromosome is missing; and since nearly related species — in Metapodius even different individuals of the same species — show the same or similar secondary male characters whether this chromosome be present or absent, it seems probable that these characters are in general determined in some other way. But if, as I have suggested, sex-limited heredity may arise through a modification of the Y-constituents of the X-element, it follows that the YY-pair thereby becomes heterozygous. In such case, the free Y-chromosome, being confined to the male line, should continue to represent characters that are no longer present in the female, and hence would be indistinguishable from secondary male characters otherwise determined. It has further become evident (as is indicated below) that the chromosome-groups are so plastic that their specific composition may vary widely from species to species. It may very well be, therefore, that Castle's suggestion may apply to some forms.
6. Modes in which the chromosorne-number may change
The constant and characteristic duality of the 'd-chromosome' in the second division suggests a series of questions regarding the mode in which the chromosome-number may change that have an important bearing on those already considered. The appearance of this chromosome must suggest to any observer that it is a compound body, consisting of two closely united components that are invariably associated in a definite way ; but it is especially noteworthy that its duality does not certainly appear before the last division.^ This case must be added to the steadily increasing evidence that chromosomes which appear single and homogeneous to the eye may nevertheless be compound bodies. An important part of it is derived from the modifications of the Xelement reviewed above; but the evidence is now being extended to the 'autosomes' or ordinary chromosomes as well. The double chromosome of Nezara suggests either the initial stages of a separation of one chromosome into two or the reverse process — in either case that we have before us one way in which the number, and the composition, of the chromosomes may change from species to species. This is supported by the recent observations of Miss E. N. Browne ('10) on Notonecta. In N. undulata there are always, in addition to a typical unequal XY pair, two small chromosomes that appear in all the divisions as separate elements. In N. irrorata there is always but one such chromosome, the total number in each division being accordingly one less than in N. irrorata. N. insulata presents a condition exactly intermediate, there being sometimes one and sometimes two such small chromosomes. When, however, but one seems to be present, the second may often be seen closely adherent to one of the larger chromosomes; and the latter may positively be identified, by its size, as always the same one. It can hardly be doubted, as the author points out, that we here see three stages in a process by which the chromosome-number is changing, either by the fusion of two originally separate chromosomes, or by the separation of one into two. It is of the utmost importance that this process affects a chromosome that can be positively identified as the same in each case; for this proves that the change is not an indefinite fluctation but the expression of a perfectly orderly process. While there is here (as in the case of the d-chromosome of Nezara) no way of knowing in which direction the change is taking place, either alternative involves the conception that the individual chromosomes may be compound bodies, whether as a result of previous fusion or as possible starting points for a subsequent segregation.
The facts now known indicate at least four possible ways in which the chromosome-number (and in three of these also the composition of the individual chromosomes, may change from species to species.
One is that suggested by the foregoing phenomena, i.e., the gradual fusion of separate chromosomes into one or the reverse process.
A second mode may be the gradual reduction and final disappearance of particular chromosome-pairs, as was suggested by Paulmier ('99), and afterwards by Montgomery and myself, in case of the microchromosomes, or ' m-chromosomes' of the coreid Hemiptera. In respect to the size of these chromosomes, a graded series may be traced from forms in which they are very large (as in Protenor) through those where they are of intermediate size down to cases where they are very small (as in Archimerus) and finally to such a condition as that seen in Pachylis (fig. 9 j-l) where they are almost as minute as centrioles and may almost be regarded as vestigial. Four of these stages are shown in fig. 9. In Protenor {a-c) the wi-chromosomes are so nearly of the same size as the next smallest pair that they often can not be positively identified in the spermatogonia! groups. In Leptoglossus phyllopus {d-f) they are always recognizable, though not much smaller than the next pair. In L. oppositus or L. corculus they are a little smaller. In Anasa (the form in which they were first discovered by Paulmier) they are of middle size {g-i) , representing perhaps a fair average of the group. Several other genera {e.g., Metapodius) show intermediate stages between this condition and that seen in Archimerus (figured in my second 'Study,' and more recently by Morrill) where the ??i-chromosomes are almost as small as in Pachylis. It is most remarkable that throughout this whole series the m-chromosomes exhibit essentially the same behavior (Wilson, '056, '06), usually remaining separate throughout the entire growth-period and only conjugating in the final prophases of the first spermatocyte-division, to form a bivalent which with rare exceptions occupies the center of the metaphase group; in some forms, also (e.g., Protenor, Alydus) they show a marked tendency to condense at a much earlier period than the other chromosomes. The m-chromosomes of Pachylis, excessively minute though they are, exhibit a behavior in all respects as constant and characteristic as even the largest of the series. In the Lygaeidae they seem to be present in some
Fig. 9 Comparison of the m-chromosomes in Hemiptera. (In each horizontal row are shown at the left a spermatogonial group, in the middle a polar view
of the first spermatocyte-division, at the right a side-view of the same division.)
a-c, Protenor belfragei; d-^, Leptoglossus phyllopus; g-i, Anasa tristis; j-l,
Pachylis gigas.
STUDIES ON CHROMOSOMES 103
species (Oedancala, t. Montgomery), in others absent (Lygaeus). In the Pyrrhocoridae (Pyrrhocoris, Largus) they are absent as far as known. So characteristic is the behavior of these chromosomes as to leave not the least doubt of their essential identity throughout the whole series ; and this series may be regarded as a progressive one, in one direction or the other, with the same reason as in case of any other graded series of morphological characters. The series thus shown in case of the m-chromosomes is as gradual and complete as in case of the Y-chromosome, and may with the same degree of probability be regarded as a descending one. Thirdly, it is probable that the chromosome-number may change by sudden mutations that produce extensive redistributions of the chromatin-substance without involving any commensurate change in its essential content. Were gradual changes, chromosome by chromosome, the usual mode of modification, we should certainly expect to find such conditions as are seen in Nezara, in Notonecta, or in the Coreidae, more frequently. In some groups, however, we find wide differences of chromosomenumber between species even of the same genus, and even between those that are very nearly related, without any accompanying evidence of a gradual process of transition — for instance, among the aphids and phylloxerans (Stevens, Morgan) or in the heteropterous genera Banasa and Thyanta. (Wilson, '09d.) In Banasa dimidiata the diploid number is 16 in both sexes, in the nearly related B. calva 26. Of the two races of Thyanta custator described above, apparently identical in other visible characters, one has in both sexes the diploid number 16, with a simple X-chromosome, while in the other the diploid number of the male is 27 and that of the female 28, and the X-chromosome consists of two components. It is improbable that the differentiation of these two forms has been accomplished by gradual modifications, chromosome by chromosome. It seems far more likely that the change took place by sudden mutation, involving a redistribution of the nuclear material which changed its form but not in an appreciable degree its substance. In the well known case of Oenothera gigas, derived by sudden mutation from Oe. Lamarckiana, a change by sudden mutation is known to be
104 EDMUND B. WILSON
PLATE 1
EXPLANATION OF FIGURES
All the figures from photographs of sections. Enlargement 1500 diameters.
10, 11 First spermatocyte-division (N. hilaris)
12, 13 The same (N. viridula)
14, 15 Second spermatocyte-division (hilaris)
16-25 Side views of second division (hilaris). The XY-pair shown in 16-23, the
d-chromosome in 16, 17, 20, 24, 25; the small chromosome is evident in
10, 12, 13, 14, 15, 17, 18.
22 Initial separation of X and Y
23 Early anaphase, X and Y separating near the center (hilaris)
26-28 Nuclei from the growth-period, showing chromosome-nucleolus and plas mosome (hilaris) 29 Corresponding stage (viridula)
a fact (Lutz, '07; Gates, '08), though it may be due in this instance to a simple doubling of the whole group. Such cases led me several years ago to the conclusion that the nucleus consists of many different materials that segregate in a particular pattern . . . and that the particular form of segregation may readily change from species to species" (Wilson, '09d, p. 2).
Such changes must involve corresponding ones in the morphological and physiological value of the individual chromosomes; and we must accordingly recognize the probability that these individual values, though constant for the species, may change from species to species as readily as does the number. Despite the conformity to a -general type often exhibited by particular genera or even by higher groups, the individual chromosomes are therefore per se of subordinate significance; and it may often be practically impossible to establish exact homologies between those of different species. How closely this bears on the origin of the diverse conditions seen in the composition of the XY-pair is obvious.
Lastly, it is almost certain that changes of number may sometimes arise as a result of abnormalities in the process of karyokinesis, such as the passage of both daughter-chromosomes, or of both members of a bivalent, to one pole. In Metapodius I found ('096) direct evidence of this in case of the XY-pair itself, and endeavored to trace to this initial cause the remarkable variations of number that occur in this genus. Many other observers have recorded anomalies of this kind, in both plants and animals. It seems entirely possible, as has been suggested by McClung ('05) and by Gates ('08) that definite mutations may be traceable to this cause; though probably such abnormalities may in general be expected to lead to pathological conditions.
CONCLUSION
Some of the suggestions offered in the foregoing discussion are admittedly of a somewhat speculative character; but they are not, as I venture to think, mere a priori constructions, but are forced upon our attention by the observed facts. The time has come
JOURNAL OF MORPHOLOGT, VOL. 22, NO. 1
106 EDMUND B. WILSON
when we must take into account more fully than has yet been done the new complexities and possibilities that have continually been unfolded as we have made better acquaintance with the chromosomes. In this respect the advance of cytology has quite kept pace with that of the experimental study of heredity; and it has established so close and detailed a parallelism between the two orders of phenomena with which these studies are respectively engaged as to compel our closest attention.
Studies on the chromosomes have steadily accumulated evidence that in the distribution of these bodies we see a mechanism that may be competent to explain some of the most complicated of the phenomena that are being brought to light by the study of heredity. New and direct evidence that the chromosomes are in fact concerned with determination has been produced by recent experimental studies, notably by those of Herbst ('09) and Baltzer ('10) on hybrid sea-urchin eggs. But the interest of the chromosomes for the study of heredity is not lessened, as some writers have seemed to imply, if we take the view — it is in one sense almost self-evident — that they are not the exclusive factors of determination. Through their study we may gain an insight into the operation of heredity that is none the less real if the chromosomes be no more than one necessary link in a complicated chain of factors. From any point of view it is indeed remarkable that so complex a series of phenomena as is displayed, for example, in sex-limited heredity can be shown to run parallel to the distribution of definite structural elements, whose combinations and recombinations can in some measure actually be followed with the microscope. Until a better explanation of this parallelism is forthcoming we may be allowed to hold fast to the hypothesis, directly supported by so many other data, that it is due to a direct causal relation between these structural elements and the process of development.
A second point that may be emphasized is the remarkable constancy of the chromosome-relations in the species, and their no less remarkable plasticity in the higher groups. The scepticism that has been expressed in regard to constancy in the species finds, I think, no real justification in the facts. It is perfectly true that
STUDIES ON CHROMOSOMES 107
individual fluctuations occasionally are seen in the number of the chromosomes, in the process of synapsis, in the distribution of the daughter-chromosomes, and in all other cytological phenomena. It is, however, also true that most observers who have made prolonged, detailed and comparative studies of any particular group, have sooner or later reached the conviction that in each species all the essential relations in the distribution of the chromosomes conform with wonderful fidelity to the specific type. So true is this that the species "may often at once be identified by an experienced observer from a single chromosome-group at any stage of the maturation-process. No one, I believe, who has engaged for a series of years in the detailed study of such a group, for instance, as the Hemiptera or the Orthoptera, returning again and again to the scrutiny of the same material, can be shaken in the conviction that the distribution of the chromosomes follows a perfectly definite order, even though disturbances of that order now and then occur. But it is equally important to recognize the fact that this order has undergone many definite modifications of detail from species to species, and that while all cases exhibit certain fundamental common features, we cannot without actual observation predict the particular conditions in any given case. It is now evident that the larger groups vary materially in respect to specific conditions. For instance, in the orthopteran family of Acrididae (McClung) the relations seem to be far more uniform than such a group as the Hemiptera, where great specific diversity is exhibited, the details often changing from species to species in a surprising manner — witness the species of Aphis or Phylloxera (Stevens, Morgan), those of AchoUa (Payne) or of Thyanta (Wilson). In these respects, too, the cytologist finds his experience running parallel to that of the experimenter on heredity; and here, once more, we find it difficult not to believe that both are studying, from different sides, essentially the same problem.
December 13, 1910.
108 EDMUND B. WILSON
LITERATURE CITED
Arnold, G. 1908 The nucleolus and microchromosomes in the spermatogenesis of Hydrophilus piceus. Arch. Zellforsch., vol. 2,
Baltzer 1909 Die Chromosomen von Strongylocentrotus lividus und Echinus microtuberculatus. Arch. f. Zellforsch., Bd. 2.
1910 Ueber die Beziehung zwischen dem Chromatin und der Entwicklung und Vererbungsrichtung bei Echinodermenbastarden. Habilitationsschrift, Wurzburg. Engelmann, Leipzig.
Boring 1909 A small chromosome in Ascaris megalocephala. Arch., f. Zellforsch., vol. 4.
BovERi, Th. 1909 " Geschlechtschromosomen" bei Nematoden. Arch. f. Zellforsch., Bd. 4.
Browne, E. N. 1910 The relation between chromosome-number and species in Notonecta. Biol. Bull., vol. 20,1.
Castle, W. E. 1909 A Mendelian view of sex-heredity. Science, n. s., March 5.
Cook, M. H. 1910 Spermatogenesis in Lepidoptera. Proc. Acad. Nat. Sci., Philadelphia, April.
Dbderer, p. 1908 Spermatogenesis in Phyllosamia. Biol. Bull., vol. 13.
Edwards, C. L. 1910 The idiochromosomesin Ascaris megalocephala and Ascaris lumbricoides. Arch. f. Zellforsch., vol. 5.
Gates, R. R. 1908a The chromosomes of Oenothera. Science, n. s., vol. 27, Aug. 2.
1908b A study of reduction in Oenothera rubrinervis. Bot. Gazette, vol. 46,
1909 The behavior of the chromosomes in Oenothera lata x O. gigas. Ibid., vol. 48.
Gross, J. 1904 Die Spermatogenese von Syromastes marginatus. Zool. Jahrb. Anat. u. Ontog., vol. 20.
GuTER, M. 1910 Accessory chromosomes in man. Biol. Bull., vol. 19.
Herbst, C. 1909 Vererbupgsstudien, VI. Die cytologischen Grundlagen der Verschiebung der Vererbungsrichtung nach der miitterlichen Seite. Arch. Entwicklungsm., Bd., 27.
LuTZ, A. M. 1907 A preliminary note on the chromosomes of Oenothera La. marckiana and one of its mutants. Sci., n. s. 26.
McClung, C. E. 1905 The chromosome complex of orthopteran spermatocytes. Biol. Bull., vol. 9.
STUDIES ON CHROMOSOMES 109
Montgomery, T. H. 1901 A study of the chromosomes ofMetazoa. Trans. Am. Phil. Soc, vol. 20.
1906 Chromosomes in the spermatogenesis of the Hemiptera Heteroptera. Trans. Am. Phil. Soc, vol. 21.
Morgan, T. H. 190Ga A biological and cytological study of sex-determination in phylloxerans and aphids. Jour. Exp. Zool., vol. 7,
1910 Sex-limited inheritance in Drosophila. Science, n. s. 32, July 22.
Morrill, C. V. 1910 The chromosomes in the oogenesis, fertilization and cleavage of coreid Hemiptera. Biol. Bull., vol. 19.
Patjlmier, F. C. 1899 The spermatogenesis of Anasa tristis. Jour. Morph., vol. 15, Suppl.
Payne, F. 1909 Some new types of chromosome distribution and their relation to sex. Biol. Bull., vol. 16.
1910 The chromosomes of Acholla multispinosa. Biol. Bull., vol. 18.
Randolph, Harriet. 1908 On the spermatogenesis of the earwig, Anisolaba maritima. Biol. Bull., vol. 15.
Sinety, R. de 1901 Recherches sur la biologie et I'anatomie des phasmes. La Cellule, t. 19.
Spillman, W. J. 1908 Spurious allemorphism. Results of some recent investigations. Am. Naturalist, vol. 42.
Stevens, N. M. 1906 Studies in spermatogenesis, II. A comparative study of the heterochromosomes in certain species of Coleoptera, Hemiptera and Lepidoptera, etc. Carnegie Inst. Pub. 36.
1908 A study of the germ-cells of certain Diptera, etc. Jour. Exp. Zool., 5, 3.
1910 The chromosomes in the germ-cells of Culex. Jour. Exp. Zool., vols.
Wallace, L. B. 1909 The spermatogenesis of Agalena nsevia. Biol. Bull., vol. 17.
Wilson, E. B. 1905a Studieson chromosomes, I. The behavior of the idiochromosomes in Hemiptera. Jour. Exp. Zool., vol. 2.
1905b Studies on chromosomes, II. The paired microchromosomes, idiochromosomes, and heterotropic chromosomes in Hemiptera. Jour. Exp. Zool., vol. 2.
1906 Studies on chromosomes, III. The sexual differences of the chromosomes in Hemiptera. Jour. Exp. Zool., vol. 3.
1909a Studies on chromosomes, IV. The accessory chromosome in Syromastes and Pyrrhocoris. Jour. Exp. Zool., vol. 6.
no EDMUND B. WILSON
1909b Studies on chromosomes, V. The chromosomes of Metapodius, etc. Jour. Exp. ZooL, vol. 6.
1909c Secondary chromosome-couplings and the sexual relations in Abraxas. Science, n. s. 29, p. 748.
1909d Differences in the chromosome-groups of closely related species and varieties, etc. Proc. Seventh Internat. Zool. Congress, Aug. 1907.
1910a The chromosomes in relation to the determination of sex. Science Progress, no. 16, April.
1910b Studies on chromosomes, VI. A new type of chromosome-combination in Metapodius. . Jour. Exp. Zool., vol. 9.
1910c Note on the chromosomes of Nezara. Science, n. s. 803, May 20.
THE TRANSPLANTATION OF OVARIES IN CHICKENS^
C. B. DAVENPORT
From Carnegie Institution of Washington: Station for Experimental Evolution
Dr. C. C. Guthrie ('08) has reported the results of transplanting ovaries from black to white hens and vice versa. A blackplumaged hen furnished by transplantation with 'white' eggs and mated to a white cock gave about equal numbers of white and spotted" chicks. Guthrie thinks that these black spots indicate that the black-plumaged foster-mother infected the engrafted 'white' eggs. So far Guthrie. But a person familiar with the results of hybridizing v/ill appreciate that Guthrie's result is better explained on the assumption that the engrafted ovary was absorbed and that the white sperm fertilized the regenerated 'black' eggs of the black hen. For the white by black cross gives white offspring with black spots in the female chicks only, i.e., half of all, as Guthrie found.
In a second set of experiments, Guthrie found that when a white hen carrying a 'black' ovary was mated to a White Leghorn male, the offspring were either white or black or spotted. Guthrie says : "The black, therefore, must have come through the black ovary." But the student of hybridization on poultry will recognize at once that, if the white-plumaged cock produced only ' white ' germ cells, none of his offspring would be black even if the eggs were 'black.' Hence, the cock must have had 'black' germ cells and, very likely, the hen also, since 'White Leghorn' hens that carry 'black' germ cells are very common and frequently show, in adult life, a pure white plumage.
If two 'White Leghorns' with 'black' germ cells be mated expectation is that in four chicks one shall be black; one spotted, and
' A preliminary paper covering these results was read before the Society for Experimental Biology and Medicine, June 1910.
Ill
112 C. B. DAVENPORT
two white; Guthrie got five chicks, one black, one spotted and three white.
Guthrie found that a black hen containing a 'white' ovary, mated with a black cock gave black-plumaged chicks, of which two out of six had white feet. He concludes that the white condition of the feet must have come from the engrafted eggs of the White Leghorn. In criticism it must be pointed out that the cross, white egg X black sperm, normally gives offspring whose plumage color is white, either pure or with black specks. The fact that all the offspring had black plumage proves that the eggs were the normal ' black ' eggs regenerated by the black hen. The white toes are frequently found in the offspring of two black birds. Thus in my pen 1041 two extracted blacks (Sumatras) mated give ten black chicks in six of which white toes are recorded. The results of this cross of Guthrie's confirm the conclusion that the transplanted ovaries were not functional and that the normal ovaries had regenerated.
To -test the possibility of such regeneration of ovaries I removed the ovaries of some hens in the autumn of 1909 and transplanted into them eggs from dissimilar hens. The operated birds were then mated to cocks resembling the soma of the so-called 'fostermother.' Were there regeneration of the ovary the offspring should be of the straight breed; but if the 'grafts' persisted and became functional the chicks should be hybrids.
Experiments 1 and 2, operations: The protocol of the grafting operations is as follows :
No. 11379, pure-bred Dark Brahma bantam, hatched February, 1909; made to fast two days. On September 29, 1909, injected with 0.005 grain of atropin in 1 cc. of water, etherized in about twelve minutes and opened up between two left intercostals. Large ovary, badly torn in removal, removal tolerably complete. One piece of ovary from no. 11605 fastened by cotton thread to inesentery near attachment of ovary. Sewed up.
No. 11605, hatched March, 1909, from White Leghorn,-Houdan ancestry. Clean-footed, with five toes on each foot, V-comb, modified high nostril, plumage color white (with black recessive). On September 29, 1909, injected 0.005 grain atropin in 1 cc. of
THE TRANSPLANTATION OF OVARIES IN CHICKENS 113
water. Etherized in twenty minutes. Plucked feathers and opened body wall between last two ribs. Large ovary completely removed or nearly so, in three or four pieces. Hemorrhage slight. Stitched in small piece of ovary of no. 11379 to peritoneum near attachment of old ovary. Sewed up. Bird recovered rapidly. Some Dark Brahma in ancestry, but its characters had become eliminated.
Results, Experiment 1. Mated in pen 1027, no. 11605 9 (with engrafted ovary from no. 11379, Dark Brahma) and 11291°", straight Dark Brahma. Table 1 gives the juvenile characteristics of 1, the male; 2, the White Leghorn-Houdan, so-called foster-mother; 3, the hen from which the ovaries were transplanted ; 4, expectation on the hypothesis that the graft succeeded; 5, expectation on the hypothesis that the graft failed and the proper ovary was regenerated ; and 6, the observed characteristics of the young offspring.
An examination of table 1 shows at once that it cannot be true that the engrafted ovary replaced the hen's proper ovary, for if it had, columns six and four should agree. On the contrary, column six agrees essentially with column five and supports the hypothesis that the engrafted eggs did not become functional.
One discordant fact there is, however, namely, the occurrence in column six of three cases of cinnamon offspring. Such offspring are to be expected on the hypothesis that some eggs of the graft became functional. If that hypothesis be true, then the other characters of the same individuals should be like those of the pure Dark Brahma. Of the three the first has extra toes, split comb and a boot of one row; it is no Dark Brahma; the second has extra toes, wide nostril and a two rowed boot; it is not a Dark Brahma; and the third has really black down with some red at the tips, five toes on the right foot, a split comb and one row of feathers on the shank; so it is not a Dark Brahma. These therefore, are not from the engrafted Dark Brahma eggs. They represent cases of imperfect dominance of the black down over cinnamon. The conclusion to be drawn from this experiment is that the engrafted eggs did not mature in the foster-mother.
s ^
1
white; white + specks (or smoked) 21
black, white below ISl
cinnamon above, white below 3J
3 rows, 12; 2 rows, 15; 1 row, 1; row, 1; unrecorded, 3
pea, 16;
split, 23; unrecorded, 3
4 toes, 18; 5 toes, 24
narrow, 12; Intermediate and wide, 12; unrecorded, 18
1
X
50 per cent white
or white +
smoky; 50 per
cent black back
intermediate,
3 to rows pea, 50 per cent;
split pea, 50 per
cent
4, 50 per cent
5, 50 per cent narrow, 50 per cent Interm. and wide,
50 per cent
CO
X
cinnamon; light
below
heavy pea
4 narrow, gr. 2
light cinnamon; gray below
heavy, 5 to 7 rows pea
4
narrow, gr. 2
K
a 1
white (+ smoky)
[black back]
absent
Y [V, 1]
5 [4] wide, grade 7
Pi
cinnamon; gray
below
heavy, 5 to 7 rows pea
4 narrow, grade 2
!
3 ,
a ^
1
X
s
1
"0
5
1
THE TRANSPLANTATION OF OVARIES IN CHICKENS 115
Experi7nent No. 2. No. 11379, Dark Brahma with engrafted ovary from no. 11605 (White Leghorn-Houdan) was mated in pen 1050 with 14122^, a single-comb Black Minorca. Table 2 gives the juvenile characters of 1, the male parent; 2, the Dark Brahma, so-called foster-mother; 3, the hen from which the ovaries were transplanted; 4, the expectation of offspring on the assumption that the graft succeeded; and 5, that the graft failed and the proper ovary was regenerated; also, 6, the observed characters in the offspring.
Without exception the characters of the offspring are clearly those of the Dark Brahma X Minorca cross and none of the White Leghorn or Houdan differential characters enter into their composition. The grafted ovary produced no eggs that developed, the extirpated ovary was regenerated.
Experiments 3 and 4, operations. The protocol of the grafting operations is as follows:
No. 11541 9 is a white-plumaged hen derived from a cross between 8681 9 ,aWhiteLeghorn-Minorca-Polishbird, and 7811 cf, a Houdan cross hatched (in pen 905) in February, 1909; fasted two days. On October 2, 1909, injected with 0.005 grain atropin in 1 cc. of water; etherized and opened. Ovary very large, two large pieces (60 per cent) of ovary removed. Strong hemorrhage. Two small pieces of ovary from no. 11383 9, Dark Brahma, sewed with peritoneum close to ovarial artery. Sewed up. Bird slow in recovery.
No. 11383 9, straight Dark Brahma, hatched February, 1909, from mating 907: 7549. On October 2, 1909, injected with atropin, etherized and opened, ovaries small, incompletely removed. Two large pieces of ovary of no. 11541 sewed into peritoneum. Sewed up. Bird recovered rapidly.
Results, experiment 3. No. 11541, White Leghorn-Black Minorca-PoHsh-Houdan hybrid, with engrafted ovary from No. 11383, Dark Brahma, was mated in pen 1027 with 11291 o^, a pure bred Dark Brahma. The results of this mating are given in table 3.
Experiment 4- No. 11383 9, pure-bred Dark Brahma with engrafted ovary from no. 11541 (White Leghorn-Black Minorca
z
o
I
1
1
5
1
i
2
2
1
DO
fl
o
i
1
i ^
m \ "*
1
X
1 i
1 1" i i
X
°? . £ S &a 1 i
If 1 ss ss 1
>
O
El
IM f li
SI
i i
is "^ -i 5
F 1 i i:
1 !
ill i:
a
I
S
II
c
o
i
i
S £
■SI"
S 8
s S
I i ■Is
„• t
S « "^
s fe g
g ft
a -^
S. S
a a ftS
I - « 1 1
as
><
<
I
«
i
i
£
1 1
1
^
1
- 1
i
2.
"foster-mother"
« i
. > 1 1 ,
1 1 1 1
i
5
ill " I .
1" 1 1 § ^
- 2
i i
g
' .
1^
5
1
1
118 C. B. DAVENPORT
Polish — Houdan hybrid) was mated with a Black Minorca 14122 d". The results of this mating are given in table 4.
Experiment 5. No. 11693 9 , used in this experiment, is a white bird that had ' smoke ' on down when hatched. It is of somewhat complex origin. Its mother was an Fi hybrid between a Black Spanish cock and a White Leghorn; its father had the same elements and also white Silkie in its ancestry. No. 11693 has, consequently, black recessive. It has a single comb, is free of the skin pigment of the Silkie, is clean-shanked and has four toes on the right foot and five on the left.
On September 19, 1909, this pullet (which was hatched March, 1910) was treated with atropin, etherized during half an hour and opened as usual between the last two ribs. All of the ovary, as far as could be seen, was removed. Pieces of ovary from no. 11280 ? (a straight-bred Dark Brahma bantam) were placed in contact with the peritoneum, near the removed ovary, but not stitched in, as the bird showed signs of succumbing. The cut was sewed up and the bird set aside where it lay quiet for half an hour.2 The Dark Brahma from which the ovary (whose eggs measured 0.5 mm. in diameter) was removed died in consequences of hemorrhage.
Later No. 11693 was mated with 11291 o^ (in mating 1027: 11693). He is a straight-bred dark Brahma bantam cock, used also in experiments 1 and 3. The results are shown in table 5.
Experiment 6. No. 11826 9, hatched March, 1909, a pure bred Dark Brahma was opened October 2, 1909, and ovary imperfectly removed. Ovary of no. 12550 (a White Leghorn-Minorca-PolishHoudan hybrid) sewed on to peritoneum at point of removal. The ovary had been kept out of body of hen about ten minutes, but covered and moist.
In the late winter of 1910 no. 11826 9 was mated in pen 1050 with 14122 cf, a single-comb Black Minorca. The results are given in table 6.
See postscript.
All black; white below
intermediate, to 3 rows
pea, modified
low, grade 1 or 2
X
white, 25 per cent;
white, black
specks, 25 per
cent; black and
white 50 per cent
absent
Y
Intermediate grade 5 to 2
I 5
>
o
i
white [black and
white]
absent
V
wide, grade 8
. s
cinnamon above;
gray below
heavy, 5 to 7 rows
pea
low, grade 1 or 2
1.
FATHER
black; white below
absent
single
narrow, grade 1
i
i
S
i
i
1
11
1
1
t
1
CONCLUSIONS
In the six experiments described above there is no evidence that the engrafted ovary ever became functional but all results are in accord with the conclusion that the more or less completely extirpated ovary regenerated and produced an abundance of eggs. With the results the data of Dr. Guthrie's paper are not in disaccord. His data, like ours, furnish no evidence for the survival of the engrafted ovaries, far less of an effect of the soma of the foster-mother on the introduced germ plasm.
Cold Spring Harbor, N. Y. September 26, 1910.
POSTSCRIPT
On January 4, 1911, No. 11693 ? was killed and opened on the left side. An ovary of fairly typical size for a hen entering her second year of laying was found. It contained numerous eggs, 4 to 5 mm. in diameter. Slightly ventrad of the main artery of the ovary is an irregular mass 5x4x2 mm. of cheesy consistency, imbedded in and covered by peritoneum. Its general appearance is that of a dried, hardened ovary, with clear traces of follicles. It doubtless represents the engrafted ovary, entirely encysted in the peritoneum.
January 30, 1911.
LITERATURE CITED
Davenport, C. B. 1906 Inheritance in poultry. Publication no. 52, Carnegie Institution of Washington.
1910 Inheritance of plumage color in poultry. Proc. Soc. Exper. Biol, and Med., vol. 7, p. 168.
Guthrie C. C. 1908 Further results of transplantation of ovaries in chickens. Jour. Exp. Zool., 5, pp. 563-576.
THE EFFECTS OF INBREEDING AND SELECTION
ON THE FERTILITY, VIGOR AND SEX RATIO
OF DROSOPHILA AMPELOPHILA
W. J. MOENKHAUS
Indiana University, Bloomington, Indiana
CONTENTS
Introductory 124
Material and methods 124
Inbreeding and selection on fertility and vigor 126
1 Introductory 126
2 Sterility 127
a Character of the sterility 127
b Degrees of sterility 128
3 Inbreeding and vigor 131
4 Sterility and selection 134
6 Discussion of results 138
Sex-ratio and selection 141
1 Introductory 141
2 The normal sex-ratio 141
3 Control of sex-ratio by selection 142
a History of strain 206 143
b History of strain 207 147
c Discussion 147
4 Influence of male and female in determining the sex-ratio 148
5 Discussion of results on sex-ratio 151
Summary 153
Literature cited 154
123
124 W. J. MOENKHAUS
INTRODUCTORY
The present report includes the results of two series of experiments on the fruit fly — -Drosophila ampelophila. One set concerns itself primarily with the effects of inbreeding and the other with sex-ratios. The experiments on inbreeding grew out of work I had been carrying on on hybridization. In these hybridization experiments the effects on the developmental processes of hybrids between species too remotely related were especially emphasized. The converse of these experiments was, naturally, to study the effect upon the young between individuals too closely related. Fishes, upon which all my experiments in hybridization were made, do not lend themselves for purposes of inbreeding without elaborate breeding facilities. Mice seemed suitable for this purpose but, both at the outset of these experiments and since, these creatures have proven miserable failures in my hands. Among the insects, I tried the common willow beetle but this proved to throw only one generation annually in this latitude. It was desirable to have an animal with a brief life history, whose food could be easily obtained at all seasons and in which the sexes could be readily distinguished. In these respects the fruit fly is almost ideal. The facts herein considered confine themselves to this species.
The experiments on sex-ratio suggested themselves in connection with the inbreeding experiments and so were carried out along with the latter and after they were completed.
MATERIAL AND METHODS
The strain which is mostly under discussion in my inbreeding experiments came from a well-filled female that was taken from the window of my residence in Bloomington. Other strains were started at the onset. Some of these came from the banana bunches at the various groceries and others came from fruit which I had laid out for this purpose. None of these were carried further than two or three generations excepting two, called 6 and 7 in my records. The latter was discontinued after the tenth generation
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 125
since it had been from the beginning apparently less prolific. The strain 6 was carried for over seventy-five generations and is the one on which the experiments in inbreeding of this report are based.
For vivaria, tall stender dishes, tumblers, quinine bottles and lamp chimneys were given a trial. They were discarded in favor of 8-dram shell vials. These were compact, so that a large number of matings could be kept in a small space, and they were most convenient in manipulating the pairs during the frequent changes to new cages that was necessary all along. The open end of the shell vial was closed with a plug of absorbent cotton, not too compact, so as to afford some ventilation. The flies are strongly positive to light, so that the vials could be laid with their bottom toward the light and the cotton plug removed with safety for the introduction of food etc. Small trays holding fifteen of these vials were used and in this way the experiments could be readily and compactly stored in the incubator, or they could be packed into a valise to be taken along wherever I went. The food was exclusively well-ripened bananas. To prevent the larvae ;rom pupating in the food, narrow strips of blotter or filter paper were introduced in which they seemed to be especially fond of pupating. It is, of course, apparent that the greatest care had to be taken to avoid contamination from flies without. The stock food had to be scrupulously watched and the instruments kept clean to avoid the introduction of eggs laid on them by extraneous females. The bananas, especially, as they come from the stores, are likely to be infected with eggs and larvae if the skin be in any way bruised or split.
The brothers and sisters were paired off, always within the first ten or twelve hours of their life as imagos. Up to this time mating has not occurred. In fact I have never found a pair that copulated during the first twenty-four hours or, if so, that produced fertile eggs.
126 W. J. MOENKHAUS
INBREEDING AND SELECTION ON PERTILITY AND VIGOR 1. Introductory
That continued inbreeding acts deleteriously on the fertility and vitahty of a race is a beUef so firmly and generally established that it is seldom questioned. This has its origin largely in the common experience of breeders whose observations, unfortunately, are too often unreliable. There are not wanting experiments such as those of Van Guaita ('98) and Bos ('94) and others, scientifically conducted, which bear out this conclusion.
On the other hand, it is refreshing to encounter in the literature such reports as that of Gentry ('05) who believes not only that inbreeding is not necessarily harmful, but also that it maybe beneficial to conserve and intensify the good points in his breed. Gentry's experiments were made on Berkshires. The most prolonged tests of close inbreeding that have been recorded were made by Castle ('06) on the same species with which the present paper deals. He inbred (brothers with sisters) for fifty-nine generations. He concludes that such close inbreeding does not necessarily result in a loss of productiveness and of vigor; at least that inbreeding cannot be regarded as a causal factor. Some of his results so nearly parallel those of the present writer that further reference to his results will be made in the body of the paper.
During the early part of October, 1903, a number of pairs were started breeding. These came from various sources in Bloomington. These different pairs were reared for the most part only a few generations, excepting pair No. 6 which was continued for about four and one-half years. During this time over seventy-five generations were produced. Toward the close of this period no exact count was kept of the generations so that only an approximate figure can be given. Five pairs of brothers and sisters were mated in each generation to insure against accidents that might terminate the strain if but one mating were made.
Along at the fifth and sixth generation it became more and more difficult to keep the strain alive with the five pairs of brothers and sisters that were mated each generation. The failure of an
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 127
occasional pair to produce young had hitherto been attributed to accidental conditions of food, etc., but this no longer seemed a satisfactory explanation of all the failures to produce young. This condition, was, therefore looked into more thoroughly. This was done by laying out instead of five pairs a much larger number from thf: offspring of a given productive pair. The greatest care was taken with the food, temperature etc. and it soon developed that a variable per cent of the pairs were sterile. These sterile pairs were to all appearances normal. It was clear now that, while inbreeding had not reduced the general vitality of the strain thus far, there had appeared a high degree of sterility.
2. Sterility
- a. Character of the sterility. Examination of all the matings
brought out the fact that in all cases eggs were present in large numbers. This seemed to suggest that the difficulty lay in the larvae either failing to emerge from the egg envelope or, succeeding in this, failing to carry themselves through the feeding stage or the transformation.
By a careful search of the food of the sterile pairs, after sufficient time for the larvae to mature had been allowed, it became evident that the difficulty lay at a time earlier than the pupal stage for none of the latter could ever be found. The food supplied these sterile pairs was the same as that of the fertile ones since it could not be foretold which pairs were going to prove infertile. Furthermore, the infertile pairs were usually kept for from twenty to thirty days, the best of food being supplied them from time to time. The same search showed that no larvae were present, at least so far as direct inspection of the food under a dissecting microscope could be depended upon.
It was always possible, of course, that the larvae failed to carry their development very far, and, thus, being small when they first emerge from the egg, might have been overlooked. It became necessary, consequently, to take the eggs as they were laid from time to time and keep them under observation to see whether the larvae ever emerged. This was done by placing a piece of banana
128 W. J. MOENKHAUS
in the vial with a sterile pair and from time to time removing the eggs one by one with the point of a needle and placing them on a piece of moist filter paper in a separate vial. Usually twenty were placed in each vial and some food added for the larvae, should they emerge. Inspection of the eggs after twenty-four, forty-eight and seventy-two hours would readily reveal the number of eggs that had produced larvae. I have laid out thus at a great expense of time literally thousands of eggs from many infertile pairs, in many cases all the eggs that a given pair produced during the first twenty-five days of its life, but I have never seen a single egg that had hatched. Eggs of fertile pairs thus laid out will readily hatch so that all the larvae will have taken to the food twenty-four hours after the eggs are deposited.
Such infertile pairs copulate frequently and it would seem that impregnation should follow. I have never sectioned the eggs to see whether spermatozoa enter the eggs or whether they contain partially developed larvae which fail to hatch. I have, however, been able to determine in this strain which of the sexes is at fault. This was done in the following manner. After a pair by sufficient trial had proven itself infertile, the male was mated to a virgin female of a fresh strain that had not been inbred and possessed a high degree of fertility, and the female was similarly mated with a male, usually one whose fertility had been established. Sixty-four such cases were tried and in no case did the females fail to produce young and in no case did the males produce any although repeated copulations took place. It is evident from the foregoing, that, in this strain, the sterility lies exclusively in the male and that the female has lost, apparently, nothing in fertility. Castle (p. 735) reports, on the other hand, that either sex may be sterile. However, Castle took no account of the eggs and larvae but merely the production of pupae, so that his sterility cannot be with certainty compared to mine. It would seem, however, that in some strains infertility may be strictly confined to the males and in others to both sexes. That sterility is complete for all males, when it occurs, is shown by both our results.
h. Degrees of sterility. The foregoing experiments concerned themselves with such pairs as were completely sterile. Other pairs
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 129
of brothers and sisters from the same parents, however, were fertile. Judging from the productiveness of these, there was often a wide divergence. It would seem that, as a result of inbreeding, we had a condition of fertility ranging from absolute infertility to comparatively high fertility among the different pairs of brothers and sisters from any given pair of parents. To test this the following experiment was carried out: About two-hundred eggs from each of fifteen pairs of flies were laid out after the fashion indicated above. Ten of these pairs had been inbred for seventeen generations while five belonged to fresh stock that had not been inbred. Of the ten pairs of the inbred strain, five belonged to a strain which had arrived at a very low degree of fertility, namely only 36 per cent of the forty- two pairs tested were fertile (table 3, seventeenth generation, strain, A) . These five pairs were brothers and sisters to many of the sterile pairs considered in the preceding section.
The other five pairs (of the ten inbred) were from a strain which had been held by selection to a high degree of fertility, namely 97 per cent of the thirty-four pairs tested were fertile. Both of these strains were descended from common great grandparents (table 3, seventeenth generation, strain B).
We have, thus, for comparison three conditions, namely, (1) eggs from a highly infertile inbred strain; (2) eggs from a highly fertile inbred strain; and (3) eggs from a presumably norma strain that had not been inbred. It should be added that the five pairs were taken at random and were not selected. Approximately the first two-hundred eggs of each pair were laid out in batches of about twenty to twenty-five to the vial. The number of eggs that hatched was noted in each case and also the number that emerged as imagos. Table 1 gives the summary of results.
From this table it appears that from the eggs which were taken from the inbred pairs with low fertility practically as large a per cent (97.27) hatched as from the eggs that came from the inbred pairs that showed a high fertility (98.2). The same is true in regard to the number that produced imagoes, 86.8 per cent and 85.1 per cent respectively. The fact clearly brought out here is that when infertility arises in this strain it arises suddenly and
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 1.
130
W. J. MOENKHAUS
does not present all intergradations. In other words, one does not find that among a large number of brothers and sisters some pairs whose eggs only partially hatch and other pairs that range in this respect, on the one hand, to perfect fertility and, on the other, to complete sterility. The fertility is either completely lost or it is of a high degree. Furthermore, when we compare the inbreds with the normals (not inbred) in regard to the percentage of eggs hatched no essential difference is observable. It would seem, therefore,
TABLE 1
Inbred {low fertility)
■ PAIRS
NUMBER OF NUMBER OP ^^,^,^^n^^ PER CENT OF
EGGS PLACED EGGS HATCHED 1 EMERGED ^°^^ HATCHED
PER CENT OF
IMAGOS
EMERGED
A
193 184 160
95.3
94.0
98.0
100.0
100.0
82.9
B
200 188
201 197 198 198 123 123
169
182
180
104
84.5
c
90.5
D
90.9
E
84.5
Total
915
890
795
97.27
86.8
Inbred (high
fertility)
A
201
173
204
197 175
198
172
200
193
169
182
156
161
165
145
98.5
99.4
98.0
97.9
96.5
90.5
B
90.1
c
78.9
D
83.7
E
82.8
Total
950
932
809
98.2
85.1
Normals (not inbred)
A
215
70
153
224
158
146
223
211
70
152
218
155
127
222
193
48
132
144
144
109
205
98.1
100.0
99.9
97.3
98.1
87.7
99.9
89.7
B
68.5
C
D
86.2
64.2
e". .:
91.1
F
74.6
G
91.9
Total
1189
1155
975
97.2
82.0
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 131
that the pairs that had not completely lost then- fertility, in so far as hatching their eggs is concerned, had suffered no deterioration whatever as a result of seventeen generations of closest inbreeding.
A fact of further importance brought out by table 1 is that of the percentage of eggs that successfully produced imagos. This does not differ essentially in the two groups of inbreds nor do these differ essentially from the normals. Castle used as his measure 'productiveness,' meaning thereby the number of pupae that were successfully produced. Making allowance for some pupae which do not emerge, the imagos produced in my experiments were an approximation to his 'productiveness.' Inbreeding, consequently, does not affect adversely the productiveness of pairs that show any fertility at all.
Castle found that his strains showed an annual fluctuation in productiveness, the period of least productiveness falling in the late autumn and early winter. My own experiments extended over about four and one half years and, although I have been on the lookout for this, I have never observed it. As Castle himself suggests, this fluctuation was probably a function of the temperature of the room. My flies were kept in a room which varied from 60 to 80 degrees and, when this was not possible, they were placed in an incubator kept at about the same range of temperature. It may also be that the productiveness of his strain ran low at this time of the year because they were placed in new hands at the opening of the college year. My observation has been that it takes some time for a new man to learn all the conditions that make for a favorable rearing of these creatures so that Castle's low productive periods may be merely a measure of the training period of the experimentor.
3. Inbreeding and vigor
At the outset of the experiments it was the expectation of the writer that such rigorous inbreeding would early and violently show itself in the vigor and fertility of the animals. In this, however, he was largely disappointed. In the strain that is here under consideration no untoward results could be detected during the
132 W. J. MOENKHAUS
first five or six generations. As previously stated, up to this time the method consisted in placing pairs of brothers and sisters in each of five vials to insure against mishaps. These mishaps consisted of drying up of the food, attacks of fungus and in some cases the escape of the flies themselves during the process of feeding etc. Those pairs that produced young were regarded as having escaped these various possible mishaps and were taken as indications of the vitality and productiveness of the strain. The expectation at that time was that any deleterious effect of the inbreeding would show itself in the offspring of any of the pairs. Consequently, when a given pair would produce offspring that was numerous, all well formed^ vigorous, and in no apparent way differing from normal offspring, to see whether some slight influence might not be present that could not be detected by ordinarj^ observation a definite measure was taken of (1) their rate of reaction to light and gravity, (2) the total number of eggs produced and (3) the percentage of eggs which hatched and emerged. An attempt was made to determine their length of life but this proved too prolonged to allow one to carry it out together with all the other incidents of the already too laborious experiments.
The reaction of this animal toward light and against gravity is well known. To get a measure of the rate of reaction the animals were made to travel through a glass tube that had been blackened for 16 cm. on the inside. This tube had a light placed at one end and was inclined about twenty-five degrees. From a glass vial the flies were admitted, one at a time, into the tube and the time from the moment of entrance into the blackened portion of the tube to their emergence was recorded. It was found essential that the two batches of flies (inbreds and normals) should be of the same age, be reared under the same conditions and that the temperature of the room be the same for the two batches. The results are as foflows: at a temperatureof 27.2° C. 133 normals took 16 seconds, average, to travel the distance, and 140 inbreds took 15.4 seconds. The two sexes in these two groups were about equal in number. In both groups the males travel the distance on an average in three seconds less time. It is clear from this that the normals and inbreeds are equally responsive to these two
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA
133
agents and that the latter have not suffered in this regard as a result of inbreeding.
In order to determine the total number of eggs produced it was necessary to isolate the pairs and twice each day pick off all the eggs that had been deposited in and around the food provided. This proved to be a most laborious process, for the eggs are too small to be followed safely with the naked eye and had to be removed individually with the point of a needle. Too much value must not be attached to this measure for the reason that the rate and, therefore, probably the number of eggs deposited seems to depend somewhat, at least, on the condition of the food present, and for the
TABLE 2
Strain 6
Number of generations inbred 2
Number of days eggs were counted 27
Total number of eggs laid 433
3
30
617
5
34
480
6
34
724
8
23
455
9
32
516
Strain 7
Number of generations inbred ....
2
26 654
3
33 662
5
29
539
6
23
498
9
33 907
10
Number of days eggs were counted
Total number of eggs laid
28
429
reason that only the eggs deposited during the first twenty-five or thirty days were counted. These creatures live to he very much older. We have kept females alive 153 days, but after the first twenty-five or thirty days the eggs come only in small numbers. Table 2 gives the actual counts of several females of both strains 6 and 7.
We see from the above counts that no material reduction has occurred in egg production during nine and ten generations of inbreeding. Such variations as occur may, of course, represent individual differences in the females.
The data given in table 1 of the relative hatching and emerging qualities of the young of normals and of pairs inbred for seventeen generations shows that there is no difference in this respect.
134 W. J. MOENKHAUS
In so far as the above determination may be taken as a measure of the vitality of this species we are justified in concluding that from six to seventeen generations of inbreeding no appreciable deterioration has resulted. No such exact determinations were made in later generations, and it is possible that eventually the effects of inbreeding would manifest themselves, but my observations during seventy-five or more generations does not lead me to believe this.
Jf.. Sterility and selection
Along at the thirteenth and fourteenth generations the sterility had become very pronounced. Of the offspring of some of the pairs, more than 50 per cent of the males were sterile. On the other hand, while practically all pairs showed at least some degree of sterility this varied very much in the different brothers and sisters of the same brood. That this sterility was a direct physiological result of the inbreeding seemed to me very doubtful. To find the effects of inbreeding showing itself in such a specific way upon the males only, did not, to say the least, meet expectations. Furthermore, sterility was not wholly wanting in forms that had not been inbred.
It was highly desirable to continue the experiments on inbreeding, and yet to keep the strain alive, it was necessary to find some way to eliminate this high degree of sterility. The process that was most effective was selection. By continuing the strain of those pairs whose offspring showed the highest degree of fertility but at the same time continuing the rigorous inbreeding, it was possible almost completely to eliminate the sterility. This at the same time gave one of the severest tests as to whether inbreeding was the responsible factor, for if the sterility could be eliminated by continuing the very process of inbreeding the latter could not well be held to be the cause of it.
This was done as follows: In the fourteenth generation three fertile pairs of brothers and sisters from the same brood were isolated and mated. The offspring of each of these were mated in pairs to determine the degree of sterility. By reference to table
3, it will be seen that the pair marked A produced offspring out of
which nine of twelve pairs tested were infertile; pair' 5 produced
offspring of which four pairs out of fourteen tested were infertile
and pair C threw offspring with five pairs out of fifteen infertile.
We have here, then, three pairs showing a wide variation in the
degree of fertihty of their offspring. Pair A showed 75 per cent of
the pairs infertile and pairs B and C approximately the reverse
ratio. In the further progress of the experiment pair C was dis
TABLE 4
Strain A
NUMBER
PAIRS TESTED
NUMBER
PAIRS FERTILE
NUMBER PAIRS
INFERTILE
PER CENT PAIRSPE
FERTILE
^ CENT PAIRS
INFERTILE
18 (1)
52
51
52
66
28
27
37
37
46
19
25
14
15
11
9
51
72
71
80
69
49
28
29
20
31
18 (2)
18 (3)
18 (4) . . .
18 (5) . .
Average for 238 pairs 69 per cent.
Strain B
18(1)... 18(2)... 18(3).^.
18(4)... 18(5)...
100 100 100 68 100
Average for 93 pairs 92.5 per cent.
continued so that only pairs A and B were used. I shall in the further description of the experiment refer to the descendants of A as strain A and of B as strain B.
Before entering upon the experiment of selection it was necessary to ascertain whether, without selection, the descendants of pairs A and B continued to show a low and high fertility respectively. Accordingly, a single one of the fertile pairs of the 15th inbred generation of strain A and B was tested. Reference to the table shows that in strain A 27 pairs or 57 per cent of the forty seven pairs tested were infertile, while in strain 5 none of the thirty
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 137
seven pairs tested were infertile. The same process was repeated with a pair of the sixteenth generation of the two strains. Strain A showed twenty-seven or sixty-four per cent of the forty-two pairs tested infertile and strain B one or three per cent of the thirty-six pairs tested.
Up to this point in the experiment only a single pair in each generation was tested as to the fertility of its offspring. It might well be that by chance in each case a pair of low fertility was taken in strain A and a pair of high fertility in strain B. To eliminate this possible error five pairs w^ere taken in each strain and the fertility of their offspring determined. It was further desirable to obtain an estimate of the variability in the fertility of the pairs in the two strains as well as to get a more correct estimate of the average fertility of both. In the diagram these five pairs are designated as 18 (1), 18 (2), etc. Table 4 shows the number of pairs of offspring tested for each pair and the number and percentage of pairs fertile and infertile.
The fertility thus varied in strain A from 51 per cent in 18 (1) to 80 per cent in 18 (4), with an average fertility of 69 per cent. In strain B the fertility was much less variable in the different pairs, the only exceptions being 18 (4), the average fertility being 92.5 per cent.
We now have definitely established two strains, one of low and another of high fertility. The important part to be emphasized here is that this was produced by the process of selection from among the variable offspring of generation fourteen of the inbred strain. To make the experiment more complete it was now necessary to obtain a highly fertile strain out of the one with low fertility. Accordingly strain B was discontinued at this point and attention restricted to strain A. Five pairs, 19 (1), 19 (2), 19 (3), etc., were taken from among the offspring of 18 (4) because this showed the highest percentage of fertility. These were tested in the same way as in the preceding generation. Table 5 gives the details.
By selection it will be seen that the average fertility was raised from 69 per cent in the 18th generation to 75 per cent in the 19th generation. Among the five pairs used one 19 (2) showed an unusually high fertility (96 per cent). This pair was accordingly
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 1
138
W. J. MOENKHAUS TABLE 5
19(1). 19(2) 19(3) 19(4) 19(5)
53
Average fertility of 239 pairs 75 per cent.
taken to select from. Five pairs were again taken as before. The results appear in table 6.
Thus it will be seen that all five pairs showed a uniformly high degree of fertility. The average fertility of all the pairs was raised to 93. 8 per cent.
5. Discussion
From the above series of experiments a number of important facts are birought out. 1. Sterility, as it appeared in the strain under consideration, is strongly transmissible through inheritance. 2. It is readily controlled by selection. 3. Inbreeding is probably not the physiological cause of it.
That this sterility is transmissible cannot be doubted. The faithfulness with which this occurs appears in the strains A and B. Both were derived from a common pair that showed a variability with respect to this character in the three pairs of its offspring
20(1). 20 (2) . 20 (3) . 20(4). 20 (5) .
Average for 211 pairs 93.8 per cent.
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 139
tested. One of these possessed a high degree of sterility, while the two other pairs showed a low degree. The descendants of the latter constituting strain 5, retained this low degree of infertility throughout. Similarly the descendants of the former, constituting strain A, retained their high degree of infertility up to the time when selection away from this condition was introduced. In the latter process the transmissibility of the character is again emphatically revealed. In the eighteenth generation, pair 4 showed a lower degree of sterility than any of the remaining four pairs of brothers and sisters. Breeding from this pair at once showed offspring with a decided decrease in sterility, compared with the eighteenth generation, the average of the nineteenth generation being 75 per cent of the pairs fertile as compared to 69 per cent of the latter. Again, in the nineteenth generation, pair 19 (2) showed a much lower degree of infertility than the other pairs. Continuing the strain from this pair, this character is faithfully reproduced in the offspring in that they average fertility of the latter is raised to 93.8 per cent.
It is important to note in this connection that Castle, in his experiments upon Drosophila, found that productiveness (which as previously noted is quite a different thing from the sterility here considered) was similarly transmissible and amenable to selection. Furthermore, Castle's experiments would seem to indicate that this character of productiveness behaves, in inheritance, after the Mendelian fashion, low productiveness acting as the recessive character. We have evidently to do here, both in the productiveness in Castle's experiments and in the sterility in my own, with characters that are germinal for they behave as such. In the strain upon which my experiments Vv^ere made we have the further remarkable condition that the infertility is inherited only by the males.
It is clear that whatever the causal factor or factors to which the sterility may be attributed, it is relatively insignificant compared to the effect of selection upon it. Furthermore, the modification is a germinal one. That inbreeding may be responsible for its prevalance in the strain seems probable, but that it is responsible
140 W. J. MOENKHAUS
for its origin is not believed. We have seen that the general vitality of the strain, as measured by its productiveness and its reaction to light and gravity, did not suffer as a result of seventeen generations of closest inbreeding. Failing in this, it is not probable that its effect would show itelf in so specific a way as the sudden and complete sterility in certain males of the strain. The improbability is further supported by the fact that the inbreeding may be continued unabated if only care be exercised in the selection of the brothers and sisters to be mated, thereby even eliminating practically what sterility may have existed.
It is much more probable that the sterility arose spontaneously in this strain or that it is present to a varying degree in this species. With the character present and highly transmissible and subject to selection it is only necessary to carry on indiscriminate breeding to have the character appear in varying intensities depending upon the chance combinations. The rule of inbreeding would be only to intensify the chance combination of the character and to insure the more or less continued presence in the successive generations.
That this character of sterility is not unique to this inbred strain is evident from its rather frequent presence in pairs not inbred. In my own experience this sterility nearly always showed itself in the males. In one instance I found among a brood, besides a sterile male, two females that failed to deposit eggs although eggs were evidently present in the oviducts. Similarly Castle found in his strain a considerable amount of sterility, and this in some cases among the females. We see, therefore, that sterility is not altogether rare even in broods that were not inbred.
The same facts doubtless hold for the character of productiveness. Castle has shown this to be transmissible and amenable to selection. Inbreeding does not produce it but is instrumental, with indiscriminate mating, in intensifying it, or, if the strain be not eliminated thereby, of preserving it in the strain.
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 141 SEX-RATIO AND SELECTION 1. Introductory
The once rather generally accepted notion that nutrition was an influential factor in the control of sex, based on the experiments of Yung ('85), Born ('81), and others, has given place to the now as commonly accepted idea that sex is determined prior to or at the time of fertilization and is independent of the food. The experimental work of Cuenot ('99) King ('07) and others, and the splendid cytological researches of Wilson and his students are largely responsible for this change of view and have been so frequently reviewed in the various recent discussions of the problem of sex that they need not be further detailed here.
The writer tried some starvation experiments on Drosophila in 1904. During the past year more extensive experiments were carried on under his direction by Mr. Claude D. Holmes, on the effects of starvation during successive generations upon the sexratio. These are published under a separate title ('10). It will suffice in this connection, to state that the results coincide with those of recent workers, namely that nutrition does not affect the sex-ratio.
2. The normal sex-ratio
One fact was very apparent in these earlier tests and in all subsequent experiments, that, under the varying conditions in these creatures were reared, there was the same persistance of the predominance of females over males. Below (table 7) is given the
FOOD
TOTAL.
NUMBER
BEARED
NUMBER OP
MALES
NUMBER OF
FEMALES
RATIO
10506
2161
4048
10218
4972
995
1943
4757
5534
1166
2105
5461
1:1 113
Grapes
1:1 171
Tomatoes and grapes
Bananas . . . .
1:1.083
1*1 14
Total
26933
12667
14266
1-1 126
142 W. J. MOENKHAUS
summary of four determinations on a large scale to obtain the normal sex-ratio. The flies were reared in the following manner. Mason jars containing a large quantity of food were exposed to flies in nature. The jars were left open until the larvae began to pupate when all flies were excluded by tying a guaze over the top. As the imagos emerged from time to time they were preserved and the sex-ratios determined. For 26933 individuals, the ratio was one male to 1.126 females.
In regard to these determinations only one question, so far as I can see, can be raised. This is the academic one of the greater mortality of the males during development or, to push the matter back a little further and to make it applicable to recent developments in our idea of sex, the greater mortality of the male determining sex cells. In reference to this it may be pointed out that the developmental conditions were as nearly normal as one can imagine. There was an abundance of food, air, light and moisture, and the larvae pupated in the remnants of the food in much the same manner as one finds them doing in nature. In this connection the experiments of Miss King ('07) on the influence of food on the sex ratio of Bufo are of importance. In this she finds that the mortality among the males is not greater than among the females. From these facts and from the knowledge that has come to me from the extensive rearing of Drosophilas for six years I am convinced that the sex-ratio in this species is not one of equality.
3. Control of sex-ratio by selection
If the sex-ratio of this species, then, is that of 1 male to every 1.126 females, this should be regarded as specific just as any other of the specific characters of the species. It should, therefore, be subject to fluctuations and to control like other specific characters.
Starting with this conception of sex-ratio, I wished to see whether it were possible to control this, within limits, of course, by the process of selection. The results of these experiments I propose to detail below.
To apply the selective process on the sex-ratio, the following simple method was employed. Two pairs were selected from
INBREEDING AND SELECTION IN DROSOPHII,A AMPELOPHILA 143
nature, the one showing a high, the other a low female ratio. These were selected as the parents of the two strains to be developed. From among the offspring of each of these two pairs a number of single matings were made. From among these the pair that showed the most favorable ratio in the desired direction was selected to continue the strain. The same process was repeated as often as desired.
From a number of pairs taken from a banana bunch in Bloomington June 12, 1907, two such pairs were obtained. These two pairs go by the numbers 206 and 207, showing the following ratio :
206— 52 cf: 135 9 or 1:2.59 207—84 d^ : 75 9 or 1 : 0.89
A. Strain 206 {high female ratio). The 206 strain will, for convenience, be called the female strain and the 207 strain the male strain, although, as will appear, the latter never developed into a predominantly male strain. In tables 8 and 9 are given in diagramatic form the results of selection for five generations in the former and six generations in the latter. At the margin the generations are numbered 1, 2, 3 etc., and the sex-ratios are indicated.
The sex-ratio of the eleven pairs of brothers and sisters mated from the first generation of the female strain (206) varied from 1 :93 (76 d'; 71 9) to l:-7.00 (8cf:56 9).
The unusually high female ratio in the latter is probably attributable to the small number of individuals obtained from this pair. Two of the pairs threw a predominance of males (table 8 nos. 4 and 8). With the exception of no. 5, all the remaining pairs threw a high female ratio. The ratio for all the pairs was 1:1.67 (578 &: 969 9). We have here a female ratio very much higher than that characteristic of the species (1:1.14) and yet considerably below that of the parent pair (1 :2.59). This may be regarded as a regression toward the normal ratio. It should be pointed out here that too much emphasis should not be placed upon the exact figures representing the ratios in the different pairs, since the number of individuals at best are rather small. In most cases, however, when the number of offspring obtained is fairly large, the ratio approximates the true one, so that in any given
144
W. J. MOENKHAUS
S=
S^
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 145
tS
1-1 02
146 W. J. MOENKHAUS
pair from which a fairly large number of offspring has been obtained shows a high female ratio for instance, this may be taken as a pretty safe indication that the female ratio would be high if all or a much larger number had been obtained.
For the next generation ten pairs were taken from brood 9 with a ratio of 1.1.94. Brood 3, with a ratio of 1 :2.31, would have been a more favorable one to select from, but this is not always possible since the matings must be made before all the offspring have emerged and therefore all the data for the complete ratio is obtained. Onh^ four pairs of this series of matings came through safely, due purely to the lack of time to give them the attention they should have had. The four pairs threw the following sexratios: 1: 2.33; 1:2,29; 1:1.27; 1:1.81. The ratio for the entire brood was 1 :1.82 (215 d" : 391 9 ). This ratio was somew^hat more predominantly female.
Pairs were now selected from the brood 8 with a ratio of 1 :2.29. Of the seven pairs mated the offspring of only four was obtained and the number of young in each case was quite small. The ratio for all the offspring of the generation was 1:2.17 (93 cf to 201 ?). The total number here involved is so small that not too much importance should be attached to the increased female ratio.
For the matings of the next generation there is little doubt that an unfortunate selection was made. The brood from which the matings were taken showed a ratio of 1 :2.46 but this ratio was based on numbers so small (52) that it probably did not represent the true ratio of the pair. This may account for the drop in the ratio for all the broods of the 4th generation to 1:1.36 (354 d' 483 9).
Two sets of matings were now made from as many broods of the fourth generation. One of these series was again taken from the brood showing the most favorable female ratio 1:1.90 (85 cf 162 9), but the other series was taken from a brood showing a relatively low female ratio, 1:1.04 (64 & 67 9). From the former the ratio of five pairs was obtained showing a ratio of 1 :1.39 (372 cT — 518 9) and from the latter the ratio of 7 pairs, showing a ratio of 1 :1.07 (496 c^ : 535 9 ).
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 147
b. Strain 207 {low female ratio). From pair 207 with a ratio of 1 :0.89 (84 o" : 75 9 ) it was hoped to develop by selection a strain showing a low female ratio. Seven matings from the first generation produced 536 d" and 579 $, or a ratio of 1:1.08. The range of ratios of the individual pairs was from 1 :1.22 (99 d" : 121 9) to 1:0.86 (79 cT: 68 9). This selection was continued for four generations, the matings being made from broods with a low female ratio. The ratios of all the offspring in the successive generations were 1:1.06 (220 cf : 223 9) 1:1.10 (581 c^ : 640 9); 1: 1.04 (142 o^ : 147 9); 1:1.17 (518 d' : 607 9) for the second, third, fourth and fifth generations respectively (See Table 9) . This low female ratio showed itself rather uniformly in all the individual matings, a notable exception occurring in the fifth genertion (see Table 9, pair 3.) with a ratio of 1:2.53 (45 c^r 144 9). On the other hand no pairs threw a great preponderance of males, the most notable among those from which a large number of progeny was obtained being pair 2 in the third generation in which the ratio was 1:0.87 (115 o-: 101 9). For the sixth generation two sets of matings were made as in the fifth generation of the strain 206. One of these was made from a brood with a ratio of 1:2.53 (45 d" : 144 9) and the other from a brood with a relatively low female ratio, 1:1.36 (72 c?: 98 9). From the former the total progeny of eight matings gave a ratio of 1:1.42 (461 d : 654 9 ) and from the latter the ratio of eleven matings was 1.1.05 (944 d^:997 9).
c. Discussion. It seems from the above experiment that the sexratio in this creature is a strongly transmissible character. Starting with a pair that throws an offspring showing either high or a low female ratio it was possible to maintain, by selection, a strain maintaining the respective ratios. The offspring from a given pair, when mated in pairs, show^ a considerable variation in the sex-ratio of their children. It is thus possible to develop a strain with a low female ratio from one with a high female ratio, or the reverse, as is shown in the fifth and sixth generation of experiment 206 and 207 respectively (tables 8 and 9). The sex-ratio is clearly amenable to selection like any other character.
148 W. J. MOENKHAUS
It is an interesting fact that it is possible to develop a strain with a high female ratio much more easily and pronouncedly than a male strain. I have repeatedly tried to hold the sex-ratio to or below that of unity but without success. Not infrequently pairs will throw a predominance of males but it has not been possible to hold them there. The best I have ever been able to do is to hold it considerably below that of the normal, but never as low as unity. On the other hand, it is relatively easy to select in the direction of females even to the extent of 1 to 2.
It should be observed that in the breeding of these strains the most rigorous inbreeding was practiced. It might, therefore, be that the difficulty of selecting for a low female ratio results from the possibility that inbreeding tends toward the elimination of the males. My extensive experience in inbreeding these creatures, however, does not bear out this explanation. Furthermore, in the sixth generation of the high female strain it, was possible in two generations to reduce this ratio to near unity notwithstanding that the same rigorous inbreeding was continued.
4. Relative influence of male and female in determining the sex-ratio
Having thus produced two strains showing a decided difference in the sex-ratio of their offspring I wished to determine two further points. First, whether the maternal or the paternal elements had an equal share in the control of this ratio, and second, whether this ratio was determined in the process of fertilization. To this end reciprocal crosses were made between the two strains and the proportion of the sexes in the offspring ascertained. Three experiments were performed in the following manner. From among a brood of each of the two strains a large number of individuals were taken. Before sexual maturity a number of males and females were isolated, while the remainder were allowed to reproduce. The latter gave a control for each of the strains. The isolated virgin females of one strain were mated with the males of the other. Each experiment thus consisted of four multiple matings. (1) A number of brothers and sisters belonging to the male strain. This furnished a control for the male strain. (2) A number of brothers and sisters belonging to the female strain. This furnished a control
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 149
for the female strain. (3) Females from the male strain mated with males from the male strain, and (4) the reciprocal of '(3)'.
In crossing two strains as in the above experiment three possibilities might obtain. First, that the two sexes have an equal influence in determining the sex-ratio; second, that either sex have a predominant influence and third, that a ratio result unlike that obtaining in either of the parental strains. While the first is probably the expected result, the experiments show in a most decided way that the male has little or no influence in determining the sex-ratio in this species (tables 10, 11 and 12). In most of the cases the ratio of the offspring falls pretty closely around that of the strain from which the females were taken. In two instances the ratios exceeded 100 per cent influence. The remaining ones, with the exception of strain 244 in which the male influence amounted to 35 per cent show the female influence almost near enough to 100 per cent to justify one in regarding the differences merely as fluctuations incident to the small number of individuals involved. The unusually great influence of the male in strain 244 might be accounted for in two ways. First the number of individuals involved in this experiment are relatively small so that the ratios of both the control and the crossed broods are not as reliable as in the other experiments. Secondly, the flies used for this experiment were taken from the earlier generations of the two strains, before, we may believe, any considerable selection had been appUed to fix the character of the respective strains. Indeed, this seems to be borne out in the other experiments.
The materials of the three experiments were not all taken from the same generation but were taken from different generations in the development of the strain. Thus, in experiment 1 the broods were taken from the first generation of strain 206 and 207. In experiment 2 the broods came from the second generation of strain 206 and the third of 207. The third experiment was made from the fourth and fifth generations of strains 206 and 207 respectively. Arranging these experiments in a series, based on the length of time that selection had been practiced on the broods used, we see that the male influence decreases as the selective time increases.
150
W. J. MOENKHAUS
TABLE 10
Experiment 1
No. of strain mated |
No. 242
2122 X 2122
No. 245
2122 9 X2149d'
No. 243
2149 X 2149
No. 244
214g9 X212CC?
^
9
&
• 9
&
'
&
9
Number of individuals
Sex-ratio (actual)
208
1.00
194
0.98
463
1.00
1.00
475
1.03
1.288
171
1.00
273
1.60
225
1.00
1.00
311
1.38
Theoretical ratio
1.288
Influence of male parents.... Influence of female parents..
7 . 3 per cent 92.7 per cent
35 per cent 65 per cent
TABLE 11 Experiment 2
No. of the strains mated. . |
No. 271
252io X 252io
No. 274
252,0 9 X2558cf
No. 272
255s X 2558
No. 273
2558 9 X252iocf
d'
9
&
9
cf
9
&
9
Number of individuals
332
1.00
545
1.69
589
1.00
1.00
919
1.56
1.365
739
1.00
818
1.106
680
1.00
698
1.026
Theoretical sex-ratio
1.00
1.365
Influence of male parents... .
Influence of female parents. .
22 per
78 per
cent
cent
100
per cent (—13)
per cent (1 . 13)
TABLE 12
Experiment 3
No. of strain mated [
No. 279
2758 X 2758
No. 281
2758 9 X2787cf
No. 280 No. 282
278r X 2787 i 278; 9 X 2758c:f
cf
9
cf
9 &
9 ' C
9
Number of individuals
Sex-ratio (actual)
289
1.00
42-,
1.477
382
1.00
1.00
551 1022
1.50 1.00
1.249:
1
1044 752
1.021 1.00
1.00
825
1.083
Theoretical ratio
1.249
Influence of male parents....
Influence of female parents..
per cent 13 per cent
100 per cent 87 per cent
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 151 TABLE 13
PER CENT OF FEMALE INFLUENCE
PER CENT
OF MALE
INFLUENCE
Experiment 1-from broods selected for one/ 212 214' 92.7 i 7.3
generation \ 214 212; 65 [ 35
Experiment 2-from broods selected for 2 and 3/ 252 255' 78 22
generations \ 255 252' 100 I
Experiment 3-from broods selected for 4 and 5/ 275 278 100
generations \ 278 275j 87 13
This fact of the prevaihng or exclusive influence of the female in determining the sex-ratio occurs in some other species of animals. Phylloxerans (Morgan '09) and Dinophilus apatris (Korschelt '82) . On the other hand, Whitney ('09) seems to have shown that in rotifers certain eggs which will produce males if unfertilized are changed to females, if impregnated. In the case of Drosophila, we can not be certain that the sex-ratio is established before fertilization since the experiments do not with certainty entirely exclude the male influence.
5. Discussion of sex-ratio
It is not the intention to enter into an elaborate discussion of the problem of sex control. The literature is certainly already sufficiently burdened with such. The writer wishes merely to point out briefly a few conclusions about sex in this species which his results seem to warrant.
The property of sexuality possessed by this species expresses itself not in the equal production, numerically, of its two states, male and female, but in an unequal production. Studies in normal sex-ratios involving a sufficiently large number of individuals are not numerous. The unequal production of the two sexes in the human species is well established. Montgomery ('08) has given the data of a large number of individuals of Theridium and finds a marked inequality in the sexes. The general assumption seems to be that an equal sex-ratio is the rule. It is not improbable,
152 W. J. MOENKHAUS
however, that, as careful determinations upon different species multiply, the condition of unequal ratios will be found increasingly common. Any theory of sex must take into consideration this normal inequality in the sex-ratios.
Sex-ratio like color, size etc., is a character belonging to a species. Sexuality of course is not, for it is common to all species reproducing by the sexual method. The particular form of sexuality, however, the proportion of the two sexual persons to which it gives expression in the process of differentiation, this is specific. For Drosophila ampelophila, the ratio of one male to 1.126 females is a specific character. This is not a ratio of merely the present generations but has been transmitted from generations remote. It is inherited. It is the expression of the physiological condition to which the species has been developed by its environmental demands.
Like other specific characters this ratio should be subject to modification, but this should not be more easily done or by other methods, in general, than those used in the modification of other characters. From this view point it should not be expected that the sex-ratio in an animal could be materially changed by such agents as food, temperature,etc. A change in the proportion of the sexes involves a much more fundamental modification than simple starvation or the reverse is likely to induce. In regard to other characters, we have long ago ceased to regard them as modifiable by such methods, but in the case of sex, it is only recently that their futility is being entertained. The most potent factor and the one most generally used to modify a character is selection. If the experiments herein recorded prove what they are held to prove, this process of selection is a potent factor in the modification of the sex-ratio also. It would be interesting to try to line this fact up with the chromosomal conception of sex. However, the writer regards this as the task of those who are engaged in these interesting and important investigations.
INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 153 SUMMARY
1. Drosophila ampelophila may be inbred (brothers and sisters) for seventy-five or more generations.
2. Inbreeding in itself is not deleterious to the fertility or vigor of this species.
3. ' Infertility normally occurs to a varying degree among the offspring of any pair. Promiscuous inbreeding among such offspring may perpetuate and even intensify this character. When sterility appeared in the strain experimented with, it was always complete, appeared suddenly and was confined to the male.
4. By the judicious selection of the brothers and sisters to be mated from a brood that shows a high degree of infertility, this infertility can be eliminated by selection although continuing the inbreeding in the closest possible way.
5. There is a wide divergence in the fertility and productiveness among the different pairs taken in nature, but by the proper selection and closest inbreeding these may be readily brought to either a high or low state with respect to these characters.
6. Many generations of closest inbreeding does not necessarily cause any loss in size, perfection of form, rate of reaction to light and gravity, egg production or length of life and sex-ratio.
7. The normal sex-ratio of this species in nature when reared under diverse conditions of food is one male to 1.126 females.
8. Different pairs in nature show a wide divergence in the sexratio of their offspring.
9. When the offspring from a pair with a given ratio are mated in pairs their offspring will show a wide range in the sex-ratio but in the aggregate will tend to reproduce the ratio of the brood to which they belong.
10. Sex-ratio is therefore a character that is strongly transmissible. By the proper selection of pairs tending to throw a high female ratio on the one hand or a low female ratio on the other it is possible to develop strains characterized by high or low female ratios.
11. In this species it is comparatively easy to develop a strain with a female ratio considerably higher than the normal but very
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 1
154 W. J. MOENKHAUS
difficult to develop a strain with a female ratio much lower than the normal or even one in which the sexes are equal in number.
12. Sex-ratio is one of the qualities that is, like color, an inherent characteristic of this creature, strongly transmissible and amenable to the process of selection.
13. The female is almost wholly responsible in the transmission of the sex-ratio. For, if females from a strain possessing a high female ratio be mated with males from a strain possessing a low female ratio or vice versa, the offspring will show a sex-ratio which is wholly or very near that of the strains from which the females were taken.
14. Sex is probably very little, if at all, influenced at fertilization in this species, but is probably determined much earlier and by the female, but there seems no reason why this may not be influenced by various factors and in some species at fertilization.
LITERATURE CITED
Born, G. 1881 Experimentelle Untersuchungen iiber die Entstehung der Geschlechtsunterschiede. Breslauer iirtzliche Zeitschrift. Bd. 3.
Bos, J. R. 1894 Untersuchungen fiber die Folgen der Zucht in engster Blutsverwandtschaft. Biol. Centralbl. pi. Bd. 14.
Castle, W. E. and others. 1906 The effects of inbreeding, cross-breeding and selecLion upon the fertility and variability of Drosophila. Proc. Amer. Acad. Arts and Sciences, vol. 41 .
CuENOT, L. 1899 Sur la determination du sexe chez les animaux. Bull, scientif. de la France et de la Belgique, t. 32.
Gentry, N. W. 1905 Inbreeding Berkshires. Proc. Amer. Breeders Association vol. 1.
Guaita, G. VON 1898 Versuche mit Kreuzungen von verschiedenen Rassen des Hausmaus. Ber, ub. d. Verhandl. d. Naturforsch. Gesellsch. zu Freiburg, Bd. 10.
Holmes, Claude D. 1910 The effect of starvation for five successive generations on the sex-ratio in Drosophila ampeloph'.la. Indiana University Studies No. 2.
King, Helen D. 1907 Food as a factor in the determination of se> in Amphibians. Biol. Bull., vol. 13.
KoRSCHELT, E. 1882 tJber Bau und Entwickelung des Dinophilus apatris. Zeitschrift f. wiss. Zool. Bd. 37.
Morgan, T. H. 1909 A biological and cytological study of sex determination in Phylloxerans and Aphids. Jour. Exp. Zool. vol. 7.
Whitney, D. D. 1909. Observations on the maturation stages of the parthenogenetic and sexual eggs of Hydatina senta. Jour. Exp. Zool. vol. 6.
Yung, E. 1885 De 1' influence des \ariations du millieu physico-chimique sur le development des animaux. Arch, des Sci. phys. et naturelles, t. 14.
CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE. E. L. MARK, Director. No. 23L
THE MECHANISM OF LOCOMOTION IN GASTROPODS
G. H. PARKER INTRODUCTION
The snail's foot in locomotion is so striking and so easily observed that it has excited the interest of naturalists for a long time and yet a complete solution of even the mechanical problems connected with its action seems not to have been attained. Within recent times a number of investigators have attacked the problem of locomotion in snails, but their efforts have been directed chiefly toward the elucidation of the action of the neuromuscular mechanism rather than toward an understanding of the external mechanical conditions that accompany locomotion. It is the object of this paper to consider, in the light of the more recent investigations and from the standpoint of renewed observation, the external mechanical factors involved in the movements of the gastropod foot.
The observations recorded in this paper were made partl}^ at the Bermuda Biological Laboratory, at the Harvard Zoological Laboratory, and at the Biological Laboratory of the United States Bureau of Fisheries at Woods Hole. I am under obligations to the directors of the laboratories mentioned for the materials and opportunities for carrying on these studies.
TYPES OF MOVEMENT
When the locomotor movements of the foot in many species of gastropods are compared, a surprising diversity is found. These different types of movement have been well classified by Vies
155
156 G. H. PARKER
('07) and are apparently characteristic not only for species but for larger groups of gastropods. In the majority of species thus far examined, the pedal waves course forward over the foot, thus agreeing in direction with the animal's locomotion. Vies has appropriately designated this type of movement as the direct type and has given the following gastropods as examples; the pulmonates (including Onchidium), Aplysia, Aeolis, Doris, Haliotis, Trochus, Cyclostoma, and certain small species of Littorina. I can confirm this statement for such of these molluscs as I have examined, namely, many pulmonates, including Onchidium, and I can add to this list Crepidula fornicata. In other gastropods the waves pass over the foot from anterior to posterior and this type has been designated by Vies as retrograde. As examples he has given Acanthochites fascicularis, Littorina littorea, and L. rudis. Besides confirming Vies' observation on Littorina littorea, I can add to this list Dolabrifera virens Verrill, Tectarius nodulosus Gmel., Nerita tessellata Gmel., and Chiton tuberculatus Linn. According to the observations of Jordan ('01, p. 99) Aplysia belongs under this head and not under that of the direct type as given by Vies.
In both chief types of movement several subtypes can be distinguished as determined by the lateral extent of the pedal waves. In some gastropods each wave extends over the functional width of the foot and thus the foot is occupied by only a single series of waves. This subtype has been termed by Vies monotaxic, and is exemplified by the pulmonates and chitons. In addition to these gastropods, Dolabrifera virens also has a monotaxic wave. In other gastropods the foot is functionally or even structurally divided along the median plane and exhibits a double system of waves, one right and the other left. This subtype has been designated ditaxic by Vies and is exemplified by Haliotis, Trochus, and Cyclostoma among the direct types, and by Littorina littorea among the retrograde types. Besides confirming Vies' statement as to Littorina littorea, I can add Tectarius nodulosus and Nerita tessellata as ditaxic gastropods. In Tectarius the waves on the two sides of the foot usually alternate and they are
LOCOMOTION IN GASTROPODS 157
SO extensive that never more than two waves can be seen on one side of the foot at once. The foot, therefore, moves forward in alternate steps, first on the right side and then on the left, the motion resembUng that of a person in a sack walk. In Nerita the wave begins anteriorly as a single wave whereupon it breaks and passes down the right and left sides of the foot to unite as one wave again at the posterior margin. These two conditions of alternate waves, as in Tectarius, and opposite waves, as in Nerita. will probably be found exemplified in other ditaxic gastropods. In certain small species of Littorina with direct movements, Vies has described four parallel sets of waves, fulfilling the requirements of a tetrataxic subtype. This occurs, according to Vies, only in connection with the retrograde type of movement. I have seen no example of it.
Among those snails that I have examined, one species, Ilyanassa obsoleta (Say), seems to find no place in Vies' classification. This snail is a vigorous, active creeper. Its foot covers a large area compared with the size of its body. Anteriorly the foot is truncated and auriculate; posteriorly it is bluntly rounded. Its ventral surface is whitish, flecked over with irregular grayish splotches. In resting, the snail uses chiefly the posterior part of the foot, the anterior part being sometimes more or less withdrawn into the shell. In locomotion the anterior part seems to be the more active. Notwithstanding the fact that this snail is very easily observed in active creeping and that its foot is marked in a most favorable way for exhibiting wave-like movements, I have never been able to discover any evidence of such movements. When in locoixiotion, the whole foot seems to glide at a uniform rate over the surface of attachment such as that of a glass plate. Only along the anterior edge and over a small portion of the anterior ventral surface of the foot, can slight variations in the rate of movement be discovered and these variations are so local and scattered that they can in no sense be regarded as forming a wave. The movement of the foot of Ilyanassa has a most striking resemblance to that of the foot of a planarian in which cilia may be the chief motor organs, but on testing the foot
158 G. H. PARKER
of Ilyanassa with carmine suspended in seawater. not the least evidence of cilia could be discovered. I therefore believe that Ilyanassa moves by a form of muscular activity that does not appear as pedal waves and it is not improbable that other gastropods will be found that have the same peculiarity. That Vies recognized something of this kind may be inferred from his statement that no changes in color can be seen in the creeping foot of Na-ssa, Buccinum, Aeolis, etc., and that the direction of the waves in these instances can be judged only by the deformations produced at the edge of the foot. As Vies makes no further mention of Nassa in his subsequent account, I suspect that it is more or less like Ilyanassa and is capable of little or no pedal-wave movement. The locomotion of such gastropods I should designate as due to arhythmic pedal movements as contrasted with rhythmic pedal movements, such as have been fully classified by Vies.
It is a significant fact that all gastropods, irrespective of their type of movement (direct or retrograde), are restricted to forward locomotion. None, so far as I am aware, can reverse and move backward as, for instance, an earthworm can. Whatever differences these various types of pedal movements possess, they still lead to but one result, the forward locomotion of the snail.
THE GASTROPOD FOOT AS A HOLDFAST
The snail's foot subserves the double function of attachment and locomotion. As means of attachment snails secrete a bed of mucus, and use the foot as a sucker. Both methods are commonly employed by the same species, but in a given form one method is usually developed much in excess of the other. For instance, in Helix pomatia, Limax maximus, and other allied species, the moist surface of the expanded foot will stick with some tenacity to glass. But if such an animal be allowed to creep its length over a glass surface and thus spread a bed of mucus on which it can rest, it will be found to have multiplied the strength of its attachment many times. The mucus adheres to the glass and the surface of the foot to the mucus very much more power
LOCOMOTION IN GASTROPODS 159
fully than the foot alone can adhere to the glass. That this attachment is due chiefly to the adhesive properties of the mucus and not to the sucking action of the foot, is seen from the fact that the attachment can be completely accomplished over a minute hole in a plate of glass. When a snail in such a position is seized and drawn off, air is sucked in through the hole in the glass as the middle of the foot rises, showing that under these extreme circumstances, the foot does act as a sucker, but in the ordinary resting state of the snail no such suction is exerted. All snails with which I am acquainted deposit more or less mucus and though this is sometimes so small in amount that it can be demonstrated only by means of powdered carmine, it serves, I believe, in so far as it is present, as a means of attachment. This production of mucus is highly developed in the pulmonates. Its relation to creeping on the surface-film of water, as exhibited by many fresh-water snails, has long been recognized.
In some snails the foot serves as an organ of attachment chiefly through its power of suction. The general surface of the foot is applied closely to the substrate after which the central portion is lifted thus converting the foot into a sucker. This kind of attachment is well exemplified in Patella, Crepidula, etc. Crepidula fornicata can be made to creep over a surface of glass and can move with ease and security over a minute hole in the glass. If, however, the snail is disturbed by being touched several times when its foot is over the hole, it will actually dislodge itself by endeavoring to suck firmly to the glass, for in so doing it will fill to repletion the forming concavity on the underside of its foot by sucking water through the underlying hole. When one contrasts the difficulty with which Crepidula is dislodged from its natural surface of attachment, particularly after it has been induced to exert full suction, with the ease with which it can be made to dislodge itself when over a small hole, the magnitude of its power of suction becomes apparent. The action of the foot of Aplysia as a suction apparatus has already been demonstrated by Jordan ('01). These two methods -of attachment, suction, and adhesion through mucus are the chief means by which snails hold to the surfaces on which they creep.
160 . G. H. PARKER
THE GASTROPOD FOOT AS A LOCOMOTOR ORGAN
Locomotion by the gastropod foot, is not dependent upon ciliary action but is a muscular operation as shown by Dubois and Vies ('07) . The precise way in which the movements of locomotion are accomplished can best be made out by examining good examples of direct and retrograde movement. The first is well exemplified in Helix pomatia and Limax maximus; the second in Chiton tuberculatus and Dolabrifera virens.
In an expanded and actively creeping Helix pomatia, the foot may measure as much as seven to eight centimeters in length by two and a half in width. Over this a succession of transverse, dark-brownish waves run from posterior to anterior. At any instant there may be as many as ten or a dozen such waves on the foot. Each wave is separated from its neighbor by a space equal to about three-times its own thickness. The waves travel over the foot in about thirty seconds, or at a rate of a centimeter in seven to eight seconds. These records, taken from a normal individual, agree fairly well with those given by Bohn ('02) and by Biedermann ('05).
As the snail creeps, it spreads from the mucous gland at the anterior edge of its foot a broad path of slime over which it makes its way. An active snail marks its course in this manner by a long track of slime. A somewhat exhausted snail, when placed upon an appropriate substrate, will almost always creep far enough to lay a mucous path that will subtend the whole of its foot, after which it will cease creeping. If it is removed to another position, it will usually repeat this operation, but it will seldom creep farther. This habit is doubtless connected with the effectual attachment of its foot to the substrate.
Locomotion in Helix, like that in other pulmonates (Kiinkel, '03), is apparently inseparable from the wave movement of its foot. When a snail is placed upon a glass plate preparatory to creeping, it lengthens and expands its foot; almost immediately thereafter pedal waves appear and the animal begins to move forward. Such a snail will creep over a perforation in a glass plate
LOCOMOTION IN GASTROPODS 161
without isuc'king air through the perforation, thus demonstrating that its attachment in locomotion/ as in rest, is due to adhesion and not to suction. In fact in a creeping Helix the foot not only does not suck but actually presses on the substrate. If, as the snail creeps, a bubble of air is introduced under it by a capillary tube or other means, this air will usually escape at the edge of the foot in such a way as to show that it was under considerable pressure. The action of such bubbles demonstrates that the foot as a whole is firmly attached to the mucous substrate, in fact presses against it. Locomotion in Helix pomatia, then, has to overcome under ordinary circumstances only the adhesion of the foot and this is accomplished apparently by the pedal waves. In snails in which the attachment is due to suction as well as to adhesion, locomotion requires that both attractive forces shall have been overcome, but, as suction is muscular, it seems likely that this would be relaxed somewhat, as seems to be the case in Crepidula, before locomotion begins.
How the pedal waves accomplish locomotion is still a disputed question. According to von Uexkiill ('09, p. 181), who has followed Jordan ('01) and Biedermann ('05) in many particulars, each pedal wave is formed by the contraction of the longitudinal muscles of the foot and takes the form of a slight swelling on the underside of the organ. Such a wave, as von Uexkiill rightly remarks, would effect nothing by way of locomotion unless some portion of the foot were fixed. Von Uexkiill ('09, p. 187) believes that the foot is provided with some such mechanical device as the setae of the earthworm, which, resist backward movement while they allow forward motion and that, therefore, the region in front of each wave may be regarded as a fixed region. Hence the contraction waves would always draw that portion of the foot where they temporarily were forward over the substrate toward the fixed point in front and as a result forward locomotion would be accomplished.
Although this explanation is free from mechanical objections, it is doubtful whether it really applies to the case in hand. Von Uexkiill has maintained in support of this view, that a snail can
162 G. H. PARKER
be slipped over a glass plate more easily forward than backward, just as an earthworm can be drawn over an appropriate surface more easily headward than tailward. I must confess that I have not been able to convince myself that there is any difference in this respect in Helix pomatia or Limax maximus; both seem to slip over the glass forward and backward with equal ease.
Moreover, the view advanced by von Uexklill is based upon what I believe to be a somewhat erroneous conception of the pedal wave. Biedermann ('05, p. 11) pointed out that the foot of Helix pomatia has great advantages over that of many other gastropods for studies of this kind because of the numerous small specks contained in its outer layer. These specks can be discerned clearly by means of a hand lens and they give a true picture of the movements of the foot. As watched through a plate of glass over which the anuTial is creeping, they can be seen, as Biedermann has described, to move momentarily forward, then come to rest, and then again to move forward. This is best demonstrated on a sheet of glass on which there are numerous scratches. Such scratches serve as landmarks and by them it can be seen that the minute specks in the foot do remain essentially fixed in position and then momentarily move forward to assume again for a brief period a position of rest. When this motion is examined in relation to the foot as a whole, it is evident that the forward motion takes place in the dark waves and that quiescence is characteristic of the intermediate lighter portions of the foot. Each wave, then, is a pulse of forward motion and the rest of the foot is momentarily quiescent. The area covered by the waves is probably a fourth or a fifth of the total area of the foot. At any moment, therefore, about three-fourths to four-fifths of the surface of the foot is stationary and about one-fourth to one-fifth is moving forward. In other words the snail stands on the greater part of its foot while it moves forward with a much lesser part.
Essentially the same conditions as have been described for Helix pomatia can be demonstrated in Limax maximus. If particles of carmine be driven into the substance of the median, active band of the foot of this slug, they can be seen to exhibit exactly
LOCOMOTION IN GASTROPODS 163
the same type of movement as has been described for the specks in the foot of HeUx. In Limax the waves, however, are light in color, instead of being dark as in Helix, and their surfaces, .as seen in the air, are marked with fine wrinkles transverse to the longitudinal axis of the animal. These wi'inkles show that the waves are regions of longidudinal contraction, as has been maintained by most recent writers on this subject.
The chief error in most previous accounts of the locomotion of the gastropod foot is found in the physical configuration ascribed to the underside of this organ. Biedermann ('05, pp. 10, 17) states that the waves are convexities on the surface of the foot and that they press more firmly against the substrate than does the rest of the foot. This view was adopted by von Uexkiill ('09, p. 187) in his discussion of gastropod locomotion. In Helix pomatia it is by no means easy to determine whether the waves are convexities or not, for the reason that they are at most only very slightly different in level from the general surface of the foot. On inspecting by reflected light the free ventral surface of a part of a Helix foot over which waves were running, I was unable to tell with certainty whether the surfaces of the waves were convex, concave, or flat. If, however, the creeping foot be closely studied through glass, evidence of a conclusive kind can be found. If, under these circumstances, a very minute air bubble entangled in the mucus under the snail is watched, it will be seen to change its form and position slightly as each wave passes over it. As the wave approaches it, it will elongate slightly on its face next the wave and at times move a little towards the wave, and as the wave leaves it, it will elongate slightly in the opposite direction and at times follow slightly the retreating wave. The motions of the bubble are exactly those that should be expected provided the wave exerted a slight suction in its passage and the reverse of what would occur supposing the wave pressed upon the bubble. The evidence, though slight, is clear and I, therefore, believe that each wave on the underside of the foot of Helix pomatia is a slight concavity.
Although the configuration of the surface of the wave in Helix pomatia could be determined only indirectly, in Limax maximus
164 G. H. PARKER
it can be seen with distinctness. If the anterior part of the foot of this slug be appHed to a glass surface, the pedal waves appear quickly over the whole foot. On inspecting the portion of the foot not yet in contact with the glass, the waves can be identified as dark bands alternating with light areas. On examining from the side the portion of the foot not yet in contact with the glass, it can be clearly seen that the waves are concavities in the foot as compared with the areas between the waves. I am, therefore, entirely convinced that, contrary to the opinion expressed by Biedermann and others, the pedal waves of the gastropods are concavities and not convexities on the foot. In these concavities, which are probably filled with the more fluid portion of the mucus, the foot moves forward, the rest of this organ being temporarily at a standstill.
The mechanical advantage of this arrangement must be obvious. The snail is attached to the substrate chiefly by adhesion to the denser mucus. This attractive force is overcome by drawing certain parts of the foot, the region of the waves, away from the substrate. These parts are then in a position to move with reduced resistance and are momentarily shifted forward while the snail supports itself on the rest of its foot. As this release from adhesion is propagated as a wave over the whole of the foot, this whole organ, together with the rest of the snail, is eventually moved forward. At first thought it might seem that such a wave movement could not produce so uniform a motion as snails show, but it must be remembered that the uniformity of this movement is seen only in parts of the animal some distance from the foot. On the foot itself the operation is alternate movement an4 rest, which becomes more and more continuous motion as points on the body more and more distant from the foot are reached. The locomotion is in many fundamental respects like that of the human being. In our locomotion each foot is alternately at rest and in motion and yet distant parts of our body, like the head, show a motion which in comparison with that of our feet is almost continuously uniform. In fact, a ditaxic gastropod with alternate, direct, single waves on the foot would almost exactly reproduce the method of locomotion found in the human being.
LOCOMOTION IN GASTROPODS 165
Tectarius, as already noted, practicalh^ fulfills these conditions except that its waves are retrograde. This general theory of the mechanics of gastropod locomotion is an elaboration of the views already set forth by Jordan ('01).
It is not my purpose in this paper to enter into an account of the musculature by which the movements already described are carried out, for I have made no observations on this part of the subject. It is, however, pertinent to show that the elements of motion implied in the preceding description are not inconsistent with the general structure of the snail's foot. The work of Jordan ('01), Biedermann ('05), and others shows conclusively, I believe, that the musculature of the snail's foot works against the elastic-walled, fluid-filled cavities of the animal's interior and that these cavities are often temporarily closed from one another. It is these spaces which, acting collectively as a vacuolated, erectile tissue, give rise to such rigidity as is possessed by the expanded foot of the snail. In this tissue two sets of muscles, longitudinal and dorso-ventral, have been identified. The dorso-ventral muscles lift the foot locally from the substrate. They are imbedded in the vacuolated tissue already mentioned and when they contract, their dorsal ends, being more firmly set than their ventral ones, serve as relatively fixed points and the ventral ones, therefore, move. The mechanical support that these muscles receive comes primarily from the tissue adjacent to their dorsal ends which in turn gets its support from other tissues reaching to the parts of the foot fixed on the substrate in front and behind the region of elevation. The "action of the ventral end lifts the foot locally and overcomes adhesion in the given region. When the muscle relaxes, the portion of the foot that was elevated is returned to its former level chiefly by the elastic action of the vacuolated tissue and the muscle recovers its original length and position. This action of the dorso-ventral muscles takes place in sequence from behind forward and thus a concave wave runs on the surface of the foot from tail to head.
The second element in the pedal wave is the forward movement of that portion of the foot which is temporarily lifted from the substrate. This must be accomplished by the contraction of the
166 G. H. PARKER
longitudinal muscles and can be best pictured by reference to the accompanying diagrams. These diagrams represent steps in the passage of a concave wave over the foot of a snail from an anterior position to a posterior one (left to right in the diagram) whereby the pointlc is temporarily released from full adhesion to the mucous surface, moved forward, and brought to full adhesion again. The point x is supposed to be associated with a particular longitudinal muscle fiber, number 2, through whose action it is moved. In A, this fiber is shown in its relaxed condition with the wave approaching. In B, the wave has released the point x from full
I 2 3
A ' ' — '
B i 1
C ' i
D ' '
K ' ' '—~
?-^- ^-^-^ ' ^<^
adhesion. In C, fiber 2 has contracted and since the posterior end of it is over a released part of the foot and the anterior end over a fixed part, the posterior end with the underlying point x has been moved anteriorly. In D, the fiber remains contracted and the point x has come again to adhere to the substrate. In E, the wave has reached the next longitudinal fibre anterior, number 3, which has contracted and drawn out the relaxing fiber, number 2, to its original length and position in reference to point x. The contraction of each longitudinal fibre then serves two purposes: it moves the foot forward as the releasing wave passes over the region and it extends the relaxing posterior fiber. In this way each
LOCOMOTION IN GASTROPODS 167
point on the foot is lifted, moved forward, and set down again and thus the foot, and with it the animal as a whole moves forward. From this theoretic consideration, it is evident that the theory of pedal-wave action advanced in the preceding paragraphs is entirely consistent with such an arrangement of muscles as has long been known to occur in the gastropod foot.
Vies ('07) has called attention to the fact that the majority of theories as to the locomotor action of the gastropod foot apply only to the direct type of movement and do not take into account the retrograde type. The theory put forward in this paper is believed to apply equally well to both types. Among retrograde gastropods. Chiton tuberculatus is an excellent example. This mollusc uses its foot as a sucker, but nevertheless can creep with considerable rapidity. It exhibits, as a rule, not more than two waves on the foot at a time; these course posteriorly at the rate of about a centimeter in five seconds. In a Chiton creeping over a glass plate, the wave when viewed from the side can be seen to be an area lifted well off the substrate. This feature is much more conspicuous in Chiton than in any other mollusc that I have examined. As in the pulmonates, the surface of the Chiton foot in direct contact with the substrate is motionless ; that in the wave area moves forward. At any moment about one quarter of the Chiton foot is moving forward while the animal supports itself on the remaining three quarters.
In Dolabrifera the foot is pear-shaped in outline with the rounded end posterior. It is about 8 mm. in length. In creeping, one to two waves can be seen on its surface at once; each wave sweeps the length of the foot in about seven seconds. As in Chiton, the waves can be clearly seen to be areas in which the foot is lifted completely from the substrate to which the rest of the foot is firmly applied. The pedal surface is mottled and in the wave area it can be seen to be moving forward, whereas on the rest of the foot it is motionless. The total wave area is about one-half the total area of the foot.
The conditions in Chiton and in Dolabrifera are essentially similar to those in the pulmonates, except that the pedal waves progress posteriorly instead of anteriorly, i.e., the dorso-ventral
168 G. H. PARKER
muscles contract in sequence from the anterior to the posterior end instead of the reverse and the longitudinal muscles follow the same sequence; otherwise they act as they do in the direct type. It is evident from this brief discussion of the nature of the waves in the retrograde type that the theory developed in connection with the direct type applies perfectly to this second type.
It remains still to point out that what I have called the arhythmic form of pedal locomotion, a form well exemplified in Ilyanassa, may be explained on the same general basis as that which has just been given for the two types of arhyhmic locomotion. If the foot of such a snail as Ilyanassa be thought of as composed of a multitude of small areas, each one of which can be lifted from the substrate, moved forward, and set down again separately, and that this action takes place irregularly and without reference to any sequence, it can easily be seen how the animal could move forward but without the formation of pedal waves. It is my belief that this is the condition in the foot of the arhythmic gastropods, but because of the small size of Ilyanassa, I have not been able to subject this opinion to experimental test.
Before closing this paper, I wish to add a word concerning the very remarkable method of locomotion observed by Carlson ('05) in Helix dupetithouarsi. The movement carried out by this snail is appropriately described as a gallop, both from its rate and configuration. The snail on strong provocation lifts the head and projects it forward, and eventually brings it to the ground, thus initiating a giant wave which proceeds backward over the length of the body. Several such waves may be present at once. Carlson suggests that this movement is only an exaggerated form of the ordinary locomotion, but I am inclined to agree with Jordan ('05, p. 104) that this is probably an entirely different type of locomotion and I suspect that this snail also possesses the typical pedal wave. In fact it seems to me likely that the gallop was, so to speak, superimposed on the pedal wave system and, had the snail when in gallop been examined from below, the pedal waves would have been seen in operation in conjunction with the body waves. I am the more inclined to the view that the gallop is an independent form of locomotion as compared with the pedal
LOCOMOTION IN GASTROPODS 169
waves, because in the gallop the body waves of this species, as reported by Carlson, were retrograde whereas the pedal waves in all Helices thus far reported are direct.
SUMMARY
Ordinary gastropod locomotion is accomplished either without pedal waves (arhythmic) or with pedal waves (rhythmic). In rhythmic locomotion the waves may run from posterior to anterior (direct) or the reverse (retrograde). The foot may exhibit one (monotaxic), two (ditaxic), or four (tetrataxic; series of waves. In the ditaxic foot the waves may be alternate or opposite.
The gastropod foot is an organ of attachment through adhesion (mucus) or suction, or both.
The pedal wave is an area of the foot that is lifted off the substrate as compared with the rest of the foot and thereby freed more or less from aahesion. It is also the region of the foot that moves forward, the rest of the foot remaining temporarily stationary. Locomotion is the cumulative result of local forward motion on the part of one section of the foot after another till the whole foot has been moved. The same type of muscular movement as that seen in rhythmic locomotion can be present in a diffuse form (not wave-like) in a gastropod foot and will result in locomotion.
170 G. H. PARKER
BIBLIOGRAPHY
BiEDERMANN, W. 1905 Studien zur vergleichenden Ph}^siologie der peristaltischen Bewegungen. II. Die locomotorischen Wellen der Schneckensohle. Arch. f. ges. Physiol., Bd. 107, pp. 1-56, Taf. 1-2.
BoHN, G. 1902 Des ondes musculaires, respiratoires et-locomotrices, chez les Annelides et les Mollusques. Bull. Mus. Hist. Nat., Paris, tome 8,pp. 96-102.
Carlson, A. J. 1905 The physiology of locomotion in gastropods. Biol. Bull. , vol. 8, pp. 85-92.
Dubois, R., et Vles, F. 1907 Locomotion des Gasteropodes. Compt. rend. Acad. Sci., Paris, tome 144, pp. 658-659.
Jordan, H. 1901 Die Physiologic der Locomotion bei Aplysia limacina. Zeit. f. Biol., Bd. 41, pp. 19&-238, Taf. 2.
1905 The physiology of locomotion in gastropods. Biol. Bull., vol. 9, pp. 138-140.
KtJNKEL, K. 1903 Zur Locomotion unserer Nacktschnecken. Zool. Anz., Bd. 26, pp. 560-566.
Uexkull, J. V. 1909 Umwelt und Innenwelt der Tiere. Berlin, 8vo, 261 pp.
Vles, F. 1907 Sur les ondes pedieuses des Mollusques reptateurs. Compt. rend. Acad. Sci., Paris, tome 145, pp. 276-278.
THE REGULATORY PROCESSES IN ORGANISMS
C. M. CHILD
Rull Zoological Laboratory, University of Chicago
Introduction 171
The organism as a physico-chemical system 173
1. The relation between metabolism, and structure 173
2. Physiological correlation and the physiological system or individual . . 179
3. The basis and nature of physiological correlation ISl
The nature of regulation 182
1. Organic or physiological equilibrium and equilibration 182
2. Regulation as equilibration 188
The regulatory processes 199
1. The relation between form regulation and functional regulation 199
2. The inducing conditions and the results 199
3. The provisional classification of the regulatory processes 200
a. The two methods of regulation 200
h. Regulatory compensation 202
c. Regulatory transformation 205
The nature of reconstitution 207
1. Restitution or reconstitution? 207
2. The initiating factor in reconstitution 211
3. The process of equilibration in reconstitution . 212
4. The complexity of reconstitution .215
5. The limits of reconstitution 217
Reproduction in general as a form of reconstitution 218
Conclusion 221
Bibliography 222
INTRODUCTION
Of late years the term 'regulation* has come into such general use and has been applied to so wide a range of organic phenomena, that it seems desirable to attempt a general consideration and analysis of the regulatory processes from the present viewpoint of physiology. Tne biologist who takes the position that there is at the present time, when the investigation and analysis of the physics and chemistry of the organic processes is still only
JOURN.\L OK MORPHOLOGY, VOL. 22. NO. 2
172 C. M. CHILD
at its beginning, no adequate basis for 'vitalistic' interpretations of regulatory phenomena finds but little satisfaction or enlightenment in Driesch's 'entelechy' or in other assumptions of the neovitalistic school.
In the present state of our knowledge these views are and must remain expressions of personal opinion. Driesch's first two Beweise der Autonomie der Lebensvorgange" (Driesch, '01, '03 etc.), which are based on certain phenomena of form regulation, constitute proofs only when we accept Driesch's premises, and as I have pointed out (Child, '08b) these premises are pure assumptions. Neither Driesch nor anyone else has placed them on a foundation of fact. The existence of the 'harmonious-equipotentiai system,' for example, which is of so great importance to Driesch, is a matter of assumption, not of fact. So far as the systems, which according to Driesch belong in this category, have been thoroughly examined, they have shown themselves to be neither harmonious nor equipotential in Driesch's sense, and to the extent which he has assumed. It is of course easy to assume, as Driesch has done, that the harmony of these systems is due to entelechy and their limitations to physico-chemical factors, but such assumptions, since they are so manifestly invented ad hoc, do not carry conviction to the minds of most biologists, what ever, their effect upon their author.
Much the same is true of other modern vitalistic hypotheses: as expressions of personal opinion, they are of great interest in the histor}^ of scientific thought, but none of them thus far has presented any convincing arguments in its own support.
Furthermore, with the exception of Driesch's analytical consideration of the regulatory phenomena in organisms, most of the recent published works of general character, which concern themselves primarily with the regulations which involve the visible morphological features of the organism, e. g., the books of Morgan ('07), Korschelt ('07) and Przibram ('09), have been devoted chiefly to the descriotive, rather than the analytical and interpretative aspects of the subject.
In view of these facts, an attempt at physiological analysis of the regulatory processes or of some of them can scarcely be re
REGULATORY PROCESSES IN ORGANISMS 173
garded as superfluous. The further my own investigations in this field proceed, the more completely I am convinced that those phenomena, which we are accustomed to call regulations are among the most characteristic, perhaps it is not too much to sa}-, the most characteristic phenomena of life.
The views of different authors concerning the relation between regulatory and ' normal' or ' typical' phenomena are very different. Roux ('95, II, pp. S43-4), for example, makes a sharp distinction between typical and regulatory development, though he admits that the distinction is analytical rather than practical. For Driesch regulation is areturn approach to the normal condition, after this condition has been disturbed by some external factor. Many physiologists, on the other hand, have used the term 'regulation ' in a much broader sense, as applying not only to the extra-normal or extra-typical but at least to many of the most typical phenomena of life.
Because there is no general agreement concerning the real basis and nature of these processes, and because the regulations are of great importance for any interpretation of life, it seems worth while to undertake a brief analysis of them and particularly of the regulations which involve form and structure to a large extent. The present paper is concerned with such an analysis.
THE ORGANISM AS A PHYSICO-CHEMICAL SYSTEM
1. The relation between jnetdbolism and structure
The structural basis of living organisms consists primarily of colloids. These colloids, together with water, make up the greater portion of what we are accustomed to call protoplasm and in this protoplasm the various reactions and processes which characterize life occur. The universal association of colloids with life suggests that these substances play an important part in some manner in determining some of the characteristic features of life.
Metabolism has been commonly conceived in the past as consisting, on the one hand, of the synthesis of an exceedingly complex
174 C. M. CHILD
and highly labile molecule, and on the other, of the breaking down of this molecule in functional activity. According to this view the colloid structure built up is used in function and must be continually replaced.
During recent years, however, many facts have been discovered which seem to make necessary some modification of this view of the relation between metabolism and colloid structure. In the first place, most proteids, and indeed most organic colloids, are relatively inactive chemically. The common interpretation of this fact has been that death involved, or perhaps consisted in a change from lability to relative stability of the proteid molecule. But the recent work of Fischer and others upon proteids makes it highly probable that the proteid molecule, although very large, is not so complex as has been supposed, but may be polymeric in high degree. Thus the assumption of extreme lability of this molecule in the living organism becomes even more difficult than before.
Work along other lines has demonstrated that the nitrogen metabolism is only a fraction of the total metabolism of the organism and that it does not necessarily increase in proportion to functional activity. Moreover, the nitrogen requirement for maintenance in animals is apparently much smaller than had been supposed. All of these facts seem to indicate that in the living organism, as in vitro, the proteids, or many of them, are relatively inactive chemically; that after they are formed, they are excluded to a large extent from metabolism simply because of their relative inactivity. If these conclusions be correct, the accumulation of proteids in the organism is a process not very different from the deposition of other forms of inactive substance in or about the cell, e. g., chitin, insoluble salts, etc. Indeed it is highly probable that the accumulation of most or all structural substances in the organism is due to the fact that they are relatively inactive under the existing conditions. After they have arisen in metabolism they persist or disappear much more slowly than other substances, simply because under the exi.sting conditions they do not enter chemical reactions as readily as other substances.
But the proteids and other substances deposited in the cell are not absolutely inactive and undoubtedh' do enter metabolism
REGULATORY PROCESSES IN ORGANISMS 175
to some extent at all times. Under certain conditions, however, i. e., in the absence of certain other substances, they, or some of them, may reenter metabolism to a much larger extent and furnish energy. It is a familiar fact that in the absence of nutritive material from without the organism uses up its own substance, ^. e., the relatively inactive substances which under other conditions had accumulated in it. In certain of the lower organisms this process may continue until the organism is reduced to a minute fraction of its original size. These facts do not, however, conflict with the suggestions made above as to the relative inactivity of structural substance, but serve rather to confirm the idea that the accumulation of these substances, when other nutritive material is present, is due to their relative inactivity.
Various authors have attempted to distinguish between a morphological and a functional metabolism, but it is doubtful whether such a distinction is valid, except as an expression of the fact that substances of different degrees of chemical activity arise in the course of metabolism and that certain of the less active substances constitute the structural basis of the organism, while others undergo chemical transformation and elimination.
It is of course not merely the nature of the substances themselves, but the existing conditions as well, which determine the degree of activity or inactivity. Under certain conditions a cell or an organ may accumulate certain substances and so acquire a certain characteristic structure, while under altered conditions these substances may rapidly disappear and others be accumulated. Thus, for example, the oocyte, during its growth period, accumulates yolk, which under the existing conditions is almost wholly inactive chemically and so appears as a structure-building substance. But when fertilization occurs the conditions within the cell are so altered that the accumulated yolk rapidly reenters metabolism and serves as nutritive material. In fact we may say that the egg does not produce yolk because it is to develop into a new organism, but that it develops as it does because it has accumulated yolk. In the periodic changes in the cell connected with growth and division there is also abundant evidence for the occurrence of changes of this character.
176 C. M. CHILD
If we may accept this view of structure in the organism, and all the facts are in its favor, then it is actually very similar in its relation to the energy current to the morphological characteristics of a river system except of course that the latter are mechanically produced. The constructed islands and bars, the depositions of the river, represent those particles or masses which have, under the conditions existing at a given time and place, been left behind by the current. Under certain conditions the river may produce structure of a certain kind at a certain point in its course, while under different conditions this structure may disappear and give place to structure of a different kind.
But the most important fact for present purposes is that in the organisms, as in the river, structure, as soon as it appears, begins to influence the metabolism, the energy current. From this time on the metabolic processes, like the flow of the river, occur in a certain structure and here the mutual interactions begin. Of the character of these interactions in organisms we are only beginning to obtain some vague conceptions, but that they occur, it is impossible to doubt.
Perhaps a few words will not be out of place concerning the bearing of these facts and suggestions upon the theory of * formative substances ' which has played a considerable role in embryological investigation during the last few years. Most of the supporters of this theory have attempted to identify the so-called formative substances with visible granules or other accumulations in the cytoplasm, without considering the fact that the appearance of these substances in visible structural form indicates that they are, at least for the time being, relatively inactive, and that they are first of all products or incidents of metabolism (Child, '06b). Of course some or all of these substances might reenter metabolism under altered conditions and so play a part in determining its character, but the important point is that they are indications of a difference in metabolism already existing in the different regions where they are formed. We might expect that the differences in metabolism, which are certainly more important as formative factors than these accumulations of granules, would persist in the different regions, even if the granules could
REGULATORY PROCESSES IN ORGANISMS 177
be removed. Fortunately the embryologists themselves have now a method of removing these granules from their usual position, i. e., the method of centrifuging the egg, and the results of recent experiments along this line indicate, as was to be expected, that the granules are quite unessential to regional localization and differentiation of the embryonic structures.
It is evident from the above suggestions that our fundamental conceptions of the relation between structure and function in organisms must be intimately connected with our ideas concerning the nature of colloid substances and their significance as a substratum or medium for chemical reactions. Within recent years it has been pointed out repeatedly that these substances afford various means for the partial or total isolation, of different chemical reactions in organisms and that their mere presence may bring about such isolation, e.g., by the formation of semipermeable membranes. Here then we have a physico-chemical basis for localization and differentiation. Moreover, the changes in the physical aggregate condition of colloids, together with the possibility of the simultaneous existence of different phases of high and low water-content, must play a part in determining the degree and place of dissociation of various substances, and therefore in determining the speed of reactions in different regions, as well as the occurrence or non-occurence of certain reactions at particular points. It is then, to say the least, highly probable that the possibilities of localization, physiological specification and the accompanying possibilities of physiological correlation of parts and of regulation are very closely connected with the fact that the formation of colloids is a component of the reaction complex known as metabolism. Moreover, as I have attempted to show in another paper (Child, 'lib), the accumulation of relatively inactive substances, particularly colloids, in the cell is undoubtedly a factor in senescence, in that it constitutes an obstacle to the metabolic interchange and so brings about a decrease in the rate of metabolism.
If the conclusion be correct, that the visible structural elements of the cell are, at least for the time being, relatively inactive chemically, then it follows that these elements do not represent the
178 C. M. CHILD
'living substance' in the stricter sense, but are really the least 'alive' of any part. The reactions which furnish the energy of life undoubtedly occur, at least in large measure, in the more fluid parts of the cell, the parts which present the least characteristic structure. The so-called living substance is actually then, so far as it presents a visible structure, chiefly a substratum or m€dium in which the reactions occur, and is itself the product of past reactions. That these structural elements, as they accumulate, must modify the rate and character of the reactions to an increasing extent, cannot be doubted. The advancing specialization of metabolism in different organs and cells is probably closely connected with the fact that these parts produce different structural elements, either in consequence of an original specification or in consequence of different correlative or external conditions which induce specification.
If we accept this view of the relation between function and structure in the organism, we must give up the idea of a definite 'living' substance in the chemical sense, and the basis of life becomes, not a specific substance, but a series of reactions in a field or medium of a certain complex constitution, which is itself the product of past reactions. We can agree with Driesch ('01, p. 140), as regards the absence of a specific living substance, though we cannot follow him in his further conclusions along this line. The life process has become individualized, not because of entelechy, but because it forms its own field or medium of action, as the river forms its channel, particularly in the later stages of its course, where deposition exceeds erosion. To put the matter briefly, life as we know it consists not in metabolism alone nor in a specific substance or structure alone, but in the physiological correlation of processes in a structural medium or substratum of a certain constitution, which makes possible localization and correlation of processes.
It is then the existing relation between the processes and the structural substratum, the mutual interaction and dependence of both, that forms the basis for the phenomena of regulation or equilibration which occur in the organism.
KEGULATORY PROCESSES IN ORGANISMS 179
2. Physiological correlation and the physiological system or individual
Organisms in general appear in the form of more or less sharply defined physiological systems or individuals and in the more complex organisms we can distinguish systems or individuals of various kind and degree. What is the basis of this unity?
The experimental investigation of orgaaisms has led those who are not yet ready to accept vitalistic hypotheses to the conclusion that two factors are chiefly involved in the formation of a living system or individual, viz. constitution and physiological correlation. In its grosser aspects the first of these is the morphological, the second the physiological factor. Most of us believe, however, that the morphological features of organisms are essentially visible expressions of dynamic processes past or present and that sooner or later we must interpret constitution in dynamic terms. The factor of physiological correlation in the organism is essentially the problem of physiology, for in the final analysis function is impossible without such correlation.
Wherever in the universe unity can be recognized, there some sort and some degree of correlation must exist, either conceptually or as a datum of nature, between the elements which compose the unit, and vice versa, wherever correlation between conceptual or phenomenal elements is recognized or established, there a unity of some sort and some degree exists. On the other hand, the character of the unity is determined by the nature or constitution, however this may have arisen, of its elements.
There is at present no adequate ground for believing that organisms differ from other phenomena in these respects. We cannot conceive an organic individual without correlation of some sort between the parts which compose it, nor can we conceive it without elements or parts of a certain more or less characteristic constitution.
The development of morphology and its separation from other fields of biology during the latter half of the nineteetith century has led, particularly in the field of zoology, to the consideration of the problem of constitution apart from that of correlation. But the
180 C. M. CHILD
introduction of the experimental method into zoology has alreadydemonstrated the limited scope and value of pure morphology for the interpretation of life. In the organism as we find it, the two factors, constitution and correlation are mutually determining. We cannot alter the constitution without altering the correlation of parts, neither can we alter correlation without changing constitution to a greater or less extent. The cases of so-called selfdifferentiation constitute no real exception to this statement,^ which has the value and significance of a law of nature. Morphology and physiology are inseparable except analytically and their artificial separation can lead only to the formulation of many pseudo -problems and to uncertain or false conclusions and hypotheses.
In so far as the organism is a physiological, i. e., a physicochemical individual or unity, in so far must physiological correlation exist between its parts in the form of actual physical and chemical processes, conditions and substances. Until it is proven by the profoundest investigation and the strictest analysis that physiological correlation does not suffice to account for the organic individual, there is no need of turning to the vitalistic hypotheses for an interpretation.
Indeed our knowledge of physiological correlation is in its earliest stages. One need only refer to the work on the conduction of stimuli in plants and through protoplasm in general and to the investigations of recent years on the thyreoid, the adrenals, the reproductive organs, the pancreas, etc., as organs of chemical correlation and to the work on hormones, to become aware of the advances in knowledge along this line within the last few years. At present we are willing to believe, in fact we find it difficult not to believe, that every metabolically active organ in the body is an organ of chemical correlation. And we also know that many
1 As Roux ('95, p. 822 etc.) has pointed out development depends primarily upon correlation and absolute self-differentiation cannot occur. In cases where parts differentiate in a relatively high degree of independence from each other, we must believe, and in some cases, e. g., the nemertean egg, we know that this condition is preceded by, and is the result of an earlier condition in which the parts are in much closer correlation.
REGULATORY PROCESSES IN ORGANISMS 181
of these correlative factors show a high degree of specificity. Moreover, the chemical factors are by no means the only factors in correlation: mechanical and other physical factors also play a part. And finally, the experimental investigations themselves have demonstrated the importance of physiological correlation in morphogenesis.
In short, there is at present every reason to believe that the existence and continuity in time and space of organic individuality are essentially dependent upon physiological correlation, i.e., upon processes and conditions which are accessible to scientific investigation and analysis.
3. The basis and nature of physiological correlation
That Dhysiological correlation is in general dependent upon the physical and chemical processes and conditions in the various parts which make up the individual cannot be doubted. These in turn are deDendent upon the constitution of the parts, which itself depends in part upon preexisting correlation and to a greater or less extent upon conditions and processes in the extra-individual environment. At every step in our consideration we recognize the mutual interdependence of constitution and correlation.
But if we consider the organic individual only as it exists at the present time, then we may say that the existing physiological correlation between parts is dependent upon the conditions and processes in the parts, however these may have been brought about.
In general we can recognize at present three main groups of correlative factors : first, mechanical or mass correlation (Roux, '95, II, p. 240), which results merely from the existence of mass without respect to constitution; second, substantial or material correlation, which consists in the actual transference or transportation of substance possessing a certain physical or chemical constitution, e. g., chemical correlation; and third, dynamic correlation, of which the essential feature is the transmission of energy rather than the actual transportation of material over any appreciable distance. None of these forms of correlation can be sharply
182 C. M. CHILD
separated from the others in the final analysis, but in its extreme forms each type is readily distinguishable.
The physiological correlative effect of a part upon others is then the result of all that that part is and has been in the past, of its physical and chemical constitution, its position, its relation to external factors and of the changes which are occurring in it. It is apparent that there exists in physiological correlation the possibility of an almost infinite variety and specificity. Driesch ('09) has recently maintained that the specificity of the 'Restitutionsreiz' together with the specificity of the reaction to it constitute' an 'Individualitat der Zuordnung' which is inexplicable on a physico-chemical basis and which therefore constitutes a new and independent 'proof of the 'Autonomic der Lebensvorgange.' Comment seems scarcely necessary. One sees here merely an assertion, a jump at conclusions, but no proof, where proof of the most convincing character is absolutely essential. If vitalism can present no more convincing arguments than this its future prospects in science are not bright.
THE NATURE OF REGULATION 1. Organic or physiological equilibrium and equilibration
One of the most characteristic features of organisms is, as Roux ('95, I, pp. 145, 154, 392, etc.) has said, their continued existence as individuals, their 'Dauerfahigkeit' amid changing internal and external conditions. On the other hand this 'Dauerfahigkeit' is only relative, not absolute, i. e., it is limited. The organism is constantly changing, and- so far as our knowledge goes, never twice the same, yet the continuity of individuality is obvious.
Nevertheless the continuity of the existence of individuality must not be emphasized to the exclusion of the fact that under certain conditions this individuality may disappear, at least in the simpler organisms, and be replaced by other individualities in larger or smaller number. Certain factors concerned in this physiological disintegration will be discussed below, but for the present we are concerned with the individual, the system as we see it
REGULATORY PROCESSES IN ORGANISMS 183
in the organism or part which constitutes a unity cHstinct to a certain extent from others.
When we investigate the processes in the organism, we find that they are very intimately connected with one another: a change in one conditions changes in others. Moreover, and this is an important point, the physiological specification of different parts is not in most cases absolute. In the highest, most complex forms absolute specification is doubtless approximated more or less closely in certain organs, but in general we find that the processes in different parts of the organism are not fixed in character. The characteristic series of reactions in a part does not represent the only possible series, but rather the particular series determined by a particular complex of conditions. A certain process may occur at one time in a certain part, at another time in others. In short there is more or less possibility of substitution among the different parts.
Let us suppose, for example, that a certain correlative factor x originates in a certain part. Under certain conditions this factor may influence various other parts, a b c — n,of which one, a, let us say, reacts with greater speed or intensit}^ than others. The reaction of this part may itself produce new correlative factors and so alter conditions in the others that their reactions are changed. But if we suppose that the receptivity of the part a to the correlative factor X is decreased, or that the part a is itself removed or rendered incapable of reaction, then the reactions of a b — n or of some of them are not altered or prevented by the effect of the reaction of a, and these parts may take the place of a in the system, though perhaps at first reacting more slowly or less intensely than a, until their constitution has become altered by repeated or continued reactions.
Through such a series of reactions the individuality of the organism is maintained, or restored, even though it may have lost a part. We see exactly such reactions in various organisms, and we can devise physico-chemical systems which show similar correlation between parts and a similar method of maintenance or restoration of something approaching the preexisting condition. In the system which we devise such processes are processes of equili
184 C. M. CHILD
bration. We find that the system is capable within certain limits of attaining or approaching a condition of equilibrium after a disturbance of a previously existing equilibrium.
And again in the law of mass action and the general principles of chemical equilibrium, together with what we know of katalysis, we have the possibility of accounting for a great variety of processes of equilibration in the organism. Until we have exhausted these and other physico-chemical possibilities and found them inadequate, we have no adequate reason for oelieving that organic individuality and its maintenance are anything unique.
The fact that physiological correlation exists between different parts of an organism must necessarily determine a certain relation, a certain proportionality in the activities of the different parts. It is this relation, this proportion in activity determined by correlation which constitutes what we call organic or physiological equilibrium in the organism. This equilibrium is dynamic, not static, it is an equilibrium of processes, not of masses and it must be dependent either upon physiological correlation, or upon something else which controls the supply of energy to this or that part in very much the way in which the man in charge controls the workings of a complex machine, e.g., a steam-shovel, turning the steam into this or that cylinder as required for the harmonious working of the whole. Driesch's entelechy is comparable to the man in charge of the engine.
But it is not the mere existence of an organic equilibrium which constitutes the real problem; it is the apparent power of adjustment, of equilibration, the harmony of action of the parts, as in the engine, which has been regarded as the strongest argument for vitalism. How, the vitalist asks, is it conceivable that a machine with such capacity of adjustment, of equilibration as the organism, which can even repair itself, can be constructed and continue to exist and work unless there is something comparable to the man in charge concerned in these processes.
As a matter of fact this question is based on a wrong conception of the organism. The organism as we see it, i. e., morphologically, is not the machine whose action constitutes life, but rather simply a part of the products of that machine, which accumulate during
REGULATORY PROCESSES IN ORGANISMS 185
its action and as they accumulate, alter and determine the character of its activity. In other words, as the products are formed they become a part of the machine. Starting with the egg, the organism is not, as Driesch asserts that it must be according to the 'machine theory,' a machine developed /or function (Driesch, '05, p. 790), but rather a machine developed by function. The result at any stage represents morphologically the products of a preexisting machine and physiologically the action of the machine as altered from the preceding stage by the products of its own activity. Each stage of development is the result of the machine plus the product of the preceding stage. Our experiments have shown that physiological correlation, not predetermined harmony is the basis of development, and that where a predetermined harmony appears to exist it is certainly in some cases, probably in others, the result of an earlier condition of correlation.
On the basis of this conception of the organism it is inconceivable that processes of adjustment of the parts to each other, i. e., processes of equilibration, should not occur, both in development in nature and under experimental conditions. The parts are what they are, not simply because of their original constitution, but because they have been acting in correlation with each other. From the moment the organic machine began to work in the first organism 'adjustment' of the parts to each other began and it has continued ever since. Could we but read it completely, every part is a record, an epitome, more or less complete according to circumstances, of what has been going on, not merely in itself, but in the whole organism. Moreover, in different parts this record is written in different characters, in different languages, according to the constitution of the part.
The distinction which Roux makes between the formative and the functional periods of development (Roux, '95, II, p. 281), is, according to this view, not fundamental in character. The formative period is functional and the functional period is formative. But this distinction is based upon the fact that at a certain more or less sharply defined stage of development the accumulated products of the activity of the machine begin to play a more or less definite role in its further physical and chemical activity. The adult
186 C. M. CHILD
organism is not then to be compared with a machine constructed of certain definite Darts, which have been put together in some way, and which, after completed construction, begin to function. It is much more nearly comparable to a river, which molds its banks and bottom, forming here a bar, there an islaad, here a bay, there a point of land, but still flowing on, though its course, its speed, its depth, the character of the substances which it carries in suspension and in solution all are altered by the structural conditions which it has built up by its own past activity. In such a system a wide range of equilibration exists and we see both the adjustment of function to form and of form to function. The relation between structure and function in the organism is similar in character to the relation between the river as an energetic process and its banks and channel. From the moment that the river began to flow it began to produce structural configurations in its environment, the products of its activit}^ accumulated in certain places and modified its flow% but just so long as the flow continues the process of equilibration goes on. If we consider merely a certain region of the river with the water containing certain substances in suspension and in solution entering at one end, depositing some of these substances and taking up others as conditions determine in the course of its passage, and finally passing out at the other end bearing certain substances more or less different from those which it brought in, the analogy becomes even more complete. In fact this region of the river, together with its bed, shows a real, though chiefly a mechanical rather than a chemical metabolism.
I believe that this comparison between a river with its channel and the organism is far more than a fanciful analogy. Theindividdual organism is merely a section from that current of energy which constitutes the essence of life, and in the individual we see the mutual correlation and interaction between the current and the conditions under which it finds itself, between the energetic process
- Rignano ('07) has referred briefly to this analogy between the river and the organism, using the case of a river equilibrating itself in connection with the piers of a bridge to illustrate the process of equilibration in organisms. See also Delage, L'Heredite, etc., 1903.
REGULATORY PROCESSES IN ORGANISMS 187
and the structural features which its activity has produced. As the banks and the channel are 'adjusted' to the activity of the current, and the current to the morphological characteristics of the banks and bed, so, and in no otherwise are structure and function in the organism, correlated with each other. It is absolutely inconceivable that 'adjustment,' equilibration should not occur. So long as the current flows, equilibration must take place in one way or another.
The organism has often been compared to a flame. Roux particularly has carried out this comparison in detail (Roux, '05, p. 109, et seq.). Although this analogy contains much that is valuable and on the chemical side is much closer than that of the river, yet on the other hand the morphological features of the river are more nearly comparable to those of the organism, In their localization, their often complex structure and their modifying effect upon the activity of the current. For these reasons I have chosen the river rather than the flame as a physico-chemical system with which the organism may be compared.
When we take the view of the organism suggested above, I believe that Driesch's first two proofs of the autonomy of vital processes (Driesch, '03, p. 74, etc. Cf . also p. 197 below) appear in their proper light. They apply only to the morphological conception of the organism as a machine constructed for function, i. e., to the banks and channel without the river. In the organism the current is working from the beginning, the organism is functioning in one way or another, and the real machine is the process, the function, plus the existing structure which past processes have produced, just as in the case of the river the real machine is the current plus the banks and channel. The process of development in the organism is comparable, not to the digging of a channel into which, after its completion, the water is turned, but to the formation of a channel with certain characteristics determined by a variety of conditions, by the activity of the current itself. From the moment the current begins to flow, structure and function become mutually interdependent and mutually determining, but there can be no river -structure without the current. Machines like the steam engine, constructed by man and considered without
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
188 . C. M. CHILD
their motive power, are comparable rather to the dead than to the living organism. They are merely the conditions under which the energy acts, but the living organism consists from the beginning of these conditions plus the energy. Development is not comparable to the construction of such a machine by man, but rather to its action after the steam is turned on. Every steam engine possesses a certain power of equilibration dependent upon its constitution, and the only reason its powers in this direction are so narrowly limited is because the energy current and the structure have not been working together from the beginning.
The only possible basis for a scientific, as opposed to a philosophical vitalistic hypothesis is the proof that the energy of organic life is something essentially different from the energj^ of the physico-chemical world. When the vitalists shall succeed in proving this or even in making it probable, then their views will be given more general consideration. But even the most extreme among this school at the present day do not attempt such proof. If we admit that the energy of the organism is not different from that in the physico-chemical world, then I believe we are forced to regard the organism as a physico-chemical system, for as I have shown above, physico-chemical systems exist in which the relation between structure and function, between the conditions of action and the energy itself, are of the same character as in the organism itself and give rise to a power of equilibration of the same character.
2. Regulation as equilibration
From what has been said it will be at once apparent that the processes which we commonly call regulatory are processes of equilibration in the organism (Holmes, '04, '07, Child, '06, '08a). They enable the organism to persist and to maintain its individuality under changing conditions, although it cannot be supposed that the condition of dynamic equilibrium is the same for different conditions, and indeed we have evidence that it is not. But within certain Umits, and for certain factors, the organism is capable of a greater or less degree of equilibration, when a change in external conditions occurs.
REGULATORY PROCESSES IN ORGANISMS 189
The question at once arises as to whether all processes of equilibration are to be regarded as regulations, or only certain of them. By zoologists the term 'regulation' has been applied mostly to processes occurring under experimental conditions outside the usual range of conditions in nature and the regulations of form and structure have been the chief, though not the only objects of investigation. Jennings ('06) has used the term with reference to phenomena of behavior which are characteristic features of life and not of abnormal or pathological conditions. Among the physiologists also we find the term often used as referring to various changes in metabohsm and reactions of different kinds in response to conditions to which every individual is subjected repeatedly.
If we define regulation as a return or approach to a condition of dynamic equilil^rium in a living organism after a previously existing condition has been disturbed by some external factor (Child , '06) , we shall include all the above phenomena as well as many others. According to this definition, the simplest reflex as well as the restoration of a missing part is a regulation, the simplest correlative compensation in metabolism, as well as the development of a whole from an isolated blastomere of an egg.
Moreover, when a complex part of an organism undergoes an equilibrating change in reaction in response to altered correlation with another part or other parts, a regulation occurs as truly as when the whole organism responds to some change in conditions outside of it. In short, regulations are equilibrating reactions to changes external to the reacting system, whether this system be a part or a whole of an organism.
And finally, regulation is not limited to the return or approach to the preexisting condition, but may be an approach to a condition very different from that, i.e., the organism or the part may become something more or less widely different from what it was originally. In every case of regeneration of lost parts some of the cells become something different from what they were before the part was removed, and their change is a reaction to altered conditions and specifically to altered correlation.
But that the regulatory process is always and necessarily of advantage to the organism does not follow from the definition.
190 C. M. CHILD
So far as it enables the organism to persist, it may be of advantage, and I see no escape from Roux's argmnent (Roux, '95, 1, p. 145, 154, etc.), that systems possessing such reactions will persist longer than others. But not all regulatory processes are of advantage to the organism and many of them, e. g., the so-called axial hetermorphoses, lead to its destruction or its disruption, but they are no less regulations because of this result.
According to Driesch ('01, p. 92), Regulation ist ein am lebenden Organismus geschender Vorgang oder die Anderung eines solchen Vorgangs, durch welchen oder durch welche eine irgendwie gesetzte Storung seines vorher bestandenen 'normalen' Zustands ganz oder teilweise, direkt oder indirekt, kompensirt und so der 'normale' Zustand oder wenigstens eine Anniiherung an ihn wieder herbeigefiihrt wird."
If we accept this definition, then the processes which do not constitute a return or approach to the previously existing 'normal' condition are not regulations. This normal condition is nothing but the condition which corresponds to a certain complex of external factors or to changes within certain limits. Under changed conditions a new equilibrium, not the old, is established. In short, if we accept such a definition, we not only exclude many processes which are as truly regulatory as any, but we are forced to assume the existence of an entelechy or other similar principle to account for the 'normal' condition and its maintenance.
Regulatory processes are determined in character and direction by the nature of the organism, on the one hand, and the nature and amount of the external change, on the other. Under the given conditions, the organism or part is capable of doing only the one thing; under other conditions, or with a different constitution, the regulation may occur in a different manner and may often lead to a different result. In Planaria, for example, the course and result of regulation differ according to the size of the piece, the region of the body from which it is taken, the temper iture, the nutritive conditions and other factors. To say that the pieces always produce a whole under all these conditions means but little, for the wholes which they produce are not alike. In plants the character of the external change often plays a very large part in deter
REGULATORY PROCESSES IN ORGANISMS 191
mining the character of the regulatory processes. In many cases, however, in both plants and animals, the action of the external factor is so indeterminate, or the external conditions are so complex that equilibration may occur in various ways under what seem to be, but are not actually similar conditions. Thus, as Jennings has pointed out, in the regulation of behavior the disturbance, the stimulus, may merely bring about reactions of an indeterminate character, which sooner or later, in one way or another lead to equilibration. Evidently then the relation between the character of the external change and the character of the regulatory process differs very widely in different cases.
The initiating factor in regulation is the external change, the disturbance of the preexisting condition. This change brings about changes within the organism or the part and these in turn lead to changes in the correlative factors, and so to equilibration or to disruption and death, in case the external change is such that equilibration of the system as a whole is impossible. But so long as the energetic processes of life continue in the system, equilibration of some sort must occur. To return to the analogy of the river, so long as the water flows, equilibration of some sort occurs, whatever the changes and whatever the obstacles. The river may alter its course, it may transform its banks and its channel so that they bear little or no resemblance to those existing before the change, it may divide into a number of streams, each of which pursues its own course, according to the conditions under which it finds itself, and builds up its own structural characteristics. In all cases, however, unless the conditions are such as to stop the flow of the water, equilibration takes place in some manner.
The range of regulatory capacity in the organism represents then merely the range of possibilities within which the flow of the current of energy which constitutes metabolism and which is the essential feature of life, is possible. Within these limits it is absolutely inconceivable that regulation or equilibration should not occur. The nature of the process depends upon the nature of the orgaaism and the conditions which it meets.
Equally inconceivable is the occurrence of regulation as a process of life under conditions which stop the metabolic current.
192 C. M. CHILD
Organisms which meet such conditions are simply eliminated. The survival and elimination of organisms is determined primarily, not by their morphological characteristics, but by their capacity for regulation of one kind or another under the conditions in which they find themselves. To attempt to understand the course of evolution from morphological characteristics alone can only lead to confusion and failure. Only a knowledge of the nature of the metabolic current in organisms and the possibilities of its equilibration under different conditions can lead to a theory of evolution and heredity which will stand.
For example, the evolution of animals and plants, like fevery other evolution, is based primarily upon differences in the metabolic processes. These undoubtedly originated as regulations and as soon as they had arisen, gave different possibilities ot further regulation : in the course of the realization of these different possibilities in accordance with the conditions of existence, animals and plants with their different morphological characteristics have arisen. In each case the visible structure represents merely a partial record of the realized possibilities. All the structural 'adaptations' in both animals and plants are based upon the processes of equilibration of the energy current and must sooner or later be expressed in terms of this current and its environment. They are not the primary and essential features of the organisms, they give us merely an outline, a diagram of the most characteristic activities of the energy current. As the banks and channel of the river, even after the water has ceased to flow, enable us to gain some conception, though a very incomplete one, of what the river has done in the past, so the structure of the organism is merely a rough sketch of what the current of life has done in the "way of deposition, arrangement and removal of materials along its course. Many of the past activities of the current are not distinguishable in the structure because their effects were slight or transitory, or because they have been masked or altered by later activity of a different character. As the river in some process of equilibration, e. g., in a flood, a period of increased energy, may sweep away many of the records of its previous activity, so the
REGULATORY PROCESSES IN ORGANISMS 193
organism, in a period of increased metabolism may remove the structural evidences of past metabolism.
In the present state of our knowledge we should think it absurd to attempt to account for the configuration of the banks and bed of the river without taking into account the action of the current. It would remain a miracle, which we could ascribe to the caprice or other quality of a personal creator, or to some other mysterious natural force. In the same way, when we attempt to interpret the structure of organisms without direct reference at every step to the current of energy of which the structure is evidence, we must neces-' sarily go astray or end in confusion or in the most bizarre hypotheses. We can do as Driesch has done and shift the burden to the shoulders of entelechy, to which we can ascribe such qualities as may please us. Or we can speak of biophores and determinants, pangens, or whatever we please to call them, or we may pin our faith to the visible chromosomes, but these are nothing but creators of a type which appeals to certain minds.
On the other hand, when we take as our starting point the process of metaboHsm, we are proceeding as the physiographer has learned to proceed in his study of rivers. As we learn how metabolism produces structure we shall be able more and more completely to interpret the nature and the past history of the organism from its structure, but at every step we must return to the process, the current, in order to understand, and we can never hope to understand all through structure, simply because structure is an incomplete record. Life is first of all an energy process, a flowing current. All that is relatively stable, all that persists as visible form and structure, represents merely some past action of the current occurring under certain conditions. Almost sixty years ago Huxley said concerning the cells: They are no more the producers of the vital phenomena than the shells scattered along the sea-beach are the instruments by which the gravitative force of the moon acts upon the ocean. Like these, the cells mark only where the vital' tides have been, and how they have acted. "-^ And even yet the truth of these words is not recognized as it should be by biologists.
- British and Foreign Medico-chirurgical Review, vol. 12, p. 314, Oct. 1853.
Cited from Whitman, the inadequacy of the cell-theory, Jour. Morph., vol. 8, 1893.
194 C. M. CHILD
Only when we take into consideration the motive power and the method of its action under the given conditions, can we hope really to advance in our knowledge of how things come to be as they are in the organism, or to determine and predict what they shall be. Man has attained his present position by acquiring knowledge and control of energy in nature. Can he hope to advance in his insight into the problems and his control of the processes of life in any other way?
According to this point of view, life, like every other continuous energetic process, is essentially a series of equilibrations, of regulations. When regulation shall cease, evolution and li'e will also cease. The power of regulation in organisms is nothing unique, but is something which they possess in common with all energetic processes in nature,which continue for any appreciable time. In fact, strictly speaking, all energetic processes in nature are equilibrations.
As was suggested above, the range of regulatory capacity in organisms is undoubtedly due in large measure to the fact that the process of metabolism produces certain colloid substances, among which the proteids and lipoids are the most characteristic. With the first proteid synthesis under certain conditions in nature the processes of regulation of the type which we find in organisms began. Perhaps we may say that life began here also. The reaction which was concerned in the first synthesis must of course have preceded the completed synthesis, but as water apart from the channel which it forms for itself in its environment is not a river, so a given chemical reaction, or a series of reactions, apart from the conditions which it produces where it takes place, is not life. We may say if we please that life began as a chemical reaction, but we must recognize the fact that the occurrence of that reaction produced certain characteristic conditions, which played a part in determining the course and character of further reactions : in short, the reaction determined the existence of structure and the mutual interrelations between structure and function: and finally, with the existence of structure of colloid nature, the possibility of regulation of the organic type also appeared, and regulation began.
REGULATORY PROCESSES IN ORGANISMS 195
My purpose in laying special emphasis upon the point that regulation is an essential characteristic of life and that life must cease when regulation ceases, is merely to show that the extreme forms of regulation, which occur under experimental or accidental conditions, are in no way different from the processes of life apart from experimental or accidental interference. The capacity for regulation is not something secondary or something acquired in the course of evolution, but it is as inseparable from life itself as the power of equilibration from the flow of the river. Not only life but the universe is an unceasing series of regulations. Every experimental investigation performed with living organisms is, so far as it does not lead to the death of the organism, an investigation of organic regulation, and death itself is an equilibration, though of another type.
To set the regulations ofT as a special category of phenomena, occurring only in organisms and of secondary or incidental significance in these, must of necessity lead to conclusions of the same character and value as those which would be reached by one who should attempt to investigate the phenomena of equilibration in the river, without considering the flow or the resistance of its banks and bed. Such a one would doubtless marvel at the wonderful harmony of action displayed by the simultaneous disappearance of a part of the bank and the encroachment of the water upon it, or by the appearance of an island and the division of the river into two channels. He would doubtless call attention to the remarkable fact that both the channel and the river were narrow and deep at some points and broad and shallow at others. He might wonder why stones moved along where the bed was steep and only fine particles where it was nearly horizontal. If he were of an investigating turn of mind, he might throw stones into the river and observe the consequences, or he might dig a ditch and turn part of the water into it. Thus he would observe further remarkable harmonies of action. If he were inclined to look for causes, he would probably conclude that the complex of phenomena was determined and controlled by some mysterious being or principle, which, judging from his own ability to bring about harmony of action between different things in his world, he would
196 C. M. CHILD
conceive to be more or less like himself, though greater, more perfect and more powerful. Doubtless also he would give it a name.
But when once the idea of the flow of the river as a motive power has entered his mind, his whole attitude toward what he has seen is altered. He sees that it is the current which carries partic.es away from the bank or stones and mud along the channel. On the other hand he sees that the banks confine the river, that the island, which it has formed divides it, that it accommodates its form to the ditch which he has dug and at the same time begins to change it. He begins to realize that the remarkable harmonies which he has observed are the result, on the one hand, of the flow of the river, i. e., in i further analysis, of the characteristics of water, and on the other, of the nature of its banks and bed. He will also realize in time, that just as long as the flow continues these harmonies of action will continue to occur. Then he may begin to investigate the characteristics of currents and of water in general, and later we find him devising water-wheels, dams, pumps etc., i. e., bringing about the most various harmonies of action between the flow of water and other phenomena.
His conception of what he saw was at first more or less similar to that of the vitalist concerning organisms and all his investigation could only end in speculation, which did not advance his real knowledge. But when he once began to reahze the action of the current as an energetic and a constructive process, then he saw that the harmonies of action were only apparent, not real, because he was dealing with mutually dependent phenomena rather than with those which were independent and predetermined.
Driesch, for example has maintained in criticism of some of my own earlier statements, that development is for function (Driesch, '95, p. 790) and the same view is apparent in his repeated comparison of the organism to a machine constructed by man. This is as if our hypothetical man should maintain that because he could dig a ditch and turn water into it, therefore the channel of the river must have been constructed by some 'entelechy,' or other principle for the water, and then the water turned in. And more specifically, Driesch's 'proofs' of the autonomy of vital proc
REGULATORY PROCESSES IN ORGANISMS 197
esses, ^ which are based on the phenomena of regulation are not proofs at all, because the 'machine' which he has in mind is comparable to the dead, rather than to the living organism, to the river frozen solid, rather than to the river flowing. Tf we could separate a portion of this frozen river with its channel from the rest it would of course remain what it was, i. e., a part, so long as it remained frozen. But if we divert any sufficient quantity of water from the flowing river it is capable of forming a whole which shows all the essential characteristics of the original river, though not identical with it. In short each flowing river, with its banks and bed is a 'machine' according to Driesch's definition, eine typische chemisch-physikalische Spezifitatskombination" (Driesch, '01, p. 187), and it may become whole when parts are taken from it or when their relative position is changed; moreover, when it is divided, each part may form a whole essentially similar in its processes and structure to the original whole. The existence of such a 'machine' is therefore a sufficient refutation of these 'proofs' of Driesch's. So long as the current flows such regulatory processes are not only possible but necessary, when the conditions arise. Neither the organism nor the river 'remain w^iole' when parts are taken from them, but they become new wholes, which under similar conditions, may become more or less like the original whole (Child, '08b), but which under other conditions, may be different.
Driesch's error is two-fold : although his general definition of a 'machine' is sufficiently broad, his argument in the 'proofs' is based only on a certain type of machine, viz., that constructed by man for function, a type which is wholly passive during its con
- A brief statement of the first two ' proofs' is as follows :
"Erstens: Eine Maschine bleibt nicht dieselbe, wenn man ihr beliebige Telle nimmt oder ihre Telle beliebig verlagert; deshalb kann das sich auf Basis harmonisch-aquipotentieller Systeme abspielende Formbildungsgeschehen kein maschinelles chemisch-physikalisches Geschehen sein.
"Zweitens: Eine nach den drei Dimensionen typisch spezifisch verschiedene Maschine bleibt nicht ganz, wenn sie geteilt wird, deshalb liegt der Geneseaquipotentieller Systeme mit komplexen Potenzen im Bereiche des Formbildungsgeschehen kein maschinelles chemisch-physikalisches Geschehen zu Grunde." Driesch, '03, p. 74.)
198 C. M. CHILD
struction, rather than on a type in which, as in the organism, the structure at each stage is determined by the function and the structure in the preceding stage. The organism is comparable not to the constructed 'machine' alone but to the machine plus the constructing activity, and since Driesch has confined his argument to the type of machine constructed by man for a definite purpose, he is very naturally and logically led to the assumption of a constructor. His 'proofs' are equivalent to the argument that because a ditch built by man for a particular purpose and possessing a specific structure but containing no water does not remain whole or the same when we take away parts of its banks or bottom, therefore the river, as we see it in nature cannot be a physicochemical system.
Similarly in his consideration of the organism he has failed to take account of the constructive activity of the continuous flow of energy in a given environment. The organism is, he says, constructed for function. His position is identical with that of our hypothetical man who concluded that the channel of the river must have been constructed for the water, and like him, Driesch has given his imagined constructor a name, or rather has adopted an old one for it, viz., entelechy. Most of us have concluded from our observations and experiments that the channel of the river is constructed by the activity of the current and we have some rather conclusive evidence upon that point. Before he can hope to see his views accepted, our man must actually prove or make it at least probable that this is not so. The burden of proof lies wholly upon him. And similarly, until Driesch can make it at least probable that the organism is constructed for and not by function, instead of merely assuming this to be the fact, he cannot expect to find wide acceptance for his views. Nowhere in Driesch's work do we find any convincing evidence upon this point: Driesch has simply chosen to assume that it is so. I am of course aware that Driesch regards entelechy as in constant connection with physicochemical factors and as working with these as means. But I see no reason why, if we postulate an entelechy for the organism, we should not at least be consistent and postulate another for the
REGULATORY PROCESSES IN ORGANISMS 199
THE Rl<](iULAT()RV PROC^KSSES
1. The relation between form regulation and functional regulation
A distinction between regulations of form and regulations of function has \'ery commonly been made. As might be expected from his conception of the organism, Driesch ('01) has attempted to draw the line very sharply. But if we adopt the point of view suggested above the distinction becomes apparent rather than real. First of all every regulation in organisms is primarily an energetic process and secondly, it occurs in a certain structure and must affect that structure to a greater or less extent. On the other hand, every change in structure must lead to a regulation of function. Structural and functional regulation are in fact inseparable in organisms. If we go further and interpret structure in terms of the constructive energy, we may say that all regulations are essentially functional, i. e., energetic.
It is sufhciently evident from what has been said, that the flow of energy in the organism is essential for regulation, and that the structure must play a part in determining its character. It should be possible, therefore, to interpret the regulatory processes in terms of the energy current, i. e., metabolism, and the preexisting structure. To refer again to the analogy of the river, both the channel and the current are involved to a greater or less extent in each equilibration in the system. The distinctioa between form regulation and functional regulation is then in part conventional and connected with the separation of morphobgy and physiology from each other, and in part a matter of convenience, since some regulatory processes involve the visible structure to a much greater extent than others. As in the classification of other natural phenomena, we separate for convenience of thought or reference a graded series into a number of (in this case two) different classes.
2. The inducing conditions and the results
As already noted, the first factor in regulation is a change of some sort in the external conditions affecting the system. In the
200 C. M. CHILD
case of a part of an organism this change may be a change in physiological correlation resulting from changes in other parts, however produced. This change, external to the system concerned, produces an internal change of some sort in some part or parts, and this in turn alters the physiological correlation between the components of the system affected. So long as life continues, these correlative changes must result in equilibration in one way or another. The processes of equilibration may be very different in different cases : they may bring about sooner or later a return or approximation to the preexisting condition — according to Driesch, this alone constitutes regulation and only when the preexisting condition was the 'normal' condition. On the other hand, the correlative changes may result in the establishment of, or approach to a condition of equilibrium more or less widely different from the preexisting, and I believe that most, if not all regulations which we usually regard as an approach or return to the preexisting condition actually represent an approach to a new equilibrium, often only slightlj^ different from the old ; it seems at least doubtful whether the organism ever really returns to a preexisting condition in the strict sense.
The new ecjuilibrium may diffei- (quantitatively or qualitatively from the old, or it may even result in the separation of the system into a larger or smaller number of systems, more or less completely isolated from each other. In all these cases, as the rate or character of the metabolic processes are changed, changes in structure as well as in function occur to a greater or less degree. The following suggestions foi- a classification of the regulatory processes are based primarily upon the metabolic processes concerned.
3. A provisional classification of the regulatory processes
a. The two methods of regulation. It is evident that any really analytical classification which is based upon the conception of regulation suggested above must take account, not merely of the visible features, but of the character of the different energetic processes, since regulatioji is, according to this view, essentially a complex of energetic processes in a substratum of a certain
KK(aJLAT<)RY I'UOCESSES IN OlUiANISMK 201
constitution. Such a classification must also be available for both the so-called functional and form regulations, since every regulation probably involves both to some extent.
Driesch's classification of the regulations (Driesch, '01, p. 95 et seq.) is based upon a conception of regulation so widely different from the one developed in this paper that it does not assist us in distinguishing the processes involved. From Driesch's point of view, the physico-chemical processes in regulation are to a large extent of secondary importance and therefore cannot serve as a basis for classification.
At present, however, we are practically unable to attain the proper basis for classification, since our knowledge of the processes involved is incomplete. Nevertheless we can distinguish with more or less certainty the resemblances and differences between different equilibria, and the following suggestions are based upon the character of the equilibria.
We may distinguish two chief type of regulatory processes, first, quantitative equilibrations or compem^alions, and second, qualitative changes in equilibrium or frans/orma/zons. In compensation the rate or intensity of the processes, their continuation in time or their extension in space are concerned: in the transformations their character as energetic processes, i. e., the nature of the chemical reactions and the physical changes. In the compensation the system remains much like that previously existhig, as regards its character and the processes of equilibration are quantitative. In the transformations a new system, qualitatively different from that previously existing, arises as the result of equilibration.
Most regulations, as they occur in nature and experiment, involve both compensations and transformation in various degrees. This is especially true of the regulatory processes which follow the removal of a part. Here some parts of the organism undergo transformation in consequence of altered correlation, while compensatory processes of various kinds are evident, both in the increase in size of the new part and often also in a decrease in old parts.
Moreover our use of these terms will depend upon the particular processes to which we have reference in a given case. The process of compensatory growth, for example, is highly complex in charac
202 C. M. CHILD
ter aad coasists in a variety of both compensatory and transformatory processes, but when an increment in all of these processes occurs, the change is quantitative in character and when it constitutes a process or part of a process of equihbration we are justified in calling it a compensation.
It is of course evident that a classification of regulatory processes must finally become identical with the classification of processes occurring in the organism, for, as I have pointed out above, the regulations are not a peculiar form of organic activity; they represent merely the equilibrations resulting from the existence of physiological correlation between parts. But we shall probably always have occasion to refer to the organism, the system, as a whole undergoing equilibration or, relatively speaking, in equilibrium, consequeatly some means of distinguishing between the different methods of equilibration is useful. This is the chief significance which any classification of the regulations can possess.
h. Regulatory compensation. Several different types of compensation can be distinguished, though they do not of course in most cases exist in nature or even in experiment apart from other processes. The following divisions under this head are suggested :
Incremental compensation: The system shows an increment as compared with that previously existing.
Decremental compensation: the system shows a decrement as compared with that previously existing.
Reversional compensation: an increment or a decremeat in some part of the system, indaced by some external factor is correlatively more or less completely elimiaated and the system approaches its previous condition.
Alterative compensation: an increment or decrement in one part produces change in the opposite direction in another or in others, so that the proportional relations in the system differ from those previously existing.
The first step in all compensations is of course a change in some part {a) induced by some factor external to the system. What particular form of compensation shall occur depends upon the degree of the change in the part and upon the character of the correlation existing between it and other parts (6, c, d-7i). If for
REGULATORY PROCESSES IN ORGANISMS 203
example the part a dominates the other parts, or if the change in a is so great that the correlative factors resulting from it become dominant, then an increment in a may bring about an incremental compensation, a decrement in a a decremental compensation. On the other hand, if the parts b, c, d-n, or certain of them, dominate a, then they may inhibit or reverse the incremental or decremental change in a and reversional compensation results. A.nd finally, alterative compensations occur whenever changes in one part induce correlatively changes in the opposite direction in others.
An incremental compensation occurs when increased metabohsm and growth follow the ingestion of food, a decremental compensation, when decreased activity of a sense organ or a muscle induces a correlative decrease in activity and perhaps atrophy in parts with which it is connected, e. g., the center in the case of the sense organ, the tendon, or even the bone in the case of a muscle. In various temperature regulations in warm-blooded animals we have reversional compensations, and finally, in many cases of regeneration and probably also often in normal development, alterative compensations occur, e. g., when increased growth of one part retards correlatively the growth of another or perhaps induces reduction in it.
The chemical substances which arise in the course of metabolism in certain parts very often produce compensations of various kinds in other parts. A good example is the correlative effect of increase in the carbon dioxide in the blood through the nervous system upon the rate of respiration. The recent work of Bayliss and Starling and others on 'hormones' gives us some insight into various other cases of compensation and other regulatory reactions ; the distribution of nutritive substances in the starving animal and under various other conditions also constitutes compensations of various kinds; the correlative changes in so-called functional structure are in many cases very characteristic compensations.
Incidentally it may, be pointed out that the view of the relation between metabolism and structure suggested above affords a basis for interpretation to a certain extent of the processes of functional hypertrophy and atrophy from disuse.
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
204 C. M. CHILD
We have seen that structure-formation of some sort, i. e., the accumulation of relatively inactive substances in or about the cell is a characteristic feature of metabolism. The close relation between the syntheses and the oxidation processes has been pointed out repeatedly by Loeb as well as by others. The structural substances, when once formed, play only a relatively small part in further metabolism, provided other more active substances, i. e., nutritive materials, are at hand, and provided the general character of metabolism is not changed. Any condition, e. g., the 'functional stimulus' which leads to increased metabolic activity of the particular kind which constitutes what we call the special function of the cell or part leads, when nutritive material is present, to increased accumulation of the inactive substances and hypertrophy is the result. On the other hand, in the absence of the functional stimulus, or when its frequency or intensity is decreased, the use of nutritive material aad the accumulation of structural substance do not occur or are less rapid, and the result is that below a certain level of functional activity the gradual breaking down of the accumulated substance, which is not immediately connected with the special functional activity of the part, exceeds the constructive processes and decrease in size and atrophy occur. The constructive processes continue only, or very largely, in connection with the functional stimulus and, for the addition of new structure nutritive material must be taken in from without, but this functional activity does not under these conditions, increase proportionally the rate of reentrance of the structural substances into metabolism; in fact, if other more active substances are present in sufficient quantity, the structural substances may be spared to a large extent.
Hypertrophy and atrophy are then the result of two different kinds of processes, the one connected with the specialized function of the part in its relation to other parts, the other to a considerable degree independent of this except in starving animals. In its 'functional activity' the part builds structure, but does not destroy it to so great an extent. The destructive process is largely independent of function and goes on more or less continuously. Whether hypertrophy or atrophy shall occur in a given case depends merely
REGULATORY PROCESSES IN ORGANISMS 205
on whether the one or the other of these processes is the more rapid. Hypertrophy is then in no sense a 'regeneration in excess' ; it is merely a direct result of increased metabolic reactions of the kind which constitute or accompany the 'function' of the part concerned. Nevertheless, it is without doubt a compensation. The occurrence of one series of metabolic reactions determines the occurrence of another series according to chemical laws; the one series furnishes energy, the other forms relatively inactive substances, which persist as structure. Closely related to functional hypertrophy is the growth in size of regenerating parts after their formation: here the 'functional stimulus' is the quantitative factor in the correlative influences from other parts. This factor induces a certain rate or frequency of reaction in the small new part, which leads to rapid accumulation of material, i. e., to hypertrophy (Child, '06a, p. 407). But as the structural substance accumulates, the structure itself constitutes an obstacle to metabolism (Child, '116) the rate of hypertrophy decreases and finally equilibrium is attained.
c. Regulatory transformation. The character of many metabolic reactions is more or less definitely known, but the exact relation of the reactions to the production of a particular kind of visible structure is a much more difficult matter to determine. The visible characteristics of organic structure are by no means adequate criteria of the character of the processes involved in its formation. We are not always justified in concluding from the differences in the visible appearance of structures that the processes concerned in their formation are actually different in nature. Great differences in appearance may arise in the same colloid substance in consequence of differences in aggregate condition or phase. But when we find substances of different constitution in different cells or parts, it is evident that processes of different character were concerned in their formation. Consequently we can often determine that a transformation has occurred by the change in the character of the structural substance. Many features of correlative differentiation, whether in ontogeny in nature or under experimental conditions are undoubtedly transformations, e. g., the formation of a bud from a differentiated cell in
206 C. M. CHILD
the plant, in consequence of the removal of other vegetative tips, the formation of a hydranth from cells of the stem of Tubularia, etc.
The difficulty lies in distinguishing qualitative from quantitative regulations. In living organisms the two are evidently very closely associated, and probably in every regulation which we can observe directly both are concerned. And in the final analysis the question of the relation between quality and quantity in general is involved, though this is scarcely a biological problem.
As regards the further classification of the regulatory transformations, I think that at present the most satisfactory basis for classification is the comparison of the new system with that existing before regulation. The following division of this group of regulations is therefore suggested:
Progressive transformation: the regulatory formation of a system possessing a greater degree of complexity, more varied localization of structure and function and consequently more varied correlation than the system existing before regulation.
Regressive transformation: the regulatory formation of a system of simpler character than the preexisting.
Transgressive transformation: the regulatory formation of a system which cannot be distinguished as more or less complex, but merely as different from the preexisting.
Suh a classification serves merely to suggest the various possibilities. As our knowledge of the processes concerned in the changes of the organic system increases, the basis of classification will change. Without doubt many progressive transformations occur in normal development. The adult organism is certainly a more complex and qualitatively different system from the blastula, and we know that correlative factors have been concerned in the changes in many parts. A regressive transformation occurs when a part undergoes dedifferentiation in consequence of altered correlation, as in various cases where cells which give rise to new parts first lose their old differentiation.
The group of transgressive transformations possesses little more than a conventional significance, since it is based upon difference from the normal which is essentially merely the usual
REGULATORY PROCESSES IN ORGANISMS 207
Very probably various 'sports' and mutations can be placed under this head, perhaps also certain of the neoplasms, more specifically the malignant tumors, though this is by no means certain. But whatever the categories which we may establish for the different regulatory processes, the important point is that we should at least make the attempt to find a physico-chemical basis for our analysis. If we do this we cannot separate structure and function since both are merely different aspects of the same process-complex and are dependent upon and determine each other. Doubtless we shall still find it convenient to speak of form regulation as distinguished from functional regulation, but we must remember that the distinction is not a real one and that every regulation in the organism is undoubtedly a regulation of both form and function, of both structure and reactions. Furthermore, we must regard our experiments on regulation as means of analyzing the factors of the process. With the proper care in experiment we can do much toward determining the nature and action of various correlative factors in regulation, and every step in this direction is a step in advance in our knowledge of the system which constitutes the organism,
THE NATURE OF RECONSTITUTION '
1. Restitution or reconstitution f
When an organism 'restores' a missing part or in general when a part of an organism forms a whole, the process seems at first glance to be so obviously a restoration in which something removed is replaced, that the term 'restitution' has found very general favor. Although I have used this term to a large extent, it has always seemed inadequate, for the reason that the process is not simply one of restoration but something more. There is no case of so-called restitution known in which the changes following the removal of a part are limited to the formation of a new similar part. In every case changes of one kind or another, quantitative or qualitative, or both, occur in other parts, sometimes limited chiefly to parts adjoining the part removed, sometimes extending throughout the system. The removal of a part
208 C. M. CHILD
brings about not simply its restitution, but an equilibration extending more or less widely, and strictly speaking, probably throughout the system. The system reconstitutes itself and the whole formed is different from the original quantitatively or qualitatively^ The whole process is a complex of equilibrations, of compensations and transformations, resulting in that which we call a whole, but no two wholes are alike.
If for example, we consider this process in a piece of the Planarian body, we find that it differs in rate and character according to size of the piece, region of the body from which it is taken, temperature and other factors which influence metabolism. The animal formed as the result resembles the original in its general shape and activity, but it is far from being identical with it. It is usually smaller than the original, the pharynx may be in quite different position in the body, and the arrangement, number and form of the intestinal branches differs more or less widely, according to conditions. Moreover, under various conditions, various degrees and kinds of incompleteness appear. Some pieces develop only a single eye, or the eyes are partially fused or otherwise different from those in the original animal, some pieces develop no pharynx and no posterior end, others no head or an imperfect' one, some develop the postpharyngeal intestinal branches more rapidly and more completely, others the prepharyngeal branches, some produce a larger, others a smaller head, some show more 'regeneration,' others more 'redifferentiation' and so on. If we place the different sorts of wholes under closely similar conditions and give them food they become more or less like each other because these conditions bring about further regulations but these regulations do not properly belong to the regulatory process which resulted from the isolation of the part, but are independent of it and are such as were occurring in the original animal during its life. It is probably not too much to say that no two pieces of the Planarian body attain the same condition in the process of regulation. When we say that because some or all of them produce wholes they are all potentially alike, we are simplj- assuming that all wholes must be alike, which is obviously untrue. As each piece is different at the start from the others, so it attains a di^erent
REGULATORY PROCESSES IN ORGANISMS 209
result. Driesch's assumption of 'equipotentiality' of different parts is shown by the facts themselves to be incorrect, and I believe that his 'harmonious-equipotential systems' do not exist in nature, as systems capable of development but only as abstractions of the human mind.
For these reasons the term 'restitution' seems to me to carry with it implications which, when we analyze them, we cannot, in the light of the facts, accept. The piece does most certainly not restore what it lost : it reconstitutes itself into something more or less widely different from that of which it formed a part, and this something often possesses visible structural characteristics which we have come to regard as characteristic of a whole. Seeing these resemblances, we abstract from the differences and say that it is the same as that of which it formed a part.
Closer examination shows us that even visibly it is not the same, but different; moreover, visible characteristics are not the sole criterion of resemblance and difference in organisms. The processes occurring are just as characteristic as the visible structure. I have shown elsewhere (Child, 'lib) that in Planaria the process of form regulation results in a rejuvenation, the pieces after undergoing regulation are physiologically younger than the animals from which they were taken, and the degree of rejuvenation varies with the degree of reconstitutional change. Manifestly then these pieces are not the same after regulation as the wholes of which they formed parts. To say that they are is simply to deny the facts as they stand before us.
In many cases, moreover, the pieces do not produce anything that can be called a whole. Pieces of Planaria may produce double heads or double tails, tailless heads or headless forms. For those who with Driesch regard the formation of a 'whole' as the uniform result of so-called restitution, these cases are difficult to interpret. The process of regulation has apparently followed the wrong track, it has gone astray and so failed of the correct result. But when we take the position that the part, when isolated, undergoes a reconstitution, which differs in its results according to the existing conditions, internal and external, we see that these 'abnormalities' differ from 'normal' results simply because different
210 C. M. CHILD
conditions existed at the beginning or elsewhere in the course of the regulation, and in many cases we can determine what those conditions are. For example, pieces of Planaria which give rise to 'wholes' at a certain temperature and under certain other conditions, can be made to produce headless forms or tailless heads, according to the region of the body from which they are taken, by subjecting them to lower temperatures, by starving the animals before beginning the experiment, by placing the pieces in dilute alcohol or ether, etc. Nothing has gone astray in t'lese cases, there is no error, the same laws have been followed as when 'wholes' are produced, different conditions simply lead to different results. Elsewhere (Child, '10c) I have attempted to analyze some of the conditions which bring about so-called heteromorphosis in Tubularia and other forms, and have shown that they are similar in character to those which bring about asexual reproduction in nature.
There are many cases in which the occurrence of reconstitution as opposed to restitution is so obvious that there can be no questioning it. A piece from the body of Hydra, for example, does not restore the missing parts, but reconstitutes itself into an organism, smaller, simpler, possessing fewer tentacles and undoubtedly physiologically younger than the original animal. In Clavellina also, as Driesch himself has shown (Driesch, '02), the isolated branchial region or a part of it does not replace the missing parts, but undergoes a process of reconstitution. Tn these cases the physiological effect upon the parts remaining of the removal of certain parts is so great that these parts do not retain their original structure, and a dedifferentiation and redifferentiatioQ occurs. But the effects so apparent in these cases are simply more extreme than in cases where only a small part is removed. Przibram ('07) has called attention to a number of cases which show verj" clearly that the removal of a part results, not in the restoration of the original, but in the establishment of a new equilibrium, differing more or less widely from that.
It is obvious that the process of reconstitution is an equilibration and just as obvious that it leads to different results under different internal and external conditions. As different rivers differ
REGULATORY PROCESSES IN ORGANISMS 211
from each other, so may the results of reconstitution, even in different pieces of the same individual, differ from each other. Moreover, as a certain amount of water does not, except under certain conditions, form a river, so does a piece of an organism reconstitute itself to a whole only under certain conditions.
2. The initiating factor in reconstitution
As Driesch has pointed out (Driesch, '01), it is evident that reconstitution occurs as the result of the absence CNichtmehrvorhandensein') of something. What is this something? Is it the structure, the form, or is it activity? In plants it is possible to bring about reconstitution, i. e., the formation of new buds, roots, etc., by inhibiting the metabohc activity of the existing growing regions without their removal, e. g., by enclosing them in plaster or in an atmosphere of hydrogen, or even by applying anesthetics locally between them and the regions concerned. In another paper I have considered numerous cases of this sort and have discussed their significance at length (Child, 'Ua). These facts show very clearly that it is not the form or structure which is involved but a process, an activity, whose effect is transmitted in some way from one part to another. If we stop the metabolism of the one part for a time the effect on the other is the same as if the first part had been removed. These facts alone should be sufficient to prevent us from regarding form regulation as a process distinct from functional regulation. It is the absence of the effect of certain processes in a certain part or in certain parts, which initiates reconstitution. In short, it is the absence or decrease below a certain point of certain physiological correlative factors which were previously present, that initiates reconstitution. In the absence or decreased effectiveness of these correlative factors, the remaining parts, still being subjected to other correlative factors, which may themselves gradually change in consequence of the removal or decreased activity of the part, react in a manner different from their previous reactions, simply because they are under different physiological conditions. Reconstitution is then initiated by a change in physiological correlation. Recently Driesch ('09) has
212 C. M. CHILD
expressed himself in somewhat similar terms, but he finds nevertheless, as already noted above, that the 'Individuahtat der Zuordnung' between agent and effect cannot be accounted for on a physico-chemical basis and therefore regards it as a new 'proof of the autonomy of vital processes.
3. The process of equilibration in reconstitution
The process of equilibration in reconstitution differs in many details in different cases, but it possesses certain more or less characteristic features, and it is desired to call attention briefly to some of these. The change in ph37siological correlation is the internal factor which has disturbed the preexisting condition, whatever that may have been. This change may or may not lead to equilibration of the living organism. If the change be great, if the other parts possess but little capacity for altering their reactions, it may lead to death. On the other hand, it may lead to reconstitution in various ways according to conditions.
Let us consider first the case where a part is removed and is formed again without any great changes in other parts, e. g., the 'regeneration' of the posterior end of Planaria.
In the absence of the correlative factors which originated in the part removed (a), certain regions (6) of the remaining parts {b,c
d n), which were before prevented by these correlative factors
from reacting as their own constitution and the correlative factors from other parts would determine, now begin to react in this manner. In the region adjoining the part removed, i. e., in the cells b,
the correlative factors originating in the parts cd n are more or
less siniilar in their effect to those which affected the part removed (a). So far as they are and remain similar, and so far as the constitution of b permits, this region will be forced by the correlative factors to react more or less in the manner of a, which is no longer present, and b will replace a more or less completely and more or less rapidly, according to conditions in the particular case^ If
' The formation of a new head in Planaria or a new hydranth in Tubularia is a somewhat different process from the formation of the proximal or posterior end. In these forms the anterior or distal region is physiologically dominant over parts
REGULATORY PROCESSES IN ORGANISMS 213
the cells of the region b adjoining the part removed, be chiefly affected correlatively by the removal of a or if they react much more rapidly than others further away, the process of formation of a will be a 'regeneration.' But if parts further away from a are also affected to a considerable extent by the change and are capable of reacting as rapidly or almost as rapidly as b, then they may also take part in the process of replacing a, which then takes on more or less completely the character of a 'redifferentiation.' In some cases we can determine experimentally whether a part shall be formed chiefly by regeneration of redifferentiation. In Planaria, for example, the amount of regeneration, as opposed to redifferentiation, in the formation of a new posterior end increases with increasing distance of the cut surface from the old posterior end. The farther the level of the cut from the old posterior end, the more completely is the development of the new part confined to cells near the cut, and vice versa. The cells near the cut are those which are most affected by the removal of the part a. Even when this part does not develop anew, they react by healing the wound. That is to say, they change their reactions most rapidly of all cells, they lose their old specification, they become capable of a more rapid metabolism (Child, 'lib) and being subjected to the correlative factors of the parts cd n, they begin to develop into something more or less like a in advance of other parts. The processes in these cells establish certain correlative factors which determine that the cells farther away from the cut shall remain as they are or take other forms of reaction.
But if we decrease the rate of metabolism in Planaria by extreme starvation or by the use of anesthetics, then parts which under the
posterior or proximal to it and controls their development directly or indirectly. Briefly stated, the regulatory formation of a dominant part is a reconstitution resulting primarily from isolation, while for the formation of a subordinate part correlation with other parts of the original system is necessary. For example, a piece of the tubularian stem may reconstitute itself into a hydranth without any other parts (Child, '07a, b, c), or a piece of Planaria into a head without other parts but a tubularian stem or stolon or a planarian tail is never formed except in connection with a more distal or anterior region of the original organism. The question of the dominance and subordination of parts and its significance will be discussed more fully elsewhere.
214 C. M. CHILD
usual conditions are formed chiefly by regeneration, e. g., the head, may be formed largely by redifferentiation (Child, '10a). T.iis means simply that the cells near the cut do not react so rapidly s under the usual conditions, so that other cells further away have ime to change their reactions and take part ia the process, while ordinarily they would be prevented from doing this by the correlative factors arising from the activity' in the cells near the cut. These instances are merely special cases under the general rule that the less rapidly the missing part is replaced, the more extensive are the changes in the remaining parts, so far as their constitution permits change.
Regeneration and redifferentiation in their extreme forms represent the extreme terms of a graded series, of which all terms are essentially of the same physiological character, i. e., all consist in a change in reaction in consequence of a change in physiological correlation. The designations 'regeneration' and 'redifferentiation' serve merely as convenient descriptions of the visible phenomena.
Where all the cells of the remaining parts are so sensitive to the absence of the correlative factors originating in the part removed that they cannot maintain themselves after its removal, the old structure of all the remaining parts may disappear to a greater or less extent, i. e., a 'dedifferentiation' occurs, as, for example in the isolated pieces of the branchial region of Clavellina. In the different cells of the mass the metabolic processes become less specified and in this respect it approaches the 'embryonic' condition, and the correlative factors in the mass approach those existing in the embryo. But during this process some of the cells have been subjected to correlative factors more or less similar to those to which the part removed was subjected at some stage of development, consequently these become in some degree the physiological representatives of that part. In short the system becomes physiologically a whole, but in consequence of the rapid dediffereatiation of the old parts it corresponds to a whole in a relatively early stage of development. From this condition renewed differentiation as a whole results necessarily from continued metaboUsm,
REGULATORY PROCESSES IN ORGANISMS 215
i. e., continued life. The new whole is, however, different from the old in size, physiological conditions, number and proportion of various parts, etc.
In the process of regeneration in the stricter sense the new part is usually at first smaU and increases rapidly in size. I believe that this growth in size is essentially similar to the functional hypertrophy of organs. The part which was removed possesses a certain size in relation to other parts, because its size was determined chiefly by correlative factors. Just so far as the new developing part is subjected to similar correlative factors, it will tend to attain the same size as the part removed. Consequently it does not always attain the same size with respect to other parts. In Planaria the relative size of the new head differs according to the region of the body from which the piece is taken, to the nutritive condition and various other factors. The process of reconstitution ceases when a certain stage, differing under different conditions, is attained. This stage represents an equilibrium of physiological correlation, i. e., of interaction between the parts; it is primarily a dynamic equilibrium, a proportionalit}' of processes, not of form. We can alter this condition of equilibrium experimentally by food, by starvation, by temperature, and in short by all factors which affect the processes.
In various papers Driesch has distinguished a number of different forms of reconstitution (restitution). His distmctions are based primarily upon differences in the visible phenomena of development or dedifferentiation and for him the chief interest lies in the recognition of the different forms, rather than in the attempt to determine how they differ from each other physiologically, since from his point of view the physiological factors are in many cases only 'means' which the enfelechy employs. It is impossible to consider here these various forms of reconstitution, and since my point of view is so widely different from that of Driesch such a consideration would show merely that his basis of distinction could not be accepted for purposes of a physiological analysis. While it is convenient to distinguish different forms or methods of reconstitution, I believe that it is much more important to resolve the phenomena into processes.
216 C. M. CHILD
Ij.. The complexity of reconstitution
The process of reconstitution is not a simple process which cannot be analyzed, but rather an exceedingly complex one; in fact its complexity is of the same order and character as that of development. It consists of a series of compensations and transformations in different parts of the system. It is only when we take into account the complexity of physiological correlation between parts and the almost infinite possibility of change and variety in this correlation that we have any hope of gaining an insight into the complex series of events. The specificity of correlation and reaction does not, as Driesch apparently beheves' (Driesch, '09) constitute a physico-chemically insoluble problem except when we follow Driesch in ignoring the energy current as an equilibrating factor in the organism and as the efficient factor in construction of the visible and tangible characteristics. The energy current performs its work under specific conditions in each case and leads to a specific result. As soon as a specific condition arises in any part of the system, from whatever cause, it determines other specific conditions in at least certain other parts. From the experiments on Planaria it is perfectly apparent that the cells at every level of the body posterior to the ganglia, are capable under certain conditions of developing into a head, but under the usual conditions they are prevented from doing this because the correlative factors arising from the presence, ^. e., the activities, of a head determine their activities in another direction. As soon as the old head is eliminated from the system, those cells, which in consequence of their past correlation are most similar to it, begin at once to form a new head, provided the piece is not too small and as soon as this occurs it determines correlatively a variety of reactions in other cells. The same may be said of the reconstistitution of any part. The place where a particular part shall arise is determined by constitutional and correlative factors in the existing system — so far of course as external factors are not concerned — and as soon as one such place is determined, it determines others and so on. 'Ihus any case of reconstitution consists of a series of regulations, each of which determines others. This
REGULATORY PROCESSES IN ORGANISMS 217
is shown very clearly by the fact that isolation of a part from certain others during the course of reconstitution may alter the course of the process in the part, according to the degree and character of the isolation, i. e., according to the correlative factors which are eliminated. By means of experiments of this kind it is possible even now to determine the action of various correlative factors in different stages of the process of reconstitution.
5. The limits of reconstitution
In every case the reconstitutional processes are limited and determined by existing conditions as the river is limited and defined by its banks and channel which its own activity has constructed in the environment through which it flows. It is not true without qualification that any part of certain organisms is capable of giving rise to any part. Driesch's often repeated statement to this effect requires modification and limitation. The part is at most only provisionally capable of giving rise to any part; in other words, only when it constitutes a component of a system possessing certain characteristics, i. e., only when it is subjected to or isolated from certain correlative factors. This is apparent from every recorded series of observations on reconstitution, except perhaps the most superficial. In some cases the system may consist of a single cell, in others of a large number of pells, but the fact remains the same. The power of reconstitution is limited, not unlimited. As I hope to show elsewhere for Planaria, and as I have shown for Tubularia (Child, '07a, '07b, '07c), the investigation of these limitations is of the greatest importance in throwing light upon the nature of the reconstitutional processes. When we find that the removal of a certain part, or even a certain amount of material, determines a different result from the removal of another part or a larger or smaller amount of material, ■ we are forced to the conclusion that the part or the material removed has some connection with the character, place or other factors in the result, and furthermore, when w^e find that inhibition of the metabolic processes or certain of them in the part or the material is as effective in certain
218 C. M. CHILD
respects as the removal of the part or the material, we have attained a basis for investigation and analysis which is proof against such assumptions as those which Driesch has made, e. g., concerning the nature of the 'harmonious-equipotential system.' For Driesch the limitations of the reconstitutional processes appear to be of secondary importance, but I believe that any one who will investigate and analyze these limitations at all thoroughly will find that they are not only essential features of the regulatory processes, but that they afford us one of the best means of gaining some insight into their nature. As water does Qot constitute a river, except under certain limiting conditions, so certain substances or processes do not constitute an organism or even life except under certain limitations. The water contains the potentialities for giving rise to any kiad of a river, as well as other specific 'machines,' but none of these exist until the specific hmitations are present. The case is essentially similar as regards the organism. The specific 'machine' exists only so far as the limitations exist. And the investigation of the limitations of reconstitution affords at present one of the best methods, if not the best, for demonstrating this to be a fact.
REPRODUCTION IN GENERAL A8 A FORM OF RECONSTITUTION
In another paper (Child, '11a) I have discussed at length the significance of physiological isolation of parts as a factor in reproduction. I have shown that certain degrees and kinds of physiological isolation of parts may arise as the result, first, of an increase in size; second, of decrease in correlative control or physiological dominance of a part in consequence of decreased activity in it; third, of decreased conductivity^ or transmissibility of correlative processes, agents or conditions; fourth, of decreased receptivity, sensitiveness or irritability of certain parts to the correlative factors originating in other parts. Furthermore, we know from experiment, as I have shown, that in a considerable number of cases physiological isolation of parts serves as well as physical isolation by section to bring about reconstitution; and if it were possible to perform the experiment, it is practically certain that we should find the same to be true for many other cases.
REGULATORY PROCESSES IN ORGANISMS 219
In the paper just referred to I have also attempted to show that at least a great variety of natural and experimental forms of reproduction, reduplication of parts, etc. are essentially processes of reconstitution following physiological isolation of parts. The chief difference between them and the cases of reconstitution following experimental section is, first, that the isolation of the part or parts is brought about within the organism physiologic ally and not by the crude method of cutting the organism into pieces; and second, that this isolation is usually partial at first and differs in degree and kind in different cases.
And finally, I have called attention to certain evidence in support of the view that the formation of sex cells and the development of organisms from them are processes not fundamentally different from other forms of reproduction, i. e., that the sex cells are first physiologically parts of the organism like other organs, and that the development of a new organism from them is initiated by changes similar in character to those which occur in other parts capable of reconstitution, when they are physiologically or physically isolated (Child, '10b, '10c).
The evidence bearing upon the first point is briefly as follows; first, the sex cells always arise in, or attain by migration particular regions of the body in a particular organism, therefore, their physiological correlation with other parts cannot be purely nutritive in character, for if it were, there is no reason why they should not take the most various positions in the same species. Second, they undergo characteristic differentiations during the life of the individual, as do other organs and these differentiations begin at a certain stage of development of the organism, i. e., at or near the end of the period of vegetative growth. This cannot be accounted for by quantitative differences in nutrition at different stages, because the growth of the primitive germ cells in earlier stages often requires very large amounts of nutritive material. If this development is predetermined, then physiological correlation between the germ cells and other parts must have existed at some earlier stage, or else we are forced to a hypothesis of preestablished harmony, which amounts to some form of vitalism.
Moreover, in organisms, which show both asexual and sexual
220 C. M. CHILD
reproduction in the same individual, the asexual reproduction occurs earlier in the life cycle than the sexual, and in organisms which produce naturally both parthenogenetic and non-parthenogenetic eggs, the parthenogenetic eggs appear earlier than the non-parthenogenetic. In both these cases the earlier product is usually capable of a greater degree of regulation when it is isolated from the parent body, than the later; in other words, both the non-parthenogenetic egg and the sperm appear from their behavior to be more Highly specified or differentiated than the asexual or parthenogenetic reproductive elements. There is then considerable evidence in support of the view that the history of the germ cells, like that of other organs, is in part the result of physiological correlation.
As regards the 'stimulus to development,' I have shown by experiment (Child, 'lib) that in Planaria the process of reconstitutution after physical isolation as well as extreme starvation followed by feeding, accomplish rejuvenation and that it is highly probable that various other factors bring about similar changes. And finally, I have considered the facts which indicate that the process of fertilization and the conditions inducing artificial parthenogenesis produce changes in the egg similar in character to the rejuvenation occurring in the other cases.
From this point of view, the stimulus to development of the egg is essentially a process or the beginning of a process of reconstitution and so is similar in its physiological effect to the factors initiating the various processes of asexual reproduction. Experimental reconstitution following section is then merely a special case of reproduction occurring under certain conditions, or we may say just as correctly that each form of reproduction in nature or experiment is a special case of reconstitution occurring under certain special conditions.
If this view be correct, then the fundamental problems of development and heredity are before us in every case where a physically or physiologically isolated part of an organism produces a new organism, just as truly as they are in sexual reproduction. In fact sexual reproduction constitutes the most complex case of all, but I am convinced that a recognition of its essential similarity to the
REGULATORY PROCESSES IN ORGANISMS 221
processes following experimental section and the physiological solation of parts is of the greatest significance for our conception and solution of the problems of inheritance and development.
CONCLUSION
It is sufficiently evident from what has been said that I consider the phenomena of regulation in organisms as congtitating the essential chacteristic of life as a coatinuing process. I agree with Jennings that the problem of regulation is the fundamental problem of life. All of our experimental investigations on living organisms are directly concerned with the problem of regulation in one way or another. Tn fact there are only two possible methods of investigating and analyzing the phenomena of life : one is concerned with regulation in the living organism, the other with the observation and analysis of the results of stopping the life-processes at this or that particular point, under these or those particular conditions. In the one case we observe and control the process in its action, in the other we seek to determine the effects of its past action. As we can watch the river at work and investigate the processes of equilibration resulting from alteration of its flow in one way or another, so we can investigate the living organism. And as we can stop the flow of the stream or divert it into other channels and determine something of what it has done along its course up to a certain time by examination of its channel, so from the dead organism, we can determine something of its past activity.
But the conclusions drawn from the examination of the channel of the 'dead' river are only fragmentary at best. Only by observing and controlling the river in action is it possible to acquire any adequate conception of what it really is. And so, I believe, with regard to the organism : the living organism will teach us more than the dead one though we must work with both. And when we work with the living organism we come at once face to face with the problem of equilibration, of regulation. And finally, I believe that the further our knowledge of the processes of equilibration, in the organism advances, the greater will be the difficulty of finding an adequate foundation in biolog}'^ for vitalistic or dualistic hypotheses.
222
C. M. CHILD
BIBLIOGRAPHY
Child, C. M. 1906a Contributions toward a theory of regulation. I. The significance of the different methods of regulation in Turbellaria. Arch, f. Entwickelungsmech. Bd. 20, H. 3,
19066 Some considerations regarding so-called formative substances. Biol. Bull. vol. 11, no. 4 .
1907a An analysis of form regulation in Tubularia. I. Stolon formation and polarity. Arch, f . Entwickelungsmech. Bd. 23, H. 3. 1907b An analysis etc. IV. Regional and polar differences in the time of hydranth-formation as a special case of regulation in a complex system. Arch. f. Entwickelungsmech. Bd. 24, H. 1. 1907c An analysis etc. V. Regulation in short pieces. Arch f. Entwickelungsmech. Bd. 24. H. 2.
1908a The physiological basis of restitution of lost parts. Jour. Exp. Zool., vol. 5, no. 4.
1908b Driesch's harmonic-equipotential systems in form regulation. Biol. Centralbl. Bd. 28, nos. 18 und 19.
1910a Analysis of form regulation with the aid of anesthetics. Biol. Bull., vol. 18, no. 4.
1911b A study of senescence and rejuvenescence, based on experiments with Planaria. Arch. f. Entwickelungsmech. Bd. 31, H. 4. 1911a Die physiologische Isolation von Teilen des Organismus, Vortr. u. Aufs. ti. Entwickelungsmech. H. 11.
Driesch, H. 1901 Die organischen Regulationen. Leipzig.
1902 tJber ein neues harmonisch-aquipotentielles System und iiber solche Systeme iiberhaupt. Studien uber das Regulationsvermogen der Organismen. 6. Die Restitutionen der Clavellina lepadiformis. Arch, f. Entwickelungsmech. Bd. 14, H. 1 u. 2.
1903 Die 'Seele' als elementarer Naturfaktor. Leipzig.
1905 Die Entwickelungsphysiologie von 1902-1905. Ergebnisse der Anat u. Entwickelungsgesch. Bd. 14 (1904).
1908 The science and philosophy of the organism. Vol.1. London.
1909 Der Restitutionsreiz. Vortr. u. Aufs. ii. Entwickelungsmech. H vii.
Holmes, S. J. 1904 The problem of form regulation. Arch. f. Entwickelungsmech. Bd. 17, H. 2 u. 3.
1907 Regeneration as functional adjustment. Jour. Exp. Zool. vol. 4, no. 3.
Jennings, H. S. 1906 Behavior of the lower organisms. New York.
KoRSCHELT, E. 1907 Regeneration und Transplantation. Jena.
Morgan, T. H. 1907 Regeneration. Ubersetzt von M. Moszkowski. Leipzig.
Przibram, H. 07 Equilibrium of animal form. Jour. Exp. Zool., vol. 5, no. 2. 1909. Experimental-Zoologie. 2. Regeneration. Leipzig u. Wien.
RiGNANO, E. 1907. Die funktionelle Anpassung und Paulys psychophysische Teleologie. Riv. di Scienza, vol. 2.
Roux, W. 1895. Gesammelte Abhandlungen fiber Entwickelungsmechanik der Organismen. Bd. 1, u. 11. Leipzig.
PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM
LORANDE LOSS WOODRUFF
Fruyn the Sheffield Biological Laboratory, Yale University
ONE FIGURE
Leeuwenhoek in 1677 described^ some little animals longer than an oval" which he had discovered two years previously, and there is some reason to believe that this is the first published record of an organism belonging to the genus Paramaecium. The name Paramaecium, however, was first employed by HilP to designate certain small organisms which were more or less oblong, in contrast to others which were round or decidedly vermiform, and either the present species aurelia or caudatum is probably the animal which he designated as 'Paramaecium species 3.'
Although Hill was the first to attempt to apply scientific names to microscopic animals, it remained for O. F. Miiller to give a general classification of these forms, and to apply the Linnean nomenclature. He began this work on the infusoria as a section of a treatise entitled, Vermium terrestrium et fluviatilium Historia, which appeared in two volumes in 1773. Unfortunately he did not live to see the publication of his special work, Animalcula Infusoria fluviatilia et marina, 1786, which was edited by his friend, O. Fabricius. Miiller described a Paramaecium and applied the specific name aurelia in the former of these works. In the latter work he described and figured^ Paramaecium aurelia
^Philosophical transactions, London, 11, 133. 1677. ^History of animals, 3, 1751. 'Plate 12, figs. 1-14.
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
223
224 LORANDE LOSS WOODRUFF
together with several other forms, which at the present time are assigned to other genera. Miiller's description is as follows*: —
Paramaecium. Vermis inconspicuus, simplex, pellucidus, membranaceus, oblongus.
Paramaecium aurelia. Paramaecium compressum, versus antica plicatum, postice acutum.
Thus the organism described by Baker^ as Animalcules in pepper water, first sort/' by Joblot^ as Chausson, by Ellis^ as Volvox terebella, etc., received the name which, in spite of various vicissitudes, has come down to the present time.
The next great student of the lower organisms, C. G. Ehrenberg, in the first two of his treatises,* described several species of Paramaecium, and one of these is Paramaecium aurelia. In his third treatise^ he described still another species which he named Paramaecium caudatum.^" Five years later, in 1838, Ehrenberg brought out his monumental monograph. Die Infusionsthierchen als vollkommene Organismen, and in this work he described these two species as follows :^^
Paramecium Aurelia, Pantoffelthierchen.
P. corpore cylindrico, subclavato, antica parte pauUo tenuiore, plica longitudinali obliqua in os multum recedens exeunte, utrinque obtuso.
Paramecium caudatum, geschwanztes Pantoffelthierchen.
P. corpore fusiformi, antica parte obtusiore, postica magis attenuata.
Thus Ehrenberg described, on the basis of shape and size, the two common forms of colorless paramaecia which appear in
- Page 86.
- The microscope made easy, London, 1742. 3rd ed., 1744, p. 72, PI. 7, fig. 1.
Observat. fait, avec le microscope, Paris, 1754.
Observations on a particular manner of increase in the Animalcula of vegetable infusions, etc. Phil. Trans., London, 1769.
^Abhandl. der Akademie d. Wissensch. zu Berlin, 1830, 1831.
"Ibid, 1833.
lOEhrenberg notes that Herrmann (Naturforscher, 1784) applied the name caudatum to a form which was probably a species of Amphileptus; also Schrank (Fauna boica, 1803) used the same name.
iiPp. 350-352. PI. 39, figs. 6, 7.
PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM 225
modern systematic works as P. aurelia O. F. M. and P. caudatum Ehrbg.
Dujardin, in 1841, in his treatise on the Infusoria '^ recognized but two species of Paramaecium as follows :
Paramecie Aurelie. — Paramecium aurelia.
Corps ovale oblong, arrondi ou obtus aux deux extremit^s, plus large en arriere. — Long de 0, 18 a 0, 25.
Paramecie a queue. — Paramecium caudatum.
Corps fusiforme, obtus ou arrondi en avant, aminci en arriere.— Long de 0, 22.
His figures of the two species show clearly the characteristic form which he considered diagnostic.
Various investigators, including Stein, and Claparede and Lachmann, questioned the justification of considering these two forms as distinct species, basing their opinions, as had Ehrenberg and Dujardin, solely on external characters, and they united these two forms under one species, and applied Miiller's original name, P. aurelia. This union of aurelia and caudatum into one species was accepted by all the subsequent students of Paramaecium, e.g., Balbiani, Blitschli, Engelmann, Gruber and Kolliker and consequently all the early literature on the conjugation of this infusorian, refers to the organism as P. aurelia, although it had but a single micronucleus.
Maupas, in 1883, in his studies on the ciliates,i^ wrote:- —
Tous les auteurs jusqu'ici ont decrit Paramecium aurelia comme ne poss^dant jamais qu'un nucleole d'assez grande taille et mesurant de 0mm,005 h. 0mm,008. C'est en effet la forme que Von rencontre la plus frequemment. Mais j'ai observe aussi de nombreux individus pourvus de deux nucl^oles plus petit s et de structure diff^rente de la piecedente. lis etaient de forme spherique et composes d'un corpuscule central opaque vivement color^ par les teintures et ne mesurant que 0mm,003; enveloppe d'une couche corticale mesurant en diametre 0mm,005, claire et ne se colorant pas.
i^Histoire naturelle des Zoophytes. Infusoires, etc. Paris, 1841. Pp. 481-483, PI. 8, figs. 5, 6, 7.
"Contributions a I'etude morphologique et anatomique des Infusoires cilies, Arch, de zool. exp. et gen., (2), I, 1883, p. 660.
226 LOKANDE LOSS WOODRUFF
Thus Maupas tacitly accepted the current view that there was one large species of Paramaecium, but observed, for the first time, that certain paramaecia have a different nuclear apparatus from that previously described. This author, however, in 1888, stated that in his earlier work he, as all his immediate predecessors, had confused two species, and he wrote^^ as follows :
Ces deux formes de micronucleus constituent le caractere distinctif le plus important entre les deux esp^ces de Parameeies. La premiere forme appartient toujours et uniquement au P. caudatum, la seconde, egalement toujours et uniquement, au P. aurelia.
Pour Ehrenberg et Dujardin, P. caudatum se distingue par un corps allonge, fusiforme, obtus en avant, aminci en arriere: P. aurelia par un corps plus large, presque ovale, obtus aux deux extremites. Ces differences de contour general, tout en ^tant reelles, ne sont pas absolument rigoureuses; car, si on ne trouve jamais de Paramecie k un seul micronucleus affectant la formed trapue obtuse, il n'est pas tr^s rare d'en rencontrer a deux micronucleus, ayant pris la forme allongee k queue. Dans ce dernier cas, il est impossible de savoir a quelle esp^ce on a affaire, sans une preparation permettant de voir les micronucleus. Ce charact^re distinctif, bas^ sur le contour general, n'a done qu'une valeur relative. II est cependant bon d'en tenir compte; car lorsqu'on s'est exerce k bien distinguer les deux especes, il suffit presque toujours et trompe rarement.
Le P. caudatum paralt avoir une taille un peu plus grande que celle du P. aurelia. Ainsi, j'ai mesur^ des premiers depuis 120 jusqu'a 325 /x, tandis que les seconds ont varie seulement entre 70 et 290 /i. En outre, P. caudatum se conjugue avec une taille variant entre 125 k 220 n, et P. aurelia entre 75 a 145 fi. Pendant la conjugaison, le deroulement rubanaire, preparant la fragmentation du nucleus, s'effectue chez le P. aurelia, des le stade D, tandis que chez le P. caudatum il ne commence que vers le milieu du stade G. Chez cette derniere espece, le nucleus mixte de copulation donne naissance finalem'ent a huit corpuscules, chez P. aurelia il n'en produit que quatre; il en resulte que chez celle-ci I'^tat normal se trouve r^tabh des la premiere bipartition qui suit la conjugaison, et chez P. caudatum seulement apres la seconde.
Toutes ces differences anatomiques et physiologiques me paraissent plus que suffisantes pour justifier la distinction des deux especes. II
"Sur la multiplication des Infusoires cilies, Arch, de zool. exp. et gen., (2), 4, 1888, pp. 231-235.
PARAMAECIUM AURELIA AND PARAMAECIUM CAU DATUM 227
est fort possible que Claparede et Lachmann aieiit eu raison, en considerant la forme caudatwn comme plus typique que la forme aurelia. Si, en effet, on examine avec soin les dessins de O. — F. Miiller, on penche k croire que le vieux micrographe a vu et figure la premiere seulement. En se conformant strictement au principe de la loi de priorite, ce serait done le nom aurelia, donne par Miiller, qui devrait etre conserve a la forme fuselee. Mais, d'un autre cote, Ehrenberg et Dujardin ont distingue ce type et I'ont decomme caudatum. Si nous lui conservons la vieille denomination aurelia, il devient impossible de transmettre le qualificatif caudatum a la forme qui, le plus souvent, est obtuse k ses deux extremites. II faudrait alors creer un nouveau nom. Je crois plus simple de conserver les denominations d'Ehrenberg.
Since 1889, when Maupas^^ and Hertwig^, in studies on conjugation added further evidence for the distinction of the two forms, they have been generally accepted as 'good' species. Calkins, however, again raised the question in 1906: I personally believe that the slight differences that distinguish the two types of Paramecium are not of specific value, and hold that P. caudatum should be regarded as a mere variant of P. aurelia. "^^ He based this view chiefly on the following observations. One of a pair of ex-conjugants of P. caudatum, which he was studying by his well-known accurate culture methods, reorganized as P. caudatum and the other as P. aurelia, i.e., the latter had two small micrpnuclei, instead of one, and remained in this condition for about forty-five generations in pedigree culture, and then reverted to the caudatum type with one large micronucleus. While the aurelia phase existed, the rate of«division was comparatively slow, and when the caudatum phase was reassumed the rate of division immediately increased considerably. Calkins also considered the relative size of the two forms, and the conjugation phenomena as described by Maupas and Hertwig, and concluded that these are not of such a character as to warrant their being considered diagnostic.
i*Le rajeunissement karyogamique chez les cilies, Arch, de zool. exp. et gen., (2), 7, 1889.
i^Ueber die Konjugation der Infusorien, Abh. kgl. bayr. Akad. d. Wiss. Miinchen, 2, CI. 17, 1889.
"Paramecium aurelia and Paramecium caudatum. Studies by the pupils of W. T. Sedgwick, 1906.
228 LORANDE LOSS WOODRUFF
Jennings, in his studies on heredity in Paramaecium/^ showed that he could readily isolate a considerable number of pure lines from a wild culture, and that these pure lines breed true, i.e., there exist inherent hereditary differences in size, persisting when all other conditions remain the same. These different lines fall usually into two main groups, one group having a mean length greater than 170^, and the other having a mean length less than 140ai. But he was able finally to isolate a line intermediate in size, and thus to bridge over the gap. As Jennings points out, even if it were not possible to isolate a strain of intermediate size between the two large groups, this would not give a basis for distinguishing two species. However, he states: I may be permitted to add to the precise data thus far given a personal impression or surmise. Though, as I -have shown, intermediate lines occur, I believe that it will be found that most Paramecia can be placed in one of the two groups that we have called ' caudatum' and 'aurelia'. In other words, if my impression is correct, most lines will have a mean length either below 145 microns or above 170 microns; rarely will lines be found whose mean falls between these values. Such at least has been my experience in a large amount of work. Furthermore, I am inclined to believe that those belonging to the smaller group (mean length below 145 microns) will be found to have as a rule two micronuclei; those belonging to the large group but one micronucleus. This matter is worthy of special examination."
Jennings and Hargitt in 1909 made this examination and in a preliminary communicatioif stated^* that two sets of races could be distinguished, one set having two micronuclei, the other but one. The races with two micronuclei were all smaller than those with one. The larger races together thus correspond with what had before been described as P. caudatum, the smaller races with P. aureha. The two differ also in the size, position and
i^Heredity, variation and evolution in Protozoa. II. Heredity and variation of size and form in Paramecium, with studies of growth, environmental action and selection, Proc. Amer. Philosophical Society, 47, no. 190, 1908.
Characteristics of the diverse races of Paramecium, Proc. Amer. Soc. Zoologists, 1909 meeting, in Science, March 25, 1910.
PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM 229
staining relations of the micronuclei, in ways that correspond to the descriptions of Hertwig and Maupas. But in rare cases specimens of the caudatum races have two micronuclei, those of aurelia races but one, thus confirming the observation of Calkins on this point."
In accordance with the conclusions of Calkins, I have used the specific name aurelia to include both the aurelia and caudaturn forms; but my extended study of Paramaecia cultures has demonstrated that these two forms are remarkably constant, and I am inclined to the view that they are distinct species, in the sense in which this term is generally used in biological work. The data on which I base this conclusion are chiefly as follows: the pedigree culture of P. aurelia which I have had under daily observation for (so far) more than three and one half years, during which time more than 2100 generations have been attained, has bred practically true to the aurelia type as described by Maupas in the passage quoted. The pedigree culture of P. caudatum which I have carried for nearly seven months, and which has attained more then 300 generations up to the present time, has bred practically true to the caudatum type as described by that author.
The pedigree culture of P. aurelia was started on May 1, 1907, with a 'wild' individual which was found in a laboratory aquarium, and was carried on at Williams College during May and June, 1907; at the Woods Hole Marine Biological Laboratory during parts of the summers of 1907 through 1910; and at Yale University during the academic years from 1907 to the present time, November 30, 1910. The pedigree culture of P. caudatum was started on May 14, 1910, with a 'wild' individual collected from a pond at New Haven, Conn., and was carried on at Yale University except for a period of a few weeks in the summer when it was taken to the Woods Hole Laboratory.
The original specimen of each culture was placed in about five drops of culture fluid on a glass slide having a central ground concavity, and when the animal had divided twice, producing four individuals, each of these was isolated on a separate slide to form the four lines of the respective cultures. The pedigree cultures have been maintained by the isolation of a specimen from
230 LORANDE LOSS WOODRUFF
each of these lines practically every day up to the present time, thus precluding the possibihty of conjugation taking place between sister cells. The number of divisions of each line has been recorded daily at the time of isolation and the average rate of these four lines has been again averaged for ten-day periods (cf. fig. 1). The culture medium has consisted of materials collected practically at random from laboratory aquaria, hay infusions, ponds, etc. The infusions were thoroughly boiled to prevent the contamination of the pure lines of the pedigree cultures by 'wild' individuals. Permanent preparations have been preserved from time to time for the study of the cytological changes during the life history.
In the light of this experience with cultures I shall consider each of the characters emphasized by Maupas.
Shape. The general shape of the aurelia and caudatum forms is, in nearly all specimens, quite distinctive; aurelia is slightly more broad at the posterior than at the anterior end, while caudatum, as the name implies, is quite pointed at the posterior end as compared with the anterior end. The posterior end, in the specimens in my pure culture, is markedly pointed, and being free from endoplasmic inclusions, appears transparent and clearly delineated even under a lens with a magnification of ten diameters. I have been accustomed to allow stock material from my pedigree aurelia culture to multiply in large flasks of hay infusion, for various experiments on conjugation, etc. Frequently I have used this material for ni}^ elementary class in biology and I have found that even the novice has called attention to the fact that the shape of the ends was reversed as compared with the figure of caudatum in the text-book. McClendon, however, stated that in his study of aurelia and caudatum he found no characters of outward form" which were diagnostic.
Changes in the vitality of my pedigree lines never have been very marked, and consequently I have not had organisms, in the direct lines of my pedigree cultures, representing physiological extremes to compare. Numerous experiments, however, have been made with 'stock' material left over after the daily isolations of the pure lines, which have clearly shown that, for
PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM 231
example, even when the aurelia and caudatum cultures are subjected to unfavorable environmental conditions, as, for example, scarcity of food, the very great majority of individuals retain the shape which is characteristic of the race.
Size. As has frequently been pointed out, size alone is an entirely inadequate character on which to base species. It is significant, however, I believe, that during the long life of my pure strains, I have never observed the relative size of the individuals of the aurelia and caudatum forms, when bred under identical conditions, to change greatly during any single period. Experiments have shown that even when the two forms have been bred under diverse conditions, for example, aurelia in a medium rich in food and caudatum in a medium with a very small amount of bacterial growth, the size of the caudatum form always has remained sufficiently great to render it distinguishable from the aurelia form. On the basis of size alone, then, it has been possible, with great accuracy, to separate the two forms when mingled together. It is probable, of course, that I began my pedigree cultures with very typical specimens of the aurelia^" and caudatum groups as described by Jennings. If such be the case, then my cultures add considerable evidence in favor of the different strains which Jennings has isolated. It appears to me, however, that what that author has done for Paramaecium, can probably be done for many closely related species of infusoria, and the very fact that he did find it difficult to secure an intermediate race between the aurelia an-d the caudatum groups is a strong point in favor of the. distinctness of the forms.
Vitality. It has been customary to regard the rate of reproduction of infusoria in culture as a just criterion of vitality. Maupas wrote i^^ "Cette faculte de reproduction (aurelia) resemble beaucoup a celle de la precedente espece (caudatum)." My cultures completely corroborate this statement, for during the six and one half months of the life of the caudatum culture, 324 generations have been attained, while during the same period,
^"For further details of the culture see: L. L. Woodruff, Two thousand generations of Paramaecium; Archiv fiir Protistenkunde, 21, 3, 1911. '^'Sur la multiplication des Infusoires cilies, loc. cit., p. 234.
232
LORANDE LOSS WOODRUFF
under identical conditions, the aurelia culture has advanced from the 1785th generation to the 2117th generation, or 332 generations. This gives a difference of only eight generations in the rate of reproduction of the two forms during seven months (cf. fig. 1). These cultures obviously do not support the statement, frequently made, that aurelia is a weaker form than caudatum.
Maupas remarked that P. aurelia was one of the most common infusoria, and Jennings found that a typical wild culture could
2.5
2.0
1.5
0.0
Fig. 1 Diagram showing the comparative rate of division of the pedigree
cultures of Paramaecium aurelia and Paramaecium caudatum, when bred under
identical conditions, from May 14, 1910, to November 30, 1910. During this period P. aurelia (designated by continuous line) advanced from 1785 to 2117 generations, while P. caudatum (designated by broken line) advanced from 1 to 324
generations. The rate of division is averaged for ten-day periods. The ordinates represent the average daily rate of division of the four lines of the cultures.
be resolved into caudatum and aurelia groups. It has been my experience that it is as easy to procure one form as the other in the wild state. Certainly my aurelia culture, which theoretically would provide individuals to the number represented by 2 to the 2117th power, gives more evidence of vitality and reproductive power than has been demonstrated for any other animal.
Conjugation. I have no data in regard to the conjugation of either of these forms, for, so far, in all experiments with stock
PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM 233
material left over after the daily isolations from my pure lines, I have failed to observe a single syzj-^gy, either between aurelia lines or caudatmn lines, or between aurelia and caudatum lines. Jenning's" experiments on conjugation in Paramaecium bring out data which add further evidence that in certain strains at least a predisposition to conjugation does not exist. Maupas wrote: C'est bien certainement une des especes (aurelia) qui se recontrent les plus frequemment a I'etat conjugu^."
Maupas, as we have seen, pointed out a difference in the nuclear phenomena during conjugation which he held to be of diagnostic value, and Hertwig apparently showed that aurelia has two micronuclei at the reorganization after conjugation. Calkins, on the other hand, has shown that P. caudatum, in one case, reorganized with two micronuclei and later reverted to the uninucleate type. Such a case can readily be considered a ' sport ' which has arisen possibly by the persistence of the stage with two micronuclei immediately following the separation of the conjugants, or by the precocious division of a single micronucleus previous to the first regular vegetative division after conjugation. Although, as Calkins stated also, forty-five generations is a long time for an abnormality, if it be such, to persist; nevertheless, I believe it is very significant that, whereas during the presence of two micronuclei the division rate averaged only 0.8 of a division per day, after the loss of one of the micronuclei the division rate increased to the remarkable rate of 2.2 divisions per day, on the average for a period af four months. It is also of interest that the other exconjugant which reorganized 'normally' as caudatum failed to live.
So far as I am aware, the following statement^'* by Simpson is the only record of a possible case of conjugation between aurelia and caudatum: Out of twenty-one attempts I had but two partial successes. Conjugation took place on two slides: the period was normal. After separation each of the ex-conjugates divided once : on the third day they died off. In anticipation of something
"What induces conjugation in Paramecium? Jour. Exp. Zool. 9, 2, 1910. ^^Observations on binary fission in the life-history of Cihata, Proc. Royal Soc. Edinburgh, 1901, pp. 407-408.
234 LORANDE LOSS WOODRUFF
of this sort from analogy in higher forms, I intended to let the two pairs run their natural course, foregoing the desire to examine their nuclear condition. In view, therefore, of the incompleteness of the experiment, it is perhaps unwarrantable to draw any results regarding hybridization and infertility, or even the 'fixity of species' so far down in the animal scale." Simpson gives no data to prove that these were actually syzygies between the two forms, but if they were, it is obvious that they were not fertile. Jennings and Hargitt stated that they had been unable to induce the two forms to conjugate.
In view of the fact that, for example, Maupas studied conjugation of both P. aurelia and P. caudatum, and Hertwig studied conjugation of P. aurelia, and also that Jennings observed conjugation in both his aurelia races and in his caudatum races, it is clear that aurelia forms conjugate and caudatum forms conjugate, but there is no positive evidence that conjugation takes place between individuals of aurelia and caudatum.
Macronucleus. The normal macronucleus of aurelia was described by Hertwig and Maupas and that of caudatum agrees very closely. It is an ellipsoidal body with a smooth contour, except for a slight depression, in which the micronucleus is usually located. But the form of the macronucleus of both aurelia and caudatum frequently departs very greatly from the 'normal' condition. It is not unusual to find paramaecia of my aurelia cultures with the macronucleus resolved into several parts. These parts apparently may be gathered together into a typical nucleus for division, or the cytoplasm and micronuclei may divide, the macronuclear fragments which are in the posterior part forming the macronucleus of one daughter cell and those in the anterior part forming the macronucleus of the other daughter cell. I shall reserve the full discussion of these interesting changes for a special paper. It is - important to emphasize the fact that these are not pathological conditions, since the general vitality, as indicated by the rate of division, is not appreciably affected.
PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM 235
Calkins,-^ however, found nuclear fragmentation in degenerating individuals of caudatum, Wallengren-^ and Kasanzeff^^^ showed that various changes including fragmentation of the macronucleus occur when paramaecia are starved, and Popoff-' described a large increase in size and fragmentation of the macronucleus in degenerating caudatum which paralleled the conditions observed in specimens ripe for conjugation. He also obtained similar changes by subjecting the animals to various reagents. ^^ Mitrophanow^* emphasized the fact that the structure of the macronucleus varied considerably under the influence of diverse conditions, and he described fragmentation and figured spherical pieces which very closely resembled micronuclei.
It is evident, then, from my cultures that the macronucleus of both aurelia and caudatum is subject to great morphological variation without appreciably affecting the rate of reproduction, i.e., it is entirely normal. It is also apparent from the work of the other authors cited that degeneration changes become manifest in the fragmentation of the macronucleus. Consequently the macronucleus presents no character which is of permanent diagnostic value.
Micronucleus. Maupas, as we have seen, regarded the micronucleus as the chief distinguishing character of aurelia and caudatum, and my cultures substantiate his view. Fixed and stained individuals show that the micronuclei of the aurelia culture for over two thousand generations have conformed in a remarkable degree to the aurelia type as described by the French investigator, and the micronuclei of the caudatum culture have conformed to his caudatum type.
2*Studies on the life history of Protozoa. IV. Death of the A Series, Jour. Exp. Zool., 1, 3, 1904.
-^Inanitionserscheinungen der Zelle, Zeit. f. allg. Physiologie, I, 1, 1901.
-"Experimentelle Untersuchungen ueber Paramecium caudatum. Inaug.' — Diss., Zurich, 1901.
-^Depression der Protozoenzelle und der Geschlechtszellen der Metazoen, Archiv fur Protistenkunde, R. Hertwig Festband, 1907.
2«Experimentelle Zellstudien III. Ueber einige XJrsachen der physiologischen Depression der Zelle. Archiv fiir Zellforschung. 4, 1909.
-"L'appareil nucleaire des Param6cies, Arch. Zool. Exp. et Gen., (4), I, 1903.
236 LORANDE LOSS WOODRUFF
It is not only the presence of two micronuclei, but their peculiar morphology, as emphasized by Maupas, which is characteristic of the aurelia type. I have found one individual of the aurelia culture with three micronuclei, and a few specimens in which I have been unable, in total mounts, to distinguish a single micronucleus or more than one micronucleus. ,But when only one micronucleus could be seen it has been of the aurelia type, and other individuals of the culture at the same period have had the two characteristic micronuclei. I have observed a variation in the number of micronuclei in various pedigree cultures of hypotrichs,^" Popoff has found reduplication in Stylonychia mytilus and Paramaecium caudatum during degeneration, and Kasanzeff has observed the same in starved P. caudatum. Thus, while my cultures of Paramaecium and various hypotrichous species substantiate Wallengren's and Calkins' statement that the micronuclei are the most stable elements in the cell, and the last to be visibly affected by environmental changes, nevertheless it is apparent that they are subject to variations under certain unknown conditions. Temporary variation, therefore, cannot be considered as having weight in determining species. The essential fact is, however, that throughout the existence of my aurelia and caudatum cultures, the morphology of the micronuclei has conformed to Maupas' description for the respective species. It must be borne in mind also that P. caudatum has been the subject of more extended study by exact culture methods than any protozoon except P. aurelia, and in all these long pedigree cultures it has bred true to the caudatum type, at least with respect to the single micronucleus. Calkins, for example, in his important investigations on the life history of this form, carried three distinct cultures, by the aid of artificial stimuli during periods of physiological depression, through 379, 570, and 742 generations respectively. McClendon, also, studied mass cultures of Paramaecium for considerable periods and stated that he never found individuals with different numbers of micronuclei in the same culture. "^
^"An experimental study on the life history of hypotrichous Infusoria, Jour. Exp. Zool., 2, 4, 1905. "Protozoan studies, Jour. Exp. Zool.. 6, 2, 1909.
PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM 237 SUMMARY
Briefly stated, I am convinced from my study of paramaecia that —
1 . A very great maj ority of individuals of aurelia and caudatum can be distinguished on the basis of shape alone;
2. A very great majority of individuals of aurelia and caudatum can be distinguished on the basis of size alone;
3. The power of reproduction, or general vitality, of aurelia and caudatum is practically identical;
4. The macronucleus of aurelia and caudatum is subject to such great variation that it affords no diagnostic feature;
5. The micronuclear apparatus of aurelia and caudatum affords crucial diagnostic characters.
I have summarized the various characters of the two forms as they have shown themselves in my long pedigree cultures, and it is evident that they have conformed practically identically to the Maupasian types — such variations as have appeared not being so great as have been observed to occur in undisputed species, or as one would expect to find when the intimate relation of the unicellular organism to the environment is considered. Therefore, I believe, that since one of the crucial tests of a species is its ability to breed true to its type indefinitely, aurelia and caudatum have adequately met this test during more generations than any other animal under observation, and accordingly Paramaecium aurelia O. F. M. and Paramaecium caudatum Ehrbg. should be regarded as distinct species. ^-^ ^^
- 2In this paper I have followed the spelling of the name of the genus as given
by its founder, except in direct quotations from other authors.
'*I have the satisfaction to note that my conclusions are in accord with the final results published by Jennings and Hargitt in the last number of this journal, which was received when this paper was in press. Hargitt says, "There is cytological warrant for distinguishing caudatum races from aurelia races, and it seems probable that it will continue to be convenient to distinguishthese as two species."
MALE ORGANS FOR SPERM-TRANSFER IN THE
CRAY-FISH, CAMBARUS AFFINIS: THEIR
STRUCTURE AND USE
E. A. ANDREWS
From the Zoological Laboratory, Johns Hopkins University
THIRTY -ONE TEXT FIGURES AND FOUR PLATES
INTRODUCTION
The present paper is a contribution to our knowledge of the means that lead to the fertilization of the egg. It is part of the history of the sperm outside the body of the animal.
Sexual reproduction in most complex animals involves the transfer of the sperm from one animal to another, before the eggs can be fertilized.
Among animals the various methods by which the sperm is transferred may be grouped under the three heads, diffuse, direct, indirect. By diffuse sperm transfer we mean the discharge of the sperm into the water, where it may meet the eggs outside of the female, as in certain coelenterates, echinoderms and annelids, or may be drawn into the body of the female, as in certain lamellibranchs. By direct sperm transfer we mean the method found in the majority of complex animals, in which there is more or less direct application of the terminal parts of the passages leading the sperm to the exterior to the passages leading from the exterior direct to the eggs. In this group there is commonly a true copulation.
By indirect sperm transfer we mean those peculiar complex methods of getting the sperm from the testis to the eggs that are found in a few cases amongst the great groups of animals, as in
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
239
240 E. A. ANDREWS
earthworms, spiders, some cephalopods and leeches. The essence of these cases of indirect sperm transfer Hes in the fact that while the sperm is transferred by organs, and not by floating, yet these organs either do not put the sperm into the egg passages, or else if they do they are not organs directly concerned with the discharge of sperm, or both may be true. In indirect sperm transfer there is no true copulation, or intromission, but at most conjugation or clasping.
The three methods are not always sharply separable, and may be regarded as only convenient groupings of physiological processes that occur here ^nd^here among animals without reference to their systematic positions. The diffuse method is obviously the one open to the most hazard in the sperm and eggs meeting; the direct method by intromission the best assured method; the indirect method most perculiar and needing special explanation in each case.
In the crayfishes and lobsters most interesting cases of indirect sperm transfer occur, and it is the purpose of the present paper to describe the organs of the males that are used to transfer the sperm to the receiving organs that have been described in previous communications (1, 2, 3). In these animals the male transfers the sperm to the outside surface of the female where it remains till the eggs are laid, when fertilization takes place outside the body. In the American lobster and the sixtj^ and more species of crayfishes of the genus Cambarus that are found in all but the most western parts of North America, the sperm on the shell of the female is stored in a special pocket, or receptacle, but in the other genera of crayfishes, all the world over, there is no such receptacle and the sperm is believed to be distributed over the shell of the female in separate spermatophores. While the sperm pocket has been described (1,2, 3) the organs of the male that fill the pocket have had only such consideration as was necessary for the systematist, who found them to be of the greatest value in distinguishing species and in forming subgenera.
In the present paper the anatomy of the male organs is examined and their use as organs of sperm- transfer is explained.
ORGANS FOR SPERM-TRANSFER 241
CAMBARUS AFFINIS
While the sexual habits of all the species of Cambarus, agree in the main, the species afFmis has been more studied, and as in describing the female organs concerned in sperm transfer we first considered this species, we will also give chief attention to the male organs of sperm transfer in this species.
As elsewhere described (4, 5) conjugation is here a long series of activities of the male accomplishing the accurate adjustment of the essential transfer organs of the male to the receptacle of the female. The receptacle of the female is a single pouch in the shell, but the transfer organs of the male are three pairs of outgrowths. On each side of the body there is a papilla, or special termination of the sperm duct, and two limbs, those of the first and second segments of the abdomen, which we may call the stylets.
To introduce the sperm into the receptacle the papilla must be adjusted both to the first and to the second stylet and both sides of the body play a necessary part in the process of sperm transfer.
THE PAPILLAE
One external instrument concerned in the process of sperm transfer is the modified end of the sperm duct that emerges from the base of the fifth leg, on each side of the animal. These organs are the papillae.
Since systematic work has been done largely upon preserved specimens it is not so generally known that the sperm duct ends in life, in a soft, turgid protuberance, which may be so collapsed after death as to leave only the rounded hole in the firmer shell as the apparent ending of the sperm duct. These papillae lie concealed by the stylets, at rest, but on raising the stylets the papillae are seen as conspicuous, clear, tubes about 3 mm. long and 1^ mm. wide, jutting out from the base of each fifth leg.
At the time of conjugation the papillae are also concealed from view since the necks of the first and the second stylets form a nicely adjusted frame about the papillae and this frame is fitted in between the bases of the fifth legs. Certain in and out move
242 E. A. ANDREWS
ments of these bases seem to adjust the papillae so that they fit accurately into the orifices of the first stylets (figs. 30, 31) and by that means the sperm discharged from each papilla is passed into the cavity of the stylet.
The papilla (P. fig. 1) is on the under side of the large first segment of the leg and projects downward and toward the median plane; but its tip turns away from the middle line of the body. The papilla is a cone with bent apex. It is translucent and distended with colorless blood. When directly injured, or upon lessening of the blood pressure from injury elsewhere, the papilla collapses, being but a thin uncalcified protrusion of the skin, kept turgid, or erected, by blood pressure. Within the papilla one can see a large central tube passing toward the tip and also chalky white masses suspended between the central tube and the thin outer wal!s.
On the shell at the base of the papilla there is, anteriorly, a single row of very long setae (fig.l ) that form a sort of protective screen over the anterior face of the papilla.
Sections show that the papilla is a continuation of the deferent duct, blood cavity and skin, so constructed that the bent, conical apex, with its soft walls can be adjusted to the hard opening of the stylet so as to fit hermetically, as a tense rubber bag might. Moreover the bent tip can be opened to discharge the sperm, when special muscles remove the obstructing valve that holds the tube closed.
A lengthwise section through this delicate papilla (fig. 2) shows that the central tube is a direct continuation of the deferent duct that leads the sperm from the testes to the tip of the papilla. Between this duct and the outer cuticle there is a large space fall of blood, traversed by little connective tissue and in it are the white bodies just mentioned, now seen to be small tubular glands, opening into the central duct. The central duct presents two strikingly distinct parts; the one continued from within the leg has the thick muscular wall and peculiar secreting lining of the deferent ducts, the other is lined by the thin cuticle inflected at the orifice at the tip of the papilla, and lacks muscle. In place of muscle the wall has only epidermis, which extends irregular!}'
ORGANS FOR SPERM-TRANSFER
243
Fig. 1 Posterior face of the left fifth leg of a living male 95 mm. long to show
the translucent papilla. (P.) 2. Qq.
Fig. 2 Longitudinal section through the papilla and the base of the fifth leg,
showing the orifice of the sperm duct, the valve, the muscles and the glands.
2. 90 mm. A.
244 E. A. ANDREWS
into the blood space as the tubular glands alluded to above. The orifice at the tip is small and is not closed by any muscle, but apparently by blood pressure only. The part of the tube lined by cuticle has its lumen much reduced by a valve, or great longitudinal ridge, which extends out as far as the abrupt bend at the orifice. In a cross section (fig. 3) this ridge is well seen, as is also the fact that some muscle fibres run into it and that the glands are chiefly on the side opposite the ridge. The ridge appears to act like a valve to hold this part of the tube closed, while contractions of the muscle would tend to open the tube wide and let the sperm pass to the orifice, which would then be forced open by the internal pressure of the sperm squeezed by the muscles of the wall all along the length of the duct, or some extent of it at all events.
The upper part of the duct, as seen in the cross section fig. 4, has its thick muscles arranged chiefly in transverse fibres and is lined by an epithelium that evidently in large measure breaks down to furnish a great mass of secretion about the sperms. It is probably this secretion that envelops the sperm in the form of macaroni-like tubes, when they pass out in a slow stream.
THE STYLETS
The most complex of the organs concerned in sperm transfer are the modified limbs of the abdomen which we will call the stylets. In the male the sixth pair of abdominal appendages form the large side parts of the tail fan while the third to the fifth inclusive are the simple and apparently rather useless swimmerets. The first and second pairs are specially constructed to serve as transfer organs for the sperm.
These appendages of the first and second somites are much stouter and longer than the following swimmerets and have a very firm attachment to the abdominal sterna. The calcified ridge across the middle of the sterna is much more developed in the first and second somites, and where the appendages are fastened it rises up as a decided elevation which remains as a stump when the appendage is cut off. On the second somite these stumps are far apart, (some 10 mm. in a male of 100 mm.) while on the first
ORGANS FOR SPERM-TRANSFER
245
Ex. m. ',
Fig. 3 Cross section of the sperm duct and valve along the line 3 of fig. 2, showing the duct closed by valve ridge. 2. A.
Fig. 4 Cross section of the sperm duct along the line of 4- of fig. 1, showing the muscular wall and the lining epithelium disintegrating in secretion. 2. A.
Fig. 5 Extreme tip of right first stylet, showing the groove bottom coming to the surface, posterior face. 2. A. Ex. m.— the external mass. M.m.— the internal mass. IS' — the level of the section, fig. 13.
Fig. 6 Diagram of stylet as in plate i, fig. i, to show location of glands in the interior, and the location of the sections, 7 to 13, shown in figs. 7 to 13.
24G E. A. ANDREWS
somite they are in contact at the median Une of the abdomen. The eUiptical transversely elongated stumps of the first appendages are 5 mm. long and those of the second about 3 mm.
Commonly these appendages are carried forward horizontally under the thorax between the thoracic legs in a deep depression of the thoracic sterna. The first pair lie close side by side with their median faces in contact. The second pair lie over and largely conceal the first, since their form enables them to come to the middle of the body beneath the first pair in spite of the fact that their bases are attached to the sterna, so far from the middle line.
In a dead male one may move the appendages upon their attached bases as follows:
The first may be moved upon its base from the horizontal up toward the vertical only about 45°. The membrane on the anterior face of the joint at the base of the appendage is stretched to its limit when the appendage is pulled up a little beyond sixty degrees, so that this appendage is never vertical and cannot swing back and forth through a wide arc as do the ordinary swimmerets. The distance traversed by its tip is some two cm. The appendage may also be rotated a very little at its base and moved from side to side a little so that ts tip travels some 5 or 6 mm.
The apex of the second may be drawn from the horizontal up a little beyond the vertical; but neither the basal protopodite nor the endopodite travels more than 90°. They are set together at a large angle, so that while the main length of the appendage is horizontal the basal part never is, and when the base goes back some 90 degrees the horizontal part is swung past the vertical line. The tip traverses some 3 cm. The base may be rotated a little and moved from side to side so that the apex travels 6 or 7 mm.
STRUCTURE AND ANATOMY OF THE FIRST STYLET
The first abdominal appendage of the male is a very stiff calcified mass of the general shape of an awl, some 17 mm. long, but having two tips. There is a groove along more than half its length and the base is articulated to the ventral shell of the anima' so
ORGANS FOR SPERM-TRANSFER 247
that the appendage has very Httle mobihty back and forth through some 45°. The normal position of the stylet is pointed forward under the thorax, where it lies horizontally in a deep groove, but in use it is dropped down and backward toward a vertical position. It has an anterior face, which is usually carried as the dorsal side, a posterior face which is usually the ventral aspect, and an outer and an inner or median face.
The general appearance of the stylet is seen in the photographs, figs. I, II, III, IV, which represent respectively the posterior, median (or rather median and posterior somewhat diagonally), the anterior and the outer faces of the same left stylet. Fig. i, the posterior face, is the view got by looking at the under side of the crayfish, after lifting up the second stylet, which lies over the first and largely conceals it.
The first pair of stylets do not spring from the sternal surface far apart as is the case with the common, unmodified swimmeret, but they arise very close together; in fact the median faces, (fig. II,) of the two come into contact so that these two appendages really form one mass. If looked at from the dorsal side, the two are seen to lie in contact at the base and all along the distal half, leaving between the constricted parts of the two a square opening that is occupied, in rest, by part of the second stylet.
In describing the stylet we will distinguish the base, the neck, and the scroll or spiral that contains the groove. The scroll ends in two tips, the more slender, side outgrowth, or spatula, and the real end bearing the groove, the canula
The base is some 6 mm. wide and long and only 2 thick, being flattened from before back. The posterior or ventral face of the base, fig. I, presents a wide groove bounded on the median side by a rounded knob and on the outer side by a long ridge which, as it passes on to the neck, bears a tuft of long, finely plumose setae, that are seen again in profile in figs, ii, iv. In this deep groove the second stylet lies when not in use, so that the two appendages are firmly packed together under the thorax of the male.
The part of the base joined to the sternum of the animal is an oblique eUiptical area, around the edge of which the hard shell gives place to the soft articular membrane that makes it possible
248 E. A. ANDREWS
to cut the whole appendage away from the sternum. In this membrane there is an articular, whitish plate that is seen in figs. Ill and IV. The whole base is pyramidal and except the posterior all its faces (figs, ii, iii, iv) are convex and rounded.
The neck is the narrowest part, before the sudden enlargement of the spiral part; it is the smallest of the three regions; and is best seen in figs, ii, iii, iv. The neck passes gradually into the base and ends abruptly at the spiral. It is some 3 mm. long and 2 wide and thick. It has an angle along the ventral face that continues the ridge of the base up to the outer part of the spiral.
The spiral or scroll may be likened to a long triangular plate with its edges rolled in together so as to leave a groove between them, but it is a plate some 8 mm. long, with the edges greatly thickened, so that the resulting mass is apparently solid. The groove begins on the median side, fig. ii, and passes in a sinuous course to the ventral side and along this diagonally to the very tip. The apparent bifid nature of the stylet is due to an outgrowth from the median part, quite separate from the real end of the organ, in which the groove is continued through its entire length. We have then to describe a sinuous groove and its two boundaries, which we will call the median mass and the external mass ; and also the two tips. The external mass, seen from the ventral side on the right of fig. i, shows a proximal part about 2 mm. long and 1 mm. wide, bearing a marked ridge parallel to its sides and continued up from the neck. And then it suddenly turns at a large angle and becomes a rounded and gradually tapering terminal part, something less than ^ nam. wide at first, and 6 mm. long. This passes behind the slender protuberance of the median mass to end as a flattened, horny tip together with the like ending of the median mass. In other words both external and median masses unite as the horny tip that we will call the canula. The sudden change in direction of the mass is accompanied by a like change in the groove whose edge it forms; this change of the groove we will call the angle of the groove.
Seen from the outer face, fig. iv, the external mass is widely swollen proximally, some 2^ mm. deep, and gradually narrows into the distal part. The round canula is bent somewhat, ventrally.
ORGANS FOR SPERM-TRANSFER 249
On the dorsal face, fig. iii, the external mass is confluent with the median mass, without boundary line. Thus the distinction between the two masses is useful chiefly on the anterior face where they form the two sides of the groove.
In fig. Ill, the long triangular region running from the notch that marks off the neck from the spiral region and ending distally in the rounded and pointed canula, is to be regarded as made up chiefly of the median mass, but the depressed part along the left edge is part of the external mass.
On the median face, fig. ii, (which is unfortunately turned so that part of the posterior face shows) the external mass shows only its prc-ximal end along the side of the diagonal groove, and into this groove the external mass here sends a narrow horny shelf, dimly seen as light in fig. ii. The external mass has an angular projection, or lip, at the very beginning of the groove which will be described in connection with the orifice of the groove. At the tip, part of the external mass is seen making the lower part of the canula, to the right, that is, the curved strip of external mass seen is flat and on a lower level than the median mass.
On the ventral face, fig. i, the median mass looks like a long rounded white bone that begins suddenly without apparent connection with the neck and, after running nearly straight for some 6 mm., turns externally across the external mass as a flat, curved process that we will call the spatula. Beyond the spatula, w^hich stands out freely as the second tip of the appendage, the median mass continues as the narrow median edge of the canula. From the external view, fig. iv, the visible part of the median mass, the spatula is back of the external mass.
In the dorsal view, fig. iii, the main part of the spiral region is median mass, forming a long triangle, beginning at a deep notch near the neck and extending in the foreground as the vis ble part of the canula and back of that as the spatula. At the notch may be seen part of the lip on the external mass.
The median face best shows the median mass, but, fig. ii, being not an exactly median view, does not do justice to it. In reality this face is markedly flat where it comes against the like face of of the other stylet of the pair. This flat face is a long ellipse, 2
250 E. A. ANDREWS
mm. wide and 5 long, and is smooth except for a roughened area near its proximal end where there is a long tuft of finely plumose setae which bend abruptly downward, that is, posteriorly, as if an adjustment to the fact that they are pressed in between the two stylets. These setae are so long as to be visible from all points of view, cf. figs, i, ii, iii, iv.
The groove itself is seen only from the median and ventral views. It is some 7 mm. long and begins as the orifice on the median face where it meets the ventral, fig. ii. The orifice is a conical opening bounded by that depression of the neck that makes the notch so conspicuous in fig. iii, by the rounded origin of the median mass, fig. ii, and by the overhanging lip of the external mass. It is of such shape that the tip of the spout, fig. 1, can fit into it. The groove leads from the orifice obliquely outward and distally between the external and median masses some 3 mm. and then turns to make a rounded angle, fig. i, toward the median line some 3 mm. more. In this part of its course it is soon concealed behind the median mass that is rising to form the base of the spatula, but it still exists there and emerging again runs the entire length of the spatula as a very narrow slit with horny edges. The groove is thus a long double curve, bending abruptly outward, then forward and slightly inward and finally outward again, as seen from the ventral side. But it also bends in the vertical plane, passing downward, then forward and upward and finally a little downward at the tip. While the walls of the groove seem to be merely hard rounded bone there projects into the groove from the side a narrow shelf of horn that springs from the external mass only. This will be seen in sections.
The spatula is a flat flagellum-like process some 2 mm. long, | wide and perhaps | to iV thick. It is curved and pointed as seen in the figures. It springs from the median mass where this suddenly narrows to help form the canula, fig. ii. In life the spatula is milky white and pliable, not bony, more like leather. At its base it passes suddenly into the bony walls of the median mass and there can be bent as if in a socket. After drying it looks more like a thin chitinous membrane over a dried contents. It is somewhat concave at the base on the dorsal face. With methyl
ORGANS FOR SPERM-TRANSFER 251
green the horny tip of canula and the shelf in the groove stand out clearly as distinct from the substance of the spatula.
The canula is some 3 mm. long and at base f mm. wide and thick. It is a long cone, flattened somewhat from before back, bent upward dorsall3\ and ending in a rather sudden point that bends outward from the median side. The canula is made up of both external and internal masses. Most of the length of the canula is clear, yellow, horny matter, but at the base this is continued as the white calcified material of the rest of the stylet. The bone of the external mass stops rather suddenly, while that of the median mass is continued in the midst of the horny cap as a central area, as seen from the median view. An enlarged view of the tip of the canula, fig. 5, shows that both external and internal masses make about the same amount of the canula, since the groove continues sinuously almost to the exact tip of the organ, but yet there is a greater prolongation of the external mass to form a short ungrooved apex. This sketch is from a canula of the opposite side of the body from that in fig. i. The two canular tips flare away from one another.
The groove may be said to begin and to end on the median face and to be shoved away from it through most of its course by the ridge that we have called the median mass (fig. i.).
INTERNAL ANATOMY
When the stylet is macerated some days the entire contents may be drawn out of the hard shell; such a cast of the shell has its general long conical form with a short conical tip that came out of the canula and a short flat plate that came out of the spatula. It is made of connective tissue and blood covered with epidermis with some red pigment cells and shows at the base some muscles and at the middle some glands.
The muscles, as made out by dissection of fresh and preserved crayfishes, are weak and run from the base of the stjdet into the adjacent ridge of the sternum upon which the stylets articulate. There is a wide thick fan of muscle that passes from the bony articular plate of the anterior face of the stylet, fig. iii. When
252 E. A. ANDREWS
this is pulled the stylet is raised dorsally into its position of rest. Since it lowers the organ into the groove on the thorax it may be called the depressor, though it really swings the appendage forward.
This depressor muscle is lodged in the protruding ridge of the sternum from which the stylets spring, and its fibres are made fast to the posterior wall of this ridge. There is also a smaller muscle attached to the base of the stylet at its external edge which would seem to antagonize the other and to tend to swing the stylet backward, that is, to raise it up from its horizontal position of rest into the erect position of use; it may be called the erector muscle.
The internal anatomy of the stylet as well as the character and mode of use of the groove, were made clear from sections.
The diagram fig. 6 shows the ventral view of a left stylet as if transparent, the extent of the glandular area being shaded; the glands occur in both external and internal masses, but not in the base of the stylet, and the}' extend from the neck to near the origin of the spatula, filling most of the cavity of the region in which they occur.
The sections, (figs. 7 to 13, inclusive), were taken across the stylet along the planes indicated by the like numerals in fig. 6.
The transverse section, (fig. 7) shows in black the exceedingly thick shell with the depression on one side that forms part of the orifice of the groove, overhung by the solid lip. Through the thickness of the shell that forms this part of the orifice are seen many fine tubes, passing from the internal glands to discharge on the surface. The interior of the stylet is a delicate mass of connective tissue, chiefly blood sinuses, crossed by few strands of tissue, and bounded by the thin epidermis against the shell.. Scattered all through this are the tubular glands that bend and are cut at various angles. These glands ultimately discharge by the numerous fine ducts that penetrate the shell. In this section the sharp angle above is the ridge {R) seen in figs. 6 and I passing along the external mass. The angle to the right is the line between the ventral and external faces of the external mass.
ORGANS FOR SPERM-TRANSFEK
253
Sections 8 and 9 show the orifice passing into the groove; they are cut obUquely transverse and, in addition to the section of the first stylet, show also the section of the second stjdet as it lies locked in the first. Disregarding for the present all but the lower part of the sections we see that the stylet has widened out from the constricted neck into a wide flattened mass sub-divisible with reference to the groove into the external and median masses. In
Fig. 7 Section across the stylet, in the region of the neck, just below the orifice, on the level 7 of fig. 6. 72— the sharp ridge on the external mass, fig. 6 and i. 2. 90 mm. A.
Fig. 8 Cross section of the same at the level 8, showing the groove above the orifice filled by the head of the accessory stylet, which is the separate. mass lying to the left and above. 2. 90 mm. .4.
fig. 8 the orifice is so overhung by the lip as to be in section a Cshaped bay, embracing the head and neck of part of the second stylet. Here again the shell is remarkably thick, but is penetrated by the ducts of the glands discharging on the surface that lines the orifice. In fig. 8 the lower straight side to the left is the flat face that is normally applied against its fellow on the outer side of the body. Above is the angle {R) that represents the ridge of
254
E. A. ANDREWS
the external mass, just as in fig. 7. In the interior some of the glands are very large. The section distal to this, (fig, 9) shows the bottom of the groove receded from the surface and constricted from the rest bj^ the continuation of the lip so that it forms a rather elliptical hole with only a very narrow slit opening into the deep groove that is seen from the surface. This surface groove is bounded on the left by the greatly thickened shell substance of the median mass and on the right by the thick shell of the external
M.m,
Fig. 9 Cross section of the same at the level 9, showing also the grasping second
stylet, above, and its wedge, to the left, where it is entering into the groove of
the spiral. 2. 90 mm. A.
Fig. 10 Cross section of the same at the level 10, showing the bottom of the groove cut off by the shelf from the external mass. 2. 90 mm. A.
Fig. 11 Cross section of the same, at the level 11, showing, above, the base of the spatula. 2. 90 mm. A.
ORGANS FOR SPERM-TRANSFER 255
mass. The cavity within the shell of the external mass is reduced to a narrow space and the glands have become few.
Further along the stylet, (fig. 10) the groove has passed from opening to the left (fig. 8), through the position shown in fig. 9, to open more toward the right. The groove is a deep and narrow one. Into it still open some few gland ducts from the remaining glands of the median mass. As before the side walls of the groove are made of very thick shell. The most unexpected fact is that the bottom of the groove is shut off as a very minute hole overhung by the continuation of the lip, which is now a horny shelf passing all along the groove, near its bottom, and so nearly meeting the opposite side as to practically shut off the bottom of the groove as a special tube. This figure shows the form of the stylet at the level, 10, of fig. 6. The flat side to the left is the flat face of the median mass, while the rounded edges of the groove are the two narrow parts of the external and median masses seen from the ventral side in figs, i and 6, just proximal to the base of the spatula.
A section through the base of the spatula, (fig. 11) shows the groove above overhung by the rising spatula that conceals it from surface view, (figs, i and 6) but still allows access to the groove from the right, in under the spatula base. The external mass {Ex. m.) is now the greater, but it contains no glands, while the median mass is reduced to a nearly solid shell prolonged as the slightly hollow spatula. The tube at the bottom of the groove is still there, overhung by the little chitinous shelf.
Near the apex of the organ, (fig. 12) the groove is again open above, as we have passed beyond the base of the spatula, only the tip of which is cut, lying well over to the right. This figure being magnified twice the diameter of the preceding figures, shows plainly the shelf that cuts off the bottom of the groove. The median mass is a narrow and nearly solid shell that forms the left wall of the straight, deep groove. The external mass is the main part of the section and contains much very watery connective tissue, covered with epidermis. In this section, the calcified part of the shell is represented in black, as in the other sections, while the chitinous or horny parts are dotted. From the surface this region of the canula looks to be only chitin. Farther on the
JOURNAL OF MOBPHOLOGY, VOL. 22. NO. 2.
256 E. A. ANDREWS
calcified part of the shell fades away and only pure chitinous matter is left, so that a section at the very tip of the canula, (fig. 13) is only chitin. This view is enlarged four times as much as the preceding one and shows the disappearance of the superficial part of the groove though the bottom, which is now close to the surface, is still overhung by the shelf from the external mass. That is the tube at the bottom of the groove can now discharge by a slit to the surface at the tip of the canula; see fig. 5, where the surface slit of the groove is represented by the black line and the bottom of the groove, or the tube, is represented by the dotted line, which comes finally to the surface at the tip as seen in the section across the level 13. As fig. 5 is of *a right stylet and the section 13 from a left stylet it shows the parts reversed; the main bulk of the section is really of the external mass, as in fig. 12.
The specialization of the bottom of the groove had not been expected till sections revealed it and suggested some special use. Sections of stylets taken when being used in conjugation soon showed that the tube at the bottom of the groove is the channel for the transfer of sperm. Along this minute tube all the sperm passes from the papilla to the sperm pocket of the female. A section across the stylet where the median surface bears a tuft of setae, between the levels of 8 and 9 of fig. 6, when sufficiently enlarged, shows that the sperm is contained inside the tube of the groove, as in fig. 14. This shows only the part of the shell about the tube, with the sharp edge of the shelf above, jutting out to almost meet the wall of the median mass (see fig. 9). The cavity of the tube is full of a secretion containing at its centre a pearshaped mass of the peculiar sperm of the crayfishes. As was shown (6) these sperm do not assume the star shape they have in books as long as they are in the male and not even when in the sperm receptacle of the female when normally protected from the water, and in this section, where they are seen in transit, they are still spherical, clear bodies with the peculiar bowl-shaped central part that, as represented in the sketch, might be thought a central nucleus. All along the groove above the orifice there is thus a strand of sperm surrounded by a paste-like white mass that fits tightly into the tube.
Fig. 12 Cross section of the same near tip, at level 12, showing the spatula cut
off to the right. 2. A.
Fig. 13 Cross section of the very tip of the stylet, at the level 13, of figs. 5 and 6, showing the groove coming to the surface of the horny canula. 2. D.
Fig. 14 Enlarged view of section across the tubule, between the levels 8 and 9 of fig. 6 showing the sperm cells enveloped in a secretion and shut in by the shelf above. 2. D.
Fig. 15 Enlarged view of section of tubule and bottom of groove, about the level 10 of fig. 6 showing the fewer sperm and little secretion in the tubule, surrounded by a thick horny layer. The calcified skeleton is represented as black. g. D.
257
258 E. A. ANDREWS
That this mass is run in under pressure seems indicated by the way it tends to flow out at the narrow sUt leading up from the tube into the groove and by the form of the sperm mass that tends hkewise to copy the shape of the cavity that is filled, being pointed toward the slit (fig. 14) . In successive sections this sperm mass is found all along the length of the groove, always in the bottom of the tube only, while the enveloping secretion for the most part disappears. Thus in the fig. 15 from the level 10, where there are still some secretion tubes coming through the heavy shell of the median mass, (fig. 10) there are a dozen or so sperm enclosed in the minute tube together with very little secretion and the sperm seem to come into contact with the shell. At this level, however, the thick and well-calcified shell (fig. 10) is covered by a thick layer of horny substance that ntiakes the shelf and continues on up the face of the external mass bounding the groove, (fig. 15). The discharge of the milk-white sperm from the tip of the canula, (figs. 5, 6 and 13) was seen in some males separated from females in conjugation.
The anatomy of the stylet thus shows it to be a more refined and specialized tool for sperm transfer than had been expected. It is essentially a very fine tube receiving sperm at its larger base and discharging it at its attenuated tip; but it has walls that give it great strength and rigidity while allowing the tip some elasticity. Moreover the receiving part of the tube is provided with glands of problematic value.
In looking for further light upon the nature of this sperm transfer organ we turn to its development in the individual.
ONTOGENY OF STYLET
We find that in Cambarus affinis the first and second larval stages are externally alike, in both sexes, while the third shows the male openings on the fifth legs, or the female on the third legs. In the first stage, there are no abdominal appendages on the first somite and but a crowding of epidermal nuclei under the shell where the appendage will be. In the second stage, these appendages are slight papillae. These indifferent stages are fol
ORGANS FOR SPERM-TRANSFER 259
lowed by the third, in which the external openings are differentiated but the appendages of the first somite are still simple papillae, alike in both sexes, unless they be longer in the male.
In the fourth stage, which is about 11 mm. long, the pleopods of the female still are simple papillae but little longer than in the third stage, while in the male they are long, simple spines, pointing toward one another and but slightly forward, as indicated in fig. 10 p. 127, Andrews, Ontogeny of Annulus, Biol. Bull., 1906.
The ventral face of the left spine or slightly specialized first pleopod, of a male 11 mm. long, is seen in fig. 16, magnified 430 in the camera sketch. This is from a larva killed July 1st, from late spring hatching. The organ is like a club ; it is very simple, nearly cylindrical and very blunt. It is not jointed, although there is a faint groove marking off the base from what will be the neck and spiral.
On the base there is a slight ridge with depressions on the median side of it. Internally there are two muscles from the base into the sternum of the abdomen. The distal part of the appendage is slightly grooved along its ventral face, thus marking off an external from a median mass. In cross section, fig. 17, the shell is not very thick and beneath it is a well formed epidermis with large nuclei, from which connective tissue strands traverse the large blood space in which blood corpuscles float. This section shows the groove on the lower side. The appendage is articulated to a slightly elevated stump on the sternum that holds one of the articular muscles and part of the other and ends in an elliptical orifice into which the base of the stylet fits. This articulation is so oblique that the stylet lies down and cross-wise towards its fellow and is but little elevated or directed forward.
In the male of this stage, the openings on the fifth legs are short slits, not a third of the width of the above simple stylet, and to each slit there leads a strand of nuclei that represents the efferent duct.
In males of 15 to 18 mm., in the fifth stage, the stylet (fig. 18) is about 1 mm. long and is somewhat more specialized. The base is set off from the terminal part by a more pronounced fur
260
E. A. ANDREWS
Fig. 16 Posterior face of left first stylet of male 11 mm. long. Enlarged 21S
diameters.
Fig. 17 Cross section of the same stylet. 2. D.
Fig. 18 Posterior face of left stylet of male 18 mm. long. Length of stylet 1 mm. f. A.
Fig. 19 Section of stylet of a male 12 or 15 mm. long. 2. D.
ORGANS FOR SPERM-TRANSFER 261
row, but there is no movable joint. The organ is more pointed and the groove is very deep from the rising up of its sides. Thus in section fig. 19, the narrow median mass, {M.m) to the left, rises high up beyond the groove and the groove itself is a narrow space between the wide external and the narrow median masses. In the surface view, (fig. 18) the bottom of the groove is indicated by the broken line; it is already twisted so that the groove looks towards the median side along its proximal part and then for a short distance toward the observer, that is toward the ventral side, and finally at the tip toward the median side again. Where the groove is open ventrally the median mass is rising up as a protuberance that will form the spatula. As yet the canula is only the spoon-shaped end of the organ.
In a male 22 to 21' mm. long and probably in the sixth stage, (fig. 20) killed October 4th, we find the same stage as in other males of this size killed in July, this being an exceptional male that failed to grow as the average do to be nearly two inches long in October. Here the spatula is quite evident as a blunt rounded finger-like elevation that crosses over the groove. As shown by the dotted line the bottom of the groove is to the right of its mouth along the proximal part of its course and to the left along under the base of the spatula; that is, the sinuousness of the groove is exaggerated by the fact that the sides not only rise up but grow over the groove, the external mass overhanging toward the median line proximally and the median mass growing over away from the median line, distally. The base of the stylet now bears a few short acicular setae and is provided with three muscles at its attachment to the sternal elevation upon which it stands. By this time the stylets point forward under the thorax. The canula is now a short rounded blunt termination of the stylet in which the groove is no longer widely open but reduced to a slit by the upgrowth of its walls.
In an autumnal male 38 mm. long, (fig. 21) the stylet has become much longer and more modeled but still shows the stiff joint between the base and the partly-formed neck. The few setae extend along the ridge of the base on to the proximal part of the external mass. The median mass sticks out abruptly at
262
E. A. ANDREWS
Fig. 20 Posterior face of left stylet of male 22 mm. long in October. S. A.
Fig. 21 Posterior face of left stylet of male 38 mm. long in October. Enlarged 25 diameters. 2. 90 mm. A.
Fig. 22 Anterior face of left accessory stylet, somewhat turned to show part of the external face: a view between vii and viii. 2. ao- On the left an enlarged sectional view of the cup at the end of the radius and the wedge cut off.
ORGANS FOR SPERM TRANSFER 263
the notch, or orifice, and bears a tuft of short setae. The spatula is long, flat and pointed. The canula is bluntly pointed and turned outward.
Later when the animal is 64 mm. long, the false joint of the stylet has disappeared and the tips become more sharp and long. Even before this size the males are known to conjugate, when about two inches long.
We thus find that the complex stylet of the adult starts from a slender papilla that becomes slightly flattened and grooved so as to form a very clumsy spoon with its depression rather more median than ventral. Then the sides of this groove grow up and make the groove into a cleft, which opens as before toward the median face proximally and distally; but along the middle of its course is forced to open ventrally and even externally by the overhanging growth of the median mass. The organ might be imitated by taking a long strip of clay with a slight length-wise groove on it and rolling the sides up over the groove, the median side tending to roll over outside the other. How the shelf from the external mass first grows out over the groove to cut off its inner part as a tube was not made out, but it is evidently a secondary specialization of the shell made by some special activity of the epidermis in a line near the bottom of the groove after the groove has becoriie deep.
THE SECOND OR ACCESSORY STYLET
The accessory stylets (figs, v-viii) are evidently specializations of the common type of abdominal appendages, (fig. 26). They are elevated only when in use in conjugation; and at rest are carried forward under the thorax, horizontally, where they rest upon the first stylets and are closely packed in with them inside the special sternal groove of the male thorax.
Figs. V, VI, VII, VIII, represent the left second stylet as seen from the ventral or posterior, the median, the anterior or dorsal, and the exterior faces, respectively. Like the unmodified pleopods this has a basal protopodite, an exopodite, an endopodite. The exopodite is a slender offset with setae, while the endopodite is
264 E. A. ANDREWS
the complex large part of the appendage that bears a terminal fiabelum and the remarkable side protuberance, found on no other limbs, which may be called the triangle.
Describing the entire stylet from the base outward, we see that the protopodite is chiefly a very strong flattened bony mass extending diagonally inward so that while the endopodite and exopodite are about parallel to the median line of the animal the protopodite forms an angle of 45° with it. This makes it possible for the endopodites of the two stylets to come together at the median line and for the endopodite of each side to lie upon the groove of the base of the first stylet, like a lance in its rest, although the bases of the two second stylets are fastened to the sternum of the second abdominal somite some distance from the median line. The protopodite is not entirely one-jointed but at its base is a soft membrane where it is joined to the sternum and in this are two large calcified plates, (figs, v and vi) besides two minute ones, (fig. vii) all of which together make a narrow basal section of the protopodite. Dissection shows there are muscles passing from this base of the protopodite into the sternum that may depress and elevate the appendage.
The protopodite is some 6 mm. long, 2 wide and 1^ thick. The exopodite is a slender filament some 9 mm. long and ^ mm. thick; a slightly flattened tapering cylinder set with long setae on External and median face. The setae are really plumose and together form a sparse brush. The exopodite is obscurely divided into some twenty segments. The basal 2 mm. is partly calcified, the rest membranous. It articulates freely with the outer distal corner of the protopodite so that it may be moved from the position of rest parallel to the endopodite, outward through 90° and swung back and forth some 45°. The tip of the exopodite often lies dorsally within the cavity or hollow of the triangle, and may have some use as a cleaning brush.
The endopodite is the stout calcified mass, roughly cylindrical but flattened from before back, some 9 mm. long on the median (fig. vi) and 7 mm. on the external face (fig. viii), and bearing at its distal end a flagellum on the external side and the flat triangle on the internal side. This bony mass is set on the protopodite
ORGANS FOR SPERM-TRANSFER 265
by a very stiff oblique joint at about 45° and allows of very little lateral and rotative motion. It may be forced outward and inward through but few degrees, its tip traveHng only 4 mm. It may be twisted so that the triangle, from being almost concealed dorsal to the end of the bony mass (fig. v), may be turned outward a few degrees toward a horizontal position and present more of its median face, somewhat as in fig. vi. The movement is comparable to that of a stiff arm that should allow only a little sidewise movement and a verj- little twisting at the elbow with the end result that the triangle, or hand, at the end, accomplishes a little adjustment to the orifice of the first stylet. This is done as if by supination, though done by the above twist at the elbow.
The flagellum is the real termination of the endopodite; it is some 3 mm. long, 1 mm. wide and rapidly tapering, also flattened, being a long triangular terminal tip to the essentially flat endopodite. By the presence of white lateral areas in the otherwise membranous flagellum, it is obscurely divided into 9 or 12 joints. At the tip and along the sides it bears long plumose setae that are often sparse or worn off along the outer side. The flagellum springs from a socket in the bony shell of the wide end of the endopodite. The external angle of the edge of this socket, figs, v and VIII, forms a hard protuberance at the end of a bony ridge (the Guide). The setae along the flagellum as well as those along the exopodite do not stand out horizontally, right and left, but slant ventrally, or posteriorly, (fig. viii).
The most novel and characteristic part of the second appendage of male crayfishes is the lateral outgrowth which we will call the triangle. It is a form of the Decapod appendix masculina of Boas. The triangle stands up dorsally so that at rest, it, with its fellow of the other side of the body, fits into the squarish cavity left between the two necks of the first stylets. It is not well seen normally from the ventral view, (fig. v) but it may be pulled outwards through 90° and then looks as in the median view (fig. vi). It is a flat triangular outgrowth, partly calcified and partly membranous. The edges are calcified and the centre membranous, so that the whole suggests a bent arm or wing with skin stretched across it. Each long side of the triangle is about 3 and
266 E. A. ANDREWS
the shorter base about 2 mm. The bony rhns of the triangle as seen in fig. vi may be called the humerus and the radio-ulna.
The distal free part of the apparatus, ,(figs. vi, vii) is a trihedral mass set with long plumose setae and might be likened to a sort of hand at the end of the fore-arm. We will call it the wedge from its appearance and use as seen in sections (fig. 9).
The humerus articulates at each end; proximally loosely with the side of the exodopodite mass, (fig. vi) ; distally at the elbow, firmly with the other firm edge of the triangle, the radio-ulna. On the external or concave face of the triangle, (fig. viii) the humerus is not as well separable from the membranous part of the triangle, and between its proximal end and the bone of the main mass of the endopodite there is more or less expanse of membrane. On this outer face, (fig. viii) we find that all the concave aspect of the triangle is membranous.
The humerus is wide and smooth and flat on the inner face, fig. VI, but on the outer face forms only a narrow edge to the membrane, fig. VIII,
The soft hollow face of the triangle in life is swollen with contained liquid. The soft area is not only the outer face of the triangular protuberance but also half of the dorsal face of the distal part of the main trunk of the endopodite.
The whole darkened area of fig. viii might be compared to the soft inside of the palm of a hand and it is this which comes against the neck of the first stylet, in conjugation.
While the humerus is wider toward the base and slender at the elbow end, the radio-ulna is the reverse; that is, it begins narrow at the elbow and widens to the hand or terminal part. . The radio-ulna is a thick plate-like mass that is not in the same plane as the humerus, but about 45° with it, so that it has the appearance of a scroll rolling in over the depressed membranous outer face of the triangle, (figs, vii, viii). The radius part is the free rounded edge, (fig. viii) and this ends abruptly opposite the base of the hand, which is back of it in the figure, while the ulna plate runs on continuously in the background of this figure and passes imperceptibly into the hand, or wedge, (fig. vii).
The radius stands free, away from the membrane, as a rounded
ORGANS FOR SPERM-TRANSFER 267
bony ridge much thicker than the ulna plate from which it is faintly marked off by a suture. Thus in sections (fig. 8), the radius looks like a head on a slender neck. The abrupt termination of the radius is very like the elbow end of the human radius, a shallow cup. The actual cup is made by clear horny matter of considerable thickness and is prolonged as a horny sharp ridge all along the radial edge of the pyramidal wedge. The head of the radius stands out as wider than the neck (fig. vii).
The ulna is but a vaguely defined thick area of the general shell and it continues as the hand or wedge, which is, next to the head of the radius, the most peculiar part of the triangle. This wedge is a hard horny pyramid of three faces. One is rounded and setose, two flat, meeting at a sharp edge, (see small sketch, fig. 22). Its exposed rounded face (figs, vi, vii) is set with a dense brush of plumose setae. The external or ulnar face (fig. vii) is smooth bone, bearing setae along its right edge and ending, to the left, in the sharp horny ridge that runs up from the head of the radius and is shown as a dark shade in fig. v. The concealed innermost face is bony and contains orange pigment; along its left edge it bears setae (fig. viii), and its right edge is the sharp horny membrane that runs up from the head of the radius. In the union of this face with the soft membrane of the concavity of the triangle there is a bony articular plate.
The photographs do not represent one feature of the triangle and that is the small tuft of some five or six, or so, very wiry bent plumose setae that spring from the elbow of the triangle and, for the most part, curve so as to lie down close to the soft membrane. These setae are roughly shown in fig. 22 at the elbow. This also gives in the side sketch, an end view of the head of the radius as seen when the base of the wedge was cut off and the stump of the ulna and free end of the radius viewed from the face where the wedge had been. This is intended to show the head of the radius as a rounded saucer with flat bottom, not deep, but with flaring and rounded sides that form a rim thicker than the neck of the radius below. The cut off setae in this figure are the bases of those on the union of ulna and wedge, just above the level of the line 23 in the main fig. 22.
268 E. A. ANDREWS
INTERNAL ANATOMY OF THE SECOND STYLET
Dissections and sections showed the presence of the same general structures as in the case of the first stylet, with the important difference that the special glands of the tube of the first stylet are absent and on the other hand the intrinsic muscles that are absent in the first are well developed in the second stylet. The muscles are arranged as in the younger stages (figs. 27, 28). Besides the three muscles at the base that pass into the sternum of the second abdominal somite a very short distance there are long muscular strands within the stylet itself.
The protopodite springs from a considerable elevation of the sternum and in the adult two muscles were found within this elevated articular region. Pulling one tended to depress the stylet into its position of rest while the smaller muscle was thought to be probably concerned with the erection of the stylet. Pulling all the basal muscles made the stylet not only lie down but also move toward the median line, which would enable it to fit in nicely with the first stylet. Some of these extrinsic muscles extend a distance into the protopodite itself, to be attached to the shell. There are also long strands arising from the shell of the protopodite and running to the exopodite and the endopodite. Those of the exopodite seem associated with the basal muscles, so that pulling the muscle in the sternum made slight twitching movements of the exopodite, simulating those seen during conjugation, which may thus be caused by contractions of the muscles that hold the entire appendage in position. Pulling the muscles that are in the distal part of the protopodite made both exopodite and endopodite move dorsally and also away from the median plane.
The muscles that move the exopodite are better developed than those of the endopodite. Within the exopodite there is a long intrinsic muscle that would seem fit to bend the slender filament slightly. Inside the endopodite, beside the slight muscles of the base concerned with the movement upon the protopodite, there are in the adult two slight threads that represent the muscle seen in early stages (figs. 27, 28) passing from the terminal flagellum down
ORGANS FOR SPERM-TRANSFER
269
into the region whence springs the triangle. These muscles are seen in the sections of the triangle (figs. 23, 24) as two black dots. These sections, with those in figs. 8, 9, show the anatomy of the triangle. Fig. 23 is a section along the Une 23 of fig. 22. The great thickness of the calcified shell is shown by the black mass. The membranous parts are shown by the thin black, as to the left in fig. 24. The cavity within is blood space traversed by connective tissue strands and faced by epidermis against the shell. In
Fig. 23 Cross section of the triangle on the level 23 of fig. 22.
Fig. 24 Cross section of the same about the level 24 of fig. 22.
90 mm. A. 90 mm. A.
fig. 23 the triangle and the distal part of the protopodite are cut across with the hollow face to the left. The dense shell mass to the left above is the guide ridge, (fig. viii) which somewhat overhangs the cavity of the triangle and bears on its median face some setae, (fig. vi), which are connected at the root with the epidermis by the long canals of which one is seen to the left (fig. 23) penetrating the shell. Opposite this on the median face of the endopodite there are also a few setae which do not appear in the
270 E. A. ANDHEWS
photo^ruph (lig. v) but present one of their canals in the shell of lig. 23, to the right. In contrast to the excessive thickness of the shell of this main stem of the endopodite, the triangle, as represented by the lower part of this section, is relatively thin shelled. The radius is th(^ thick knob in the lower left corner. The shell to the I'ight is the ulna, the thick mass against the concavity of the triangle is in reality more membranous than calcified, but as yet thick. But further toward the elbow (fig. 24) along the hne 24, (fig. 22), the corresponding region is a thin membrane reaching from tlu^ n(^ck of the radius across to the thick guide ridge. In reality the elbow stands out more as in fig. viii so that the width of the section 24 is nmch greater than fig. 28. Fig. 24 shows clearly, on the riglit, the hinge-like line of demai-cation between the outstanding triangle and the main stem of the endopodite, being in fact cut at the edge of the proximal articulation of the humerus (fig. vi), wliere there is a sudden change in level in passing from the humerus to the main stem. In sections 23 and 24, the small black dots above within the connective tissue, are the muscles that iim up int-o the flagellum, much as hi fig. 28.
Section 8 shows the radius standing out from the flat triangle with the thick mass of the humerus above in the figure, wliile fig. 9 shows the thick end of the endopodite above and in the groove of the first stylet the cut off wedge, as will be described below in considering the :idjustm(Mits of the first and second stylets during conjugation.
ON'iXKVKNY OF TITi; A(XM<]SS()HV, Oil Sl'X'ONI), STVLl'^T
Between the individual development of the first and the second stylets there is this imi)ortant diffeivnce that while the first never at any time looks like one of the ordinary pleoiK)ds but is of late appearance and is also a dwarfed, specialized, or reduced appendage from the first, the second appendage is present as soon as the others are and is at first like the ordinary appendage and becomes specialized by the addition of an outgrowth and not by the loss of parts.
ORGANS FOR SPERM-TRAN8FER 271
The plcopods of the second, third, fourth and fiftli somites of both males and females are represented at the time of hatching and all alike have the appearance seen in i\g. 2() whi(;h is magnified 75 diameters and represents the anterior face of the third left pleopod of a male 18 mm., in July, when in the fifth larval stage. The pleopod is flat and translucent; the endopodite (En.) is longer than the exopodite {Ex.) and both are fringed by long setae that are really plumes, though not so figured. Both endopodite and exopodite are obscurely joint(Hl and the proto]x)dite has a short annular segment as well as a long main segment. Through the thin shell may be seen the muscles, represented by the dotted lines. At the base are three large and one minute muscles; two of the main three are posterior and one anterior, and apparently the movement of the entire appendage would be a more powerful backward swing and weak forward recovery, as in swimming. Within the main segment of the protopodite are three long muscles that would seem to aid in bending the appcnulage at its base, while distally there are two muscles which both go to the exopodite to move it. The endopodite is left with only intrinsic muscles to move it at its base and with a long branched muscle that can act only to bend the endopodite itself. Th(^ exopodite has also intrinsic muscles at its base as well as i\w musck^s of the protopodite to move it. There is likewise a long branched muscle to bend the exopodite.
In the early stages the second appendage of the male is quite like this third pleopod, but in a male of 21 mm. (probably in the same larval stage as the male having the third appendage shown in fig. 26) we find th(^ pleopod of the second somite modified as in fig. 27, that is, there has been added to it the excrescence seen on the median side of the endopodite. This is to become the triangle or appendix mascuUna of the adult.
The first discovered trace of this outgrowth was seen in a larva of the fourth stage, 11 mm. long, in July. This first beginning of the triangle is the slight elevation {x) seen in fig. 25, on the side of the endopodite. This figure represents only that part of the endopodite which is not well jointed and forms a sort of base beyond which is the more flabelliform distal part, (fig. 26). It
jotniNAi, OK Moupiioi.ofiy, vol,. 22, NO. 2
will be noted that the row of plumes on the right, or median side
of the endopodite (fig. 25), is interrupted distally so that there is
a blank space where one would expect one or two setae, and in
this space there protrudes to the right a rounded elevation. The
position of this slight elevation with reference to the muscles leaves
Fig. 25 Posterior view of basal part of the endopodite of the accessory stylet
of a male 11 mm. long. Enlarged 215 diameters.
no doubt that it is the same thing as the larger elevation of the next larval stage (fig. 27) . In the preparation the epidermis, not here shown, grew out to form this elevation as a hollow outgrowth, leaving no question as to the possible artificial nature of the bulging of the cuticle shown in fig. 25.
OKGANS FOK SPERM-TRANSFER
273
In the fifth larval stage (fig. 27) the protopodite has become wider and stouter and the basal part of the endopodite is much expanded distally where the protuberance arises from it. The
26 27
Fig. 26 Anterior face of third left pleopod of a male 18 mm. long. 2. D. Fig. 27 Anterior faoe of left accessory stylet of male 21 mm. long. Enlarged 75 diameters.
result is that the exopodite begins to take on that relative insignificance in size, characteristic in the adult accessory stylet.
274 E. A. ANDREWS
The new growth on the median side of the basal part of the endopodite (fig. 27), is a sort of knob set on a neck and indined at about 45° to the axis of the endopodite. Its form is not spherical but rather more that of a short cylinder on a slightly shorter neck. The long axis of the cylinder and of the neck is at an angle of 45 degrees to the side of the endopodite. Not only this protruding knob must be reckoned as part of the future triangle but also the neighboring widened area of the endopodite which is depressed as indicated in the shadow in fig. 27 and which will be the depressed anterior face of the future triangle. In fact this depression is accentuated by the position of the knob, which not only stands out as represented in the figure but also rises up toward the observer; that is, anteriorly away from the general plane of the endopodite. The base of the flabeUiform distal part of the endopodite is continued on to the external distal corner of the basal region of the endopodite as a ridge standing up above the depressed area, and forpiing what will be the guide ridge of the perfected organ.
In a small male, 38 mm. long, in October, the second pleopod had advanced to the state of perfection shown in fig. 28, which is an external view of a left accessory stylet, which was about three times as long as the one shown in fig. 27. The muscles in the protopodite remain as before, though not so well seen from this point of view, and the same is true of the endopodite and the exopodite. The protopodite and the exopodite have grown so large and massive that the slender exopodite is much subordinated. The great increase in the basal part of the endopodite, along with the enlargement and specialization of the triangle, leaves the plumose terminal part of the endopodite as a slender palp-like remnant of the original end of the endopodite. The triangle is now so much longer at its free edge than at its attached part that it has the adult triangular form when seen from the median face ; or more explicitly, the obliquely set cylindrical knob of fig. 27 has grown so much longer at its free edge than at its attachment that the length between its ends about equals the distance of the proximal end or elbow from the main mass of the endopodite, which
ORGANS FOR SPURM-TRANSFER
275
29 28
Fis. 28 External face of left accessory stylet of male 38 mm. long in October, enlarged 25 diameters.
Fig. 29 External view of the united first and second stylets of the left side of an adult male, 110 mm. long. 2. 90 mm. Oq.
Fig. 30 View of the median face of the same. 2. 90 mm. Cq.
Infigs. 29, 30, 31: I| = 1st stylet II = 2nd stylet 5 = crossed fifth leg seen in section
5' = other fifth leg not crossed. C =canula Sp = spatula.
276 E. A. ANDREWS
gives the wide scalene triangle as seen from the median side (fig. vi). The proximal elongation of the cylinder makes the elbow of the triangle, while the distal elongation has made the pyramid or wedge that runs up toward the flagellum of the endopodite. As yet no setae were seen on the wedge. The triangle, however, is not merely a flat plate that grows out diagonally, but from the first it is thick through in the anterior-posterior direction, thus producing the cylindrical edge seen in fig. 27, where the thick edge is restricted and marked off by a less thick neck; moreover the thickening of the cylinder is toward the anterior face. By the stage shown in fig. 28 there is great thickening toward the external face. Moreover the external free edge of this thickened cylinder is now itself thickened as a ridge hanging out from the ventral rim over the depressed area as indicated by the broken line in fig. 28. This rounded thick edge is the future radius. (Compare figs. 28 and viii.) From this state it is an easy transition to the more sculptured form of the appendage seen in adults.
The second pleopods of the male thus owe their special structure to a gradual emphasis of the endopodite and protopodite with the addition of an outgrowth peculiar to these appendages, the triangle. The triangle at first is a mere blister on the median side of the endopodite but soon becomes an oblique plate that is surmounted by a thickening. The plate grows anteriorly and the thickening of its free edge becomes longer than the base of the plate, with a resulting triangular form as seen from the median face. The thick ridge grows out externally and this extension itself acquires a thickened rim, posteriorly, which is the radius.
The triangle is thus a triangle only as seen from the median face of the pleopod, in its entirety the triangle is a curved object like a half open hand, and as such is capable of being applied to the rounded surface of the first stylet. It is made of a cylinder obliquely set along the edge of a plate and curving over it, like fingers over the palm. A slip of paper if cut of angular form and bent twice at right angles may be made to represent the stylet.
ORGANS FOR SPERM-TRANSFEK 277
USE OF STYLETS IN CONJUGATION
The way in which the various parts of the stylets are used in the process of conjugation and sperm transfer has been found out partly by direct observation, partly by experiment, and partly by more indirect inferences that still leave some questions unanswered.
The phenomena of conjugation in general have been described elsewhere (5) and we will here consider chiefly the use of the stylets. There is a stage in the early part of conjugation, where the male has seized the female and clasped all her claws, when he rises up away from her sufficiently to allow the pleopods to swing back and forth. In this swinging the long stiff stylets and accessory stylets take part and then are soon locked together, after which the stylets are held by the crossed fifth leg so that henceforth they make a rigid mass which cannot be folded down against the thorax again by any pressure until that fifth leg is removed. The process of locking together of the stylets is as follows :
The swinging of the pleopods is caused by their basal muscles ; and likewise the muscles in the bases of the stylets move them slightly backward, or erect them, and forward, or depress them. While both first and second generally move together and right and left alike, they have been seen to move independently. By a special movement of the second stylets they are clasped against the first in such a way that the triangle is applied to the neck of the first stylet. By arching the abdomen, cat-like, the second stylet is drawn up dorsally along the first, and then, by partial relaxation of the arch of abdomen, the second is shoved distally, along the first, while held tight against it; the result is that the wedge glides along in the groove of the stylet and the radius enters into the inner tubule through the flaring orifice and is shoved in so far that it remains fast. In sections (fig. 8) it is seen that radius fits into the groove as in a socket and, all the walls being thick and solid, the radius cannot be forced out again without running it back along the orifice. The fact is that the locking is very firm and when one tries to pull the second stylet backward the first is dragged with it and only by pulling the second dorsally toward
278 E. A. ANDREWS
the base of the first can one separate the two, as by that means the radius is brought to the orifice out of which it readily passes.
When the two stylets have been erected by their own erector muscles and locked together by their muscular movements which lead to this mechanical fastening of the edge of the triangle within the groove, they form one organ, physiologically, which is to transfer the sperm without any further muscular activity within it.
The appearance of the two locked organs is indicated in the somewhat diagrammatic sketches 29, 30. In 29 the external view of the left stylets and part of the fifth thoracic legs is shown. The second stylet, to the right of the figure shows the solid tip region of the endopodite applied closely against the most protuberant part of the posterior face of the spiral of the first stylet, while the terminal flabellum runs along parallel to the canula and spatuula. In fact .the tip of the bony endopodite seems to overlap the contours of the spiral and this is due to the soft nature of the depressed region of the median face of the end of the endopodite as is seen in fig. viii. The guide ridge is the part seen external to the spiral in fig. 29, while the soft surface is squeezed against the rounded face of the spiral and the triangle is applied close against the median face of the spiral so that it can be seen only from the median view.
Turning to the median view we see, (fig. 30) the triangle lying over the neck and extending out along the groove. The elbow of the triangle lies over the orifice. The radial edge of the triangle conforms with the obliquity of the groove since both the wedge and the radius are firmly inserted in the groove.
Figures 29 and 30, show the supporting fifth leg in section, as a rounded cross-hatched area. It will prevent the locked stylets from being shoved forward, or closed up against the sternum anteriorly. It is also obvious that the movement backward toward a vertical position will be hindered, not only by the inclination and rigidity of the basal joint of the first stylet, but by a like joining of the base of the second stylet, since one cannot move back without the other, for the radius and wedge will go no further
ORGANS FOR SPERM-TRANSFER 279
toward tip of groove. The second forms a mechanical brace tending to hold the first from going backward.
In order to separate the two the second must move toward the animal and glide along the first till free from it. And this motion is actually seen. The locking is not always done without trial and may be broken and renewed during conjugation, so that we often see two positions of the stylets, that of perfect locking,
Fig. 31 Same view when the accessory is drawn back into position of recession
showing the papilla at the mouth of the groove.
as in figs. 29 and 30 when the triangle is most advanced toward the tip of the spiral, and a preliminary and alternate position of recession when the triangle is applied against the base of the first style proximal to the orifice. This position of recession is shown in fig. 31. The triangle goes as far toward the basal end of the first stylet as possible, till stopped by the knob on the base (fig. ii). In this recession the orifice with the papilla meeting it, is exposed and the ventral lip is seen.
It should be borne in mind that the back and forth play of the triangle on the first stylet is limited not only by the knob basally and the narrowness of the groove that prevents the radius from going into it dorsally beyond the position of figure 30, but it is limited laterally by the fact that the triangles of the two sides
280 E. A. ANDREWS
of the body are in contact and are held together by being placed in the squarish hole between the necks of the first two stylets.
The two triangles play back and forth like two hands with bent fingers, back to back, in a narrow space between the first stylets and, like hands, each runs its palm or soft flat surface along the median constricted part of the first stylet and the firm guiding ridge — its thumb, as it were — along the external face of the stylet (fig. 29). In one case from 3 to 4 seconds were taken to glide the triangles back from the normal position to the recession (fig. 31) ; there they remained four or five seconds and advanced strongly in two seconds. Another recession took 12 seconds, but the advance occupied 2 seconds.
If we imagine figure v apphed to i, vi to ii and vii to iii, VIII to IV, we will appreciate how nicely all the surfaces adjust themselves. The oblique ridge of the external mass of fig. i is overlaid by the soft depressed area, (figs, viii, 22) so that the thumb-like guide shows external to the ridge as in fig. 29.
In life the two sets of appendages, right and left, are so closely applied together that the median face of neither can be seen, directly, without mutilation experiments on one side, but the presence of the guide ridge along the external face of the spiral (fig. 29) enables one to judge where the triangle must be at any stage of advance or recession, a matter of importance in deciding as to its use in sperm transfer.
That an application of the second, or accessory stylet, to the first is necessary for the completion of normal conjugation and the filling of the sperm pocket by transferred sperm, was determined not only by the above facts of structure and use but by the following experiments. The instincts of the male are so strong that, when in the process of conjugation the second stylet on one side was cut off, there was no immediate visible effect, except the escape of some blood from the stump of the appendage. And when on the next day all the stylets, both first and second, were cut off, the male seized and turned a female and carried the conjugation as far as possible in the absence of the organs of transfer. The instincts thus go on without the means of carrying them to completion.
ORGANS FOR SPERM-TRANSFER 281
It was then easy to get males to begin conjugation when the accessory stylets had been removed from both sides. Three such males made conjugation experiments with several females, successively, but in no case was there an evidence that the annulus had been filled by these mutilated males, through in one case the union lasted for eight and one-half hours. In these attempted conjugations it was not evident how the absence of the second stylet prevented perfect sperm transfer. In one case the male let fall three or four sperm masses, or pseudo-spermatophores, about 1 mm. long on the telson of the female but it was not determined how this happened. Apparently this was from failure to have a close union at the orifice, which would lead one to think the failure due to absence of the triangle that normally holds the papilla tight to the orifice. But the failure may have been due to the absence of piston like movements of the radius. More experiments should show what the uses of the different parts of the triangle really are.
HOW THE SPERM IS FORCED ALONG THE TUBULE OF THE STYLET
The adjustment of the papillae, whose anatomy has been described, to the stylets must now be considered in order to appreciate the final use of the stylet.
As seen in fig. 1, the papilla juts out toward the median plane so far that it can be placed across the narrowest part of the first stylet where the notch is (fig. in); that is across the dorsal face of the first stylet. But its tip turns abruptly inward far enough to reach along the median face (fig. ii) as far as the orifice, into which its tip fits. In figs. 30, 31, this position of the papilla is crudely represented; in reality the tensely swollen translucent spout is very nicely applied to the rounded faces of the entrance .to the groove. The papilla is seen in this position when the triangle is receded (fig. 31) and in the advance of the triangle its tip becomes concealed, but it doubtless remains as before.
Returning to the actions of the combined stylets which embrace the papillae we note certain 'tamping' movements. Besides the advance and recession of the second stylet along the first, the first and second together when locked, are seen to execute quick jerks
282 E. A. ANDREWS
that carry the tips of the first back and forth a part of a millimeter only. \Yhen the tips of the stylets have gained entrance into the annulus, these thrusts may serve to introduce the tip farther into its cavity. As in the movements of recession the force here must be exerted by the muscles of the abdomen, as the stylets themselves have no telescopic power; and actual twitching of the anterior part of the abdomen were seen.
SPERM EMISSION AND CONDUCTION
In normal conjugation nothing is seen of the sperm so that its transfer from the deferent duct to the cavity of the annulus is a matter of inference. The papilla is applied to the orifice of the tube of the first stylet so that it may discharge into it and sections show the tube full of sperm, (figs. 14, 15) ; moreover in some abnormal cases the sperm is seen to issue from the tip of the canula into the water, and, as the tip of the canula is normally inside the sperm pocket, it is evident that the sperm must pass along the stylet from the papilla. The force that propels the sperm is no doubt muscular contraction, but it is not clear at first what muscles are concerned; there are none within the first stylet which acts merely as a passive tube.
From such figures as 2, it is evident that the deferent duct has powerful transverse muscles that could squeeze out the sperm with force and this seems the main if not only motive force to carry thie sperm through the papilla and all along the tube of the stylet into the annulus.
The force necessary to propel the liquid sperm through a tube that is only some 20 to 40 ^ in diameter (figs. 13, 15) is great and attempts to force india ink through the tubule of the stylet with a small hypodermic syringe failed. When the specially ground canula was inserted into the orifice, while the radius was engaged in the tubule, no ink could be forced out of the tip of the stylet. It was inferred that the radius blocked the way, as it fits in so as to nearly occlude the lumen (fig. 8) , but the same failure was met with when the triangle was removed from the stylet, but then the ink jetted out along the proximal part of the groove where the
ORGANS FOR SPERM-TRANSFER 283
triangle had been. Apparently the wedge of the triangle is well fitted to hold the liquid in the tubule since it fills up the groove external to it (fig. 9), where the sides of the groove are not as close together as they are distally (fig. 10), which is beyond the wedge. When the ink had been introduced into the tubule and not forced out of the tip of the stylet the triangle was applied to the stjdet and the radius worked back and forth like a piston in the tubule with the result that some of the ink issued from the tip' of the canula of the stylet.
This suggested that the radius might act like a piston in normal sperm transfer and thus propel the sperm from the papilla along the tubule to the annulus. We also saw that when a pair was separated in conjugation the sperm that issued from the tip of the canula of the stylet was mixed with bubbles of air when held out of the water, which suggested some action at the base of the tubule (at the orifice) to draw the air into the tubule. However, this might be movements of the triangle or simply failure of the triangle to hold a tight joint around the tip of the papilla and orifice, for thus air could be drawn in by the stream of sperm advancing, driven by pressure of the muscles of the deferent duct. When the radius was inserted into the orifice and shoved along in the tubule, sperm was forced out of the tip of the canula, which seemed to demonstrate the ability of the radius to act as a propelling piston.
We failed to detect any such piston motions during conjugation, but they would be of very slight extent and not readilj'^ observed. The movements of advance and recession described above are of a much grosser magnitude than the piston movements that might be supposed to take place. The movements 31, 30 are only for getting right adjustment of the enveloping triangle over the papilla tip and the entrance of the radius into the tubule so that the hand-like triangle may make such tight binding of the papilla to the orifice that no sperm escapes or comes into contact with the water. Yet the piston may then presumably be in position to advance or recede a little. When we thrust the triangle strongly so far along the stylet that the elbow was at the orifice, (fig. 30), the triangle tended to spring slowly back out of the groove
284 E. A. ANDREWS
till only half of the length of the radius remained in the groove, owing apparently to the elastic side walls of the groove shoving against the wedge (figs. 9, 10) as these walls are the closer together toward the tip of the stylet.
By this mechanical means the piston might tend to recede, while the movements of the muscles of the abdomen might make the entire second stylet advance enough to shove the piston along the groove again. We can easily pump the radius back and forth in the groove by moving the whole second stylet. The muscles of the abdomen make the slight twitching back and forth jerks of both first and second stylets above mentioned as tamping movements. Now after the first stylet, with the second locked to it, is introduced into the cavity of theannulus as far as possible, these movements of tamping, if they be continued, could not advance the first stylet but may push the second further along the first and so cause the piston to act on the sperm. The dish-like head of the end of the radius (fig. 22) receives explanation upon the assumption that it is useful in shoving the sperm along in the tubule, in fact, the solid bone-like piston with horny cupped tip provided with elastic flaring edge seems a remarkably well made apparatus for pushing liquid along in a tube that it fits so well.
Some such piston movements might be expected from the statements of Coste (C. R. 46, 1858), that Gerbe in his laboratory saw the male Astacus apply the foliacious part of the second stylet to the first stylet and by reiterated back and forth motions during the passage of sperm, keep as he thought, the trough of the first stylet free from sperm that might harden there else. Schillinger, states that the second stylet is used to push the spermatophores out of the first stylet.'
The groove and its concealed inner part that forms the tubule are of course open to the water and if the sperm is to pass free from contact with the water to the cavity of the annulus the assumed piston movements of the radius may serve to clean out the tubule and fill it with harmless secretions. The source of such secretions may be surmised to be the glands in the tip of the pap ' As reported by Ortmann in Bronn's Klassen und Ordnungen.
ORGANS FOR SPERM-TRANSFER 285
ilia (fig. 2) or those along the tubule itself (fig. 8). Possibly this preparatory action of the radius is all that it has to fulfill and that the pressure of the muscle of the efferent duct is all sufficient to cause the sperm to run through the length of the stylet. In connection with this question we have to bear in mind that the sperm is in some way freed from its envelope of secretion made in the efferent duct before it is laid away inside the sperm pocket where it exists pure (1).
This separation of sperm from enveloping secretion takes place in the tubule of the stylet. In the proximal part of the tubule the secretion of the deferent duct (fig. 2), is still all around the strand of sperm (fig. 14), but distally the sperm is almost pure inside the tubule (fig. 15).
We found also that in one case a male, fallen on the side while still holding a female, had the stylets only partly erected so that they were free in the water and from the tip of each canula a very fine stream of sperm, finer than the tip of the spatula, issued slowly and coiled up in a small mass. From one canula the sperm then slowly sank in ten minutes down in the still water as a fine thread with a coil at the tip. Another male showed faint sperm jelly on the tip of the flagellum of the endopodite of the second stylet and this was pure sperm becoming modified by the water; there was no secretion.
There are however besides these escapes of pure sperm, escapes of sperm inside of secretions that resemble spermatophores. In a male, in which the triangle was in the position of recession, (fig 31), there were such white sperm threads, | to 1 mm. long, about the orifice of the groove. The pseudo-spermatophores that in abnormal or interrupted conjugations were sometimes seen were soft, paste-like tubes containing a central mass of sperm. The short pieces of tube stick by their ends to the inside of a pipette used to pick them up and to the shell of the crayfish on which they fall.
The wall of these tubes is a very thin layer of secretion which is vesiculate and stringy like dough and can be drawn out into clear threads with minute droplets along them. These would seem to be not normal spermatophores, which in Astacus have
286 E. A. ANDREWS
thick walls, but only rods of sperm enveloped in some slight secretion from the deferent duct, or possibly that of the papilla or of the glands of the spiral. The thin walls of these tubes break open, hernia like, and sperms ooze out.
The separation of the sperm from the secretion of the deferent duct may be due merely to the diminution in diameter of the tubule; the pressure of the duct would drive the central part of the current faster than the envelope and thus the central sperm might flow out of the very narrow tip of the canula and leave the envelope of secretion behind in the wider parts of the tubule. Finally, when enough sperm had passed along to fill the annulus, the enveloping secretion might be forced out and this would make that wax-like mass that fills the external parts of the annulus and projects in excess from its mouth as the so-called sperm plug. Possibly again the piston movements of the radius might come into play to clean out the secretion from the stylet tubule and ram it into the annulus. In fact in the last stages of conjugation of one pair slow and repeated movements of advance and recession of the triangle were seen which may be interpreted as concerned with plug making.
The use of the glands of the spiral is not known. Possibly their secretion cleanses the surfaces to be used in sperm transfer and aids in keeping water from the sperm. Possibly the secretion may help the enveloping secretion of the deferent duct to adhere to the walls of the tubule of the spiral and thus hold it back till the sperm has passed on into the annulus.
THE RIGHT AND LEFT DUPLICATION OF STYLETS
The striking fact of the exact duplication of both first and second stylets right and left suggests questions as to the use of right and left in conjugation. Are both sides used at each conjugation?
Again the remarkable dimorphism of the females of C. affinis and of probably all other species of that genus, which expresses itself in the occurrence of females with the vestibule of the sperm pocket opening a little to the right of the middle line and of females with the pocket opening to the left, so that the symmetry
ORGANS FOR SPERM-TRANSFER 287
of the two i.s reversed, raises the question as to whether the males are adjusted, in habit, to these two kinds of females, so as to use the right set of sperm transfer organs for a left-handed female and vice versa, or not.
The crucial experiments to determine whether males actually use the one stylet for right and the other for left-handed females have not yet been made. However, some facts and considerations make it improbable that a male is obliged to do so and indicate that a male may adjust his stylets so as to use either right or left on any form of female annulus, leaving the question still open as to what is the normal habit of the males with reference to the two forms of females.
In the first place we found that though the two first stylets seem to be in the annulus they are never both firmly inserted. One is fixed firmly by its tip while the other may be drawn away by a pair of forceps. Moreover the one that is inserted has its tip some f to If mm. in advance of the other and its base is locked against the base of the other, diagonally, the abdomen being advanced more on one side than the other.
Observations showed that not only were there cases of the right stylet in the left annulus but of right stylet in the right annulus and of left stylet in left annulus and of left stylet in right annulus. Whether in these cases the sperm was actually transferred was, unfortunately, not made out. It is possible that a male may insert one stylet and afterwards the other till finally the actual sperm transfer takes place with some more definite reference to the symmetry of the annulus than the above observations would indicate.
That there is any alteration in the advance of the stylets was not made out, but there is often an alteration in the use of the fifth leg, right and left. At any one time many males will be found with the left and others with the right leg crossed, but continuous observations show that the male will change from right to left in difficult cases especially, till a better adjustment is obtained.
It was at first thought that there was a relation between the fifth leg and the advance and use of the first stylet so that these were on the same side, that is, the stylet being advanced by the
JOURNAL OF MORPHOLOGY, VOL. 22, SO. 2
288 E. A. ANDREWS
use of the leg of that side of the body, but cases were recorded in which the advanced stylet was on the opposite side from the crossed leg. Males crossed the right leg with either right or left stylet advanced and males crossed the left leg with either right . or left stylet advanced. Here again there is the possible objection that the condition observed was not permanent or the one employed in actual sperm transfer. More minute observation of several normal cases are necessary.
One good case seems, however, rather conclusive. In this a male, in November, crossed the left fifth and advanced the left stylet, but after an hour of attempts to enter the annulus, crossed the right fifth and five hours later the right stylet was one mm. in advance of the other and the female had a sperm plug in a right annulus. Here the leg and stylet used did coincide, but the annulus was not the one to be expected.
Again in some conjugates killed by boiling while united it was found that in one a right stylet was advanced to a right annulus and in others a left stylet to a left and to a right annulus.
As far as the evidence goes it gives the impression that the male is free to use either right or left stylet with either right or left fifth legs till successful in getting some one tip of the stylets into the vestibule of the annulus, which may be a right or a left one, indifferently. Yet future observations may show that the lines of least resistance are for the male to use the left stylet for the right-handed female, and the reverse, and that this actually takes place, in nature as the normal, though we doubt if it be at all necessary. Observations show that both papillae are ready to discharge sperm at the same time and it should be determined by experiment whether the male uses both right and left sets of sperm transfer organs, alternately, at each conjugation or not.
When the first and the second stylets were cut off from one side of some sixteen males and, either at once or some weeks after, these males were given females, the unexpected result followed that in spite of many repeated attempts, one lasting nine hours, the numerous conjugations of these unilaterally mutilated males did not result in any clear cases of successful sperm transfer. In
ORGANS FOR SPERM-TRANSFER 289
many cases the annuli of the females were artificially cleared out so that any new plugs would have been seen.
Among these cases there were males that alternately used the fifth left and right legs in crossing, though some had only the left series of stylets and others the right; the leg being crossed on the side where there was no stylet and on the side where there was a stylet. And these same cases were attempting conjugation with females that were of both kinds, right and' left forms, so that there was no agreement between the kind of annulus and the fifth leg used.
In only one case was there any sperm seen and this was seen twice in successive conjugations of the same male that seems to have been peculiar. This sperm lay in pseudo spermatophores, 8 mm. long, upon the telson of the female under the left stylet, and probably escaped from some imperfection of the closure of the triangle.
While it was not found out why there was this apparent inability to complete sperm transfer while the stylets of one side were missing it is thought that this is not due to the need of using sperm from both sides of the body at each conjugation but rather to the mechanical factor that the two sets of stylets are always applied to one another so firmly as to hold the tips of the stylets at the annulus, so that when one is absent the tip of the remaining one lacking the usual support cannot be readily brought to the middle line of the body. Moreover it is possible that the triangle will not be well applied to the orifice unless the fellow triangle be there to shove against it, as both are packed in side by side between the necks of the first stylets.
SUMMARY
Though the sperm of the crayfish, Cambarus affinis, is injured by exposure to water, it is transferred from the male to the female under water and stored up in an external pouch.
The part played by the female in this insurance against injury in transit has been elsewhere described.
290 E. A. ANDREWS
The present paper describes only those organs of the male that are combined to form a safe conduit for the sperm from the male to the receptacle on the female.
The actual sperm transit apparatus of the male consists of three organs on each side of the body. The anatomy and use of these three organs are here described in detail.
The 'papilla' or end of the deferent duct is provided with glands and a valve. It is distended by blood and applied to fit accurately to the beginning of a tube.
This tube is the innermost part of the groove of the first stylet, or limb of the abdomen, and hitherto its existence and use has not been described.
The first abdominal limb is, in action, a duct leading the sperm uninterruptedly from the deferent duct into the receptacle of the female. It contains large glands of problematical use, and relies for mechanical support upon the habit of the male in using the second abdominal limb as well as one of the fifth thoracic limbs to insure the entrance of the first stylet into the receptacle of the female.
The second stylet is accessory to the first in applying its handlike outgrowth over the papilla and insuring a tight joint. It also gives mechanical support to the first stylet. How much it may also serve as a piston for cleaning the tube or even for aiding in sperm transfer is left undecided.
The ontogeny of the first stylet shows that it begins after the other abdominal limbs and is from the first a simple unbranched outgrowth which becomes a tube by the depression of its central and elevation of its lateral parts to form a deep groove, the bottom of which is ultimately isolated by a shelf.
The morphology of the organ, based upon its use, anatomy and development, gives the basis for its utilization in defining species and subgenera. The tip or canula that is inserted into the receptacle to discharge sperm is the real tip of the organ and all other tips are to be referred to lateral outgrowths from one or the other side of the original groove.
The ontogeny of the second stylet shows that in the first larva it is just like the following abdominal limbs; but its subsequent
ORGANS FOR SPERM-TRANSFER 291
fate is to add on a lateral outgrowth {appendix masculina) wliich becomes the useful part of this organ when acting as a necessary part of the sperm transit apparatus.
The duplication of all three organs, right and left, seems necessary in, as far as removal of one set leads to the lack of necessary mechanical support for the perfect functioning of the opposite set.
The evidence is against the conclusion that the right and left openings of different receptacles upon different females are necessarily met by the males employing the stylets of one side rather than an other. In each case the male may by trial obtain the entrance of some one of the two stylets into the receptacle of the female.
The extreme solidity of the shell of the stylets is to be correlated with the amount of force exerted by the male in making a water tight passage for the sperm from the deferent duct into the receptacle of the female.
While all six organs are necessary for sperm transfer, most of them may be removed without preventing the males from carrying out many of the stages of conjugation that would normally lead up to sperm transfer.
Many of the peculiarities of the form and structure of the transfer organs are demonstrated to be of use, or even necessary.
The accurate interadjustment of the six organs is necessary for the perpetuation of the species.
It is difficult to believe that in the evolution of Cambarus the increasing perfection of these organs could have been decisive in eliminating the less perfect organs. Astacus survives with more simple organs and the majority of genera of crayfish. have no stylets at all. The perfection of the organs, characteristic of Cambarus may have been brought about from laws of change that it will require much experimentation to discover.
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
292 E. A. ANDREWS
, IJTKRATURE CJTP:D
1 Andrews, K. A. U)()B The luumluw vcntralis. Proc. Boston 8oc. Nat. Hist.
vol. 32.
2 1908 The aiinulus of a Mexican crayfish. Biol. Bull. vol. 14.
3 1908 The sperm receptacle of the crayfishes Cambaius cubensis and C. para
doxus. Proc. Wash. Acad. Sci. vol. 10.
i 1904 Breeding habits of crayfish. Am. Nat. vol. 38.
o 1910 Conjugation in the crayfish C'ambarus^affinis. Jour. Exp. Zoo!, vol. 9.
6 1904 Crayfish spermatozoa, .\natoin. .\nz. vol. 2h.
PLATES 1, 2. 3, 4
EXPLANATION OK FIGXTRES
I. Photograph taken with a magnification of about ten diameters, of the posterior face of the first stylet of the left side.
II. Photograph of the same, taken from the median side, but diagonally, so that the posterior side is also shown in part.
III. Photograph of the same from the anterior face. I\'. Photograph of the same from the external face.
V. Photograph taken enlarged about ten diameters, of the second, or accessory stylet, of the left side of adult male. Posterior face.
VI. The same from the median face. V'll. The same from the anterior face. VIII. The same from tiie externa! face.
OVIPOSITION INDUCED BY THE MALE IN PIGEONS
WALLACE CRAIG
Department of Philosophy, University of Maiiie
The influence of the male upon the time of oviposition is a matter in regard to which pigeons differ from some other birds, notably the domestic fowl. With regard to the fowl I have consulted a number of poultry keepers and experts, chiefly Dr. Raymond Pearl and Dr. Frank M. Surface, of the Maine Agricultural Experiment Station, where the most extensive studies of the egglaying of fowls have been, and are being carried on. Dr. Pearl and Dr. Surface tell me that the domestic hen, and also the hen of the wild Gallus bankiva so far as can be ascertained, commence their spring laying at an approximately fixed date which can neither be deferred by withholding the cock nor advanced by giving the cock before the usual time.
Pigeons differ widely from poultry in this respect. If, from the winter season onward, an old female piegon be kept unmated and isolated, she refrains from egg-laying, in evident distress for want of a mate, until the breeding season is far advanced; at length she does begin to lay, but her laying without a mate manifestly partakes of the abnormal. And a virgin pigeon, if kept isolated from other pigeons, may postpone her laying for a still longer period. On the other hand, a female pigeon, young or old, will lay very early in the season if she be early mated. Moreover, there is a pretty definite interval between the first copulation and the laying of the first egg, namely six or seven days; if the egg be delayed much beyond this time, the fact indicates some indisposition on the part of the female. And as the pair rear brood after brood throughout the season, this time-relation between copulation and egg-laying is regularly repeated.
299
300 WALLACE CRAIG
The utility of this time adjustment in pigeons seems obvious. The male pigeon takes his turn daily in the duty of incubation: hence the female must not lay the eggs before he is ready to sit. This aspect of the matter, which has to do with pigeon sociology, has already been treated elsewhere (Craig '08) and will be discussed more fully in a book dealing with pigeon behavior. The present paper is to show, not why the male should determine the time of oviposition, but how he does determine it.
The thesis of the present paper is, that the influence of the male in inducing oviposition is a psychological influence; that the stimulus to oviposition is not the introduction of sperm, for the male can cause the female to laj^ even though he does not copulate with her. This is easily proven by an experiment, which requires only pigeons, patience, and time, and I shall now recount seven repetitions of such experiment, the first two being accidental cases, the other five being trials designed and carried out on purpose to test the thesis.
Case 1 (1903). In the spring of 1903 I brought together a virgin female dove (individual female no. 7, the species in all these trials being the blonde ring-dove, Turtur risorius) and a young inexperienced male, intending simply that they should mate in the normal manner. The young male played up to the female, but due to his inexperience and to other causes which need not be discussed here, his mating behavior was imperfect and he did not copulate with her. Nevertheless, in due time (six days) she laid an egg, and a second egg, as usual, forty hours later. This was the first intimation to me that a male bird can stimulate the female to lay, without copulating with her. Such an explanation seemed so absurd at that time that I dismissed it with the assumption that the birds must have copulated unobserved, and I did not even test the eggs to see if they were fertile. Looking back on that case now, however, and considering the observed behavior of that male, I feel reasonably certain that he did not fertilize the eggs but simply stimulated oviposition through the psychic (neural) channels.
Case 2 (1904). A female dove (no. 5) had been kept alone ever since her mate had died in November, 1903, and as time wore on
OVIPOSITION IN PIGEONS 301
she showed mtense anxiety to mate. She being a very tame bird, I had often caught and held her gently, but she did not like to be held, so one day in early March I tried tickling her head and pulling the feathers about her neck somewhat as a courting male would do it, and, finding that the poor lonely bird received these attentions with intense pleasure and became still more tame, I continued to preen her neck daily. She now acted toward the hand as if it were a mate, went through a nesting performance in her seed dish, there being no nest in her cage, and to my astonishment laid her eggs in due season. The first egg was laid March 11 and the second March 13. There is no doubt in my mind that the caressing of this bird's head and neck brought on oviposition. I once tried to repeat the experiment with another female dove, but she would not accept the touch of the hand as the former dove had done. Yet there is other evidence indicating that, with a specially tamed bird, this experiment, inducing oviposition by the hand, could be successfully repeated.
This case called to mind that of 1903, and suggested an experiment to determine definitely whether the male dove can stimulate the female to lay, without actual copulation. Opportunity to try this experiment was not found till 1907 and following years, when it was planned as follows.
Method of the regular trials
The experiment requires an unmated female dove that is not laying eggs, preferably a young dove that has never laid. It is best tried early in the season (e.g., in February), especially if an old dove be used, for, as said above, if the female is kept too long without a mate she may lay without one. Side by side with this female, in a separate cage, is placed an unmated male, and the two are given several days to become acquainted. When they act toward one another like mate and mate, the doors separating them are opened and they are allowed to come together for a time, under constant supervision. When they attempt to copulate, a slender rod which can be thrust between the bars of the cage is used to keep them apart. Such attempts are made many times in a day,
302 WALLACE CRAIG
mostly in the afternoon, and are continued for several days in succession; hence it is best that the experimenter should be able to devote some hours a day for several days in succession to a single pair or at most two pairs of birds. Whenever the birds are not under surveillance they are shut apart, each in his or her own cage. But they should be allowed to come together daily until the egg is laid.
A factor which caused difficulty in one of my trials was the nest. In cases 1, 2, 3 and 6, the bird laid without any nest at all (except that in case 6 a nest was given just a few hours before the egg was deposited). But in case 4 {q.v.) the female refused to lay without a nest : it was then necessary to remove the male and make the trial again, first giving the female a nest, and waiting long enough to prove that the nest alone would not cause her to lay.
Results of the regular trials
Case 3 (1907). Female dove, no. 20. This bird had been bought recently from a dealer, and it was not known whether she had laid earlier in the season. But she was kept isolated for some time, during which she showed no inclination to lay. She was then given a male in the manner indicated. No nest given.
June 9. Male allowed in cage of female, and plays up to her.
June 15. First egg.
June 17. Second egg. (The second egg was of no special interest. After the first egg was laid, I generally left the doors open, allowing the pair to come together without surveillance.)
Case 4- (1908). Female, the same. She had not laid since the close of last season. No nest given.
February 4. Male allowed to enter.
The female was unresponsive and showed by her behavior that this time she was holding back for want of a nest. This deficiency was supplied in the following manner (vide ut supra.)
February 8. Male taken away to another building.
March 10. Nest put in cage. Female paid practically no attention to it. Many days were allowed to pass, in order to make sure that the nest alone would not stimulate the female to lay.
OVIPOSITION IN PIGEONS 303
March 21. Male (after short period in sight of female, that they might become re-acquainted) allowed to enter.
March 27. Egg laid.
Case 5 (1910). Female, the same as in cases 3 and 4. She has laid no eggs since last season (1909.)
January 20. I begin to allow male in cage, at same time putting nest in.
January 29. Egg laid.
Case 6 (1908). Female, no. 19. Virgin, has never laid. No nest given. In this case, the date on which the female was first given the requisite stimulus cannot be stated so definitely as in the other cases.
July 12. Male,- in his cage, placed close to cage of female. Cooing commences. Female so excited that she several times assumes, and maintains in extreme degree, the copulation posture.
July 14. Male allowed into cage of female, but he fights her, so that it is necessary to remove him (otherwise the female might be painfully injured), and to allow the pair a few days more of preliminary acquaintanceship.
July 18. Male allowed to begin his series of daily visits.
July 22. Egg laid.
Case 7 (1910). Female no. 19, the same as in case 6. She has laid no eggs since last summer (1909.)
For several days before contact with the male, a nest was kept in her cage; but she paid no attention to it, showing that the nest alone would not stimulate her to lay.
January 20. Male allowed to enter.
January 26. Egg laid.
SUMMARY
1. In six cases, stimulation of a female dove by a male, without copulation, was followed by oviposition; and in one other instance (case 2), stimulation by the hand of man in imitation of a male dove was followed by oviposition
2. In six of the seven cases (being all except case 3, in which the previous history was unknown), it was known that the female
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
304 WALLACE CRAIG
had laid no eggs previously during the current year. In two of these six cases the dove was a virgin aad had never laid.
3. It is true that the female may, if left without a mate, begin to lay late in the season. Hence it might be suspected that the sequence of stimulation and egg-laying in the seven cases was mere coincidence. But this is precluded, first of course by the fact that coincidences are not known to happen seven times in succession, and further by the following considerations.
4. In some of the trials it was proven that the female when stimulated by the male laid much earlier in the season than she did when not so stimulated. This is shown in the following table.
Female, no. 20.
1908. (Case 4), stimulated by male, laid March 27.
1909. (Control), without male, began to lay May 13.
1910. (Case 5), stimulated by male, laid January 29. Female, no. 19.
1909. (Control), without male, began to lay April 26.
1910. (Case 7), stimulated by male, laid January 26.
5. The interval between the first stimulation by the male, and the laying of the first egg, was as follows:
Case 1. 6 days.
Case 2. (Male not used.)
Case 3. 7 days.
Case 4. 6 days.
Case 5. 9 days.
Case 6. 4 to 10 days, depending on what is regarded as the first stimulation in this case.
Case 7. 6 days.
The average and the variation of these intervals tally closely with the average and the variation of the interval in normal breeding, between the first copulation and the laying of the first egg.
6. There were no exceptions. Ovoposition never failed to follow within nine days after the first contact with the male. (The onlj^ partial failure was that of the first trial in case 4, which was due to faulty experimental conditions.)
OVIPOSITION IN PIGEONS 305
CONCLUSION
These facts make it certain that the male dove can stimulate the female to lay, without copulating with her.
Harper ('04) mentioned the fact that ovulation in the pigeon does not take place until after the bird is mated, but he was in doubt as to how far the influence of mating was a 'mental' one and how far it was a matter of the introduction of sperm. The present paper goes to show that the stimulus to the whole process of egg development and laying is a psychic (neural) stimulus, not dependent upon the introduction of sperm.
BIBLIOGRAPHY
Craig, Wallace 1908 The voices of pigeons regarded as a means of social control. Am. Jour. Sociol., vol. 14, pp. 86-100.
Harper, Eugene Howard 1904 The fertilization and early development of the pigeon's egg. Am. Jour. Anat., vol. 3, pp. 349-386.
THE ANT-COLONY AS AN ORGANISM'
WILLIAM MORTON WHP]ELER
As a zoologist, reared among what are now rapidly coming to be regarded as antiquated ideals, I confess to a feeling of great diffidence in addressing an audience so thoroughly versed in the very latest as well as the very oldest biological facts, methods and hypotheses. I feel, indeed, like some village potter who is bringing to the market of the metropolis a pitiable sample of his craft, a pot of some old-fashioned design, possibly with a concealed crack which may prevent it from ringing true. Although in what I have to say, I shall strenuously endeavor to be modern, I can only beg you, if I fail to come within hailing distance of the advance guard of present day zoologists, to remember that the range of adaptability in all organisms, even in zoologists, is very limited.
Under the circumstances, my only hope lies in appealing to our permanent common biological interests and these, I take it, must always center in the organism. But the point of view from which we study this most extraordinary of nature's manifestations, is continually shifting. Twenty years ago we were captivated by the morphology of the organism, now its behavior occupies the foreground of our attention. Once we thought we were seriously studying biology when we were scrutinizing paraffine sections of animals and plants or dried specimens mounted on pins or pressed between layers of blotting paper; now we are sure that we were studying merely the exuviae of organisms, the effete residua of the life-process. If the neovitalistic school has done nothing else, it has jolted us out of this delusion which was gradually taking possession of our faculties. It is certain that whatever changes may overtake biology in the future, we must henceforth grapple
1 A lecture prepared for delivery at the Marine Biological Laboratory, Woods Hole, Mass., August 2, 1910.
307
308 WILLIAM MORTON WHEELER
with the organism as a dynamic agencj' acting in a very complex and unstable environment. In using the term organism, therefore, I shall drop the adjective ' living,' since I do not regard pickled animals or dried plants as organisms.
As I wish to describe a peculiar type of organism, I may be asked, before proceeding, to state more concisely what I mean by an organism. It is obvious that no adequate definition can be given, because the organism is neither a thing nor a concept, but a continual flux or process, and hence forever changing and never completed. As good a formal definition as I can frame is the following: An organism is a complex, definitely coordinated and therefore individualized system of activities, which are primarily directed to obtaining and assimilating substances from an environment, to producing other similar systems, known as offspring, and to protecting the system itself and usually also its offspring from disturbances emanating from the environment. The three fundamental activities enumerated in this definition, namely nutrition, reproduction and protection seem to have their inception in what we know, from exclusively subjective experience, as feelings of hunger, affection and fear respectively.
Biologists long ago constructed an elaborate hierarchy of organisms. Those of a speculative turn of mind, like Spencer and Weismann, postulated the existence of very simple organisms, the physiological units, or biophores, which, though invisible, were nevertheless conceived as combining the fundamental activites above enumerated. These biophores were supposed to form by aggregation the cells, which may exist as independent organisms in the Protozoa and Protophyta or unite with other cells to form more complex aggregates, for which Haeckel's term 'persons' may be adopted. The person may be merely a cell-aggregate or consist of complexes of such aggregates as the metameres of the higher animals, for the separate metameres, according to a very generally accepted theory, are supposed to be more or less modified or highly specialized persons. Somewhat similar conditions are supposed to obtain in the composition of the vascular plants. The integration both of the metameric and non-metameric Metazoa may proceed still further, the simple persons combining to
THE ANT-COLONY AS AN ORGANISM 309
form colonies in which the persons are primarily nutritive and acquire fixed and definite spatial relations to one another, whereas the more specialized animals, like the social insects, may constitute families of mobile persons with reproduction as the 'Leitmotiv' of their consociation. In man we have families associating to form still more complex aggregates, the true societies. Other comprehensive organisms are the coenobioses, or more or less definite consociations of animals and plants of different species, which the ecologists are endeavoring to analyze. Finally we have philosophers, like Fechner, stepping in with the assertion, that the earth as a whole is merely a great organism, that the planetary systems in turn are colonies of earths and suns and that the universe itself is to be regarded as one stupendous organism. Thus starting with the biophore as the smallest and ending with the universe as the most comprehensive we have a sufficiently magnificent hierarchy of organisms to satisfy even the most zealous panpsychist. As biologists we may, for present purposes, lop off and discard the ends of this series of organisms, the biophores as being purely hypothetical and the cosmos as involving too many ultrabiological assumptions. We then have left the following series: first, the Protozoon or Protophyte, second the simple or non-metameric person, third the metameric person, fourth the colony of the nutritive type, fifth the family, or colony of the reproductive type, sixth the coenobiose, and seventh the true, or human societ}'. Closer inspection shows that these are sufficiently heterogeneous when compared with one another and with the personal organism, which is the prototype of the series, but I believe, nevertheless that all of them are real organisms and not merely conceptual constructions or analogies. One of them, the insect colony, has interested me exceedingly, and as I have repeatedly found its treatment as an organism to yield fruitful results in my studies, I have acquired the conviction that our biological theories must remain inadequate so long as we confine ourselves to the study of the cells and persons and leave the psychologists, sociologists and metaphysicians to deal with the more complex organisms. Indeed our failure to cooperate with these investigators in the study of animal and plant societies has blinded us to many aspects of the cellular and personal activities with which we are constantly dealing. This failure, moreover, is largely responsible for our fear of the psychological and the metaphysical, a fear which becomes the more ludicrous from the fact that even our so-called 'exact' sciences smell to heaven with the rankest kind of materialistic metaphysics.
Leaving these generalities for the present, permit me to present the evidence for the contention that the animal colony is a true -organism and not merely the analogue of the person. To make this evidence as concrete as possible I shall take the ant-colonj^ as a paradigm and ask you to accept my statement that the colonies of the termites, social bees and wasps, which the limited time at my disposal does not permit to consider, will be found to offer the same and in some cases even more satisfactory data. I select the ant-colony not only because I am more familiar with its activities, but because it is much more interesting than that of the polyps, more typical and less specialized than that of the honey bee, less generalized than that of the wasps and bumble-bees, and has been much more thoroughly investigated than the colonies of the stingless bees and the termites.
The most general organismal character of the ant-colony is its individuality. Like the cell or the person, it behaves as a unitary whole, maintaining its identity in space, resisting dissolution and, as a general rule, any fusion with other colonies of the same or alien species. This resistance is very strongly manifested in the fierce defensive and offensive cooperation of the colonial personnel. Moreover, every ant-colony has its own peculiar idiosyncrasies of composition and behavior. This is most clearly seen in the character of the nest, which bears about the same relation to the colony that the shell bears to the individual Foraminifer or mollusc. The nest is a unitary structure, built on a definite but plastic design and through the cooperation of a number of persons. It not only reflects the idiosyncrasies of these persons individually and as a whole, but it often has a most interesting adaptive growth and orientation which may be regarded as a kind of tropism. In many species the nest mounds, which are used as incubators of the brood and as sun-parlors for the adult ants, are constructed in
THE ANT-COLONY AS AN ORGANISM 311
such a manner as to utilize the solar radiation to the utmost. In the Alps and Rocky Mountains we find the nests oriented in such a manner that the portions in which the brood is reared face south or east, and as time goes on the nests often grow slowly in these directions, like plants turning to the light, so that they become greatly elongated. This orientation is, in fact, so constant in some species that the Swiss mountaineers, when lost in a fog, can use it as a compass.
Every complete ant-colony, moreover, has a definite stature which depends, of course, on the number of its component persons. And this stature, like that of personal organisms, varies greatlj^ with the species and is not determined exclusively by the amount of food, but also by the queen mother's fertility, which is constitutional. Certain ants live in affluence but are nevertheless unable to form colonies of more than fifty or a hundred individuals, while others, under the same conditions, have a personnel of thousands or tens of thousands.
One of the most general structural pecuharities of the person is the duality of its composition as expressed in the germ-plasm on the one hand and the soma on the other, and the same is true of the ant-colony, in which the mother queen and the virgin males and females represent the germ-plasm, or, more accurately speaking, the ' Keimbahn,' while the normally sterile females, or workers and soldiers, in all their developmental stages, represent the soma. In discussing the question of the inheritance or non-inheritance of acquired characters the Neodarwinians trace all the congenital modifications of the worker and soldier phases to the queen, just as in the personal organism all the congenital somatic characters are traced to the germ-plasm of the egg. Since the homologue of the reproductive organ of the ant-colony consists of the virgin males and females, and since the males mature earlier than the females, the colony may be regarded as a protandric hermaphrodite. Some colonies, however — and this is probably characteristic of certain species — produce only males or females and are therefore in a sense gonochoristic, or dioecious. And this protandric hermaphroditism and gonochorism, like the corresponding conditions in persons, may be interpreted as a device for, or, at any rate, as an aid, in insuring cross-fertilization. The fecundated queen of the ant-colony represents the first link in the 'Keimbahn' and therefore corresponds to the fertilized egg of the personal organism. She produces both the worker personnel and the virgin males and females, just as the fertilized egg produces both the soma and the germ-cells. The colonial soma, moreover, may be differentiated as the result of a physiological division of labor into two distinct castes, comprising the workers in which the nutritive and nidificational activities predominate, and the soldiers, which are primarily protective. Here, too, the resemblance to the differentiation of the personal soma into entodermal and ectodermal tissues can hardly be overlooked.
The structure of the ant-colony thus appears to be very simple as compared with that of its component persons. The question naturally arises as to the particular type of unicellular or personal organism which it most resembles. Undoubtedly, if we could see it acting in its entirety, the ant-colony would resemble a gigantic foraminiferous Rhizopod, in which the nest would represent the shell, the queen the nucleus, the mass of ants the Plasmodium and the files of workers, which are continually going in and out of the nest, the pseudopodia.
The ant-colony, of course, like the person, has both an ontogenetic and a phylogenetic development; the former open to observation, the latter inferred from the ontogeny, a comparison of the various species of ants with one another and with allied Hymenopterous insects, and from the paleontological record. The fecundated queen, as I have stated, represents the fertilized egg which produces the colonial organism, but she is a winged and possibly conscious egg, capable not only of actively disseminating the species, like the minute eggs of many marine animals, but of selecting the site for the future colony. After finding this site she discards her wings and henceforth becomes sedentary like the wingless workers which she will produce. The whole colony rests satisfied with the nesting site selected by its queen if the environmental conditions remain relatively constant. If these become unfavorable, however, the colony will move as a whole to a new site. In most species such movements are rather limited, but the nomadic driver and legionary ants are almost continually moving from place to place and must cover a considerable territory during the year. After the queen has selected the nesting site, she immures herself in some earthen or vegetable cavity, laj^s a number of eggs, supplying them with yolk derived by metabolism from her fat-body and now useless wing-muscles, and feeds the hatching larvae on her salivary secretion, which, though highly nutritious, is, nevertheless, very limited in quantitj^ so that the offspring when mature are dwarfed and very few in number. They are in fact, workers of the smallest and feeblest caste; but they set to work enlarging the nest, break through the soil or plant tissues, construct an entrance on the surface and seek food for themselves and their famished mother. This food enables her to replenish her fat-body and to produce more eggs. Her expansive instincts and activities now contract, so to speak, and becoine reduced henceforth to a perpetual routine of assimilation, metabolism and oviposition. She produces brood after brood during her long life which may extend over a period of ten to thirteen years. Her workers assume the duties of foraging, of feeding the larvae and one another, and of completing the nest. Their size and polymorphism increase with successive broods, till the soldier forms, if these are characteristic of the species, make their appearance. Then the individuals which correspond to the reproductive cells of the personal organism, namely, the virgin males and females develop, and the colonial organism may be said to have reached maturity. Like the personal organism, it may persist for thirty or forty years or,, perhaps, even longer without much growth of its soma, since the workers and soldiers of which this consists are exposed to many vicissitudes and live only from three to four years and probably, as a rule, for a much shorter period. If the queen grow too old or die the colonj^, as a rule, dwindles and eventually perishes unless her place is taken by one or more of her fertile daughters.
This is the ontogenetic history of most ant-colonies. It is so similar to the phylogenetic history derived from the sources mentioned above that we have no hesitation in affirming that it conforms in the most striking manner to the biogenetic law. The very ancient behavior of the soUtary female Hymenopteron is still reproduced during the incipient stage of colony formation, just as the unicellular phase of the Metazoon is represented bj^ the egg. A further correspondence of the ontogeny and phylogeny is indicated by the fact that the most archaic and primitive of living ants form small colonies of monomorphic workers closely resembling the queen, whereas the more recent and most highly specialized ants produce large colonies of workers not only verj^ unlike the queen but unhke one another.
In order to complete the foregoing account it will be necessary to consider some interesting modifications of the usual method of colony formation and growth, especially as these modifications furnish additional and striking evidence in favor of the contention that the ant-colony is a true organism. In many species, after the colony has reached maturity and especially if the food-supply continue to be abundant, several of the virgin females may be fecundated in the nest, lose their wings and remain as members of the colony. This may, indeed, contain half a dozen and in extreme cases as many as forty or fifty or even more fertile queens. But often the growth of the colonial organism becomes excessive through an increase in the worker personnel and passes over into a form of colonial reproduction, when the young fertilized queens, each accompanied by a band of workers, start new nests in the vicinity of the parental formicary. In this manner a very large and complex colony may arise and extend over many adjacent nests. For some time the new settlements may remain in communication with the home-nest through files of workers, but eventually the daughter settlements may become detached and form independent colonies. The resemblance of this method of reproduction, which is essentially the same as the^swarming in the honey-bee, to the asexual reproduction of many unicellular and multicellular organisms by a process of budding, is too obvious to need further comment.
The important role of nutrition in the development of the colony will be clear from the foregoing remarks. It becomes even more striking in the methods adopted by the queens of certain parasitic species in starting their colonies. Some European
THE ANT-COLONY AS AN ORGANISM 315
observers and myself have found a number of queen-ants that are unable to found colonies without the aid of workers of allied species. These queens may be separated into four groups, as follows:
1. The queen which enters a colony of an alien species and decapitates its queen or is the occasion of her being killed off by her own workers. The intrusive queen is then adopted by the workers and a compound colonial organism arises, consisting of the germ-plasm of one species and the soma of another. The queen proceeds to lay eggs, which are reared by the alien workers, thus relieving her of all the labor and exhaustion endured by the independent typical ant-queen during the early stages of colony formation. Pari passu with the development of the worker offspring of the intrusive queen, the worker nurses grow old and die, so that the colony eventually comes to consist of only one species, the soma of the host being replaced bj^ that of the parasite. This method of colony formation, first observed among our American ants and later among certain European and North African species, I have called temporary social parasitism. Now many of the species, which behave in this manner, have extremely small queens, or queens provided with a peculiar pilosity or sculpture that tend to endear them to the workers of the alien colonies which they invade. If we regard the large fertilized queens of ordinary ants, which are supplied with a voluminous fat-body and wing-musculature, as representing eggs provided with a great amount of yolk, and the diminutive queens of the temporary social parasites as the equivalents of alecithal eggs, we have another striking resemblance between the personal and colonial organisms, for the large queens, like the yolk-laden eggs of many vertebrates, are produced in small numbers but are able to generate the colonial soma independently, whereas the small queens, which are produced in great numbers, in order that some of them may survive the vicissitudes of a parasitic life, correspond to the small yolk-less eggs of many parasites, which have to be deposited in plant or animal tissues in order that the imperfect young on hatching may be surrounded by an abundance of food.
2. The queen of the blood-red slave-maker (Formica sanguinea) adopts a different method. She enters the colony of an
316 WILLIAM MORTON WHEELER
allied species, snatches up the worker brood and kills any of the workers or queens that endeavor to dispute her possessions. The ants hatch with a sense of affiliation with their foster mother and proceed to rear her eggs and larvae as soon as they appear. Here, too, the colony is formed by a mixture of two species, but the workers produced by the intrusive queen inherit her predatory instincts and therefore become slave-makers. They keep on kidnapping worker larvae and pupse from the nests of the alien species, carry them home, and eat some of them but permit many to mature, so that the mixed character of the colony is maintained. This, however, is not invariably the case, for old and vigorous sanguinea colonies may cease to make slave-raids and the slaves may die off and leave a pure colony of the predatory species. The advantages of this method of colony formation arfe obvious, for the colonial soma, being composed of two species, grows more rapidly and is much more efficient as a nutritive and protective support to the colonial germ-plasm, which is restricted to the predatory species.
3. The colony-founding queen of the amazon ants of the genus Polyergus resorts to a modification of the method adopted by sanguinea, as has been shown by Emery's recent observations. She enters the colony of an alien species, perforates its queen's head with her sickle-shaped mandibles and permits herself to be adopted by the workers. She pays no attention to the brood but begins to lay eggs, the larvae from which are carefully reared by the workers. The Polyergus offspring inherit the pugnacity of their mother, but, like the sanguinea workers, have the ability to kidnap the brood of other ants. They are, in fact, slave-makers of a very deft and ferocious type. Like their mother, however, the}' are unable to excavate the nest, to care for their own young or to take food except from the mouths of the workers that hatch from the kidnapped larvae and pupae. The mixture of the two species is therefore obligatory, and the slave personnel, which represents the nutritive and nest-building portions of the colonial soma, has to be maintained throughout the life of the colony.
4. Certain feeble queen ants belonging to a few aberrant genera (Anergates, Wheeleriella) invade populous nests of an alien species and are adopted in the place of their queens, which are
THE ANT-COLONY AS AN ORGANISM 317
destroyed by their own workers. The parasites then proceed to lay eggs but these give rise only to males and females as the worker caste is entirely suppressed. The colony retains a mixed character, the parasitic species usurping the functions of the germplasm, while the host is purely somatic. As there are no means of prolonging the lives of the host-workers and as they do not reproduce, the whole colony is short-lived and the maturation of the parasitic sexual individuals has to be accelerated so that it will fall within the brief life-time of the worker hosts. This condition I have called permanent social parasitism.
These four peculiar types of colony-formation all lead to the formation of compoand colonial organisms, comparable to certain compound personal organisms which, with few exceptions, can be produced only by artificial means. In temporary social parasitism the colonial egg can develop its soma only when grafted on to the soma of another species. This soma eventually perishes and the colony then assumes a normal complexion. This condition reminds us of certain tropical plants, like the species of Clusia and Ficus, which develop as epiphytes on other trees but after killing their hosts take root in the soil and thenceforth grow as independent organisms. The slave-makers of the sanguinea or facultative type are also unable to develop the soma except when grafted on to the soma of another species, but in this case the cooperation of both somas in nourishing and protecting the germplasm is maintained for a much longer period. This kind of colony may be compared with a graft made by uniting the longitudinal half of one plant with that of another so that both take nourishment through their roots. To make the resemblance more complete one of the grafted halves would have to be pruned in such a manner as to prevent flowering. In the amazons or obligatory slave-makers and the permanent social parasites the alien soma alone has a nutritive function, so that the conditions are like those in ordinary vegetable grafts, in which the stock retains the roots and the scion produces the flowers and fruit.
I have dwelt on the various methods of colony formation not only because they give us an insight into colonial reproduction, but because they throw light on the colonial organism from the
318 WILLIAM MORTON WHEELER
standpoint of parasitology. That the four types of queens and their offspring are directly comparable with entoparasitic persons is not so remarkable as the fact that in ants the host and parasite form a mixed organism which could only be obtained with persons by jumbling together the component cells of host and parasite like two kinds of peas shaken in a bottle. Notwithstanding this mixture the parasitic colony not only retains its identity and the anticipatory character of its behavior but castrates the host colony and constrains its soma either to cooperate in many of its activities or to specialize as a purely nutritive or nest-building auxiliary. The host is thus reduced to the status of a nourishing or protective organ of the parasite. This behavior has many striking analogies among persons. Giard long ago called attention to the fact that when the cirriped Sacculina settles under the abdomen of a male crab and sends its rootlike haustoria into the tissues of its host, the latter undergoes castration, and its narrow abdomen expands to form a protection for the soft-bodied parasite. In other words, the parasite acts as if it were a mass of crabs' eggs and the male crab behaves as if it had changed its sex and develops an abdomen of the female type.
Not only are there ants, like those already considered, that may be regarded as colonial entoparasites, but there are also a number of species that may be called colonial ectoparasites. These form the so-called 'compound nests,' in which two or more species live amicably side by side, or may even mingle freely with one another, but rear their broods in separate nests, thus indicating in the clearest manner the integrity of the colonial organism. This is also shown by the vast number of myrmecophilous insects, which are, of course, ento- or ectoparasitic persons, and behave towards the ant colony as if it were a rather incoherent and therefore more vulnerable, or exploitable personal organism.
Finally we come to what the neovitalists regard as the most striking autonomic manifestations of the organism, namely the regulations and restitutions, and face the question as to whether these, too, have their counterpart in the colonial organism. I believe that the following facts compel us to answer this ques
THE ANT-COLONY AS AN ORGANISM 319
tion in the affinnative. If the worker personnel be removed from a young ant-colony, leaving only the fertile queen, we find that this insect, if provided with a sufficiently voluminous fat-body, will set to work and rear another brood, or, in other words, regenerate the missing soma. And, of course, any portion of the worker or sexual personnel, that is removed from a vigorous colony will be readily replaced by development of a corresponding portion of the brood. On the other hand, if the queen alone be removed, one of the workers will often develop its ovaries and take on the egg-laying function of the queen. In ants such substitution queens, or gynaecoid workers are not fertilized and are therefore unable to assume their mother's worker- and queen-producing functions. The termites, however, show a remarkable provision for restituting both of the fertile parents of the colony from the so-called complemental males and females. In ants we have a production of fertile from normally infertile individuals, but the incompleteness of the result does not disprove the existence of a pronounced restitutional tendency.
Very striking examples of this tendency are exhibited when colonies are injured by parasitic myrmecophiles. I shall consider only the case of the peculiar beetle Lomechusa strumosa, which breeds in colonies of the blood-red slave-maker (Formica sanguinea) . Though the beetle and its larvse are treated with great affection, the latter devour the ant larvse in great numbers, so that little of the brood survives during the esirlj sumAier months when the colony is producing its greatest annual increment to the worker personnel. The ants seem to perceive this defect and endeavor to remedy it by converting all the surviving queen larvse into workers. But as these larvse have passed the stage in their development when such an operation can be successful, the result is the production of a lot of pseudogynes, or abortive creatures structurally intermediate between the workers and queens and therefore useless in either capacity. It is instructive to compare this case with the regeneration of the lens from the iris in the Amphibian eye. In his recent analysis of the stimuli of restitution in personal organisms Driesch reaches the conclusion that "the specificity of what is taken away certainly forms part of the
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
320 WILLIAM MORTON WHEELEK
stimulus we are searching for, and it does so by being communicated in some way by something that has relations to many, if not all, parts of the organism and not only to the neighboring ones." He also says that "each part of the organism assigns its specific share to an unknown something and that this something is altered as soon as a part is removed or absolutely stopped in its functional life, and that the specific alteration of the something is our stimulus of restitutions." These quotations and Driesch's further discussion of the problem are even clearer in their application to the colonial than to the personal organism, for in the former it is much easier to see how each individual insect "can do more than one thing in the service of restitution" than it is to understand how each cell of the person can do more than one thing in restoring a lost organ.
I fear that I may have wearied you with this long attempt to prove that the ant-colony is a true organism, especially as this statement must seem to some of you to be too trite for discussion, but when an author like Driesch writes a large work in two volumes on the "Philosophy of the Organism" and ignores the colonial organisms altogether, an old-fashioned zoologist may perhaps be pardoned for calling attention to a well-founded, though somewhat thread-bare, biological conception.
If it be granted that the ant-colony and those of the other social insects are organisms, we are still confronted with the formidable question as to what regulates the anticipatory cooperation, or synergy of the colonial personnel and determines its unitary and individualized course. The resemblance of the ant- or bee-colony to the human state long ago suggested a naive reply to this question. Aristotle naturally supposed the colonial activities to be directed and regulated by a ^aaCkebs or riyeixuv, because these personages managed affairs in the Greek states. After the sex of the fertile individual had been discovered by Swammerdam, the word 'queen' was naturally substituted for (SaatXevs or 'king/ and as queens in human states do not necessarily govern and are often rather anabolic, sedentary and prolific persons and the objects of much flattering attention, the term is not altogether inapt when applied to the fertile females of insect colonies. It
THE ANT-COLONY AS AN ORGANISM 321
has been retained although everybody knows that these colonies represent a form of society very different from our own, a kind of communistic anarchy, in which there is neither guide, overseer nor ruler," as Solomon correctly observed. In this respect too, the colony is essentially the same as the personal organism, at least in the opinion of those who do not feel compelled to assume the existence of a 'soul' in the scholastic sense. For it is clear, that to primitive thinkers the soul was supposed to bear the same relation to the person as the iSaatXevs to the insect colony and the king to the human state. This supposition is still held though in a more subtle form, by writers of the present day. Some of these, like Maeterlinck, clothe the postulated controlling agency in a mystical or poetic garb and call it the 'spirit of the hive.' The following passage from the Belgian poet's charming account of the honey-bee will serve to illustrate this method of meeting the problem:
What is this 'spirit of the hive' — where does it reside? It is not like the special instinct that teaches the bird to construct its well planned nest, and then seek other skies when the day for migration returns. Nor is it a kind of mechanical habit of the race, or blind craving for life, that will fling the bees upon any wild hazard the moment an unforeseen event shall derange the accustomed order of phenomena. On the contrary, be the event never so masterful, the 'spirit of the hive' still will follow it, step by step, like an alert and quickwitted slave, who is able to derive advantage even from his master's most dangerous orders.
It disposes pitilessly of the wealth and the happiness, the Hberty and life, of all this winged people; and yet with discretion, as though governed itself by some great duty. It regulates day by day the number of births, and contrives that these shall strictly accord with the number of flowers that brighten the country-side. It decrees the queen's deposition or warns her that she must depart; it compels her to bring her own rivals into the world, and rears them royally, protecting them from their mother's political hatred. So, too, in accordance with the generosity of the flowers, the age of the spring, and the probable dangers of the nuptial flight will it permit or forbid the first-born of the virgin princesses to slay in their cradles her younger sisters, who are singing the song of the queens. At other times, when the season wanes, and flowery hours grow shorter, it will command the workers themselves to slaughter the whole imperial
322 WILLIAM MORTON WHEELER
brood, that the era of revolutions may close, and work become the sole object of all. The 'spirit of the hive' is prudent and thrifty, but by no means parsimonious. And thus, aware, it would seem, that nature's laws are somewhat wild and extravagant in all that pertains to love, it tolerates, during summer days of abundance, the embarrassing presence in the hive of three or four hundred males, from whose ranks the queen about to be born shall select her lover; three or four hundred foolish, clumsy, useless, noisy creatures, who are pretentious, gluttonous, dirty, coarse, totally and scandalously idle, insatiable, and enormous.
But after the queen's impregnation, when flowers begin to close sooner and open later, the spirit one morning will coldly decree the simultaneous and general massacre of every male. It regulates the worker;^' labours with due regard to their age; it allots their task to the nurses who tend the nymphs and the larvae, the ladies of honour who wait on the queen and never allow her out of their sight ; the house-bees who air, refresh, or heat the hive by fanning their wings, and hasten the evaporation of the honey that may be too highly charged with water ; the architects, masons, wax- workers, and sculptors who form the chain and construct the combs ; the foragers who sally forth to the flowers in search of the nectar that turns into honey, of the pollen that feeds the nymphs and the larvae, the propohs that welds and strengthens the buildings of the city, or the water and salt required by the youth of the nation. Its orders have gone to the chemists who ensure the preservation of the honey by letting a drop of formic acid fall in from the end of their sting; to the capsule makers who seal down the cells when the treasure is ripe, to the sweepers who maintain public places and streets most irreproachably clean, to the bearers whose duty it is to remove the corpses ; and to the amazons of the guard who keep watch on the threshold by night and by day, question comers and goers, recognize the novices who return from their very first flight, scare away vagabonds, marauders and loiterers, expel all intruders, attack redoubtable foes in a body, and, if need be, barricade the entrance.
Finally, it is the spirit of the hive that fixes the hour of the great annual sacrifice to the genius of the race: the hour, that is, of the swarm; when we find a whole people, who have attained the topmost pinnacle of prosperity and power, suddenly abandoning to the generation to come their wealth and their palaces, their homes and the fruits of their labour; themselves content to encounter the hardships and perils of a new and distant country. This act, be it conscious or not, undoubtedly passes the fimits of human morahty. Its result will sometimes be ruin, but poverty
THE ANT-COLONY AS AN ORGANISM 323
always; and the thrice-happy city is s'cattered abroad in obedience to a law superior to its own happiness. Where has this law been decreed which, as we soon shall find, is by no means as blind and inevitable as one might believe? Where, in what assembly, what council, what intellectual amd moral sphere, does this spirit reside to whom all must submit, itself being vassal to an heroic duty, to an intelligence whose eyes are persistently fixed on the future?
It comes to pass with the bees as with most of the things in this world; we remark some few of their habits; we say they do this, they work in such and such fashion, their queens are born thus, their workers are virgin, the}^ swarm at a certain time. And then we imagine we know them, and ask nothing more. We watch them hasten from flower to flower, we see the constant agitation within the hive; their life seems very simple to us, and bounded, like every life, by the instinctive cares of reproduction and nourishment. But let the eye draw near, and endeavour to see ; and at once the least phenomenon of all becomes overpoweringly complex; we are confronted by the enigma of intellect, of destiny, will, aim, means, causes; the incomprehensible organization of the most insignificant act of life.
Other authors like Driesch, give the postulated controlling agency the sharper outlines of a would-be scientific but in reality metaphysical entity and call it the 'entelechy.' It is true that the entelechy is deduced by Driesch from the autonomic peculiarities of the personal organism, but as the colony has all the essential attributes of the organism, he would undoubtedly assign it an entelechy, which according to the definition would have to be nonspacial, but working into space, nonspsychic, but conceivable only after analogy with the psychic, and non-energetic, but nevertheless capable of determining the specificity of the colonial activities through releasing and distributing energy.
I confess that I find the entelechy quite as useless an aid in unravelling the complex activities of the ant-colony as others have found it in analyzing the personal organism. This angel-child, entelechy, comes, to be sure, of most distinguished antecedents, having been mothered by the Platonic idea, fathered by the Kantian Ding-an-sich, suckled at the breast of the scholastic forma substantialis and christened, from a strong family likeness, after old Aristotle's darhng evTeXexeta, but nevertheless, I believe that
324 WILLIAM MORTON WHEELER
we ought not to let it play about in our laboratories, not because it would occupy any space or interfere with our apparatus, but because it might distract us from the serious work in hand. I am quite willing to see it spanked and sent back to the metaphysical house-hold.
But, speaking seriously, it seems to me that if the organism be inexplicable on purely biological grounds, we should do better to resort to psychological agencies like consciousness and the will. These have at least the value which attaches to the most immediate experience. And even the subconscious and the superconscious are more serviceable as explanations than such anaemic metaphysical abstractions as the entelechy. Of course, psychic vitalism is one of Driesch's pet aversions and he will have none of it, because he is a solipsist, but the fact that he is compelled to operate with a 'psychoid' and with an entelechy conceivable only jper analogiam with the psychic, shows the inconsistency of his position.
Before we can adopt any ultrabiological agencies, however, except in a tentative and provisional manner, an old and very knotty problem will have to be more thoroughly elucidated. I refer to the problem of the correlation and cooperation of parts. If the cell is a colony of lower physiological units, or biophores, as some cytologists believe, we must face the fact that all organisms are colonical or social and that one of the fundamental tendencies of life is sociogenic. Every organism manifests a strong predelection for seeking out other organisms and either assimilating them or cooperating with them to form a more comprehensive and efficient individual. Whether, with the mechanists, we attribute this tendency to chemotropism or cytotropism, or with the psychic neovitalists, interpret it as conscious and voluntary, we certainly cannot afford to ignore the facts. The study of the ontogeny of the person, i.e., the person in the process of making, in the hands of recent experimentalists, has thrown a flood of light on the peculiarities of organization, but the animal and plant colony are in certain respects more accessible to observation and experiment, because the component individuals bear such loose spacial relations to one another. Then too, the much simpler and more primi
THE ANT-COLONY AS AN ORGANISM 325
tive organismal type of the colony, as compared with that of the person, should enable us to follow the process of consociation and the resulting physiological division of labor more successfully. In the problem, as thus conceived, we must include, not only the true colony and society, and the innumerable cases of symbiosis, parasitism and coenobiosis, but also the consociation and mutual modification of hereditary tendencies in parthenogenetic and biparental plants and animals, since in all of these phenomena our attention is arrested not so much by the struggle for existence, which used to be painted in such lurid colors, as by the ability of the organism to temporize and compromise with other organisms, to inhibit certain activities of the aequipotential unit in the interests of the unit itself and of other organisms ; in a word, to secure survival through a kind of egoistic altruism. ^
2 Since this paragraph was written I have found that several recent authors have given more explicit expression to a very similar conception to the role of cooperation and struggle in the development of organisms. Especially worthy of mention in this connection are Kammerer (Allgemeine Symbiose und Kampf urns Dasein als gleichberechtigte Triebkrafte der Evolution. Arch. f. Rass. u. Ges.-Biol.6, 1909, pp. 585-608), Schiefferdecker (Symbiose. Sitzb. niederrhein. Ges. f. Natur. u. Heilk. zu Bonn, 13, Juni, 1904, 11 pp.), Bolsche (Daseinskampf und gegeuseitige Hilfe in der Entwicklung. Kosmos, 6, '1909); and Kropotkin (Mutual aid, a factor of evolution, London, 1902).
SEXUAL ACTIVITIES OF THE SQUID, LOLIGO PEALII (LES.)
I. COPULATION, EGG-LAYING AND FERTILIZATION
OILMAN A. DREW
From the University of Maine, Orono, Maine
THIRTEEN FIGURES FOUR PLATES
This account, which deals with some of the sexual activities of the squid, is based upon observation made on specimens kept in glass sided aquaria at the Marine Biological Laboratory, Woods Hole, Mass. Specimens caught in the fish traps of the immediate vicinity may, by careful handling, be kept in aquaria in fairly good condition for a number of days. Such specimens occasionally copulate and eggs are sometimes laid.
There are two methods of copulation. By one method the spermatophores ejaculate their contents so the sperm reservoirs thrown from them are attached in a special depression on the inner side of the outer buccal membrane opposite the junction of the two ventral arms (figs. 8 and 10). They then slowly emit sperm, which are carried to and stored in, a special sperm receptacle that opens near this depression and is imbedded in the tissue of the outer buccal membrane (figs. 10 and 11). In this receptacle the sperm are mixed with a secretion and are not active. How long the sperm may be retained in the receptacle is not known, but there is some reason to think that they may be retained for at least some weeks. Females with eggs that can be fertilized may be found during the four months, June to late September, that I have worked at Woods Hole. Without exception every adult female that had not spawned had the sperm receptacle filled more or less completely with sperm, although in many cases the
327
328 OILMAN A. DREW
eggs were far from mature. This, together with the dormant condition of the sperm in the receptacle, and the fact that they seem to be poured out only during egg laying, point to a possible long retention. It is certain that the same female may have sperm reservoirs attached near this receptacle a number of times after it has been filled, and it is possible that the same sperm do not continue long in the receptacle. There seems, however, to be no evidence that they are discharged except during the period of egg laying.
The other method of copulation results in fastening the sperm reservoirs of the ejaculated spermatophores near the end of the oviduct (fig. 8, s) usually directly on its walls but sometimes on the mantle, gill or visceral mass. There is no special receptacle for the sperm from these sperm reservoirs. They escape into the water, becoming active as they escape, and pass out with the water through the funnel. The escape of the sperm is rather rapid but there are vast numbers in each reservoir, from which they are constantly poured like smoke from a chimney until the reservoir is empty. It is not known how long it takes to empty a reservoir but by keeping reservoirs from spermatophores that ejaculated in dishes of sea-water, and by examining reservoirs normally attached to the oviducts and buccal membranes of females, it seems probable that the sperm do not all escape for two or more days.
In aquaria I have seen rather more cases of copulation where the spermatophores are inserted into the mantle chamber than where the sperm reservoirs are attached to the buccal membrane. This may be because of the limited quarters in aquaria. In the larger floating tanks, in which specimens are sometimes kept before they are brought into the laboratory, the buccal membrane copulation seems proportionally more common than in aquaria, but even here the mantle chamber copulation seems to be rather more frequent.
The same individuals may copulate several times in the course of a few hours. In general the male is aggressive. The female may attempt to escape or she may be quite passive. Spermatophores seem to be inserted in the mantle chambers of only those
SEXUAL ACTIVITIES OF THE SQUID 329
females that are nearly ready to deposit their eggs. In the large number of trials made it was found that the eggs of these individuals were so nearly mature they could be artificially fertilized. Females that are nearly ready to deposit eggs have the nidamental glands considerably swollen and the accessory nidamental glands are highly colored with bright red. Wherever the spermatophores were inserted in the mantle chamber these glands were in this condition.
Before copulation both female and male are usually especially active and may be known as sexually excited animals by their peculiar movements. The female in swimming seems to be nervous or excited. She throws short but rapid puffs of water from the funnel, moves the tail fin very rapidly and, leaving the arms quite limp, spreads them apart and frequently throws them to one side. This gives the arms a jerky or trembling motion not shown in ordinary swimming. Except during the most rapid movements of the female, the male solemnly swims by her side, an inch or two away, but parallel, and with his head in the same direction. He frequently manipulates his arms, spreading them apart, commonly with the two dorsal arms elevated nearly or quite to a perpendicular position, and the third arms spread far to the sides (fig. 3). This position is not infrequently accompanied by localized activity of chromatophores. A spot may appear near the base of each third arm and a smaller spot on each second arm a little further from its base. These spots do not remain continuously while the male is in this attitude but suddenly appear with each increase of activity on the part of either the male or female. Occasionally blushing is quite general over the head and anterior end of the body and sometimes includes the whole body but the bodies of both animals generally remain colorless except for the special spots mentioned on the male. The attitude of the male, with elevated and spread arms, is not continuous but is assumed every few minutes, or in some cases seconds, and the arms may be brought into the usual position of a swimming animal for periods of many minutes.
Males do not all respond equally to the presence of sexually active females. Not uncommonly one male in an aquarium containing
330 OILMAN A. DREW
several males will follow the females around by the hour while the other males remain entirely inattentive. Usually when a male begins to show sexual activity he will follow a single female although other females that show similar activities are present in the aquarium. Occasionally he may change to another individual but he nearly always returns after a few minutes to the one to which he has been paying chief attention.
A few males have been observed that were so sexually excited they followed individuals around quite indiscriminately. Under such conditions I have upon three occasions seen a male catch another male and insert spermatophores into his mantle chamber. Two of the three instances were between the same individuals, the second performance being only a few minutes after the first. In each of these cases the male seized made great efforts to get away and finally to get hold of the male that was holding him but was unsuccessful. Upon killing the male that received the spermatophores, sperm reservoirs were found attached to the base of the left gill and to the adjacent visceral mass. Such exceptionally active males may copulate repeatedly with a single female. In a few cases this has been carried so far that the female has actually been killed. Even after the female has beco-me entirely inactive and apparently dead the male may copulate with her several times. In one case, a male that had been several days without food, after copulating with a weakened female, retained his hold and killed her by eating a considerable hole through the mantle.
The male always uses the same arm for transferring the spermatophores. This arm, the left ventral, is not greatly modified, but a short distance from its tip some of the suckers, especially those in the row farthest from the midline of the body, and a ridge between the rows of suckers show modification (fig. 4, h). The peduncles of a dozen or more of the suckers of the outer row are considerably elongated and the sucking discs of a few, (six or eight) are greatly reduced in size or entirely absent. In both directions from these, the discs become increasingly normal until no modification is apparent. The suckers of the row toward the midline of the body are somewhat modified, the peduncles being somewhat shorter than those of the other suckers in the row, and the suck
SEXUAL ACTIVITIES OF THE SQUID 331
ing discs somewhat smaller, but in none of the suckers of this row are the sucking discs entirely absent. A glandular plaited ridge extends lengthwise between the suckers of this region and gives off branches that join each of the peduncles. This ridge is highest and broadest opposite the suckers that are most modified and gradually disappears as the suckers become normal. At its highest point it hag about the same elevation as the shortest modified suckers, which are adjacent. Sections of the modified portion of the arm show that the ridge and suckers mentioned are covered by a thick columnar epithelium that stains deeply. Many of these epithelial cells are filled with large rounded granules that stain with eosin. The cells that cover other portions of the arm are flattened or cubical, do not stain very deeply, and do not contain granules. It seems probable that the cells of the hectocotylized region secrete a substance that aids the arm in holding the spermatophores. The modified suckers probably make the bending and grasping necessary for the transfer of the spermatophores more easily accomplished.
The positions of the animals during copulation are rather hard to determine as the whole process generally does not occupy more than ten seconds and during this time the animals are usually swimming and the arms are changing positions, but by carefully focusing attention during different acts upon first one arm and then another, the positions and movements have been determined with some accuracy I think. Fig. 1 represents the positions of the animals while the arm of the male that bears the spermatophores is inserted into the mantle chamber of the female. This figure is the result of my conception of positions after having carefully observed copulation more than twenty times. Since drawing the figure many other observations have been made and the positions always seem to be essentially as given.
The male usually grasps the female while both are sw^imming. Occasonally the female maybe resting on the bottom in the characteristic attitude, with the tips of the arms and the posterior end of the body touching and the head and funnel region somewhat elevated. If not swimming, she usually, when grasped, starts to swim, but in a few cases that I have observed she made no effort
332 OILMAN A. DREW
and left the bottom only as she was lifted or turned by the male. In every case the male attached from the left side of the female. He frequently swims close to her and brushes the tips of his arms along her head and mantle. Just before attaching, if both are swimming, he sinks slightly beneath her and grasps her body with his arms so that his right arms are all on the right side of her body and his left arms are all on her left side. The body of the male is seldom exactly ventral to the female but usually slightly toward the left side. Attachment is evidently made as nearly as possible in the required position but when the female darts ahead, as she frequently does, the male is likely to attach too far posteriorly . In such cases he does not let go his hold but crawls rapidly forward, arm over arm, until the right position is attained. Naturally the positions of the individual arms differ somewhat but in general the arrangement is reasonably well shown in fig. 1.
For about a second after his position is attained the arms seem busy in making firm attachments, then with a rapid sweep his left ventral arm is passed by the end of his funnel and is immediately inserted into the mantle chamber along the left side of her neck, near the funnel. During the act both animals are usually quite without color and the inserted arm of the male may be seen fairly distinctly inside the mantle chamber.
The movement of the arm past the funnel is rapid and only once have I actually seen the grasping of the spermatophores and their transference to the mantle chamber. In this case while watching squid in an aquarium that was placed so the squid were between me and a window, a male grasped a female that was resting on the bottom. The female, contrary to the usual custom, did not move. As the male had attached far back on the body, opportunity was given me to get into position for observation before the male could crawl forward. As the female made no attempt to get free, the male seemed far more deliberate than usual. Just before the arm was passed by the end of the funnel, the penis could be seen protruding into it. A number of spermatophores appeared in the opening of the funnel and were grasped by bending the tip of the arm around them. With a rapid sweep of the arm they were immediatelv inserted into the mantle chamber of the female
SEXUAL ACTIVITIES OF THE SQUID 333
where they were held about five or six seconds. The arm was then withdrawn and in about five or six seconds more the empty cases of the spermatophores passed out of the funnel of the female with a respiratory jet of water. These spermatophore cases were pretty closely attached to each other by having the tubes of their ejaculatory apparatus twisted together. They were recovered and found to be 41 in number. To the cluster were attached five sperm reservoirs. Examination of the female later showed that most of the other reservoirs were attached near the end of the oviduct. While the number of spermatophores used in an act of copulation varies greatly, the observations that have been made, indicate that this may be a little, but not much above the average.
The animals nearly always separate almost immediately after the arm is withdrawn. Beside the male which started to eat the female, a very few individuals have remained attached for from some seconds to nearly a minute after the arm has been withdrawn.
After copulation the female frequently seems considerably fatigued and may settle to the bottom and rest some minutes before becoming active again. I am rather inclined to think that this is due to her struggles, for when the female remained quiet, the apparent fa,tigue did not seem so marked. The male does not seem greatly affected, but is likely to continue to be very active for some time.
The copulation that leads to the filling of the sperm receptacle on the buccal membrane does not seem to be preceded by special movements. Although I have observed it several times the absence of preparatory movement has left me rather unprepared for the observations that must necessarily be made so quickly, for in this, as in the other form of copulation, the animals are seldom in contact more than ten seconds. In the cases I have observed my attention has been attracted by the sudden dart of one squid, the male, from one end of the aquarium directly at another, the iemale. Before the dart the squid face each other, and are separated by thirty centimeters or more. The movement was always exceedingly rapid and was probably due in each case to the
334 GILMAN A. DREW
expulsion of a single jet of water. The male seemed to reach the female before she had time to move much, although she has given me the impression of attempting to dodge as if frightened. The two animals become attached head to head with their arms intermingled, each grasping the other (fig. 2). Then as in the other method, the male sweeps his left ventral arm past the end of the funnel and grasps the bundle of spermatophores. These are immediately thrust between the ventral arms of the female and held there for a few seconds. The animals then separate and examination has shown fresh sperm reservoirs attached to the receiving depression on the buccal membrane of the female. The empty cases of the ejaculated spermatophores may be held between the arms several minutes but they are finally dropped. Here, as in the other method of copulation, only the sperm reservoirs are retained for any length of time.
The spermatophores begin to ejaculate immediately after leaving the penis and the whole process is completed in a very few seconds. Pulling the filament attached to the ejaculatory end of a spermatophore is all that is needed to start its ejaculation. As the ejaculatory end of the spermatophore leaves the penis last and, as the spermatophores in the penis and the spermatophoric sac are imbedded in a viscid secretion, there is every reason to believe that the pull given the spermatophores by the arm with which they are grasped, when this arm starts to transfer them from the penis to the mantle chamber or to the buccal membrane, is sufficient to start ejaculation. The arm carries the spermatophores into the position necessary for the attachment of the sperm reservoirs while they are ejaculating and holds them there until the ejaculation is complete and the reservoirs are attached.
The structure of the spermatophores and the mechanics of ejaculation which lead to the attachment of the reservoirs will be treated in another paper. It should, however, be understood that the spermatophores are never attached as such, but they ejaculate and the sperm reservoirs are attached. As the reservoirs are attached by cement carried inside the spermatophores and liberated by the ejaculation, they may be stuck anywhere.
The sperm slowly escape from these reservoirs and may then
SEXUAL ACTIVITIES OF THE SQUID 335
become free in the water, as when they are attached in the mantle chamber, or may be stored in a special receptacle, as when they are attached in the special depression on the outer buccal membrane. They are mixed with a viscid secretion in the reservoir and probably also before entering the reservoir, although I am not certain about the latter. The epithelium of the region is abundantly supplied with goblet cells which very possibly supp'y secretion for this purpose.
The depression in which the sperm reservoirs are mostly attached is supplied with a deeply staining columnar epithelium which is covered by a mass of rather hard material, evidently secreted by these cells, that shows distinct markings parallel with the surface of the epithelium (figs. 11 and 12). These markings seem to indicate that the material is secreted intermittently and thus is formed in layers. This material forms a suitable place for attachment of sperm reservoirs and probably serves no other purpose. Reservoirs are sometimes attached to other portions of the buccal membranes or to the tentacles but they are far more abundant in the depression than anywhere else. The sperm that escape from the reservoirs that are not attached in this depression probably do not find their way into the sperm receptacle.
The sperm receptacle has the shape of a compound alveolar gland (fig. 11). It is imbedded in the outer buccal membrane and opens on the inner surface of this membrane at a point opposite the junction of the two ventral arms. Simple cubical epithelium lines the deeper alveoli of the receptacle, and cubical epithelium with many goblet cells the portion nearer the opening. Some, but not many, cilia have been seen on these cells. The killing fluids used may not have preserved them, for the tails of sperm in the reservoirs are not often individually visible in the sections. With the exception of the tails of the sperm and the possible cilia on the cells the material gives evidence of good preservation. A layer of muscle fibers surrounds the receptacle as a whole and bundles of fibers run between and around the individual alveoli.
It was not determined whether the sperm are active in the interval between their discharge from the reservoirs and their en JOURXAL OP MORPHOLOGY, VOL. 22, NO. 2
336 OILMAN A. DREW
trance into the receptacle or not. That they are not active while stored in the receptacle is shown by opening filled receptacles on dry slides. The sperm are invariably quiet, but immediately become active when sea-water is added. In specimens killed soon after copulation, sections show the sperm entering the receptacle in narrow streams and not spread out as one might expect them to be if the sperm were active (fig. 11). It was not possible to remove all the sea-water from living specimens in which the receptacles were being filled without causing disturbances in the vicinity of the reservoirs and that made it impossible to determine the normal condition of the sperm in transit from one to the other. In the sections that show sperm entering the reservoir the tails all point in the same direction, as would be the case if they were not swimming actively but were being moved by an outside force. The heads go first and the tails all trail behind. Swimming sperm usually move in all directions but there may be some directive cause that would account for their positions even if they are stored through their own activities.
As previously stated, a female that is nearly ready to deposit her eggs can be told by her peculiar nervous movements and the way she manipulates her arms. Frequently the borders of the accessory nidamental glands, which are very red at this time, may be seen through the semi-transparent mantle and thus form a further indication that the eggs are nearly ready to be deposited. The nidamental and oviducal glands of such an animal are always somewhat, and frequently greatly, enlarged. Immediately after the eggs have been deposited these glands, while still large, are soft and flabby.
As is well known the squid deposits her eggs imbedded in strings of a jelly-like substance which vary in size with the size of the animal depositing them but which probably average about 8 mm. in diameter and 90 mm. long. The jelly consists of an inner mass that surrounds the eggs and a thick, rather tough but still jellylike sheath that forms the outer covering. The inner jelly is secreted inside the oviduct by the oviducal glands. The outer jelly is secreted by the nidamental glands and is apparently moulded into shape as it passes through the funnel. The accessory nidamental
SEXUAL ACTIVITIES OF THE SQUID 337
glands, which he j List in front of the anterior ends of the nidamental glands and open by wide openings near the narrow openings of these glands, are very active during this period and secrete a viscid material. What the special function of this secretion is has not been determined. It would seem from position and activity that the secretions from both sets of glands must be mixed as the}' are poured out.
Until recently eggs have not commonly been deposited in aquaria at Woods Hole. This maybe due to the way the animals have been handled. Squid will not stand rough handling, either in capture or transportation, and live well in aquaria afterward. When captured in fish traps, quickly and carefully transferred to live cars where they are supplied with an abundance of water, and transported to the aquaria with as little excitement and as good water as possible, they may be kept several days in pretty good condition, but they wear themselves out by constantly bumping against the walls of the aquaria and are not vigorous many days. During each of the months I have worked at Woods Hole, June to late September, specimens have been obtained that have deposited eggs in aquaria. During the first three months specimens ready to deposit eggs are rather easy to get. In September only a small proportion of those captured still contained eggs.
Eggs are somewhat more frequently deposited in aquaria at night than during the day, but this may be due to the frequent if not nearly continuous disturbance to which they are subjected during the day in a laboratory where many people are working. The usual number of strings deposited by a female in what would seem to be a continuous laying period ranged from one to six. These strings were commonly delivered from fifteen to forty minutes apart, the time between any two strings being quite variable in an individual. One specimen, however, deposited twentythree strings in an hour and thirty-five minutes. These were laid during a comparatively dark day when the laboratory was quiet. Possibly the small number deposited by other females was due to disturbance.
The end of the egg string begins to protrude from the end of the funnel while the female rests upon the bottom in the attitude
338 OILMAN A. DKEW
habitually assumed by resting squid (fig. 5). When from one to two centimeters of the egg string protrudes from the funnel, the female leaves the bottom and begins to swim slowly backward. This swimming is apparently due both to movements of the tail fin and to small jets of water forced from the funnel along the sides of the egg string. The jets of water cause the egg string to be protruded gradually. The protruding end is now caught by the ends of the two dorsal arms, which are bent ventrally between the other arms for this purpose (fig. 6), and as the string is ejected from the funnel, it is drawn between the circlet of arms. It usually takes from half a minute to a minute for the egg string to pass through the funnel and to disappear between the arms. It is then held between the arms about two minutes or sometimes a little longer. While the string is held between the arms it is completely enclosed by them and their free ends keep twisting around each other. In this position they form a cone with the apex at the ends of the arms (fig. 7). At other times the arms are held so they form a dorso-ventrally flattened expansion that serves somewhat as a rudder or anterior fin. The arms while enclosing the eggs are never entirely still but move slightly upon each other and are probably busy in moving the string about. While the string is thus held the animal slowly swims back and forth, never rapidly but continuously.
Toward the end of the period during which the string of eggs is held, the animal shows an increasing tendency to turn the body into a nearly perpendicular position to bring and keep the tips of the arms in contact with the bottom (left animal in fig. 9). With the arms held quite rigid and the tail fin moving rapidly she goes bounding along on the tips of her arms, dorsal side foremost, with a movement somewhat similar to the bounces that may be obtained by pushing a lead pencil, held by one extremity and slightly inclined from the perpendicular, over a table. This action is generally repeated several times. She occasionally catches hold of objects with her suckers, finally catches some object firmly, draws down into close contact with it for two or three seconds (right animal in fig. 9) and, when she releases her hold, leaves the string of eggs fastened to the object she had laid hold of. At
SEXUAL ACTIVITIES OF THE SQUID 339
this time the jelly of the string is soft and sticky. It hardens quite rapidly and soon will not stick to objects, but at this time it adheres readily. The position of the string when taken between the arms indicates that the string is finally stuck by the end that first leaves the funnel.
After sticking a string of eggs the female rests upon the bottom some minutes before another string makes its appearance. She usually selects some protruding object like a stone, shell, or water pipe upon which to stick the egg strings. Having stuck one string she usually, but not always, returns to the same place to stick later strings. If strings are present when a female begins to deposit she usually attaches to these strings, or to nearby objects. This no doubt accounts for the very large clusters, with strings containing eggs in various stages of development, that are sometimes found. Upon several occasions clusters in fishtraps a ad live-cars have been found that would not go into an ordinary ten-quart pail. Such clusters are of course formed by many females.
It is evident that the eggs may be fertilized in the oviduct, in the mantle chamber, or between the arms. Examination of the contents of the oviduct have in no case given evidence of sperm. Eggs taken from the oviduct may easily be fertilized by placing them in sea-water containing sperm, but in no case did eggs taken from the oviduct show evidence of fertilization although many sperm reservoirs that were giving off active sperm were attached to the walls of the oviduct and to surrounding organs.
There can be no doubt, however, that eggs may be and are fertihzed in the mantle chamber and also between the arms. That the eggs may be fertilized in the mantle chamber is indicated by reason rather than by obervation. When sperm reservoirs are attached in the mantle chamber the sperm are constantly liberated in the water in this chamber as long as the supply lasts. The eggs upon leaving the oviduct also pass into the mantle chamber and, as before stated, when eggs and sperm are mixed in seawater, fertilization results.
That fertilization may be delayed until the egg string is formed and held between the arms is indicated by observations made on
340 OILMAN A. DREW
the female already mentioned that deposited twenty-three strings. She was in a rather large aquarium with a number of other squid. Copulation had occurred several times but this particular squid, . which had been under observation some hours, had not been seen to copulate. Dissection later showed that there were no sperm reservoirs attached in her mantle chamber. Because of disturbance she upon six occasions failed to get the egg string between her arms. When she reached for the string with her dorsal arms she was each time disturbed so she dropped the string and ejected it directly into the water. Four of these strings were recovered as quickly as possible after they were dropped, and placed in dishes of fresh sea-water where the proportion of fertilized eggs could be determined. From 40 to 50 per cent of the eggs in the strings developed. More than 99 per cent of the eggs m strings that had been held between the arms and then placed in similar dishes developed. As already mentioned there had been copulation among other squid in the aquarium and as the reservoirs were attached in the mantle chambers there must have been many free sperm in the water of the aquarium. It seems probable that enough of these sperm reached the strings that were dropped, before they could be removed from the aquarium, to fertilize a portion of the eggs. Microscopic examination of these strings immediately after they were dropped revealed very few sperm, but the strings that were held between the arms were swarming with them. Sperm were able to penetrate and move actively about in the soft jelly of a recently formed string, but the jelly soon hardened so fresh sperm brought in contact with it were not able to work their way in.
A "curious bit of habit reflex was exhibited by this squid each time she dropped a string of eggs. Immediately after the disturbance she took the attitude she would normally have taken had the egg string been successfully lodged between the arms. The arms were held in the form of a cone, the tips were twisted together and she passed on through each of the succeeding phases even to drawing down tight against an object as if to attach the egg string that had never been between the arms. After this she rested until the next string was formed, but she never interrupted the orderly
SEXUAL ACTIVITIES OF THE SQUID 341
sequence of her activities because she had accidentally lost a string of eggs.
The methods of copulation of cephalopods have attracted the attention of observers from very early times but the act of copulation has not been actually seen for many species and where observations have been made they have for the most part been incomplete. Aristotle makes several statements regarding the breeding habits of cephalopods. It is quite possible that he saw something of the act of copulation for some species, but his statements are hard to follow and are evidently inaccurate. The most important statements are here quoted to show the curious medley of facts and fiction. In chapter 5, book 5, he says:
1 . All the malacia, as the polypus, sepia and teuthis, approach each other in the same manner, for they are united mouth to mouth: the tentacula of one sex being adapted to those of the other; for when the polypus has fixed the part called the head upon the ground, it extends its tentacula which the other adapts to the expansion of its tentacula, and they make their acetabula answer together. And some persons say that the male has an organ like a penis in that one of its tentacula which contains the two largest acetabula. This organ is sinewy, as far as the middle of the tentaculum, and they say it is all inserted into the nostril of the female.
2. The sepia and loligo swim about coiled together in this way, and with their mouths and tentacula united, they swim in contrary directions to each other. They adapt the organ called the nostril of the male to the similar organ in the female; and the one swims forwards, and the other backwards. The ova of the female are produced in the part called the physeter, by means of which some persons say that they copulate.
Again in chapter 10, book 5, he says:
1. The malacia breed in the spring, and first of all the marine sepia, though this one breeds at all seasons. It produces its ova in fifteen days. When the ova are extruded, the male follows, and ejects his ink upon them when they become hard. They go about in pairs. The male is more variegated than the female, and blacker on the back. The sexes of the polypus unite in the winter, the young are produced in the spring, when these creatures conceal themselves for two months. It produces an ovum like long hair, similar to the fruit of the white poplar. The fecund
342 OILMAN A. DREW
ity of this animal is very great, for a great mumber of young are produced from its ova. The male differs from the female in having a longer head, and the part of the tentaculum which the fishermen call the penis is white. It incubates upon the ova it produces, so that it becomes out of condition, and is not sought after at this season.
Part of these statements, such as The sepia and loligo swim about coiled together in this way, and with their mouths and tentacula united, they swim in contrary directions to each other" would seem to be based upon such observations as could be made from above but the further statement that they adapt their nostrils (funnels) together, probably indicates the ease with which observation and supposition can be mixed. It is not necessary further to analyze Aristotle's statements. No doubt much was based upon fishermen's stories but he evidently did study the anatomy and habits of these animals and recognized the probability that one of the arms of the male is used in copulation.
While the modified arm of the male thus early received attention, the true hectocotylus that separates entirely from the male and attaches itself in the mantle chamber of the female escaped notice for many centuries. To quote from the Cambridge Natural History :
The typical hectocotylus seems to have entirely escaped notice until early in the present (last) century, when both Delle Chiaje and Cuvier described it, as detected within the female, as a parasite, the latter under the name of Hectocotylus octopodis. KoUiker, in 1845-49 regarded the Hectocotylus of Tre?noctopus as the entire male animal, and went so far as to discern in it an intestine, heart, and reproductive system. It was not until 1851 that the investigation of V^rany and Filippi confirmed a suggestion of Dujardin, while H. Miiller in 1853 completed the discovery by describing the entire male as Argonauta.
While nearly all male cephalopods show some modification of one or more arms, the only ones that have been reported with detachable arms are Argonauta, Ocythoe, and Tremoctopus.
Extended studies have been made on the modification of the arms of cephalopods, and there have been a few observations upon
SEXUAL ACTIVITIES OF THE SQUID 343
the functional activities of these arms, but most of the observations have consisted in finding sperm reservoirs recently attached to various portions of females.
In 1869 Lafont described copulation in Sepia. A translation of that portion that deals with the act itself is as follows:
In copulation the male and female precipitate themselves upon one another, hold together by their arms which are twined together, and remain thus, mouth to mouth, for a variable time, which may last for two or three minutes. This act is followed in the female by a state of very marked general prostration, while in the case of the male the general excitation is greatly prolonged and for a considerable time it keeps the splendid appearance these animals show as the result of the accomplishment of the function of reproduction.
He supposed that while the animals were attached by their arms, head to head, the male ejected a packet of spermatophores, which ejaculated while in his mantle chamber and the sperm reservoirs were then thrown from the funnel of the male into the branchial chamber of the female with the current of water entering her branchial chamber.
Sepia, like Loligo, has a receptacle for the storage of spermatozoa in the buccal membrane, and the position observed by Lafont of animals attached head to head was doubtless a true position of copulation, but it seems probable that the spermatophores were not disposed of in the way suggested, but were transferred to the buccal membrane by one of the arms of the male. Lafont found sperm reservoirs attached in the mantle chamber of the females near the mouths of the oviducts, so it seems probable that in this form, as in Loligo pealii, both methods of copulation occur.
Racovitza (1894, a) observed copulation inSepiola. The male seized the female, turned it over and inserted his first pair of arms into the mantle chamber. Copulation lasted eight minutes during which the female struggled to free herself. He speaks of the spermatophores being fixed on the folds of a large pocket situated on the left side of the pallial cavity of the female. These ejaculate and the freed reservoirs deliver their sperm into the pocket
344 OILMAN A. DREW
which m turn ejects them (from his description I take it they are not stored up in this pocket as in the receptacle on the buccal membrane of a squid) into the pallial cavity where they are supposed to meet the eggs as they are laid.
The most complete account of copulation that I have seen for anycephalopod was given by Racovitza in 1894 (b) for Octopus vulgaris. He observed copulation in an aquarium and gives a figure showing the positions of the animals. The copulation differs markedly from that of Loligo, as might be expected, for Octopus has a hectocotylized arm that is much more differentiated than that of Loligo. The animals were some distance apart in the aquarium. The male reached over with the hectocotyhzed arm, which for this species is the third on the right side and, after caressing the female with its tip, introduced its end into her mantle chamber by the side of the funnel. Here it remained for something more than an hour. During this time the female remained quiet, except for certain spasmodic movements, while the male showed only slight movements of the hectocotylized arm which were supposed to be associated with the movements of spermatophores down the longitudinal groove of this arm. Although it was not possible actually to see the spermatophores in transit, examination of the female after copulation showed numbers of the sperm reservoirs, derived from the ejaculated spermatophores, within the oviducts.
Evidently there are at least three methods of copulation practiced by cephalopods. A method of caducous hectocotylism in which the charged hectocotyl is liberated in the mantle chamber of the female; a method in which the arm does not liberate any special portion but is so modified that it can transfer spermatophores by a mechanism within itself to the region of the oviduct of the female; and finally a shght modification of the arm that simply enables it to grasp the spermatophores which are then transferred directly to the female by moving the arm. Where the latter method is employed there may be two kinds of copulation, as in Loligo pealii.
Racovitza, (1894, c) in commenting on the copulation of Rossia believes that, although special receptacles are found outside the
SEXUAL ACTIVITIES OF THE SQUID 845
mantle chamber of this species, they cannot be considered as normally functional. He seems led to this conclusion by finding sperm reservoirs attached to various portions of the bodies of the animals as well as in the immediate neighborhood of the mouths of the oviducts. It would seem more likely in the light of the observations here recorded for Loligo, that a copulation that leads to the filling of these receptacles is normal and that the sperm so stored may be used in fertilizing the eggs.
It is certainly hard to conceive by what steps a complicated method of transferring sperm that has led to the formation of a hectocotylized arm and complicated spermatophores might be perfected. The modification of different arms for copulation by different cephalopods further increases the difficulty in understanding the history of hectocotylism as a whole.
While evidence that bears directly upon the history of the hectocotylism seems to be lacking, such complications are so frequently considered to be impossible to explain by known evolutionary factors that it may be well at least to consider the great difficulties presented by such structures. It must not be supposed that in so doing I put myself in the position of defending a thesis. This would be too much like the methods employed by many of the Greek philosophers who needed little or no basis of fact upon which to build. My only reason for considering the matter here is to show that, with all the difficulties, the condition of hectocotylism among modern cephalopods cannot be considered beyond the possible range of evolutionary factors.
Among the Dibranchiata the arms that show hectocotylism are the first, the third and the fourth on both sides of the body. Sometimes more than a single one is affected. In such cases the modified arms may be symmetrically placed on the two sides of the body, or they may be adjacent arms on the same side of the body. Steenstrup attempted to base the classification of cephalopods upon their hectocotylized arms but Brock and Hoyle have shown that forms whose general body structure would seem to indicate relationship, do not always have homologous arms modified.
While the arm is usually constantly on one side for all members of a genus, unless both sides are modified as not infrequently hap
346 GILMAN A. DREW
pens, a genus whose general body structure indicates near relationship may have the similar arm of the other side modified. The position of the arm on the right or left side of the body is not generally considered very significant. The somewhat frequent occurrence of genera showing hectocotylism of arms symmetrically placed on the two sides of the body may indicate a primitive paired condition that has been replaced among the majority of existing cephalopod genera, by specializing on one side and dropping out on the other. Whether this accounts for the condition or not, the shifting of a modification from one side of the body to the other, sometimes involving modifications of other body structures and sometimes apparently not, is not uncommon among animals, and even if not easily explained, evidently has no very great phylogenetic significance. Shifting in series is not so common and when we find in the same family, genera with the fourth and others with the first arm hectocotjdized it becomes difficult to imagine ancestral conditions that made this posisble.
Wherever known, male cephalopods use one or more of the arms to transfer sperm to the female. Copulation has not been described for many of the species but the presence of more or less modified arms in more than half the recogaized families may be taken as an indication that either these animals or their ancestors used their arms in copulation.
If the spadix of Nautilus is used in copulation we have a possible indication that a number of arms may have been employed in the transfer of sperm by primitive cephalopods. It is of course possible that all the arms were used for this purpose and that the present diversity can be accounted for by the specialization of one or the other of the arms involved in this primitive condition. This, however, does not seem reasonable when the diversity within the limits of a single family is considered.
The arm that is used, and the way in which it is used, is associated with the character of the spermatophores and the position of their final discharge. The Octopoda show the greatest structural modification in their hectocotylized arms. While two of the families of this group give no evidence of hectocotylism, none of the genera of the remaining families are known to be free from it.
SEXUAL ACTIVITIES OF THE SQUID 347
and wherever found it is always the third arm that is involved. Sometimes this arm is on the right and sometimes it is on the left side. In three genera it is known to be caducous and in a fourth (Alloposus) it is supposed to be. In the remaining genera in which the hectocotylized arm has been studied, the modifications, while not resulting in the actual separation of the arms, are of an extensive nature. In Octopus, for instance, they involve not only changes. in size, form, and the condition of suckers, but a special groove is present through which the spe matophores are supposed to be carried from the base, presumably from the penis to the tip. The tip in turn is modified so it is supposed to function in placing the spermatophores in position for ejaculation.
The Decapoda do not show such extensively modified hectocotylized arms. The changes are here chiefly confined to some of the suckers and their immediate vicinity. In Loligo this modification apparently serves to aid the arm in grasping the spermatophores, which are then transferred by the movement of the arm. While the actual grasping of the spermatophores has not been previously observed, there can be little doubt that other forms of the Decapoda use the arms in a similar manner. Where copulation has been observed the movements of the arms indicate that they are used in the transfer, and the positions of the sperm reservoirs that have been found attached to the females indicate that some arm must have functioned in getting them into position. As there is no special transferring mechanism, this must have been accomplished by the free movements of the arms.
Where structural modification is shght and the placing o the spermatophores is due to dexterity, there is less difficulty in understanding how the function may be shifted from one arm to another in response to changes in the position of the attachment of the reservoirs on the female, than would be the case were great structural changes involved. It would be much more difficult to understand how there could be a shifting in series of arms as highly modified as those of the Octopoda, where only the modified arm could possibly perform the function.
It mast not be understood that habit formation requiring such dexterity is considered easier to originate than modification in
348 OILMAN A. DREW
structure that will perform similar acts. When, however, the habit and dexterity have been acquired, it is not inconceivable that they might be shifted to another closely similar appendage if the position of this appendage becomes more suitable for the purpose. The modification is so slight in the arms of most of the Decapoda, and the modification varies so greatly in the different genera, that it may have been functionally acquired in each case. So far as can be seen it would be mechanically quite possible for a squid to use an unmodified arm, instead of the one that shows the modification, for the transfer of the spermatophores. The spermatophores might not be so tightly or compactly held but the normal suckers would hardly seem to interfere greatly in the performance of the function.
There is still another question involved. Is there any genetic relation between these two methods of transfer and if there be, which, if either, most probably came first?
A special method of copulation that requires the use of arms and complicated spermatophores is not found among animals often enough to make it at all probable that it has arisen in this group more than once, so we can hardly doubt that the two methods are genetically related.
At first sight the squid's method of grasping the spermatophores and transferring them directly might be considered the simpler process, but there is some reason to doubt that this method was at the beginning of the series. While it would be hazardous to say that Octopoda were the ancestors of Decapoda, there is much reason to believe that the ancestors of the latter lived upon the bottom and were far less active than the modern animals. Such animals would not seem to be so well adapted for the transfer of spermatophores by dexterous movements as the more active, freeswimming forms. It is at least certainly true among modern cephalopods that those that show great modifications in the structure of the hectocotylized arms are found entirely among the less active bottom forms. If the method of transferring sperm by means of the arms originated before the Decapoda became free-swimming animals, and this seems the only explanation of its prevalence
SEXUAL ACTIVITIES OF THE SQUID 349
among both Decapoda and Octopoda in modern times, it would seem that structural modification probably came early.
Possibly this modification was based upon the use of one or more arms as guides for the transfer of the sperm. It is possible that having first used the arms as guides, structural modifications and dexterous movements were developed as divergent methods. If the two methods form a linear series, there is some reason to think structural modifications came first. It would seem much easier to explain modifications that lead to the change in the structure of appendages for the transfer of spermatozoa, as the grooved hectocotyhzed arm of Octopus or the modified abdominal appendages of certain Crustacea, than to explain a sudden change that would result in a practically unmodified arm functioning by grasping spermatophores of a very specialized kind, transferring them quickly and accurately to the required position and holding them there until they have had ample time to ejaculate and fix their contents. It seems more reasonable to suppose that an arm modified as a machine to perform this process, with its tip serving to place the spermatophores in position, might in time acquire the necessary dexterity and then lose the modifications previously acquired, than to look at this as the beginning of the series. Again we find that in such cases as the squid, where the arm is little modified but very dexterous, there is a special receptacle at some distance from the opening of the oviduct that is norma ly filled with sperm during the breeding season. This would certainly seem to be a comparatively recently acquired receptacle, so the copulation leading to its being filled would also be considered comparatively recent. That this receptacle is concerned in the fertilization of the eggs is shown by observations made while the eggs were being laid.
With no personal knowledge of the breeding habits of other cephalopods than the squid, it would seem more reasonable to consider the method of using the detachable hectocotyl of such forms as Tremoctopus as one extreme, the method used by Loligo in grasping spermatophores and transferring them directly as another extreme and the condition shown by Octopus as the modern greatly specialized product of a modification such as early cephal
350 GILMAN A. DREW
opods probably developed. This would mean that the detachable hectocotyl is an extreme specialization in structure and that the modification shown by the squid represents possibly a degeneration in structure but a remarkable specialization in habit.
Why a form should have two methods of copulation is not at all clear. Certainly the introduction of the spermatophores into the mantle chamber to a position near the oviduct is to be considered more primitive than their being placed in a position to fill a receptacle outside of the mantle chamber, but why mantle chamber copulation should be retained after the receptacle has been perfected is not clear. That mantle chamber copulation is not absolutely necessary for the fertilization of the eggs I think is proved by my observations; that it is common is certain. That the sperm receptacle is an improvement over the free attachment of the sperm reservoirs in the mantle chamber is evident from the longer possible retention of the sperm in the receptacle. In a limited period after the sperm reservoirs are freed from the spermatophores, as when deposited in the mantle chamber, the sperm all escape and are wasted unless oviposition takes place in the meantime.
SUMMARY
Squid have two methods of copulation. By one method sperm reservoirs are attached in the mantle chamber on or near the oviduct and immediately begin to discharge their contents freely in the water. By the other method sperm reservoirs are attached to the outer buccal membrane and the sperm become stored in a special receptacle in the membrane, which is placed opposite the junction of the two ventral arms and opens on its inner surface.
The left ventral arm of the male is always used in transferring the spermatophores, which are grasped by the arm and transferred by its free movement. Ejaculation of the spermatophores is evidently started by the pull given their filaments when the arm starts to transfer them from the penis to the mantle chamber or buccal membrane. The transfer requires rapidity and dexterity and the spermatophores are held in position until ejaculation is complete and their sperm reservoirs are fastened. As many as forty spermatophores may be transferred at a time.
SEXUAL ACTIVITIES OF THE SQUID 351
The egg strings are composed of two kinds of jelly. One kind is supplied by the oviducal gland and the other by the nidamental and probably accessory nidamental glands. The string is apparently molded into shape by passing through the funnel. The jelly is at first soft and sticky but soon becomes tough and loses most of its stickiness.
From the funnel the egg string is drawn between the circlet of arms, where it is held two or more minutes. In sticking the string the female grasps some object with her arms and draws down tight so the string is evidently crowded against it. When she releases her hold the string is left sticking to the object.
Fertilization evidently does not take place inside the oviduct. It doubtless may take place in the mantle chamber when sperm reservoirs are present there, and is known to take place while the egg string is held between the arms. The sperm are hberated from the receptacle while the eggs are between the arms.
Notwithstanding complications, the conditions of hectocotylism shown by cephalopods need not be considered beyond the influence of factors of evolution.
LITERATURE CITED
The cephalopod literature is very extensive. Only those papers directly referred to are here given.
Aristotle History of animals. Trans, by Richard Cresswell. 1891.
Brock, J. 1882 Anat. u. Syst. d. Cephalopoden. Z. f. wiss. Zool. 36. 1884 Mannchen d. Sepioloidea lineata. Z. f. wiss. Zool. 40.
HoYLE, W. E. 1907 Presidential address of Zoological Section. Rept. Brit. Ass. Adv. Sci.
Lafont, M. a. 1869 Observations sur la fecundation des Mollusques Cephalopods der Golfe de Gascogne. Ann. des Sci. Nat. (5) 11.
Racovitza, Emile. G. 1894a Sur I'accouplement des quelques Cephalopods Sepiola rondeletii (Leach), Rossia macrosoma (d. Ch.) et Octopus vulgaris (Lam.). Comp. Rend. I'Acad. des Sci. 118.
1894b Notes de Biologic. I. Accouplement et Fecondation chez rOctopus vulgaris Lam. Arch. d. Zool. Exper. et Gen. (3) 2. 1894c Notes de Biologie. III. Moeurs et Reproduction de la Rossia macrosima (D. Ch.). Arch. d. Zool. Exper. et Gen. (3) 2.
Steenstrup, J. J. S. 1856-57 Hectocotyl. hos Octopodstsegterne. Vid. Selsk. Skr. (5) 4, Translated Ann. N. H. (2) 20. 1881 Sepiadarium og Idiosepius. Vid. Selsk. Skr. (6) L. 1887 Nota; Teuthologica? 7. Overs. Vid. Selsk. Forh.
JOURXAL OF MORPHOLOGY, VOL. 22, XO. 2
352 OILMAN A. DREW
EXPLANATION OF FIGURES
All of the figures that represent the attitudes of squid were drawn from memory after repeated observations. While each figure is thus really a composite, and must represent impressions received rather than the actual positions of particular individuals, much care has been given to the preparation of the figures and it is believed that the general attitudes are reasonably well represented. Sexually mature squid are usually as much as 15 cm. and may exceed 40 cm. in length.
ABBREVIATIONS
hni, inner buccal membrane n, nidamental gland
bmo, outer buccal membrane na, accessory nidamental gland
d, depression in which sperm reservoirs o, oviduct
are attached r, rectum
g, gill s, sperm reservoirs (ejaculated from sper h, modified (hectocotylized) portion of matophores)
arm sr, sperm receptacle
j, jaws sro, opening of sperm receptacle
PLATE 1
EXPLANATION OF FIGURES
1 Copulating squid showing the positions taken by the animals when the spermatophores are inserted into the mantle chamber. The figure shows the animals during the period the arm of the male is inserted in the mantle chamber of the female. Drawn from memory after many observations.
2 Copulating squid showing the positions of the animals when the spei'matophores are placed so that their reservoirs become attached to the outer buccal membrane. The figure .shows the male in the act of grasping the spermatophores with the tip of his arm as they are ejected through the funnel. Drawn from memory after many observations.
3 A common attitude of a sexually excited male. The arms are not kept rigidly in a set position, but are frequently spread as shown in the illustration and held thus for from a few seconds to a minute or more at a time. The drawing is based upon sketches made of active animals.
• 4 Photograph of the two ventral arms of a male squid, showing the slight modification (h) consisting of enlarged peduncles, reduced sucking discs and a ridge between the suckers, toward the tip of the left arm. The wrinkles on the arms are due to shrinkage. A bit of the outer buccal membrane shows between the arms. The arms from which the photograph was made are 9^ cm. long.
PLATE 2
EXPLANATION OF FIGURES
5 A female at rest with the egg string beginning to protrude. Drawn from memorj' and hurried sketches after many observations.
6 A female after she starts to swim, reaching for the egg string with her dorsal arms. With these arms she draws the string between the circle of arms as it is ejected from the funnel. Drawn from memory after many observations.
7 A swimming female, showing the positions of the arms while they surround the egg string. They are held in this position, with the tips somewhat twisted together, for two or three minutes. While the arms closely surround the egg string they show slight individual movements that may be of service in moving the egg string so sperm will be more evenly distributed over it. Drawn from memoiy after many observations.
8 A female squid with the mantle cut and spread o])eii and the arms separated to show the position of attached sperm reservoirs (s) on the oviduct (o) and the sperm receptacle (sr) in the outer buccal membrane.
354
IXTAI- ACTUITIKS OF THE SQUri)
lill.MAX A. DHKW
mi'HOI.OCY, V<
PLATE 3
EXPLANATION OF FIGURE
9 The specimen on the left 8ide shows a female in the position she assumes as she bounces over the bottom on the tips of her arms just previous to selecting a place for sticking the egg string. The specimen on the right side shows the position of a female during the act of sticking an egg string to a rock. Only a few seconds are required to stick the string. The positions of the animals are drawn from memory after many observations.
PLATE I
EXPLANATION OF FICUKKS
10 Jaws and buccal moinbrane of a female squid, with the out er luemliranc [b/no) pulled ventrally to expose the sperm receptacle (the opening of which is shown at .sTo) and the adjacent depression (d). Several sperm reservoirs (.s), ejaculated from spermatophores, arc shown attached in the depression. Magnified about 7 diameters.
11 Section of the outer buccal membrane taken through the sperm receptacle (.s/-). This was taken from an animal shortly after the sperm reservoirs (.s) had been attached and shows sjierm in transit from reservoirs to receptacle. Magnified about 22 diameters.
12 Section through the e])ithelium and secretion lining the depression in which sperm reservoirs are attached. Magnified about 300 diameters.
13 Section through an alveolus of a sperm receptacle. The clear spaces in the epithelivmi are goblet cells. Traces of the Hagella on the sperm and possibly cilia on some of the epithelial cells were visible but they were not definite enough to be put in the drawing. Magnified about 300 tliameters.
358
359
STUDIES OF FERTILIZATION IN NEREIS
I. THE CORTICAL CHANGES IN THE EGG! II. PARTIAL FERTILIZATION
FRANK R. LILLIE
From the Hull Zoological Laboratory, University of Chicago
TEN FIGURES ONE PLATE
I. THE CORTICAL CHANGES IN THE EGG
In many animals one of the immediate effects of fertilization is to cause the egg to throw off a membrane, which is therefore known as the fertilization membrane. This is the case for instance in the eggs of echinids and nematodes. In other cases, where a definite vitelline membrane exists prior to fertihzation, cortical changes occur in the egg immediately after insemination, leading to the formation of a space, the so-called perivitelline space, between the protoplasm of the egg and the vitelline membrane. This is the case for instance in the eggs of at least many annehds, molluscs and vertebrates, and it is unquestionably a more common phenomenon than the formation of a fertilization membrane. There can be little doubt that these apparently different phenomena are simply varying expressions of a change in the cortex of the egg, which is of the same nature in all cases. Loeb's studies ('09) have thrown much light on the nature of these cortical changes. In the case of the egg of Nereis they are relatively obvious in their character and readily followed.
The ovocyte of Nereis is somewhat flattened in a polar direction, measuring about 87.5 x 100^; it is girdled by a double equatorial zone of large oil drops. The large germinal vesicle is central and somewhat elongated in a polar direction.
In his description of the unfertilized egg, Wilson ('92) distinguished two membranes: a delicate outer vitelline membrane,
361
362 FRANK R. LILLIE
and a subjacent membrane or layer, about 6-7^ in thickness, which he called the zona radiata. As will appear from the sequel however, the latter is not a membrane in the usual meaning of the word, but a cortical, coarselj alveolar layer of the egg. Jt is transparent and somewhat granular, and the granules tend to be arranged in radiating lines. There is no perivitelline space in the unfertilized egg.
In sections of unfertilized ovocytes fixed in Flemming's fluid, the zona radiata is seen to be a coarsely alveolar layer with homogenous alveolar contents (fig. 1). The walls of the alveoli are continuous internally with the potoplasm of the egg, and unite externally to form a protoplasmic la^'er applied to the vitelline membrane. The alveoli are closed externally (figs. 1 and 2). The zona radiata is in fact a coarse emulsion or foamstructure.
Unfertilized eggs of Nereis are entirely devoid of jelly and they lie in immediate contact in the sea-water. If India ink be ground up in the water, the particles come in contact with the vitelline membrane. ILsich fertilized egg, on the other hand, is surrounded by a thick layer of colorless transparent jelly; If many eggs are contained in the dish, fusion of the contiguous gelatinous membranes binds the eggs into a mass; the cortical layer (zona radiata) is absent in fertilized eggs, and there is a narrow perivitelline space between the vitelline membrane and the surface of the egg (fig. 3).
The jelly is formed by the extrusion, or diffusion, of the alveolar contents of the cortical layer through the vitelline membrane ; the egg of Nereis, in fact, secretes its own jelly, as may be readily demonstrated in life by inseminating under the microscope with excess of sperm. If the sperm be added to closely placed eggs and a cover glass applied so as to force the eggs into a single layer, aiid the preparation examined with no loss of time, the spermatozoa will be seen in large numbers in contact with the vitelhne membrane. In one or two minutes the spermatozoa are moved away from the surface of the membrane by some invisible repelling substance, and, if the eggs be numerous, the spermatozoa unite in three to five minutes to form lines that bound hexa
STUDIES OF FERTILIZATION
363
gonal areas with the eggs in the centers of the hexagons (fig. A). The substance that sweeps the spermatozoa away from the surface of the eggs is the jelly. Synchronously with its formation, the alveoli of the cortical layer are emptied and the alveolar walls now appear as delicate lines crossing a wide perivitelline space^ (fig. B).
However, not all of the spermatozoa are thus carried out by the secreted jelly, but in the case of each egg a single spermatozoon remains attached to the vitelline membrane. This is very prettily demonstrated if the eggs are under some pressure, so that
Fig. A. Diagram of fertilization with excess of sperm. The outflow of jelly
from the eggs has carried the supernumerary spermatozoa away from the surface
of the eggs (see text). In the case of each egg the single effective spermatozoon
remains attached. From a sketch of the living object.
the spermatozoa are prevented from reaching the eggs above or below. In this case one can discover the single spermatozoon attached to the vitelline membrane in practically every egg (fig. A). All stages of the disappearance of the cortical layer may be readily and rapidly observed. The alveolar walls, however,
1 Wilson ('92) states that "from twenty to thirty minutes after fertilization the striae of the zona suddenly become indistinct and in the course of two or three minutes the zona itself entirely disappears, leaving only the outer membrane." But inasmuch as he was under the impression that the unfertilized eggs possess a transparent, thick, gelatinous envelope like the fertilized ones, he failed to observe the interesting phenomenon of formation of the jelly described here.
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
364 FRANK R. LILLTE
remain as delicate strands of protoplasm uniting the vitelline membrane to the surface of the egg. The jelly, therefore, represents the alveolar contents only of the cortical layer of the unfertihzed egg, and the perivitelline space is nothing but the contracted alveoli of the cortical layer filled with fluid. The perivitelline space must, therefore, be regarded as intraovular with a delicate external cytoplasmic wall lining the vitelline membrane; this we may distinguish as the plasma membrane; it is comparable in some respects to the ferflization membrane of sea-urchins.
Unfertilized eggs allowed to stand in the sea-water form no jelly and retain the cortical layer; the germinal vesicle remains intact; but, if they be strongly centrifuged, the jelly forms, the perivitelline space arises, the germinal vesicle breaks down and both polar bodies are formed; but parthenogenetic development, usually at least, does not take place. Similarly, the addition of a fairly strong solution of potassium chloride to the sea-water causes formation of the jelly and the perivitelhne space while the eggs are still in the solution; maturation takes place after transfer to sea-water but cleavage does not occur (in my experiments; cf . Fischer) , though some differentiation may take place without cleavage. It would appear, then, that conditions that so alter the permeability of the plasma and the vitelline membranes as to permit the outflow of the alveolar contents of the cortical layer initiate development, but that the normal continuation of development is dependent on other factors.
In the normal fertilization of Nereis the stimulus of the spermatozoon causes the formation of the jelly and the perivitelline space iong before it has penetrated the membrane; in fact penetration does not take place until 40 to 50 minutes after insemination. However, mere contact of the spermatozoon with the membrane is apparently not sufficient; but actual attachment of at least a single spermatozoon is required ; this is shown by the fact that the effective spermatozoon is not carried away from the membrane with the unsuccessful ones by the outflow of jelly. Yet the effect of the locafized stimulus of the attached spermatozoon is practically instantaneously effective over the entire
STUDIES OF FERTILIZATION 365
extent of the membrane; it is more like an electrical discharge or some other physical disturbance than a chemical effect.
J. Loeb ('09) has formed the hypothesis that the cortical layer of the egg, especially of sea-urchins, is an emulsion which is rendered stable by a third substance consisting of lipoids, especially cholesterin. The emulsion becomes unstable on solution of the lipoids; this enables the albuminous drops, which he conceives to form one phase of the emulsion, to take up water; hence the layer liquefies and the perivitelline space arises; the fertilization membrane is thus formed. Hence, according to Loeb, the action of lipoid-dissolving substances is to cause the formation of the fertilization membrane. Without committing ourselves to these specific views of the nature of the cortical emulsion, which Loeb himself does not hold very strongly, we may admit that Loeb's hypothesis, that the formation of the fertilization membrane is due to the breaking down of a cortical emulsion, fits the case of Nereis very well. If we go further, however, we must note an important lack of agreement with Loeb's hypothesis. As Loeb himself points out, the theory implies that the membrane of the egg is permeable for sea-water and crystalloid substances, and on the other hand impermeable for colloids; in Nereis the contents of the cortical alveoli are unquestionably colloid, as Loeb's hypothesis requires, but it is perfectly certain that they diffuse through the membrane to form the external jelly ; at the same time, unquestionably, sea-water enters to take the space previously occupied by the colloid. The membrane is therefore permeable for both crystalloids and colloids at this time. I have not, however, investigated farther the properties of the egg membranes and must leave this problem to those who are better qualified as physiologists to make such a study.
It would appear that the presence of this colloid substance in the cortex is an inhibition to the maturation of the egg, because as soon as it is removed, maturation processes are set in motion and both polar bodies are formed. In what manner it inhibits is of course problematical. In the egg of Ascaris megalocephala there is a similar excretion of a cortical colloid which forms, in this case, the thick resistant perivitelline membrane. The ap
366 FRANK R. LILLIE
pearance of the fertilization membrane of echinids might be similarly due to excretion of a cortical colloid which is removed by diffusion and hence is not detected. It is a problem worthy of careful investigation whether the loss of cortical colloids is not the first step in fertilization geuerally.
II. PARTIAL FERTILIZATION
Two functions of the spermatozoon in fertilization may be sharply distinguished. The first is the initiation of the development and the second is the transfer of paternal qualities to the fertilized ovum (heredity from male parent). The first function alone is under consideration in these experiments.
We have seen that in Nereis the immediate effect of attachment of the spermatozoon is essentially the same as a mechanical shock (centrifuging), or a chemical stimulus (KCl); that is, it causes the breaking down of the cortical emulsion and consequent formation of the gelatinous envelope of the egg. But apparently the resemblance extends no farther, for in the case of mechanical or chemical stimulation the impulse to development is lost or greatly weakened after maturation has occurred; and the eggs do not segment. On the other hand the normally fertilized egg does not stop after maturation, but proceeds with its development in a normal fashion. Now the cause of this difference might be either: (a) because the stimulus of the spermatozoon is qualitatively different from, or stronger than, mechanical or chemical stimulation, or (b) because the fertihzing action of the spermatozoon is not completed with the cortical changes but continues after its entrance into the egg. If the first alternative were correct, then the elimination of the spermatozoon after membrane formation should not prevent the normal cleavage and development of ova which had once been stimulated by it; but if the second alternative were correct and the sperm nucleus were prevented from entering the egg after it had induced membrane-formation, then such ova should proceed no further in their development than those mechanically or chemically stimulated.
STUDIES OF FERTILIZATION 367
I have been able to perform this experiment on the eggs of Nereis and have found that eggs in which the spermatozoon is removed after the cortical changes have occurred proceed but little farther in their development than eggs mechanically or chemically stimulated, and they do not undergo segmentation. Fertilization is therefore still incomplete after the formation of the fertilization membrane.
It will be seen that if the results above indicated be demonstrated, the process of fertilization is obviously something more than a beginning of cytolysis or a mere alteration of permeability of the peripheral cell membrane. It would appear to be a progressive change, starting at the periphery and gradually involving the more central portions of the cell. We would, at least, have to distinguish two stages in the fertihzing action of the spermatozoon, one before and the other after penetration.
I shall consider first the evidence for the statement that in the egg of Nereis elimination of the spermatozoon after membraneformation leaves the process of fertilization incomplete. In the second place I shall note the respects in which the completely fertilized egg differs from the partially fertilized egg, and finally, shall consider the bearing of the facts on the theory of fertilization. Inasmuch as it will be necessary to make frequent comparisons with the normal fertihzation, a brief account of the salient features of this process will be given first.
A. Salient features of the normal fertilization
The egg of Nereis is difficult to fix in a thoroughly satisfactory fashion; owing, no doubt, to the presence of the large oil-drops and yolk-granules, uneven fixation with shrinkage is hard to avoid. The eggs appear likewise difficult of penetration, owing probably to the rather viscid jelly from which they cannot be separated; this also makes any considerable number of eggs very bulky and the killing fluid is apt to be much diluted if used in ordinary amounts. After considerable experimenting with picric acid, corrosive subhmate and osmic acid fixing fluids, I finally found one which gives practically perfect results in all
368 FRANK R. LILLIE
stages of maturation and fertilization. This is Meves' modification of Flemming's fluid made as follows: chromic acid, 0.5 per cent, 15 cc; osmic acid, 2 per cent, 3.5 cc; glacial acetic acid 3 drops. The eggs were left in the fluid from thirty to fortyfive minutes. Fixation in this fluid causes no shrinkage, the oil is so changed that it is not dissolved out by subsequent imbedding in parafline; the sections stain beautifully m iron haematoxylin, and certain substances are clearly differentiated which can be detected only with the greatest difficulty after fixation in any other fluid tried.
a. The penetration of the spermatozoon. It was noted in the first part of tjiis paper that a single spermatozoon becomes attached to the egg-membrane immediately after insemination, and that the breaking down of the cortical layer, secretion of the jelly and formation of the perivitelline space follow immediately, though the actual penetration of the spermatozoon is delayed forty or fifty minutes.
The act of penetration involves no motile activity on the part of the spermatozoon ; after the latter has become attached to the vitelline membrane all movement of the spermatozoon ceases and it remains absolutely immotile throughout the forty or fifty minutes that elapse before it is taken into the egg. The events of this period as seen in the living egg are as follows :
1. The jelly is formed by outflow of the alveolar contents of the cortical layer-as already noted; although a large amount of jelly is formed in two or three minutes, yet the process lasts ten or fifteen minutes before the deeper alveoli are emptied. There is then a very wide perivitelline space crossed by the alveolar walls which are attached to the plasma membrane, presenting a very striking appearance (fig. B).
2. The protoplasm of the egg immediately beneath the attached spermatozoon then forms a rounded elevation, the entrance cone, which gradually moves across the perivitelline space and comes in contact, and fuses, with the membrane (fig. B, a). This condition is usually fully attained about fifteen to seventeen minutes after insemination.
3. The entrance cone then gradually retracts, drawing the membrane down to form a depression in which the spermatozoon is included. At this stage one may easily imagine that the sper
Fig. B. History of the fertilization-cone as been in the living egg. Four camera drawings of the same egg: —
a Seventeen minutes after insemination,
b Nineteen minutes after insemination,
c Twenty-two minutes after insemination,
d Twenty-four minutes after insemination. The fertilization-cone is shown at the height of its development in a, its gradual recession and the simultaneous formation of a depression in the membrane is shown in b, c and d.
matozoon has been taken into the egg, as it is apt to be concealed in the depression of the membrane; but this is not the case. The stage of best development of the depression, corresponding to
370 FEANK R, LILLIE
the complete retraction of the entrance cone, is about twenty-two to twenty -five minutes after fertihzation (fig. B, b, c, d).
4. The perivitelhne space then narrows around the entire egg, and the depression of the membrane in which the spermatozoon is seated disappears; in consequence, the spermatozoon again becomes prominent externally.
5. It remains prominent for ten or fifteen minutes (about forty to fifty minutes after insemination), and then disappears rather abruptly within the egg as though some resistance had given away. Its penetration coincides with the late anaphase of the first maturation division; in a few cases it may be a little earlier or a little later.
The egg is changing form at this time and in consequnce the perivitelline space is often widened locally, especially in the animal hemisphere; if this happen in the region of penetration, which may be any part of the egg, strong cytoplasmic strands are drawn out between the membrane of the egg and the point of penetratio.a, showing that the egg membrane and the cytoplasm are actually fused here.
To repeat and extend the observations on the living egg several series of eggs were preserved every five minutes from the time of insemination in Meves' fluid. The study of the sections confirmed and extended the above observation on the living egg as follows :
Ten minutes after insemination the entrance cone is quite well formed and the spermatozoon is clearly seen outside, separated from the entrance cone only by the thickness of the vitelline membrane with which it is in contact.
Fifteen minutes after insemination essentially the same condition persists. The entrance cone is homogeneous, shading off into the surrounding yolk-filled protoplasm. It stains very dark in iron haematoxylin. The head of the spermatozoon appears exactly as before.
Twenty minutes after insemination the entrance cone has flattened out, but the spermatozoon is stiH external to the membrane. The substance of the entrance cone is, however, as readily recognized as when it projected above the surface of the egg.
STUDIES OF FERTILIZATION 371
About thirty-seven minutes after insemination (metaphase of first maturation division) the sperm is still readily found on the exterior of the vitelline membrane external to the substance of the entrance cone which is now lens shaped. The substance of the entrance cone is homogeneous and it stains less than before ; it is sharply marked off from the unaltered egg cytoplasm by a layer of small basophile granules. In the center of its external face is a sharply differentiated granule which stains intensely black in iron haematoxylin, and which is connected to the sperm head by a fine thread passing through the vitelline membrane; penetration has already begun.
Forty-three minutes after insemination (late metaphase of the first maturation division) the entrance cone sinks into the egg-cytoplasm, and the head of the spermatozoon begins to be drawn within the egg in the form of a thick thread, less than onethird the diameter of the sperm head, however. The sperm nucleus is being drawn through the small perforation in the vitelline membrane.
Forty-eight minutes after insemination (stages of anaphase of the first maturation division), nearly all of the sperm head is drawn into the egg in the form of a thick thread several times longer than the original sperm head. Before the head is entirely within the egg its inner end begins to swell and becomes vesicular. The entire entrance cone penetrates the egg-protoplasm always retaining its original connection with the apex of the spermatozoon, so that the original orientation of the sperm is preserved and may be readily recognized after penetration.
Fifty-four minutes after insemination (telophase of the first maturation division), the entire head of the spermatozoon is within the egg. The tail and middle piece usually remain without.
As I intend to publish a separate account of the interesting details of penetration of the spermatozoon, and as the later stages do not concern the present problem, I shall simply say, therefore, that as the united sperm-head and entrance cone penetrate farther into the egg cytoplasm, they rotate in such a way that the entrance cone which was originally in advance of the sperm nucleus comes to lie behind it. During the rotation the sperm
372 FRANK R. LILLIE
aster arises from the pole of the sperm nucleus opposite the entrance cone, thus in the position of the original middle piece.
Morgan has recently ('10), with entire justice as it appears to me, taken a stand against the current view that penetration of the sperm is due to mechanical boring into the egg. He believes that the presence of the sperm calls forth a reaction on the part of the egg that leads to the absorption of the former. There can be no question that the latter conception fits the case of Nereis much better than the former. In the first place the spermatozoon is absolutely motionless after its attachment to the membrane ; in the next place the formation of the entrance cone shows a verj^ decided reaction on the part of the egg to the presence of the spermatozoon; in the third place the retraction of the spermatozoon into a depression of the membrane is due to the retraction of the entrance cone; and finally, as I shall show in a subsequent cytological study, the inclusion of the spermatozoon within the egg appears to be due to activity of the substance of the entrance cone, and not to active penetration by the spermatozoon. The spermatozoon does not penetrate the egg, it is drawn in or engulfed.
6. The later history of the sperm nucleus. The sperm amphiaster is visible in the preparations all through the period of the second maturation division (fig 4). After the formation of the second polar body the sperm-nucleus begins to enlarge and the amphiaster gradually wanes, but it may be recognized up to the time of contact of the germ nuclei. The centrosomes of the first cleavage spindle then begin to appear. Whether or not they are continuous with those of the sperm amphiaster is a question which I shall take up in the next study of this series. The cleavage asters rapidly become very large and conspicuous (figs. 5 and 6). During the growth of the germ nuclei a considerable number of large granules staining strongly in iron haematoxylin appear in their immediate vicinity.
The main point of these observations on the normal fertilization, both in the living eggs and also in section, is to demonstrate for elucidation of the experiments following: (1) That membrane formation precedes penetration of the spermatozoon by a long time. (2) That the spermatozoon does not penetrate
STUDIES OF FERTILIZATION 373
the vitelline membrane and enter the egg until at least forty to fifty minutes after insemination, although its attachment to the membrane takes place immediately. (3) That the presence of the sperm-nucleus is readily demonstrable in all stages after penetration.
B. Removal of the spermatozoon after membrane formation
In the summer of 1909 I was studying the effects of centrifuging on the egg of Nereis with the aim of getting more data on the problem of polarity and the theory of formative stuffs. It soon became apparent that the effects of centrifuging varied with the stage of development, and so several series of experiments were made in which the eggs were centrifuged at regular intervals from before fertilization up to the time of the first cleavage.
The effects of centrifuging may be divided into three categories: (1) A certain proportion of centrifuged eggs develop approximately normally, the percentage varying greatly with the stage of centrifuging. (2) A certain proportion of eggs, varying at different stages, segment more or less abnormally, sometimes extremely so {e.g. meroblastic), and produce embryos with more or less pronounced abnormalities. (8) At certain stages of centrifuging a variable proportion of eggs fails to carry out even the first cleavage. The investigation of the causes of such failure to segment revealed the fact that it was owing to the removal of the spermatozoon after membrane-formation. It is the evidence for this statement that is now under consideration.
The results with reference to failure to segment were, in general, as follows :
1. If unfertilized eggs were centrifuged and then fertilized, all segmented, and a large percentage tended to develop quite normally.
2. A disturbing factor comes in shortly after insemination, owing to the fact that when the jelly is first secreted by the eggs it is so viscid that the eggs stick together in the bottom of the centrifuge in a mass which cannot be separated into its constitu
ent eggs. The extreme viscidity of the jelly gradually disappears, and after about twenty minutes from insemination, the
eggs no longer mat together. It is therefore difficult to investigate the effects of centrifuging on the developmental capacity
of the eggs during the first ten or fifteen minutes after insemination. However, when the viscid stage begins to pass away
and eggs can be separated from the mass for examination, the
majority are found to undergo segmentation, as many as 98 percent in one case (experimeat 2, 1910) twenty-one minutes after
insemination.
Fig. C. The effects of centrifuging on the power to segment in Nereis. The abscissae represent minutes from the time of i semination, the ordinates the percentage of eggs dividing. At position a penetration of the spermatozoon is just completed in most of the eggs. At position b the first polar body is extruded. Data from experiment 2, 1910.
3. For about the next thirty minutes (twenty-one to fiftythree minutes after insemination) centrifuging tends to inhibit the cleavage of a certain proportion of the eggs which gradually increases up to about thirty-seven minutes after insemination and then decreases again. For instance, in experiment 2 of 1910, of the eggs centrifuged twenty- one minutes after insemination 98 per cent segmented; twenty-six minutes after insemination 36 per cent segmented; thirty-two minutes, 33 per cent; thirty-seven minutes, 21 per cent segmented; forty-three minutes, 26 per cent segmented; forty-eight minutes, 52 per cent segmented; fifty-three minutes, 75 per cent segmented; fifty
STUDIES OF FERTILIZATION
375
eight minutes, 90 per cent segmented; sixty-three minutes, 90+ per cent segmented; sixty-nine minutes, 95+ per cent segmented; control eggs, 99 per cent segmented. (See Fig. C.)
4. From this time on nearly all of the eggs segment after centrifuging until, during the anaphase and telophase of the first cleavage spindle, centrifuging again tends to inhibit cleavage.
The following table (Experiment 29) gives a fairly typical series of results. There were twelve such experimental series in all, more or less complete, giving concordant results except that in some at the period corresponding to 29D, 90 to 98 per cent were so affected that they failed to segment. On either side of this critical period there is a decreasing susceptibility to such injury by centrifuging.
Experiment 29 September 8, 1909
DESIGNATION
29..
29 A.
29B. 29C.
Control (not centri fuged) Before insemination
20 minutes 30 minutes
29D i 41 minutes
29E. 29F. 29G. 29H. 291. 29 J. 29K.
51 minutes 66 minutes 79 minutes 95 minutes 114 minutes 121 minutes 127 minutes
29L 149 minutes
PERCENTAGE SEGMENTED
100 per cent
100 per cent practically Majority Majority 20 to 30 per cent 70 to 80 per cent 90 to 95 per cent 100 per cent 100 per cent 100 per cent Some unsegraented Less than majority and these irregular Most segmented further
Eggs matted loosely
Centrifuged during process of Ist cleavage
Centrifuged in 2-cell stage
Two major processes are going on in the egg at this time, viz. : maturation and fertilization. The injury is not primarily to the process of maturation, for the eggs that do not segment form the polar bodies; nor is it probable that there is a general
37(
FRANK R. LTLLIE
systemic injury to the egg protoplasm at this time not received at other times, when the fact that maturation continues and polarity is preserved in these eggs, is considered. It would, therefore, appear probable that the injury is to the process of fertilization itself, and this conjecture is completely confirmed by cytological study. The most conclusive experiment in this respect is no. 27, the details of which are as follows:
The eggs were fertilized at 9:28 a.m., September 4, 1909. Some of these were kept for control and all segmented normally. The remainder were centrifuged about 60 x 120 revolutions at a radius of 6 cm. in one minute, at the following times : 27A at 9 :58 A.M. ; 27B at 10 :03 ; 27C at 10 :08 ; 27D at 10:12; 27E at 10:16. About 20 per cent of 27A segmented, 5 to 10 per cent of 27B, 20 per cent of 27C, 50 to 60 per cent of 27D, and 75 to 90 per cent of 27E. Samples of the control and of each of 27A, 27B, 27C, 27D and 27E were preserved at 10:31 and 10:43 a.m.
EXPERIMENT 27
Eggs fertilized at 9:28 a.m., Sept. 4, 1909
TIME AFTER
DESIGNA
CENTRIFUGED
INSEMI
SAMPLES PRESERVED
LIVING EGGS
TION
NATION
27 Con
Xot centrifuged
(1) 10:311 A.M.
All segmented
trol..
(2) 10:45 §A.M.
27A . . .
60 X 120 rev. in
(1) 10:30 A.M.
About 20 per cent
1 min. 9:58 a.m.
30 min.
(2) 10:431 A.M.
segmented
27B....
60 X 120 rev. in
(1) 10:30J A.M.
About 5-10 per
1 min. 10:03 a.m.
35 min.
(2) 10:431 A.M.
cent segmented
27C....
60 X 120 rev. in
(1) 10:30iA.M.
About 20 per cent
1 min. 10:08 A.M.
40 min.
(2) 10:44 A.M.
segmented
27D...._
60 X 120 rev. in
(1) 10:31 A.M.
About 50-60 per
1 min. 10:12 a.m.
44 min.
(2) 10:441 A.M.
cent segmented
27E....
60 X 120 rev. in
(1) 10:31 A.M.
About 75-90 per
1 min. 10:16 a.m.
48 min.
(2) 10:45 A.M.
cent segmented
^Since the above was written this experiment has been repeated (Exp. 2, '10), with the added precaution of preserving a sample of the normal eggs corresponding to each stage of centrifuging, in order to make certain of the stages of fertilization in each case. The results completely confirm those of experiment 27, and the sperm was found to be external in the most critical stage (37 minutes after insemination in this experiment; see page 371).
STUDIES OF FERTILIZATION 377
Cytological studj' of the twelve lots of preserved eggs showed stages ranging from the metaphase of the second maturation spindle to the prophase of the first cleavage, the earlier stages of course being found in lot 1 in each case.
In the control lots it was easy to demonstrate the sperm nucleus at all stages to the formation of the first cleavage spindle. The sperm nucleus is rendered particularly conspicuous during the second maturation division by the large amphiaster that accompanies it (fig. 4), both lying in the yolk-free protoplasm. After the formation of the second polar body the sperm amphiaster gradually fades, but the sperm nucleus can be recognized by its position and by the remnants of radiations up to the time of union of the two germ nuclei; and in the later stages its presence may be inferred by the degree of development of the cleavage amphiaster and the number of chromosomes. There is, therefore, no time from the beginning of the second maturation division up to the formation of the first cleavage spindle when the presence of the sperm nucleus cannot be readily demonstrated.
In the study of the serial sections of the control eggs I found no egg in which, all sections being present, the sperm nucleus could not be demonstrated. In the serial sections of 27A, the sperm nucleus could be recognized in only about 37 per cent of the eggs; in 27B in only 10 per cent to 20 per cent; in 27C in about 25 per cent; in 27D in about 53 per cent; in 27E in about 76 per cent. The stages of maturation of lots A to E corresponded very closely with the stages of maturation of the control eggs killed at the same time.
It is a relatively simple matter to demonstrate the presence of the sperm nucleus, for a single positive observation suffices; but, to be sure of the absence of a sperm nucleus from any particular egg, it is necessary to examine practically every section of the egg, and the absence of two consecutive sections is sufficient reason for excluding an egg from the count. This may be one reason why the number of eggs in the different lots shown to contain sperm nuclei tends to be somewhat larger than the estimate of the number of eggs that segmented. Another reason
378
FRANK R. LILLIE
probably is that a sperm nucleus may persist to a certain extent even if injured and unable to produce the full fertilizing effect and cause cleavage.
A third reason for discrepancy in the results is that abnormalities of maturation may be produced by centrifuging which render the determination of the sperm nucleus more difficult than usual. It frequently happens that, as the first maturation spindle is driven from the surface by the centrifugal force, it divides before it reaches the surface again, producing two maturation nuclei within the egg. The two second maturation spindles may then unite to form a tetraster, one pole of which approaches the surface and a single polar body is formed, leaving three nuclei within the eggs (fig. 7). Such eggs were readily recognized by the absence of the first polar globule, and by the presence of the extra nuclei. But it was sometimes difficult to determine in certain stages whether there were only three nuclei, the sperm nucleus being absent, or four, the sperm nucleus being present.
A fourth reason for a certain discrepancy between the estimate of the number of eggs that segmented and the number determined to have sperm nuclei might be that at the time the experiment was made the importance of exact determination of the number of eggs that segmented was not realized, and the determination was made only roughly. Putting the results side by side we have :
PERCENTAGE OF EGGS OBSERVED TO DIVIDE IN LIVING CONDITION
PERCENTAGE DETERMINED
BY SERIAL SECTIONS TO
POSSESS SPERM NUCLEI
27 Control
All
About 20 per cent 5 to 10 per cent About 20 per cent About 50-60 per cent About 75-90 per cent
All
27A
About 37 per cent
About 10-20 per cent
About 25 per cent
About 53 per cent
About 76 per cent
27B
27C
27D
27E
Considering the various sources of error, the agreement is very close except in 27A. But in this case we do not have to
STUDIES OF FERTILIZATION 379
explain why eggs in which the sperm nucleus was absent segmented but on the contrary, why certain eggs that possessed the sperm nucleus failed to segment, which is a very different thing. There is no evidence that any egg in which the sperm nucleus was absent succeeded in dividing.
The general conclusion that removal of the spermatozoon at the times noted in the experiments involves incomplete fertilization, is sufficiently demonstrated by the results.
Let us call the stage at which the spermatozoon is eliminated in the greatest proportion of eggs, the critical period. The exact number of minutes from the time of mixing eggs and sperm to this stage varies of course through the season, owing to the variations of temperature. Moreover, it is not exactly determined in all experiments, for in some the stages of centrifuging fall on either side of it. This being understood, we may note that in eight experiments the critical period occurred at from 25 minutes to 40 minutes after fertilization. This is quite a wide variation, but when the time is represented as a fraction of the entire period between fertilization and the first cleavage, it is found that in all cases the period up to the critical period is between 27 and 33 per cent of the total time up to the first cleavage. It is obviously a corresponding stage in all cases, for the observed differences fall within the chances of error, viz : that the critical period is hit exactly in only very few experiments, and that the time of beginning of the first cleavage must be stated rather arbitrarily on account of the variation in rate of individual eggs.
The critical period occurs shortly before the penetration of the spermatozoon into the egg. We noted in the first part of this paper that the penetration of the spermatozoon is extremely gradual; my observations on this point, both from the study of the living egg and also of sections, show that it requires forty to fifty minutes for the head of the spermatozoon to disappear through the membrane.
As the most critical period comes in the great majority of experiments from thirty-five to forty minutes after insemination, it is obvious that the spermatozoon is in some way prevented
JOURNAL OF MORPHOLOGY, VOL.
380 FRANK R. LILLIE
from entering the egg. The explanation is comparatively simple; the spermatozoon is imbedded in the jelly by which the egg is surrounded. When the jelly is first formed it is very viscid, and adheres to the eggs during centrif uging so that they mat together in the centrifuge. However, this stage passes and the result of centrif uging is then to separate the jelly from the eggs. In many cases the jelly carries off the attached spermatozoon with it. After penetration this can of course no longer happen. The curve of variation of the per cent of eggs centrifuged before penetration that fail to segment is due to the following factors: (a) At about twenty-five minutes the sperm head is sunk in a deep depression of the membrane and hence is less likely to be torn away by the jelly; (fe) the change in consistency of the jelly presumably extends from without inwards; hence at first the innermost layer in which the spermatozoon is imbedded tends for a time (also presumably) to remain with the egg; (c) the time of penetration varies somewhat in any lot of eggs. These facts together would explain why the percentage of eggs that fail to segment after centrifuging rises to a maximum and sinks again to a minimum in the successive stages of centrifuging.
The fact that centrifuging inhibits cleavage in a small per cent of the eggs from fifty to fifty-five minutes after insemination, leads me to suspect that in some cases the sperm may be destroyed after its penetration into the egg. In experiment 2, 1910, for instance, cleavage was inhibited in 25 per cent of the eggs centrifuged fifty-three minutes after f ertihzation ; in the control eggs killed at the time of centrifuging, penetration of the sperm is completed in the great majority of eggs, though it is found external still in a very few ; it is difficult to estimate the per cent of the latter, but the impression is that it is less than 25 per cent. However, it is impossible to confirm this, and I mention the matter here to call attention to the error in my first paper read before the Central Branch of the American Zoological Society (Abstract in Science, vol. 18, p. 36, May, 1910), in which I stated that the destruction of the sperm nucleus followed penetration. A renewed study of the penetration has proved that this is not the case, usually at least.
STUDIES OF FERTILIZATION 381
We are not, of course, free to infer that fertilization is complete as the stimulus to development immediately after penetration of the spermatozoon. The experiments prove directly that in the egg of Nereis the stimulus of the spermatozoon as the impulse to development consists of two distinct parts: (1) an external stimulus that causes membrane formation and which is sufficient of itself to induce the maturation ; (2) an internal stimulus dependent on penetration of the spermatozoon. How long after penetration fertilization is still incomplete cannot be decided on the basis of these experiments.
In concluding this section, we may note that the cause of failure to segment following centrifuging during the anaphase of the first cleavage is an entirely different one. The cause in this case is the breaking up of the karyokinetic figure and function, and dispersing the chromosomes. This is readily demonstrated in sections. In the case of eggs centrifuged at the ' critical stage', the sections show that the maturation spindle receives no injury from centrifuging, but appears coherent and normal in the sections. The sections of eggs centrifuged during the anaphases of the first cleavage show the chromosomes dispersed through the cytoplasm and the cleavage spindle no longer coherent, but broken up. In the first case the cause of failure to segment is elimination of the sperm-nucleus, as shown by study of series 27. This cannot be the cause in the second case, and a sufficient explanation of the failure to segment is found in the destruction of the karyokinetic figure.
C. Comparison of co??ipletely and partially fertilized eggs in later
We have noted so far that definite proportions of eggs centrifuged at definite periods in the process of fertilization fail to develop a sperm-nucleus, and that similar proportions of the same lots of eggs when left to develop fail to undergo segmentation. The facts (1) that all the control eggs of the same lot segment, and (2) that the centrifuged eggs that fail to segment, nevertheless had formed the fertilization membrane and under
382 FRANK R, LILLIE
gone maturation, prove that the unsegmented eggs had received at least the first stimulus of fertilization. It was also shown that the critical period for suppressing segmentation by centrifuging occurs at a time shortly before entrance of the spermatozoon, and that it is due to prevention of penetration. The partially fertilized eggs, therefore, resemble the normal ones in the fact that membrane formation and the first stimulus to development are called forth by action of the spermatozoon, and they differ from the normally fertilized eggs in that the internal egg protoplasm has not received the direct stimulus of the spermatozoon. A cy tological examination of such eggs could not fail to be of interest and might give some clue to the internal function of the spermatozoon in fertilization.
Both polar bodies form regularly in such eggs as already noted, and the egg-nucleus (female pronucleus) arises and attains the same size as in normally fertihzed eggs. The chromosomes of the first cleavage spindle then form in the usual fashion and at the usual time, accompanied by disappearance of the nuclear membrane. But, whereas, in the presence of a sperm nucleus, cytoplasmic asters accompany these processes and a spindle rapidly arises during the prophases of the first cleavage, in the absence of the sperm nucleus there is absolutely no sign of c\^tasters or evidence of spindle formation. The chromosomes lie naked in the cytoplasm surrounded by a clear area (fig. 7).
Each chromosome then splits longitudinally in the usual fashion, but the halves do not separate. At the time of the telophase of the normal first cleavage there is a tendency to scattering and breaking up of the chromosomes. When the normal eggs have reached the two and four-celled stages, the scattering and breaking up of the chromosomes have progressed much farther in the unsegmented eggs, and in the course of two or three hours there remains no differentiated nucleus or chromosomes, but only numerous chromatic granules scattered throughout the cytoplasm.
The behavior of the partially fertilized eggs may be compared on the one hand with that of normally fertilized eggs and on the other with that of eggs caused to mature by centrifuging. As com
STUDIES OF FERTILIZATION 383
pared with the former, the chief difference observable by cytological methods is the entire absence of the achromatic part of the karyokinetic figure. The differences in later stages may be conceived as secondary effects of this defect or of the conditions determining such defect. When eggs are caused to mature by centrifuging the process begins as in normally fertilized eggs by the breaking down of the cortical layer and formation of the jelly; the germinal vesicle ruptures and the two maturation divisions follow. After the completion of the maturation the chromosomal vesicles of the egg nucleus usually fail to unite perfectly, and in a httle while they separate and scatter in the cytoplasm without formation of chromosomes, so that each egg appears to possess a considerable number of small nuclei. In a few cases the first indications of chromosome formation may be observed in the vesicles shortly after maturation but not later. Subsequently the chromosomal vesicles appear to dissolve in the cytoplasm liberating small chromatic nucleoli.
The partial stimulus of the spermatozoon is thus somewhat more effective than the mechanical shock of centrifuging, though both produced the same initial changes, apparently equally well. This may possibly be due to entrance of a small amount of matter from the spermatozoon; for at the critical period the perforatorium of the sperm has penetrated the membrane and is imbedded in the entrance cone.
D. General discussion
The general conclusion that the function of the spermatozoon in the stimulus to development involves at least two factors has already been clearly stated by Boveri ('07) and Loeb ('096) : According to Loeb, one factor is the cytolysis of the very thin cortical layer of egg"; but while this stimulates development, the latter is often abnormal and therefore usually comes to a halt. A second process is necessary to ensure more normal and lasting development (Loeb '096). Apparently Loeb is not very clear concerning the nature of the second factor, but is inclined to regard it as inhibiting the cytolysis which he conceives to be
384 FRANK R. LILLIE
begun as the first factor of the developmental stimulus. This conclusion was formed as a result of two kinds of experiments: In the first Loeb found that the best results in artificial parthenogenesis are obtained, in the egg of a Californian sea-urchin, by a double treatment : first using a cytolytic agent and then following it by treatment with hypertonic sea-water, or by inhibiting oxidation for a while. In the second class of experiments Loeb and Elder found that mere membrane formation might be induced in sea-urchin eggs by external contact of starfish spermatozoa, but farther development did not take place except in the relatively few cases in which the spermatozoon entered the egg (Loeb '09b, p. 249), or unless the eggs were treated after membrane formation with hypertonic sea-water. Although this experiment is complicated by the hybridizing, yet it demonstrates the same distinction between external and internal functions of the spermatozoon in fertilization that I have shown for Nereis by a different method.
Artificial parthenogenesis may be induced in the sea-urchin egg without membrane formation and this fact appears to me to indicate that the internal function of the spermatozoon is probably at least as fundamental as the external function (membrane formation), though, as Loeb points out, development without membrane formation takes place in a less normal fashion than after membrane formation. But inasmuch as we may have membrane-formation without development, and development without membrane formation, it would seem premature at least to consider membrane formation as the chief function of the spermatozoon in fertilization.
The experiments described in this paper show that m Nereis fertilization by the spermatozoon is incomplete after the formation of the membrane. The question then arises, when is the function of the spermatozoon in fertilization completed? Ziegler's and Wilson's experiments show that it is incomplete even some time after entrance of the spermatozoon. Ziegler's experiments ('98) consisted "in so constricting the egg of the sea-urchin after penetration of the spermatozoon that the one part contains the sperm nucleus, the other part the female sex-nucleus. The
STUDIES OF FERTILIZATION 385
part that contains the sperm nucleus undergoes cleavage and develops farther; in the other part the female sex-nucleus undergoes remarkable transformations, dissolving and reappearing, a process which is repeated several times." In spite, therefore, of the presence of the sperm-nucleus in a constricted part of the same egg, the part containing the egg nucleus was not fully fertilized. It made abortive attempts at division, but the karyokinetic figure was too feeble to carry the process through.
Wilson observed ('03) that if the fertihzed eggs of Cerebratulus be cut in two shortly after the penetration of the spermatozoon only a single fragment develops even though the fragments be refertilized immediately after the operation. In such cases it is almost invariably the nucleated fragments that develop, but in a very few cases I have observed that the enucleated fragment develops, ' while the nucleated one forms the polar bodies, but proceeds no further." By the nucleated fragment in this case Wilson means the fragment containing the maturation spindle. Farther on he adds The few cases in which the enucleated fragment of the bisected fertihzed egg develops are doubtless those in which the plane of section separates the sperm-nucleus from the egg-nucleus." This is indeed the only possible explanation. In such cases the fragment containing the egg-nucleus is only partially fertilized.
Boveri has also observed that if freshly fertilized sea-urchin eggs be broken into fragments by shaking, certain of the fragments contain the egg nucleus alone. If such fragments are not subsequently entered by a spermatozoon, the nucleus enlarges, dissolves and reappears again; but they do not segment ('96). Later he observed that such pieces may divide at least twice ('02).
These experiments demonstrate that fertilization is still partial even some time after entrance of the spermatozoon into the egg; but they do not show at what stage it is complete. Boveri's very interesting observations on 'partial fertilization' in the sea-urchin egg ('88 and '90) carry the solution of the problem a step farther (cf. also Teichmann '02). In the experiments which furnished the material for his observations both eggs and sperm were weakened, the former by standing for fourteen hours
386 FRANK R. LILLIE
in sea-water and the latter by treatment with KOH prolonged to a stage in which only a small percentage of the spermatozoa continued to move. Under these conditions in a large number of eggs the sperm aster separated from the sperm nucleus, which was usually left on one side, and proceeded alone to conjugate with the egg-nucleus. Thereupon the cleavage spindle formed with the egg-nucleus alone, and segmentation of the egg ensued. In the four-cell stage usually, but sometimes in the two or eightcell stage, the sperm nucleus united with one of the segmentation nuclei. Boveri concludes from this and other results that the fertihzing action of the spermatozoon consists in the introduction of a centrosome into the egg.^ When this has united with the egg nucleus, with or without participation of the sperm nucleus, fertilization would be complete; or with the sperm nucleus alone in merogony it likewise completes fertilization.
Boveri has used the term 'partial fertilization' for the phenomenon just described, although he admits that it is a misnomer. It is unfortunate that such a significance should have come to be attached to the expression, because, as has been shown, my own results and those of Ziegler, Wilson and Boveri himself prove that partial fertilization in the literal sense really occurs. The various stages of partial fertilization as shown by the results in the literature on the subject are:
1. External contact alone by the spermatozoon producing, a. Formation of the fertilization membrane, (Loeb and Elder
for sea-urchins, Lillie for Nereis).
h. Maturation and formation of the chromosomes from the egg nucleus without spindle, (Lillie for Nereis).
c. Maturation and cleavage to stereoblastula, (Bataillon: hybrid union of eggs of Pelodytes punctatus and Bufo calamita with sperm of Triton alpestris).
2. If the spermatozoon be removed shortlj after entrance, a. Maturation alone may result, (Wilson on Cerebratulus) .
^Herbst ('07 p. 202 and '09 p. 277) interprets Boveri's 'partial fertilization' as a combination of parthenogenesis and fertilization. Such an interpretation does not, however, explain Boveri's account of the behavior of the sperm-centrosomes.
STUDIES OF FERTILIZATION 387
b. An abortive karyokinetic figure may form with the egg nucleus alone, (Ziegler and Boveri on sea-urchins).
c. In some cases at least two cleavages may result, (Boveri on sea-urchins). Boveri's so-called 'partial fertilization, is really full fertilization, so it does not come in this series.
The difference in reaction of the egg-cytoplasm to its own nucleus and to the introduced part of the spermatozoon can be explained on only one of two grounds; either in the general sense of Boveri's theory on purely morphological grounds, or on the ground of a chemical difference, presumably of sexual origin, between the egg on the one hand and the sperm on the other. The latter form of interpretation seems to me preferable because it is a physiological interpretation which takes cognizance of the sexual factor in fertilization,
Boveri's theory of fertilization rests on the identification of the centrosomes of the sperm aster with a definite formed element (centrosome) introduced into the egg by the spermatozoon; but it has never been demonstrated in any case in all the literature on the subject of fertilization that the centrosomes of the cleavage spindle, or indeed of the sperm aster, are derived from any definite formed element of the spermatozoon. Boveri himself admits this in his sixth cell study ('07, page 266) ; and so long as definite proof of the continuity of the so-called spermcentrosomes from the spermatid up to the formation of the first cleavage spindle is lacking, all of Boveri's observations are open to another interpretation than the one he has given; to the interpretation, namely, that the sperm asters represent a reaction of the egg cytoplasm to a male element, or at least a foreign element, represented for the most part by the sperm nucleus.
The biological analysis of fertilization seems to me to rest now upon the problem of the origin of the sperm aster in the egg. More crucial evidence is needed on this point, and I do not believe that any refinement of cytological technique will give the result. Experimental evidence is needed; either, as Boveri ('88) suggested, the introduction of a non-nucleated spermatozoon in the egg to prove whether or not the sperm asters would arise without the nucleus and fertilize the egg, or the introduction of only the
388 FRANK R. LILLIE
anterior part of the sperm into the egg to prove whether or not the sperm nucleus without the centrosome, which is contained in the middle piece of the spermatozoon, would cause the production of asters in the egg-cytoplasm.
As Boveri, among others, has pointed out, there is not onlyone, but several stages of inhibition in the history of the egg. This may be illustrated by noting the stages at which in the eggs of various animals the need for fertilization arises. In some eggs it is before the rupture of the germinal vesicle {e.g., Nereis), in others at the time of the mesophase of the first maturation spindle {e.g., Chaetopterus and Cerebratulus), in others again after the formation of the first polar body {e.g., Amphioxus, amphibians), in others again, not until after the formation of both polar bodies {e.g., sea-urchin). There is no doubt that the last stage of inhibition is the most difficult one to overcome, both because many eggs pass by the earlier stages without apparent specific stimuli and also because it is possible to cause eggs that normally stop at the first or second stage of inhibition to pass on to the last stage by stimuli that are ineffective when this stage is reached, {e.g., Nereis as noted in this paper and Chaetopterus as noted in various earlier papers).
The nature of the inhibition that causes the need for fertilization is a most fundamental problem. Is it the same in all these cases, i.e., a gradually increasing inhibition that may be effective before maturation, but in some cases not until maturation has progressed a certain distance, or even until it is complete?'* If it is the same cause at all these stages then it is certain that the need for fertilization is not due to any defect of the egg-centrosomes, for the pause takes place in Chaetopterus (e.g.) while the egg-centrosomes are at the very height of their activity. If,
^Bataillon ('10) holds the view that the nature of the inhibition is the same whether the arrest is at the stage of the resting nucleus or in the height of karyokinesis. He holds the view that the inhibition is due to accumulation of excretory products and that the stimulus to development is essentially a process of elimination. Bataillon's paper was received after my own was completely written. His interesting results will be considered more fully in my next paper. In Part I of the present paper (last paragraph) I have presented a view similar in some respects to Bataillon's.
STUDIES OF FERTILIZATION 389
therefore, we are to hold to the theory of Boveri in its Hteral sense, we must beheve that there are different kinds of inhibition. However, it is, I beheve, simpler and more logical to hold that the inhibition differs only in intensity at these various stages; and this point of view seems to be supported by the fact that the same stimulus which at a lower intensity will cause only maturation to take place in Chaetopterus, at a higher intensity will cause differentiation also to proceed, though in this case without cleavage. Boveri ('07), however, holds that there are different kinds of inhibition, that the postulated degeneracy of the eggcentrosomes after maturation is in a sense the more primitive, and that other kinds have been secondarily acquired, a point of view that gives a more or less definitely teleological aspect to the question.
From a physiological point of view we might inquire, what are the conditions that cause the postulated sudden degeneration of the egg-centrosomes? Such a condition if found, would be nearer the fundamental cause of inhibition of the egg and it might turn out to be the same cause that conditions in so many cases an earlier arrest of activities in the egg.
The experiments on artificial parthenogenesis are sometimes regarded as involving the entire problem of fertilization. But if it be true, as many believe, that biological fertilization, (if I may be pardoned such an expression) is fundamentally a sexual reaction, then the physico-chemical analysis of fertilization must compass the entire problem of sex, which is much wider than the problem of parthenogenesis. The physico-chemical analysis of fertilization has dealt, up to the present exclusively, with the latter problem, and for this reason the earlier title of such studies 'artificial parthenogenesis', seems to me much more fitting than 'chemical fertilization' which is sometimes loosely used. From the zoological point of view, at least, parthenogenesis and fertilization are not interchangeable functions. There is a factor present in fertilization which is absent in parthenogenesis, and the latter is never the exclusive mode of reproduction among animals. The biological analysis of fertilization therefore involves problems that do not occur in the physico-chemical analysis of parthenogene>-is.
390 FRANK R. LILLIE
LITERATURE CITED
Bataillon, E. 1910 Le probleme de la fecondation circonscrit par I'impregnation sans amphixie et la parthenogenese traumatique. Arch, de Zool. exp. et gen. 5 ser. Tome 6.
BovERi, Th. 1888 Ueber partielle Befruchtung. Sitz'b. d. Ges. fiir Morph. u. Phys. in Munchen. Bd. 4, H. 2.
1890 Zellenstudien. Heft. 3. Ueber das Verhalten der chromatischen Kernsubstanz bei der Bildung der Richtungskorper und bei der Befruchtung. Jena. pp. 32 ff.
1896 Zur Physiologic der Kern und Zelltheilung. Sitz'ber. d. Phjs.Med. Ges. zu Wurzburg, (cited from Teichmann — unfortunately the original paper was inaccessible to me).
1902 Das Problem der Befruchtung. Jena, G. Fischer.
1907 Zellen-Studien. Heft 6. Die Entwicklung dispermer Seeigeleier. Ein Beitrag zur Befruchtungslehre und zur Theorie des Kerns. Jena. Gustav Fischer.
Fischer, Martin H. 1903 Artificial parthenogenesis in Nereis. Am. Jour. Physiol, vol. 9, pp. 100-109.
Herbst, Curt 1907 Vererbungsstudien V. Auf der Suche nach der Ursache der grosseren oder geringeren Ahnlichkeit der N^chkommen mit einem der beiden Eltern. Arch. f. Entw'mech. Bd. 24, p. 185.
1909 Vererbungsstudien VI. Die cytologische Grundlagen der Verschiebung der Vererbungsrichtung nach der miitterlichen Seite. I Mittheilung. Arch f. Entw'mech. Bd. 27, p. 266.
LoEB, Jacques 1909a Ueber das Wesen der formativen Reizung. Berlin, Julius Springer.
1909b Die chemische Entwicklungserregung des tierischen Eies. Berlin, Julius Springer.
Morgan, T. H. 1910 Cross and self-fertilization in Ciona intestinalis. Archiv f. Entw'mech. der Organismen. Bd. 30, ii, Theil.
Teichmann, Ernst 1902 Ueber Furchung befruchteter Seeigeleier ohneBeteiligung des Spermakerns — Jen. Zeitschr. f. Naturw. N. F. Bd. 30, p!105.
Wilson, E. B. 1892 The cell-lineage of Nereis — A contribution to the cytogeny of the annelid body. Jour. Morph. vol. 6.
1903 Experiments on cleavage and localization in the Nemertine egg. Arch. f. Entw'mech. vol. 16, pp. 417-418.
ZiEGLER, H. E. 1898 Experimentelle Studien tiber die Zelltheilung, II. Arch. f. Entw. mech. vi.
PLATE 1
EXPLANATION OF FIOURKS
1 Axial section of an unfertilizod normal ovocyte of Nereis, fixed in Flemming's fluid, weaker solution. In this fixing fluid the yolk granules swpll and tend to run together. The oil drops are dissolved out in the preparation and are represented by emptj^ spaces, v. in. vitelline membrane, c.l. cortical layer, from which the jelly is formed.
2 Section of an egg of Nereis, fixed in Aleves' fluid five minutes after insemination. The Cortical layer is already somewhat reduced in thickness. The yolk granules are not swollen. The oil drops are not dissolved out. The section is approximately horizontal, c.l., remains of cortical layer; v.m., vitelline membrane.
3 Section of an egg of Nereis, fixed in Meves' fluid fifteen minutes after insemination. The cortical layer has entirely disappeared, and the perivitelline space is formed. The germinal vesicle is breaking down and the first maturation spindle is forming, p.v. perivitelline space, v.m., vitelline membrane.
4 Section of an egg of Nereis, fixed in Meves' fluid fifty-seven minutes after insemination. Only one centrosome of the sperm amphiaster is shown.
5 Section through the first cleavage-spindle of Nereis, normal, one hour and twenty-seven minutes after insemination.
6 Tetrapolar second maturation spindle of Nereis. See text for description (p. 378). Three egg nuclei are formed in such a case.
7 Egg nucleus of egg of Nereis in which the spermatozoon was removed by centrifuging. The chromosomes of the first cleavage are formed, but there are no asters. Cf. fig. 5.
All figures drawn with the camera lucida with Zeiss comp. oc. 6 and 2 mm. horn, oil im. obj.
THE GROWTH AND DIFFERENTIATION OF THE
CHAIN OF CYCLOSALPA AFFINIS
CHAMISSO
WM. E. RITTER and MYRTLE E. JOHNSON
From the Laboratory of the Marine Biological Association of San Diego
TWENTY-FIVE FIGURES FOUR PLATES
CONTENTS
Purpose of the research 396
1. Special 396
2. General 396
Brief description of the species 398
Measurements of the zooids of the wheels and of a portion of the chain not yet
transformed into wheels 399
Treatment of the quantitative data 406
Attempt to connect the formation of wheels with morphological, physiological,
and mechanical phenomena presented by the animals 414
1. Segmentation of the stolon, and the deploying point 414
2. Position and relation of the zooids in the chain from the deploying point '
to the twist 414
a Shifting of the zooids 414
b Peduncles and foot-pieces 417
c Emergence of the chain to the outside world 417
d Twist in the chain 418
e Reduction of the foot-pieces, first break in the chain, and formation
of the first wheel 418
3. Comparison of the rate of growth of the chain as a whole with the rate of
other animals 419
Discussion of the observations from the causal standpoint 420
1. Cause of the twist * 420
2. Unequal growth of zooids and foot-pieces as a factor in the breaking up of
the chain 422
3. Impossibility that the character of the blood supply to the zooids can be
the cause of the size scheme within the wheels 427
4. Unlikelihood that the wheel arrangement of the zooids in Cyclosalpa has,
as believed by Brooks, anything to do with the position of the first four blastozooids of Pyrosoma 429
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
395
396 W. E. RITTER AND M. E. JOHNSON
The larger significance of such studies 431
1. Supplementing biological with quantitative observations 431
2. Natural periodicity in organisms and exacter methods in biological re search 432
3. The inadequacy of treating periodicity, generally, as an aspect of fluc tuating variation 440
Bibliography 444
PURPOSES OF THE RESEARCH
1. Special
One of us (Johnson, '10) has shown that the individuals of the blocks into which the chains of blastozooids of Salpa fusiformisruncinata, S. cylindrica, and S. zonaria-cordiformis become differentiated, fall into size schemes, or systems.
The question naturally arises, how general is this phenomenon among salpae? The possibility that the wheel grouping in Cyciosalpa corresponds to the block grouping in Salpa proper, occurs to one rather readily in spite of the conspicuous differences between the two. If this conjecture be right, we should expect to find a size scheme of zooids in the wheels of Cyclosalpa similar to that in the blocks of Salpa. That such a scheme exists in the wheels even more pronouncedly than in the blocks, the sequel will show.
2. General
This much more evidence is consequently adduced favorable to the idea of correspondence between the wheels and the blocks. But what do we mean by correspondence? In a general sense the blocks and wheels undoubtedly correspond: both are groups of shnilar organisms similarly located with reference to the parent zooid. This much of correspondence is recognizable to cursory inspection. Does the discovery of a similar size scheme among the zooids in the groups in the two species advance our interpretation of these organisms much if at all ? Does it amount to anything more than a recognition of one more resemblance? According to the meaning that interpretation' and 'resemblance' have in most later biological writing, we must probably say no. We
CHAIN OF CYCLOSALPA AFFINIS 397
must hold that unless the new correspondence includes somewhere what we hold as one or more 'causal factors' not much has been accomplished. If, for example, we extend the inquiry to the question of the dependence of the size scheme of the zooids upon growth and other internal factors on the one hand, and upon environmental factors on the other; and if here, too, we find further correspondence, our belief in the essential identity as we might say, unless standing for extreme exactness of expression, would be reached.
We shall see that the size scheme of the zooids in the wheels is almost certainly foreshadowed before the wheels themselves are. If this be so, then the resemblance between the Cyclosalpa chain and the Salpa chain is considerably closer before than after the wheels appear in the former. But it is difficult, if not impossible, to attribute the block production in the Salpa chain entirely to other than inherent factors, of which growth seems to be the most immediate. So far, therefore, as we can rely upon our evidence for the marking off of the Cyclosalpa chain into groups before the wheels are formed we seem to have placed the notion of correspondence between the wheels and the blocks on firm ground.
Evidence still more convincing perhaps, that the wheels and blocks correspond in a strict biological sense, in the sense that both are expressions of periodicity in growth, is found in the fact that growth and development are observed to be periodic in so wide a range of living beings. The growth of plants for example, appears to be nearly if not quite always of that nature. Finally, belief in the correspondence would, so far as we can see, reach high water mark, should it be finally made very probable that not only growth and development but all strictly biological processes whatever, are periodic. We are undoubtedly a long way from this last conception. Certain it is, though, that we now have sufficient facts to make the hypothesis of periodicity as warrantable as its opposite, namely, that certain phenomena are continuous in the sense of not being automatically interruptive and group-wise. We may now proceed to the handling of our data.
398 W. E. RITTER AND M. E. JOHNSON
BRIEF DESCRIPTION OF THE SPECIES
The species Cyclosalpa affinis (Chamisso) was taken in abundance at La Jolla from May to November, 1909, during which time most of the observational portion of this research was made. The longest chains and the largest wheels were brought in during the earlier part of this period, while the later catches yielded many specimens of the solitary form with short chains, and many medium sized single wheels.
Although the salpae do not survive for more than a day or two in the ordinary aquarium, the material has been sufficiently abundant to admit of considerable work, with the living specimens.
The two generations of this species differ markedly, as do those of all members of the genus. The intestine of the solitary form (fig. 11) is straight, extending nearly the full length of the animal, the anal opening being just back of the ganglion. The intestine of the aggregate generation, on the contrary, projects from the ventral side of the creature as a large, almost circular loop, (fig. 12) the anus being only a little to the left of the oesophageal mouth.
The solitarj^ form has eight body muscle bands, according to our system of enumeration, while the aggregate generation has five on the dorsal and six on the ventral side. The hypophysis (%p.), endostyle (end.), and gill (gi.) present much the same appearance in both forms. The orifices are also similar except that the , solitary form possesses short, tail-like appendages one on each side of the atrial orifice. Our records show the maximum length of specimens of the solitary generation to be 15 cm. and of the aggregate 8 cm. In both generations the test is thin, soft, and highly transparent, without special thickenings. In the aggregate generation, projecting from the ventral side, is the broad, thin peduncle (ped.) by which the zooids are united to form the wheel, and within the pharyngeal cavity on the right side, two thirds of the way back, is the embryo. In the young sohtary individual, the eleoblast, near the heart, and the remnant of the placenta, about one-third of the way back from the oral orifice, are both opaque, nearly spherical bodies and are very prominent.
CflAIN OF CYCLOSALPA AFFINIS 399
The stolon originates just above the heart and extends straight forward along the median line. The zooids, as in other salpa chains, are first in single file, but at a certain point, the deploying point, shift to double file. The deploying point in this species occurs close to the anterior end of the heart, 3 or 4 mm. from the root of the stolon. At a point about two-thirds of the way between the heart and the branchial orifice, the chain bends downward and passes to the outside world through an opening in the test. Immediately outside the opening, the chain doubles back under the parent and turns over so that it appears to be greatly twisted at this point. Before the twist, the zooids are arranged in two nearly parallel rows along the common stolonic blood vessel. After the twist, they are arranged in wheels which are connected tangentially and contain six to sixteen zooids each. These wheels show a gradual increase in size toward the distal end (fig. 11).
MEASUREMENTS OF THE ZOOIDS, OF THE WHEELS AND OF A
PORTION OF THE CHAIN NOT YET TRANSFORMED INTO
WHEELS
Serial measurements were made of the zooids of a number of chains with and without wheels as well as of the zooids of separate wheels. Lengths only were taken.
The measurements of the wheels were made with dividers and the results are given in millimeters. Those of the unbroken chains were made with the micrometer eyepiece used in the Zeiss binocular microscope. A unit in the tables represents 0.1 nam. of actual length.
The zooids of Chain I were separated from the chain for measurement, but the others were measured while still on the chain. It appeared that the latter is the more accurate method, since separating the zooids, besides being a tedious process, is apt to distort and mutilate them. In these measurements, the posterior extremity was taken at the atrial orifice. The intestine was not included, as a slight difference in its inclination would make a difference in the apparent length.
In table 1 are given the lengths of the zooids of the unbroken portion of Chains I-VII. Table 2 gives the same data for the
TABLE
Length measurements of the zooids of wheels of chains VII and VIII Horizontal lines indicate end of wheels. Double lines indicate one wheel lost. Unit — 1 nam.
NO.
CHAIN VII
CHAIN VIII
CHAIN VIII— continued
CHAIN vin— continued
1
7.3
7.3
6.4
6.1
41
13.7
12.6
81
20.5
20.9
2
7.6
7.7 7.7
7.6
7.8
8.2
8.1
6.7
6.9
7.1
6.8
7.0
6.7
42
43
44
45
46
13.1
12.9
12.0
12.2
82
83
84
85
86
19.2
18.6
3
18.6
19.8
20.7
22.7
18.6
4
13.0
13.9
14.0
12.9
13.0
13.4
17.2
5
7.8
8.2
7.4
7.4
6.8
6.8
19.8
6
8.3
19.8
7
8.6
8.4
7.5
7.4
47
14.2
13.5
87
21.9
20.2
8
8.4
8.7
7.6
7.5
48
14.2
14.1
88
22.5
20.4
9
8.7
8.8
6.7
7.1
49
14.3
14.0
89
21.9 20.5
10
8.9
8.7
8.8
8.7
6.8
7.2
50
i 51
52
14.3
12.8
90
91
92
19.8
21.1
22.9
17.5
11
7.2
8.7
7.8
7.6
14.9
15.2
12.4
14.0
20.0
12
8.2
8.8
21.0
13
9.5
9.7
8.2
8.3
53
15.3
14.7
93
24.2
23.2
14
9.5
10.0
8.5
8.6
54
15.7
15.3
94
22.8
23.2
15
9.7
10.0
8.8
8.2
! 55
16.0
14.3
95
24.1
21.6
16
9.8
10.2
9.4
8.4
56
16.4
15.3
96
23.0
19.9
17
9.6
10.3
10.0
9.3
9.2
9.3
10.6
8.7
57
58
t 59
15.0
14.8
14.0
97
98
99
21.3
20.6
18
9.4
9.8
8.4
9.1
15.4
15.5
21.2
23.7
20.9
19
13.9
21.3
20
10.6
11.0
9.8
8.9
60
16.6
13.2
100
24.5
22.6
21
10.9
11.1
9.7
9.1
61
16.2
15.4
101
24.6
23.4
22
10.9
11.4
10.5
9.0
62
15.3
14.7
102
23.6
23.6
23
10.8
10.1
11.4
10.9
9.7
8.8
i 63
64
65
66
15.2
14.0
15.4
14.3
13.6
103
104
105
106
24.0
23.9
22.5
24
9.8
9.6
10.2
8.9
8.9
8.5
22.1
25
9.7
11.0
10.6
10.7
16.0
16.3
23.0
25.1
22.1
26
18.3
22.2
27
11.1
10.5
9.8
8.1
67
18.8
17.9
107
26.3
23.1
28
11.4
11.4
9.8
8.5
68
19 6
17.7
108
26.1
23.7
29
11.8
11.5
11.1
10.6
11.2
11.2
10.1
9.2
9.1
69
70
71
72
73
74
75
76
77
78
79
80
18.4
18.0
15.7
17.7
17.1
14.7
14.7
109
110
Ill
26.4
25.5
22.4
23.4
30
9.6
10.9
11.8
12.1
12.3
12.3
11.6
22.9
31
10.2
9.9
10.5
10.8
10.7
11.5
21.3
32
11.7
11.1
12.8
13.0
12.9
12.2
12.5
11.3
11.4
12.4
11.8
11.8
11.7
16.2
18.8
19.4
19.7
19.5
33
34
35
36
17.3
19.4
19.4
18.3
18.4
16.1
19.1
20.0
37
38
12.0
12.7
13.4
13.6
11.5
11.4
11.4
11.6
18.6
19.3
20.5
22.2
39
40
wheel portions of Chains VII and VIII. In all, the unbroken
portions of seven chains were measured. The number of zooids
used depended upon the miminum size measurable. In four cases,
90 zooids were taken but in the other three chains only 80, 70 and
52 zooids respectively were large enough to be measured accurately.
[.The measurements of all series are given but only two are plotted, the right-hand series of Chain II in fig. 1, and the right-hand
S 3 13 17 Bl ZS t9 J3 J? 41 'ti 4a S3 37 tl 6} 69 73 77 61 63 89
Fig. 1 Plot of length measurements of the zooids, peduncles, and foot-pieces
of chain II, right series. Vertical distances represent length. Horizontal distances represent position in the chain.
series of Chain VII in fig. 2. In the latter figure, 1-90 are the zooids of the unbroken part of the chain while 91-108 are wheel zooids, the divisions between wheels being indicated by vertical dotted fines. Table 3 gives the measurements of zooids of several short chains of wheels which furnish figures for comparing graphs of wheels of various sizes. In each series of results here given except length of Chain I, two measurements were taken and these
404
W. E. RITTER AND M. E. JOHNSON
Length measurements of zooids of various small groups of toheels. Unit — 1mm.
Group A
Group B
Group C
Group D
POUB WHEELS
FOUR WHEELS
POUR WHEELS
TWO WHEELS
H.
L.
R.
L.
R.
L.
R.
L.
1
10.9
11.7
12.1
12.1
12.6
13.2
14.7
15.3
15.6
16.3
14.8
10.7
10.8
11.1
11.4
13 2
19.0
18.4 19.1 19.0 20.5 17.9
18.5
18.6
19.1
19.. 6
20.7
19.9
16.7
18.0
19.1
19.7 20.1 21.6 21.1
20.8 27.9
21.5 29.4 21.2 1 29.2
21.9 i 29.8
22.6 ' 30.1 21.9 29.3
27.2
2
26.4
3
30.0
4
30.7
5
30.8
6
13.5
13.6
14.5
15.7
15.8
15.1
29.5
7
19.6
21.1
21.6
22.0
22.5
22.2
19.8
19.9
20.2
22.1
22.4
23.0
21.8
21.8
23.5
24.2
24.3
23.1
22.0
23.4
24.2
24.2
23.7
24.2
23.0
28.9
30.9
30.8
40
30.4
30.3
30.7
8
22.3
22.6
22.6
22.5
21.0
31.8
9
32.9
10
31.9
11
33.2
12
16.6
17.0
16.9
17.1
18.5
18.1
12.7
18.5
19.4
19.4
18.9
15.1
15.1
17.1
16.7
17.9
18.5
18.9
15.9
31.3
13 '. ...
22.3
23.3
22.3
24.6
24.1
23.0
19.9
22.3
24.0
25.5
23 5
14
15
16
24.0
24.1
25.0
25.2
25.7
25. S
17
18
25.0
26.5
27.6
28.4 27.8 27.8 23.8
19
15.7
18.8 18.9 18.6 19.8
22.8
24.6
24.9
25.8
25.7
24.9
22.3
24.0
26.1
25.4
25.7
25.7
24.5
20
24.0
26.6
27.3
28.5
28.4
28.0
25.9
21
22
23
24
25
26 ...
CHAIN OF CYCLOSALPA AFFINIS
405
TABLE 3— CONTINUED
Gboup E Group F
Group G
Group H
THREE
VTHEBLS
THREE WHEBI-S
K. L.
TWO WHEELS
one WHEEL
R.
L.
r.
L.
R.
L.
1
23.4
24.6
25.2
24.7
25.1
23.6
23.0 22.3 22.9
24.8 23.8 23.3
25.7 24.4 23.9
26.1 , 24.8 1 23.9 26.5 25.0 24.7
21.3
21.9 21.6 22.3 22.4
21.0
22.2
22.6
24.2
22.6
25.4
25.1
25.0
24.9
23.9
23 1
2
24.5
3
25.1
4
25.1
5
24 8
25.3 26.3
25.1
25.0
6
21.3
22.0
22.5
24.3
25.3
25.7
23.2
25.2
25.9
26.8
26.8
26.1
24.4
24.2
7
24.9
26.8
27.6
27.2
27.7 27.2
24.1
26.9
8
248
24 2
9
10
11
27.2 1 27.2 25.9
27.5 ' 28.7 25.1
27.7 28.6 ! 26.0
27.2 28.4 26.8
12
13
14
27.4
28.8
28.7
29.7
29.2
27.6
27.5
28.3
28.4
27.2
27.9
27.2
27.2
29.2
29.8
29.9
30.3
15
16
17
18
19
30.0
29.8
29.2
30.3
28.2
26.6
29.9
29.8
29.3
30.4
406
W. E. RITTER AND M. E. JOHNSON
were averaged for the final result. Where the first and second measurement differed by more than 10 per cent, a third measurement was taken and the three figures averaged.
Fig. 2 Plot of the length measurements of the zooids of chain VII, right side, including wheels.
TREATMENT OF THE QUANTITATIVE DATA
At first glance, one sees a resemblance between the curves for the wheels of Cyclosalpa affinis and the blocks of Salpa fusiformisruncinata. In both cases, the end zooids are smaller than those nearest them, the maximum values lying somewhere between, usually nearer the distal end.
Fig. 3 Mean curves for wheels of various sizes. Vertical distances represent length of zooids. Horizontal distances represent position in the wheel.
408 W. E. RITTER AND M. E. JOHNSON
Some variation in the graphs of wheels of different sizes was noted, and to make sure of its general trend, the data for all the wheels were considered. The wheels were first grouped according to size, Group A included wheels whose zooids averaged 5-10 mm. in length; Group B, 10-15 mm.; and so on. In Group A were ten wheels. Not only does the number of zooids in a wheel vary, but the number in one-half of a wheel is not always the same as in the other half. For this reason the ten wheels were regarded as twenty half wheels.
Among these twenty half wheels of Group A were three containing four zooids; one with five zooids; nine with six zooids; and seven with seven zooids. The corresponding values of the three fourzooid half wheels were averaged, the three first zooids together, the three second zooids, the three third, and the three last zooids. The result was a typical curve for a four- zooid half-wheel whose zooids have an average length of 6-10 mm. The five, six, and seven-zooid half-wheels were averaged in the same way. Similar computations were made for the other four groups and the results plotted. The graphs were smoothed and those for each size were averaged in order to get the typical curve for that size. These curves (fig. 3) show that the size differences between the zooids of a half-wheel greatly increase as the zooids grow and that the typical form already noted becomes increasingly evident.
Passing now to the unbroken portion of the chain, we find that the zooids increase in length very slowly at first and more rapidly later; also that though the curve is fairly smooth at first, it becomes quite irregular toward the end. Upon closer examination of fig. 2 and the graphs of other chains, we surmise that these irregularities are the forerunners of the groups making up the wheels; in other words that the periodicity shown so plainly in the wheel part of the chain extends back into the unbroken part. Were this found to be true, the fact could hardly be ignored in considering the problem of the break-up of the chain and the production of wheels.
In order to test the conjecture more critically we submitted the measurements to Mr. George F. McEwen, the mathematical expert of the Marine Biological Station of San Diego for examina
CHAIN OF CYCLOSALPA AFFINIS 409
tion. Out of this examination has come the graphs shown in figs. 4, 5, 6, 7, and 8.
A curve was computed to fit the graph (fig. 2), as nearly as possible. From the equation of this smooth curve we get a 'calculated value' for each zooid; that is the length of each zooid, if the series were as smooth as our calculated curve. We next subtract the observed length of each zooid from the calculated length, and get a series of values, some plus and some minus according as the irregular graph went below or above the smooth curve. When we plot these plus and minus values above and below a horizontal line we have the graph fig. 7. It shows that the values follow the curve fairly well at first and then vary more and more; in other words, that we have a periodic curve of increasing amplitude. ^
'Mr. McEwen gives the following summary of the method used: The sizes for each of the points corresponding to the numbers 45, 50, etc., to 90 were taken as the ordinates of a curve whose abscissae were 1, 2, etc., to 10. It was assumed that the above curve corresponded to an equation of the form
y = a+bxi+cxi
and the most probable values of the coefficients a, b, and c were computed according to the method of least squares. By substituting {2x — 8) for Xi in the above equation, the equation
y = a + b (2x — 8) + c (2x — 8)'
was obtained in. which, if y^ of the number of the point is substituted, will equal the computed value of the corresponding size. (This equation was used to calculate the corresponding values of y, which were used in connection with the observed values for computing the algebraic sum of the residuals and the probable error, for the purpose of determining if the equation was a proper expression for measured values of y.)
It was assumed that this equation, determined from the 10 points was very nearly the same as if it had been computed from the 45 actual points, and therefore represented the relation between the number and the average size of all the points. This assumption was verified in one case by including all the points and comparing with the result when only 10 points were used.
The observed values of y were subtracted from the corresponding computed values and these differences were plotted as ordinates against the numbers as abscissae, thus giving a representation of the deviation of the observed values from those given by the equation. These deviations are due to errors in the measurements, and to the fact that the assumed equation was not a true expression for the relation. As the error in measurement was =*= 0.1, it is evident that the deviations are due mainly to the latter fact.
The periodic character of these curves shows that the true law is a periodic fluctuation of increasing amplitude about a mean value increasing in a regular manner with the number of the point.
Fig. 4 Plot of differences for chain IV, right side.
Fig. 5 Plot of differences for chain VI, right and left sides.
Chain IV, whose plot of differences is shown in fig. 4, is one of the smaller chains and in it one would expect to find the grouping less evident than in the larger chains. However it can be plainly seen even here. The right and left sides of Chain VI are shown in fig. 5. With Chains IV and VI, the differences were figured only for the zooids 45-90. In figs 6 and 7, the two sides of Chain VII
Fig. 6 Plot of differences for chain VII, left side
are given entire, the curves and the differences being figured separately for the two parts of the chain, since it can be fitted better when but half is considered at one time. The complete series being given, one can more readily see how the amplitude of the waves increases toward the end.
It will be remembered that in computing the differences, the observed values were subtracted from the calculated values. Hence upward curves in fig. 2 appear as downward curves in fig. 7, To make the comparison with the wheel graphs easier the signs
were reversed for fig. 8 so that values greater than the corresponding ones in the fitted curve he above the x axis while smaller ones
lie below. Vertical dotted lines have also been drawn to indicate
a possible grouping of the zooids. Irregularities appear, it is true,
but since irregularities often appear in the wheels also it is to be
expected here. Moreover, with such small values the chances
for error are so great that one would expect considerable variation.
Fig. 7 Plot of differences for chain VII, right side
What we get then from these plots of differences is the probable fact that the unbroken paH of the chain really shows a periodicity or incipient grouping closely resembling that of the wheeled portion of the chain, the groups including four to eight zooids each, which is the number found in the completed wheels.
The plots of differences brought to light an aspect of the matter which had not been anticipated, namely, the existence of another toave with a longer period, shown in all the curves. The plots in
CHAIN OF CYCLOSALPA AFFINIS
413
figs. 4 and 5 have been smoothed by averagmg for every ninth point in order to make this curve more plain. The curves are not just the same for the two sides of Chain VI, the difference probably being due to a difference in the way in which the computed curve fits in the two cases.
With what, if any, other biological phenomena in the species this newly discovered periodicity is connected we do not know.
7zoo/ets. : 7zooids \6 zooids. \S zooidj. \S zooids. • S zooids\ AS zooids. \ 4zooids
■<6-S2. \53-J9. \60-eS. \66-70. '7/-76. .77-81. \]«-fftf. '^7-90
Fig. 8 Plot of differences for chain VII, right side Inverted and possible groups indicated
Of its existence however, there seems to be no doubt; and it is certainly interesting to recognize that we have here an instance, by no means uncommon in organic phenomena, of waves, so to speak, of one size riding upon those of another size. It is highly desirable to take these cases in hand with a view to finding their connection with other phenomena.
414 W. E. RITTER AND M. E. JOHNSON
ATTEMPT TO CONNECT THE FORMATION OF WHEELS WITH MORPHOLOGICAL, PHYSIOLOGICAL, AND MECHANICAL PHENOMENA PRESENTED BY THE ANIMALS
1 . Segmentation of stolon and the deploying point
To find other factors entering into the wheel production, a study of the structure of the chain was made. The portion of the chain in which the zooids are in single file is of the same general form as that of other species. The incipient zooids, marked off by the infolding ectoderm, have their aboral ends uppermost, and the dorsal side of each against the ventral side of its neighbor, the dorsal sides being towards the base of the stolon. The blood supply passes out through one-half of the large axial blood vessel and back through the other half. The segmentation of the stolon in some cases extends to the root of the stolon, in others not quite so far. A possible significance of this variation will be pointed out in another connection. Where segmentation extends to the root of the stolon, the more proximal segmentation lines are very irregular and this fact may be of considerable interest in a way we shall not stop to consider here. Judging by some chains, one would say that the segmentation begins at the sides of the stolon, but others lead us to suppose the beginning is above and below (along the genital rod and the neural tube) while in still others it seems to be equally advanced in all parts of the circumference. The latter condition is probably the usual one.
2. Position and relation of the zooids in the chain from the deploying point to the twist
a. Shifting of the zooids. What we call the deploying point, is the point, or region where the zooids, by moving alternately to the right and left shift from single to double file. While the zooid is moving out, it also moves upward and begins to turn, so that its dorsal side faces out instead of toward the base of the stolon. These changes .take place gradually. The oral ends shove out and begin to turn and before the turn is complete, the aboral ends
CHAIN OF CYCLOSALPA AFFINIS 415
shove out and turn. The sketch of the deploy mg point will make this clearer. Figs. 14, 15 and 16 are dorsal, lateral and ventral, views of the deploying point of one chain. These drawings were outlined with the aid of the camera and much care was taken to make them accurate. Calling the zooid whose oral end has just begun to shift, no. 1, the aboral ends of nos. 8 and 9 (numbering on one side only) are beginning to do the same. No. 25 (not shown in the figure) seems to have reached the final position with the rearrangement of the internal organs complete. We find now that the right sides of the right-hand zooids (considering those to be right-hand zooids that correspond to the right-hand side of the parent) and the left sides of the left-hand zooids are toward the base of the stolon. This statement applies to the chain before it emerges from the parent. The orientation of the older, extruded part of the chain is given later.
All of the observations on the early growth and differentiation of the chain agree with those made by Brooks for C. pinnata with the exception of the orientation. Brooks ('93 p. 79) says of the single file zooids:
At this stage each Salpa is bilaterally symmetrical, and its plane of symmetry is the same as that of the stolon, while its long axis is at right angles to that of the stolon, which becomes converted into a single row of Salpae, so placed that the dorsal surfaces of all of them are toward the base of the stolon, their ventral surfaces towards its tip, their right and left sides on its right and left respectively, their oral ends at its top or neural side, and their aboral ends at its bottom or genital side.
Again in his description of the double row he says :
The single row of Salpae becomes converted into a double row, which consists of a series of right-handed Salpae and a series of left-handed ones, placed with .... the left sides of those on the right and the right sides of those on left towards the base of the stolon.
-The loop-like structures seen at the oral extremity of the zooids in figs. 14 and 15 might easily be mistaken for the intestine. They are not this structure, but indicate very nearly where the oral orifice will appear.
416 W. E. HITTER AND M. E. JOHNSON
This, as will be seen by comparing it with our description, is the opposite of the condition found inC. aifinis, since Brooks places the oral ends of the zooids uppermost while we find the aboral ends up. This mistake was probably due to lack of sufficient material for the study. He says (p. 87):
In all my preserved specimens the tip of the stolon had been so much flattened by contact with the side of the bottle, in transportation, that I have not been able to study in detail the way in which this wheel-like arrangement is acquired, and the subject should receive the attention of those who are able to stud}^ living specimens.
It is a point upon which one could easily go astray if hampered by a lack of material.
As the changes in internal organization seem to correspond with those of C. pinnata, and as Brooks'description is so clear and complete, we need not go into the subject, but refer to his account (Brooks, '93 pp. 80-106.)
When the zooid has moved into its secondary position it lies upon the stolonic blood vessel rather than to the side or around it. With this change, two small vessels develop for each zooid, one leading to it from each half of the stolonic vessel (fig. 24, ihv.). The blood flows along one-half of the main vessel (say the upper half) out through the upper small vessels to each zooid and returns by the way of the lower set of small vessels to the lower half of the main vessel where it joins the inflowing current. These currents are reversed with the reversal of the blood current in the parent. The zooids now increase in size very rapidly, lengthening out more above the upper level of the vessel than below it, so that at the twist the oral ends extend but a little way below the vessel, while the aboral ends extend far above it. As a result,the aboral ends of the zooids of opposite rows come in closer contact than do the oral ends. Since the zooids of the two rows are arranged alternately, each zooid will lie against two of the opposite row. As growth continues and the zooids, through their increased size, move outward as well as upward, they are forced farther apart, but the connection is retained through peduncles which now develop.
CHAIN OF CYCLOSALPA AFFINIS 417
b. The peduncles and foot-pieces. These structures play so important a role in the production of the wheels that they must be described in some detail. Almost all the figures show the peduncle in one stage or another. Fig. 12, best gives its relation to the full grown zooid, showing that it is a thin flap or sheet extending out from the ventral median line. The diagram (fig. 17) shows that the peduncles of the series are parallel throughout the first part of the chain, and that each by means of its 'foot-piece' {fp.) is in contact with four others, its two neighbors in each row. These foot-pieces are also well shown in the right-hand portion of figs. 19 and 22. As the zooids grow and extend out farther from the blood vessel, the peduncles lengthen, and the foot-pieces grow longer as the zooids grow wider, at least until the region of the twist is reached.
Along with the great increase in the size of the zooids and the development of the peduncles comes a change in the circulatory system. The two individual blood vessels coalesce to form one vessel with two channels (fig. 24, ihv.). The blood current has the same course as before except that the incoming and outgoing currents of each zooid pass through one vessel. Observation of the blood currents in the living animals made the task of working out the circulation much easier and more certain than it would have been if confined to preserved specimens. The cross section (fig. 25) shows well the relation of the zooids to the blood vessel, the foot-pieces joining the zooids above the vessel, (fp.) and the individual vessels leading from the zooids to the two parts of the large vessel (ibv.)
c. The emergence of the chain to the outside world. Through the first part of its course, the chain is enclosed within a definite tube in the test just below the endostyle, this tube ending at a point just posterior to the placental vessel and anterior to the first body muscle band. This first muscle bends posteriorly here so that its insertion is along the lower part of the tube opening. The placenta usually disappears before the chain reaches this point. There seems to be more or less of a cavity left in the test where the placenta was, and the chain, as it reaches this point, no longer being held in its horizontal position by the tube, following the
418 W. E. RITTER AND M. E. JOHNSON
line of least resistance, turns down into the cavity, and by the rapid growth of the zooids, soon breaks through the thin wall to the outside, the tip bending downward.
d. The twist in the chain. The general character of this part of the chain naay be seen from figs. 11 and 13, while the peduncles and blood vessels of the region are shown in the diagrams figs. 17 and 1,8. Before the twist, we have within the parent, a straight double row of zooids with oral ends down. After it, the chain is turned back under the parent, and the zooids are again found with oral ends down. Until the zooids break through the test to the outside the chain has not begun to twist, the zooids still lying symmetrically along both sides of the median line. In fig. 13 thirty-six zooids are outside and the twist is just complete. The presence of two rows of zooids in the chain makes the turn appear more complicated than it really is. The chain simply doubles back under and then turns over, this turn being almost invariably to the left. This leaves the zooids with aboral ends again uppermost, but the row that was before on the left is now on the right side of the parent.
e. Reduction of foot-pieces, first break in the chai7i,and formation of the first wheel. The first visible intimation of the break-up of the chain comes in the peduncles and foot-pieces. The footpieces (fig. 17) gradually grow longer toward the distal end of the chain, coming to their maximum length a little before the end of the unbroken part is reached. After the maximum they shrink (fig. 22). The decrease is much more rapid than the increase, there being only about sixteen to twenty-four zooids in the diminishing series. Fig. 19 shows that the first group consists of nine zooids whose peduncles have broken loose from the rest. The foot-pieces have shrunk still more and the distal ends of the peduncles have been drawn closer together. But while the foot-pieces, by which the zooids are held in the axial line of the stolon, become successively and rapidly smaller just before the beginning of the break in the chain, the zooids themselves are becoming constantly larger. A consequent crowding of the zooids results. This brings about a pushing of the bodies of the zooids forward in the chain beyond the foot-pieces. The strain to which
CHAIN OF CYCLOSALPA AFFINIS 419
the series of adhering foot-pieces is thus subjected results inevitably in a pulling apart of the foot-pieces somewhere. As a matter of fact the break produces groups and not single pairs. These groups then promptly shape themselves into the wheels.
3. Comparison of the rate of growth of the chain as a whole with the rate of growth of other animals. As a matter somewhat to one side of the main problem, we have thought it worth while to compare the rate of growth of the chain as a whole with what is known of the growth of other organisms. This was done by the method employed by Minot ('91) in his study- of the rate of growth of guinea pigs; namely, by finding the per cent of increase throughout the chain. The values of the corresponding zooids of Chains I, II, III, VI, and VII were averaged. (Chains IV and V being so much shorter were omitted.) The per cent of gain of the secoiwl over the first, third over the second, etc., was then computed and the values plotted. The result is a very ragged line showing a gradual increase through two-thirds of its length and a more sudden drop at the end. To get a graph whose course was more evident, the increment was computed again, this time taking the series in groups of five. The first value here is the per cent of increment of the second five over the first five, etc. (table 4, fig. 9). The gradual increase, maximum toward the end, and rapid decline is here plainly shown in spite of the limited data.
This result seems strikingly different from that for the guinea pig and other animals of higher order, where the per cent of increment is a diminishing one from birth on. The difference may, however, be more apparent than real since, to make the comparison more correct, it would seem that stages in the mammalian development preceding birth would have to be used.
The drop in rate of increase, when the wheel part of the chain is reached, may be significant for the comparison, but we do not consider our observations carried far enough into the life of the chain as a whole to warrant any speculation based upon them. A study of the growth of still younger and still older, larger zooids will have to be made to meet the requirements here.
Fig. 9 Percentage of increment throughout chain
DISCUSSION OF THE OBSERVATIONS FROM THE CAUSAL STANDPOINT
1. Cause of the twist
Although salpae do not move through the water very rapidly, still there is enough motion to make the end of the chain double back imder the parent, as soon as it projects into the water. The reason for its turning over is less evident. The zooids begin to pulsate some time before the twist is reached and, having been with aboral extremities uppermost in the original or normal position, we may suppose they tend to assume the same position again when the normal state of things is interfered with by the bending back of the chain. Observation of the living animals shows that the chains of wheels and the separate wheels (at least the smaller ones) usually move along with aboral extremities uppermost.
It is easier to say that zooids ' tend to assume the normal position' than to show the cause of this tendency. It may be that the specific gravity of the oral ends is greater, or that the pulsation may have something to do with it, or there may be some tropism
CHAIN OF CYCLOSALPA AFFINIS
421
TABLE 4 Per cent of increment throughout the chains
SERIAL NOS. OF THE ZOOIDS
AVEBAGE SIZE
FEB CENT -h 2
1- 6
6-10 11-15 16-20 21-25 26-30 31-35 36-40 41^5 46-50 51-55 56-60 61-65 66-70. 71-75 76-80 81-85.
9.2 9.9 10.3 9.3 11.7 13.3 13.4 12.6 15.1 14.9 17.9 16.7 17.4 14.2 15.1 16.7 14.5
Wheel portion
422 W. E. RITTER AND M. E. JOHNSON
involved; but we have no observations under this head. As the chain Hes, with aboral ends of the zooids uppermost, the propulsion of the zooids drives them away from the ventral side of the parent, while if they were with oral ends up the pulsations would drive the oral ends up against the ventral side of the parent. In one specimen in which the, chain had not yet emerged, the end of the chain was turning to get around the placental blood vessel. Had our observations been limited to this one instance, we might conclude that the twist is initiated in this way. But after looking over a large amount of material and finding that the placenta usually disappears before the chain reaches that point, it is evident that this in no way accounts for the twist.
2. Unequal growth of zooids arid foot-pieces as a factor in the breaking up of the chain
We seem to have found a cause sufficient for the present research, for the break, in some way, of the chain of zooids. This is, as already pointed out, the unequal growth of the bodies and footpieces of the zooids. The first question that arises when we attempt to push the analysis farther is, why is the break into groups rather than into single pairs of zooids? Nothing in the differential growth recognized appears to bear upon this question. So far as that is concerned we should suppose the zooids would be picked off one by one, or at most in single pairs.
Just how constant these groups are, may be seen from the frequency polygon (fig. 10). We see that of ninety-two half wheels seventy-three contained six or seven zooids each, while only two contained eight, and five contained four zooids. This constancy is to be expected when we regard the breaking apart as a growth phenomenon depending upon constant causes rather than upon chance.
In some waj^ the wheel phenomenon is clearly dependent to a large extent on the strength of the adherence among the footpieces, which are but parts of the central ends of the peduncles. As may be seen by fig. 18, the radial blood vessels, the other main connection of the zooids, break apart early in the life of the
CHAIN OF CYCLOSALPA AFFINIS
423
p
P
~^
20
i
36
/
\
18
31
/
\
16
28
1
\
14
i
s
24
j
\
IZ
1
\
ZO
1
\
10
/
16
8
/
,e
1
\
6
/
?
8
/
\
4
/
4
^
/
\
B
/
V
/
/
\
/
\
3
4
A
7
B
S
7
8
S
/e
1/
a
/J
14
15
16
No. of zoo/e/a... Ma/flV/iee/. Whole hf heel.
Fig. 10 Frequency polygon showing the number of zooids in the wheels
wheel, so that in the fully grown wheels the zooids are held together almost entirely by the peduncles. The observed facts certainly suggest group adherence among the foot-pieces themselves. Can direct evidence of any such thing be obtained?
Having proved the existence of a pronounced size grouping of the zooids in the wheels it naturally occurred to us that there may be something of the same sort in the peduncles and foot-pieces. We consequently made a considerable number of measurements on these structures. Some of the numbers are given in table 5, and in fig. 1; the graphs of peduncle lengths (dotted hne) and footpiece lengths (lower continuous line) are presented. It is doubtful if these show anything. We have not assumed that they do. The difficulties in the way of making the measurements are too great for the methods employed. It should, however, be borne distinctly in mind that these negative results prove no more than the insufficiency of the measurements. The fact that the footpieces do cling to one another in groups, and that the zooids to to which they belong are demonstrably different in size, appears to make it probable, a priori, that the adhesive power of the footpieces is of the gradational, or periodic sort, in spite of our failure to find it. The suggestion is that the graded size of the zooids is reflected in the adhesive power of the foot-pieces. Could this
424
W. E. RITTER AND M. E. JOHNSON
OiCOiOOiCOiOOOOiOCCiOiCOCOOOiCOiOifDOOOOO
si U
So
ooocoooooo<©ocoooeoio»ooooioioiooiocooo
0»000000>OOeOOOOOOiCOOOOOO»C"20fOCOOO
oooooiciooiocxicooo>oo»co>oiooococciioioioofcoo
OOiCiOOOOO»OiC»^inOiOOLOOO>OOOOOOiOiOO
I0»00i00000i0000000i00 0000»0000
is
lOOiOOOOiOOiOiOiOOOOOOiOiOiOiOOiOiOiOOiOOO
i <N CC .-I 1 '
'000»C»OiOOO»OiOi.OOiOOiOiOOOlMOiOOOOiOOO
CHAIN OF CYCLOSALPA AFFINIS 425
00 "3 00 .-i^oJ
cooo 00 t~-t>
00 o
OOiOiOOiOiOOOOiO^OOLOiOiOiOiOiOOOOOOOOOiOOOOOOOiO
0»C»Ci»00»OiOOOiO'OOOOOOiOiOOiOOOO»OOOtOOOOOOiOOiOO
<x)oot^!£!t^Mc<ic<^(^^l-Hco■*■^co■<4^co(^^coeoc<^Meoc<^<^^'-^c^^.—^,— (.-HC],— looocio
^•-ooLOiooio»ooicu^>o-*T}<Tt<-*-*TtiTti'^TtiTficocoirc>(roforoeocococc(
426 W. E. RITTER AND M. E. JOHNSON
conjecture be proved true, an exceedingly important biological point would have been made.
And now as to the evidence that a periodicity corresponding to the future wheels does exist in the chain before its break-up. In discussing the results of our treatment of the data pertaining to the unbroken part of the chain, we said the curves, as shown in fig. 2, for example, 'probably' show a periodicity. We permitted ourselves to doubt to this extent, in the interest of conservatism. We wish now to sum up the evidence for periodicity. Its strength lies in the fact that it is cumulative rather than in the sufficiency of any one piece.
In the first place, does not the undoubted fact of periodicity in the wheels themselves, and the groups that immediately precede them, make the presence of periodicity in the rest of the unbroken part of the chain probable a priori? It would seem so. In the second place mathematical treatment of the quantitative data makes it almost certain that a periodicity corresponding to theory actually does exist. Third and finally the probable extension of the periods far back into the young part of the chain, leads us to suspect that this fact is connected with another observation of quite a different order, an observation, that is, which strongly indicates that the periodicity is really established at least as early as the segmentation of the stolon itself.
One of us has shown that in Salpa fusiformis-runcinata the very early segmented part of the stolon may be interrupted by an unsegmented part (Johnson, '10, p. 154 and fig. 8). While such interruptions have not been observed in Cyclosalpa affinis attention was called, when speaking of the first stages in the segmentation of the stolon, to the fact that in some cases the segmentation reaches to the very root of the stolon, while in others a stretch of unsegmented stolon exists. May not this difference indicate a periodicity in the segmentation corresponding to the periodicity in growth that we have found?
The reader may think that the grouping, as shown in the plots of differences, is too variable and indefinite to warrant the conclusions we have drawn. True, the groups here are not as regular as the wheel graphs shown at the end of the curve (fig. 2) , but though
CHAIN OF CYCLOSALPA AFFINIS 427
the small groups appear to be more irregular on account of their riding on the secondary waves, they are of the same sort. It must be remembered, too, that the values are very small and the chances of error are large. In fact, such a uniformity of result throughout all the graphs examined, in spite of small values and difficulty of measurement, is very convincing.
The transformation of the groups of zooids into wheels is easily understood : The moment the break occurs so that the pressure of the zooids upon one another in the group can exert its effect backward as well as forward, the hindmost pair swings in toward the axial line, each of the other pairs up to the transverse middle line of the group following in its proportional amount. Since by this time the foot-pieces have wholly or almost wholly disappeared and the central ends of the peduncles have become closely appressed, the swing of the zooids disposes the peduncles in the form of the spokes of a wheel, the hub being represented by a small elliptical space. The course of things here described is illustrated in fig. 17. That the pressure tending to force the mid-zooids of the groups outward is considerable is obvious from the zig-zag form into which the axial vessel is thrown, due to the pull on the radial vessels, as seen in the second group of fig. 18. The disappearance of the axial vessel in the older wheels may be supposed to be partly due to the same cause, although probably the vessel is actually in course of degeneration.
3. Impossibility that the character of the blood supply to the zooids can be the cause of the size schemes within the wheels
No study involving the growth of the zooids could be complete without attention having been given to so fundamental a matter as that of the blood supply. For example, the question naturally arises, does not the break-up of the chain into groups so affect the common blood vessel of the stolon that the zooids do not share alike in nutriment received, and is not this inequality responsible for the disparity in size among the zooids?
The changes in the circulatory system are best shown by the diagram fig. 18. At the end of the continuous part of the chain,
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
428 W. E. RITTER AND M. E. JOHNSON
the individual blood vessels are arranged at regular intervals along the large vessel. The arrangement is the same for the first wheel, but with the second or third wheel the axial vessel begins to shrink. As the vessel remains in connection with the individual lateral vessels, while growing smaller, it comes to have a zig-zag course, due to the opposite but alternate pulls upon it by the growing zooids. The shrinkage of the vessels goes on so rapidlj^ and to such an extent that in the next wheel the vascular connection between the central zooids is lost. The portion of the main vessel which joins two wheels together persists for some time. In fact this and the transparent cellulose envelope which forms around the wheels, filling in the spaces between the zooids, are all that hold the chains together and very long chains of wheels are sometimes found. The small remnants of the individual vessels gradually disappear. Though these vessels end blindly, the blood may still be seen in them for some time, flowing out one side and back the other. After the disappearance of the main vessel at the center of the wheels, short circuits are maintained between the zooids connected at any point. Thus in fig. 18, zooids 1, 2, 12, 6^, and 7^ have a circuit of their own. Thus it would seem that if any of the zooids of the wheels have an advantage over others the end ones would be favored as against the middle ones, but the middle ones are on the whole larger. Hence inequality in blood, supply seems to be excluded from being a determining factor in the size relations observed.
If there be any communication between the zooids of the unbroken chain or of the wheels, other than by the circulatory system just described, it must be through the peduncles. The vessels in the peduncles are irregularly arranged but they are distinctly larger toward the edges and reach part way into some of of the papillae. They are easily followed in the living specimens. To test the question of blood communication between zooids, injections were made. Methylene blue in sea water was used, which could plainly be seen in the transparent peduncles and in the bodies of the salpae. The first attempt was on two wheels whose stage of development was the same as the third and fourth in fig. 18. The needle was inserted in the stolonic vessel half way
CHAIN OF CYCLOSALPA AFFINIS 429
between the two wheels. The color shot out through the small vessels to the peduncles of the zooids still remaining in contact. It went throughout the vessels of the peduncles of the zooids but stopped cleanly at the edge of the peduncle. No zooids whose connection with the main vessel had been lost showed any touch of the color. However, the wheel was again examined about fifteen minutes later. The stained zooids had died and dropped away from the wheel, the peduncle dropping away with the zooid. A slight stain was found around the papillae of the peduncles of one or two of the other zooids where they had come in contact with the stained ones. We conclude that there is no direct vascular connection here but that there is possibly some interchange by absorption through the thin ectoderm. Another injection was made in the peduncle of one of the zooids in a wheel. The color flowed throughout the peduncle and into the zooid but did not enter other zooids of the wheel. We therefore seem driven to conclude that the hlood supply is not a factor in the size differentiation of the zooids of a wheel.
4. Unlikelihood that the wheel arrangement of the zooids in Cyclosalpa has, as believed by Brooks, anything to do with the position of the four first blastozooids of Pyrosoma
Brooks was firmly convinced that the radial, or wheel arrangement of the asexually produced zooids in Cyclosalpa is homologous with the radial disposition of the first four blastozooids of Pyrosoma. This he regarded as one of the strongest evidences of the close relationship between the two genera. Thus he says (Brooks '93, p. 133) :
The opinion that Salpa and Pyrosoma are closely related does not however, rest upon superficial resemblances, but upon their fundamental identity of structure, although one of the details, the resemblance in their asexual multiplication, is so complete as to be almost enough in itself to establish their affinity.
The same view he expresses with only a little less assurance in several other connections. We had no thought, in entering upon
430 W. E. RITTER AND M. E. JOHNSON
the present study, of considering this point, nor do we propose now to go into it extensively. However, our results on the growth and mechanical factors involved in producing the wheels of Cyclosalpa seem to have so much bearing on the question, that we can hardly pass it by without notice. The resemblance between such a figure of the Cyclosalpa wheel as, for example, that given by Brooks ('93, pi. 1, fig. 2), and reproduced by Delage and H^rouard (p. 203, fig. 151) and a figure of an early Pyrosoma colony like 15, (pi. 31), by Huxley ('59) is considerable and not unnaturally suggests true heredity kinship. The moment, however, one comes to look into the details of how each group comes about ontogenetically rather than phylogenetically, he finds them so different that his imagination is balked at an attempt to interpret them as both referable to a common hereditary operation. In the first place Brooks seems never to have observed the fact that the Cyclosalpa wheel is at the outset bilateral. None of his published figures give any intimation of this, nor does he refer to it in his text. For instance, the two figures, 8 and 9, pi. 2, of his latest publication (Brooks, '08) represent wheels of C. fioridana, and C. pinnata as though they were perfect — as though the zooids were disposed in exactly the same way throughout the circuit. We would not, of course, assert that he did not draw just what he saw in these two instances, especially since we have had no chance to examine the wheels of C. fioridana, and have seen but a single one of C. pinnata. In the one specimen of C. pinnata which we have, attentive study finds that two zooids on opposite sides of the circuit have slightly different positions from the others. These probably indicate where the axis of the chain lay; but the departure from perfect regularity is so slight and of such a character that it might be easily overlooked had one not discovered, by studying the formation of the wheels, what their real nature is. In C. afiinis the bilaterality of the wheels is probably never wholly obliterated.
The first four ascidiozooids in Pyrosoma, on the contrary, stand in single file as do the Salpa zooids before the deploying point is reached and the radial order is taken on by the swinging around of the file so that number four comes to be adjacent to
CHAIN OF CYCLOSALPA AFFINIS 431
number one. Further there is no opportunity in the Pyrosoma g^roup for the differential mechanical action caused in Cyclosalpa by the growth and crowding of the zooids while the foot-pieces diminish in size. Neither is it possible seemingly, for the periodic phenomenon to play any such part in the arrangement of the Pyrosoma zooids as it appears to in Cyclosalpa.
THE LARGER SIGNIFICANCE OF SUCH STUDIES
1. Supple7nenting biological with quantitative observations
We venture to call attention to the way in which morphological and physiological observations and considerations join hands with quantitative observations in this research. Numerous structural details in the adult individuals of both sexual and asexual generations, in the chain of zooids as a whole, in the individual wheels and the individual zooids composing the wheels, and in the unbroken part of the chain both as a whole and as to its individual elements had to be attended to. On the functional side not only growth in several of its aspects, but the mode of swimming, certain facts pertaining to the circulation of the blood, and some points about nutrition have come in for consideration.
All this sort of thing is so familiar to modern biologists as to need no special mention. Not so with what we have done in a quantitative way. It seems to us that in this we have entered a region of research that biologists will be compelled to regard vastly more seriously in the future than they have in the past or do now. The case in hand furnishes a rather striking illustration of what the quantitative method can do. It can enable us to see facts we cannot see otherwise. It amounts to a great increase in the power of our eyes just as does the microscope. This statement is to be taken literally, not figuratively. One may easily imagine a magnifying instrument that would so enlarge the wheels as to make visible the size differences between the zooids. It would seem that this is what the application of mathematics in physical science very frequently does. We should never have suspected from ordinary examination size differences
432 W. E. RITTER AND M. E. JOHNSON
of a systematic character among the zooids of the chains. It was only from certain biological considerations combined with aid from this instrument, that the existence of the system was made certain. And it should be specially noted how our results would have been affected by failure to recognize this fact. The breaking up of the chain in some way, and the production of wheels from the breaking, could have been inferred from the unequal growth of the bodies and foot-pieces of the zooids; but why the breaking should be into groups rather than into single pairs would have remained with no definite answer but for the discovery of the periodicity in growth in the unbroken as well as in the broken part of the chain.
2. Natural periodicity in organisms and exacter methods of research
But promptly comes the question from some of the foremost biologists, What of it? What particular good is there in knowing that growth is periodic so long as we have no explanation of why it is so? Our real interest, they say, is in the causes not the mere facts of organic phenomena. This objection displays, in our opinion, one of the most pervasive and fundamental weaknesses in the biological philosophy of the day. Looked at critically, it is found to mean that facts of nature, in order to be interesting and deemed really worth while, must be prejudged; that an explanation of them must be ready at hand before they are observed in order that they may be attractive. The issue must be looked squarely in the face. It is in fact the old, old issue between the inductive and the deductive methods of interpreting nature; between observation and reason going hand in hand, and the power of reason alone; between the a posteriori and a priori modes of reasoning. The objection carries the implication that great numbers of facts of nature can be explained without having been themselves examined; that the unobserved causes of many observable effects may be sufficiently inferred from observations on other effects than the particular ones under consideration. In a word the meaning is implied if not expressed, that some time nature may
CHAIN OF CYCLOSALPA AFFINIS 433
be fully known without having been fully studied. This conception of nature and the knowledge of nature is always and everywhere the begetter of dogmatic assertion on the part of leaders, of subserviency to authority on the part of followers, and of idolatry to certain facts and neglect of others by everybody. This is not the place to go into the logic, or rather, the epistemology, of biology. The case under treatment does, however, justify us in a few observations and reflections on procedure in research.
Why is it that the biological sciences are designated as observational and descriptive, to distinguish them from the physical sciences which are called quantitative and exact? Surely no present-day student of nature would contend that living objects are qualitative alone and so must be dealt with in terms of quality, while non-living objects are quantitative and are to be dealt with in terms of quantity ! There is surely no structural part or activity of any organism that does not exist in some quantity or other, and hence is not susceptible of being measured in some way. Contrarywise, there is surely no inorganic body or substance that has not qualities of some sort by which it is described and defined. Yet why is it that in spite of the brave effort made by a few distinguished men of science during the last half century to introduce conceptions of quantity and the methods of mathematics into biology, these efforts have met with only limited success at best, and are ignored in practice and frowned upon in theory by many of the foremost bilogists? Only a few months ago a distinguished investigator declared in the presence of the senior author of this paper that the quantitative method in biology is dead, and this student suiting practice to theory, though working in fields where quantitative conceptions and exact determinations are particularly important, rarely attempts to measure in any rigorous way the biological phenomena with which he deals. Attention cannot be called too strongly to the extent to which much of what is esteemed the very highest type of recent biological work has laid stress on accurate quantitative determination of certain environmental factors of organisms, but has ignored almost wholly quantitative determinations of the vital phenomena themselves. There can be no question about the
434 W. E. R/TTER AND M. E. JOHNSON
importance of exactness in the determination of external factors. So far these methods are admirable; but, it appears to us, it must be recognized that when exactness has gone thus far it has gone at best not more than half the way. Nothing less than equal exactness all along the line will do to fulfil the highest demands of physical science.
Let one recall the degree of refinement with which physicists and chemists are measuring the phenomena with which they deal: the wave lengths and angles of refraction of light; the quantity of heat generated in chemical reactions; diffusion rates of gases and liquids; atomic weights and combining ratios, and innumerable other things. Then let him compare these with the ridiculously crude quantitative determinations made in nearly all departments of biology. A few aspects of physiology, as for instance, the temperature of the human body; and a number of phases of the psychology of higher animals — reaction times, for example — have been brought under mensurational treatment comparable with the standards of exactness long demanded in physics. But the vast fields of morphology, of general physiology, of individual and race growth and decline, of propagation, of variation, of automatic and responsive action, etc., have hardly been touched quantitatively as physics and chemistry would understand this term. As yet we in biology have hardly heard of anything corresponding to physical constants, units of measurement, coefficients of change, etc. Yet will any one, fully alive to the spirit of modern physical science, venture to maintain that inorganic phenomena are so utterly different from organic, that conceptions and practices so enormously fruitful in the one realm are wholly inapplicable in the other?
It is a significant fact that many biologists, the most ardent in defence of the so-called mechanistic or materialistic view of living things, are farthest away from, even most hostile to, the very methods for biology proper that have so largely made the physical sciences what they are. One looks in vain through numbers of technical writings by biologists of this school for anything like exact, comprehensive accounts, either qualitative or quantitative of organsims or parts of organisms, or even functions
CHAIN OP^ CYCLOSALPA AFFINIS 435
of organisms, dealt with. Yet how these writings bristle with such expressions as 'differs considerably/ 'constant results.' 'as a rule/ 'very similar/ 'normal segmentation/ 'normal nuclear spindle/ 'normal blastulae/ 'normal animal/ 'practically identical/ 'essential features,' 'increases in exact proportion,' and so on!
Two rejoinders are frequently made to this demand for carrying more exact methods into biology. One is on the purely theoretical ground that it is not necessary; that 'mere quantity' is of no great moment in life phenomena; that slight differences are of the purely 'fluctuating' or individual sort, so have no large significance. To answer this objection in full would take up much farther into philosophical discussion than we can go here, but it may be the more warrantably passed by because the attitude of mind that makes it is seen to be obviously hostile to the whole trend and spirit of physical science. If the history of progress in science can be relied upon to furnish any clue as to how progress is to be continued in the future, the man of science, who holds a general view of nature that makes many facts insignificant and negligible, is bound to come to grief sooner or later.
The other objection is more practically justifiable. It is that the phenomena of living beings are so complex and subtle, and that animals, especially, are so sensitive to changes in external conditions as to make it impossible to apply to them in more than a very limited way, the exacter methods of the physical laboratory. Our answer to this is two-fold. In the first place, we are persuaded that exact methods could be applied far more widely than they are, and they undoubtedly would be, did our general conceptions call for such applications. The other answer is that if it be true, as it well may be, that many life processes are too subtle and involved to submit to measurement on an exact and large scale, then the only course open for the interpretation of such processes is to iritroduce no considerations that involve the conception of accurately measured quantity. The extent to which this principle, seemingly so obvious and unescapable, has been violated in much biological theory during the last quarter century or more, is seen to be remarkable once
436 W. E. RITTER AND M. E. JOHNSON
one comes to think about the matter. For example, reflect on the extent to which theories of development and heredity have made use of the notion of equation and reduction nuclear divisions of the germ cells ; yet who has determined in any rigid quantitative way the elements that enter into the hypothetical equalities and inequalities? How familiar is the textbook statement that the chromatin of the male fertilization nucleus is 'exactly equal' to that of the female nucleus with which it fuses! But on what sort of determinations does this assertion rest? On scarcely another thread of evidence than that they ' look equal !' And here we come upon the almost incredible naivete with which biologists in most things eminently sound, have gone down before this fallacy! Only a short time ago while discussing this point with a number of biologists, one of them, a man of excellent standing and great carefulness in nearly all scientific matters, replied to my strictures, "if chromosomes look equal why are they not equal?" The words were hardly off the man's tongue when he saw what a remarkable statement he had made. The incident illustrates the straits to which one may blindly go in following a theory.
We conclude this topic with a quotation from John Tyndall. In his well-known address on the "Scientific Use of the Imagination," he says:
Let me say here that many of our physiological observers appear to form a very inadequate estimate of the distance which separates the microscopic from the molecular limit, and that, as a consequence, they sometimes employ a phraseology calculated to mislead. When, for example, the contents of a cell are described as perfectly homogeneous or as absolutely structureless, because the microscope fails to discover any structure ; or when two structures are pronounced to be without difference, because the microscope can discover none, then, I think the microscope begins to plaj' a mischievous part.
In view of the vast amount of evidence now before us from so many aspects of biolog;^^, that vital processes are periodic in their most fundamental manifestations, it appears unwarrantable to assume without proof that any whatever are not so.- But see what periodicity means ; it means that the phenomena are increas
CHAIN OF CYCLOSALPA AFFINIS 437
ing and decreasing ; that they have phases ; that the time element being considered, they change in value from moment to moment. How then can we treat any particular phase, or stage of such phenomena so as to meet the demands of rigorous science without considering each phase in relation to the other phases? So far as they are treated without such reference the procedure would seem to be of the nature of 'random observations' — of the 'grabsample' kind — that always, whether in common life, business, or science finally proves to be inadequate if not disastrous. Astronomy, physics, chemistry, and in general geology, have passed quite out of this portion of their careers.
Taking it as established that biology is allied in essential nature with these older, less complex sciences, does it not seem inevitable that it too must move on and leave its cruder, haphazard methods behind? Does it not look as though this very fact of periodicity, this gradual come-and-go of things in the operations of organisms is to be one of the chief if not the chief way out? To press the inquiry a little closer, does it not look as though the wide prevalence of repetitive parts in reproduction and growth, which though like one another still differ from one another by some regular quantity, is to be one of the most important, though only one, of these exits?
It appears to us that cell division, for example, including the division of all cell parts subject to this process will have to be looked at sooner or later from this standpoint. Take the Foraminifera, for instance, unicellular organisms (according to the current interpretation) the bodies of great numbers of which become divided into many sections called nodes and chambers. In the great majority of species, as a glance at figures enables one to see, these divisions fall into quantitatively differentiated series. To make the point more cogent we introduce figures of two species Reophax membranaceus Brady (fig. 21) and Peneroplis arietinus, Batsch. sp. (fig. 20). Now let one compare these organisms with the salpa chain, the one, for example, represented in fig. 18, and catechise himself something like this : surely there is some resemblance between these objects. Both are composed of a considerable number of sections rather regular in form and much like one an
438 W, E, RITTER AND M. E. JOHNSON
other, though obviously differing from one another in size. Both objects are hving, and both have come to be what we see them by a process of organic growth. Can we properly ignore these similarities in our efforts to interpret the organism, because on the whole the differences between them are more numerous and conspicuous than are the resemblances? Is it not at least possible that by turning to these few correspondences seriously they may serve as the starting point for the discovery of still others, and finally result in the detection of laws of organic growth and functioning that would greatly broaden our conceptions of, and hold upon, life phenomena?
One reason for selecting the Foraminif era as a group with v.'hich to make the comparison is the fact that the comparison of these organsims with higher ones in somewhat the same way has been made by several other zoologists. For instance, Schaudin ('95) speaks of the production and breaking off of parts in Calcituba polymorpha Roboz. as having "eine gewisse Ahnlichkeit mit der Strobilation."
But the most interesting comparison from our standpoint, of Foraminifera with other organisms was made by L. F. de Pourtales in 1850. At the meeting that year of the American Association for the Advancement of Science Professor L. Agassiz presented a short communication from this young zoologist in which Agassiz said:
Mr. Pourtales has, for the first time, pointed out a direct, well sustained analogy, which is to be found in the order of succession of the cells in foraminiferae of the genera Textularia, Candima, Biloculina, Triloculina, and Quinqueloculina. This succession agrees fully with the succession of leaves in plants — so fully that it can be expressed by the same fractions with which botanists are now in the habit of expressing phyllotaxis in the vegetable kingdom. This is, therefore, an important additional hnk in the investigation of the plan which regulates the normal position of parts in organized beings — a link which may lead to include into one universal formula the rhythmic movements which preside over the development of all finite l)eings. (Pourtales, '50, p. 89.)
This communication appealed strongly to at least one of those who heard it. At the next meeting of the association the presi
CHAIN OF CYCLOSALPA AFFINIS 439
dent, Professor A. D. Baehe, said in what we should now call his presidential address :
The germ of two most important discoveries in natural history was contained in papers by two of our youngest members. [The first is omitted as not relevant.] The contents of the other were thus expressed : 'The order of succession of parts in foraminiferae is identical with the successive development of leaves in plants, and can be expressed by the same formulae.' Such discoveries, just warm from the study, it may be, as in these cases, forced to light by the occasion of our meetings, are among our greatest triumphs in the way of advancement. (Proc.,vol. 4, p. 159.)
We find no evidence that these ideas of Pourtales have been carried farther either by him or by any one else, though our examination of the literature with reference to the point has been far from exhaustive. D'Orbigny, twenty-five years before, had done much work on the fundamental types of growth in the foraminiferae, though we find no reference to his having compared the arrangements here found, with phyllotaxy in plants.
Our object in calling attention to this matter is, in the first place, to show that we are not quite alone in thinking such comparisons are profitable; and in the second place, to call attention to the possibility that exact studies in the quantitative relationship existing among the members of a repetitive series as well as upon the ordinal arrangement of these members, may be profitable. But should it be found that such studies are significant when prosecuted on unicellular organisms in which the segmentation does not go to complete severance of the pieces, it would seem to follow that they should also be significant when made on species in which the severance is complete, and then to all cell division whatever.
This, of course, brings us immediately to the cyclical phenomena in the propagation of the Infusoria that has received so much attention in recent years, particularly at the hands of Maupas, Calkins, Jennings, Woodruff, and others. Concerning these researches we do no more now than remark that if the general conceptions on which we are going are sound, the phenomena of
440 W. E. RITTER AND M. E. JOHNSON
protozoon division, and of all cell division will have to be examined much more systematically and vastly more exactly, quantitatively, than they yet have been.
3. The inadequacy of treating periodicity generally, as an aspect of fluctuating variation
Here seems to be the place to point out how much more objective, more workable, more important 'periodicity' is in our conception than it is as usually conceived by biologists.
We compare our ideas on the subject with those held by only one other investigator, Hugo de Vries has dealt with certain aspects of periodicity exhibited by plants, quite at length and in several of his works. He states the general facts with clearness. (De Vries, '05, p. 721) :
This law of periodicity involves the general principle that every axis, as a rule, increases in strength when growing, but sooner or later reaches a maximum and may afterwards decrease. This periodic augmentation and declination is often boldly manifest, though in other cases it may be hidden by the effect of alternate influences. Pinnate leaves generally, have their lower blades smaller than their upper ones, the tallest being seen sometimes near the apex and sometimes at a distance from it.
There can be no doubt that the phenomena we are dealing with in Salpa and calling periodicity resemble closely those in plants thus described. The question, are they 'exactly the same' phenomena, we do not raise. Rather, we ask, in view of the closeness of resemblance ought they or ought they not to be looked at from much the same standpoint? The truth is de Vries has regarded the phenomena in plants very differently from what we have in Salpa, and his standpoint is surely inadequate for the facts we are dealing with. This dependency on local nutrition," saj^s de Vries, "leads to the general law of periodicity, which, broadly speaking, governs the occurrence of the fluctuating deviations of the organs" (p. 721). Again (de Vries, '01, vol. 1, p. 638) under the section, "Die Periodicitat semilatenter Eigenschaften," we read:
CHAIN OF CYCLOSALPA AFFINIS 441
Ueber die grossere oder geringere Haufigkeit des Sichtbarwerdens ssemilatenter Eigenschaften entscheidet nicht nur die augenblickliche Lebenslage, d.h. die ausseren Einfliisse wahrend der empJSndlichen Periode der Entwickeluiig. Fast ebenso gross ist die Bedeutung der individuellen Kraft des jungen Pflanzentheiles, diese aber ist das Ergebniss der Wirkung der ausseren Factoren in den vorhergehenden Zeitabschnitten, theils nach Wochen und Monaten, theils nach Jahren gerechnet. . . . Diese Erscheimmg tritt am deutlichsten zii Tage in der Periodicitat der Anomalien aiif der Pflanze.
Again, pushing the matter a step farther, and in a somewhat different direction:
From a broad point of view, fluctuating variability falls under two heads. They obey quite the same laws and are therefore easily confused, but with respect to questions of heredity they should be carefully separated.^ They are designated by the terms individual, and partial fluctuation. Individual variability indicates the differences between individuals, while partial variability is hmited to the deviations shown by the parts of one organism from the average stature." ('05, p. 717). . . . . The individual differences seem to be due, at least in a very great measure, to such apparent trifles. (As differences in soil, moisture, light, etc.). On the other hand partial differences are often manifestly due to similar causes. . . . The development of the leaves is dependent on their position, whether inserted on strong or weak branches, exposed to more or less light, or nourished by strong or weak roots (p. 721). Then follows the quotation already given, viz., "This dependency upon local nutrition, etc."
De Vries' standpoint seems clear: Periodicity in plants is a special form of the more general phenomenon of fluctuating variation which in turn is due to 'ausseren Factoren.' The quantitative differences that manifest themselves in the periods may be lumped together and treated according to the law of probability as first applied to organic beings by Quetelet. After illustrating the application of the method of statistics, the author says: It should be repeated once more that the empirical result is
' It would be very interesting to have deVries follow up this point critically and impartially.
442 W. E. RITTER AND M. E. JOHNSON
quite the same for individual, and for partial fluctuations" (p. 732). And: In the present state of our knowledge the fluctuation-curves do not contribute in any large measure to an elucidation of the causes." (p. 734.)
And so we come to the real issue. Certainly, as de Vries says, the differences called partial 7nay he treated en masse, so to speak. For example, we might pick to pieces ten wheels of the same dimensions of the Cyclosalpa chain, mix the zooids indiscriminately in a dish, then measure them and plot the results. The curve would be the same — the normal probability curve — but would give us no clue to the way the zooids are disposed as to size in the individual wheels. In that case the treatment would not, it is true, contribute in any large measure to an elucidation of the causes." But in our case we have seen that no evidence can be found tending to show that the size scheme as it actually does occur in the wheels is dependent on external factors. All the evidence is to the effect that it is due to the growth process itself independently of any correspondingly differentiating external conditions. In other words, the periodicity in growth occurs under external conditions, that so far as the evidence goes, are not correspondingly periodic. Viewed in this light, can we still say the curves teach us "measureably little about the cause of the phenomena under consideration?" It seems to us not. Truly they do not furnish us 'a complete explanation' of the phenomena. They do, however, tell us, seemingly, this much: That the cause is in the nature of the growth process itself; that the growth goes that way.
If now it should turn out as suggested that not only the length of the zooids falls into a size scheme, but that many of the other morphological dimensions, and functional capacities fall into similar schemes, then the instructiveness of the curves would, for us at least, be very great touching the causes of the phenomena.
Whatever view may be held as to the relation of the periodicity in plants to that in the Salpa chain, it will we believe be allowed that the general question is one of many sides and great possible importance to biological theory. We have not pretended to do more than call attention to it here.
CHAIN OF CYCLOSALPA AFFINIS 443
We conclude with an acknowledgment of our indebtedness to the work of several other biologists who have entered by one or another gate the course upon which we find ourselves. Of these perhaps the first to be mentioned is Julius Sachs whose idea of the grand period of growth in plants must, it seems to us, expand and pla}^ a much larger role in biological theory in the future than it has in the past. After Sachs, chronologically, the various investigations by C. S.Minot on the rate of growth in aminals has largely influenced our observations and thinking. Another research, that by T. Tammes entitled "Die Periodicitat morphologischer Erscheinungen bei den Pflanzen,"has had considerable to do with shaping our ideas on the strictly biological side. But by far the most important as opening up the way to the quantitative work has been Raymond Pearl's Variation and Differentiation in Ceratophyllum." Although Pearl's quantitative data in this research are entirely enumerative rather than mensural; and although his aims and results are in several rather important particulars different from ours, his fundamental problem really gave us our starting point.
JOrnXAI, OF MORPHOLOGY, VOL. 22, xc
444 W. E. RITTER AND M. E. JOHNSON
BIBLIOGRAPHY
Brooks, W. K. 1893 The genus Salpa. Memoirs of the Biological Laboratory of the Johns Hopkins University, vol. 2.
1908 The pelagic Tunicata of the Gulf Stream. In Publication 102, Carnegie Institution of Washington, pp. 73-94.
Brady, Henry B. 1884 The Foraminifera. The voyage of H. M. S. Challenger. Zoology, vol. 9, and plates.
Delage, Yves, etHerouard, Edgard. 1898 Traite de Zoologie concrete. Tome 8, Les Procordes.
Huxley, T. H. 1859 On the anatomy and development of Pyrosoma. Trans. Linn. Soc. 23 (1862) p. 193-250.
Johnson, Myrtle Elizabeth 1910 A quantitative study of the development of the chain in Salpa fusiformis-runcinata. Univ. of Calif. Publications, Zoology, vol. 6, no. 7, pp. 145-176.
MiNOT, C. S. 1891 Senescence and rejuvenation. First paper: On the weight of guinea pigs. Journal of Physiology, vol. 12, pp. 97-153. (Also numerous other writings by Professor Minot.)
Pearl, Raymond (assisted by Olive M. Pepper and Florence J. Hagle) 1907 Variation and differentiation in Ceratophyllum. Publication no. 58, Carnegie Institution of Washington.
DE PouRTALES, L. F. 1850 On the order of succession of parts in Foraminiferae. Proc. of the American Assoc, for the Advancement of Science. Third Meeting , vol. 3, p. 89. Reference to this by Prof. A. D. Bache, A. A. S., Fourth meeting. Proceedings, vol. 4, p. 159.
Sachs, Julius 1873 Lehrbuch der Botanik, Aufi. 3. (The grand period of growth is dealt with by the author in various other publications.)
SCHAUDiN, F. 1895 Untersuchungen an Foraminiferen. I. Calcituba polymorpha Roboz. Zeitsch. fiir wiss. Zoologie, 59, 2 pp. 191-232.
Tammes, T. 1903 Die Periodicitat morphologischer Erscheinungen bei den Pflanzen. Verhand. Kon. Akad. Wetensch. Amsterdam. Tweede Sectie, Deel. 9, no. 5.
Vries, Hugo de 1901-1903 Die Mutationstheorie. 1905 Species and varieties.
PLATES
A.BBREVIATIONS
atr., atrial orifice
hi., heart
emh., embryo
i.h.v., individual blood v
end., endostyle
int., intestine
/.p., foot-piece
oes., oesophagus
g., ganglion
or., oral orifice
gi., gill
ped., peduncle
gon., gonad
ph., pharynx
st.b.v., stolonic blood vessel
445
PLATE I
EXPLANATION OF FIGURE
11 Cyclosalpa affinis Chamisso, solitary generation with chain of five wheels. Natural size.
446
CIIAIX OF CYCLOSALFA AFFINIS
W. E. HITTER AND M. E. JOHNSON
It
t
»
\!^m ..^..^aiaJHBBI^^^^^^M
^/K
^J^W
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
PLATE 2
EXPLANATION OF FIGURES
12 Cyclosalpa affinis, aggregate generation. X 1|.
13 Cyclosalpa affinis, solitary generation, witli young chain of zooids ju.st emerging. X 2.
448
CHAIN OF CYCLOSALPA AFFINI ■\V. E. RITTEli AXD M. K. JtiHVSON
1- niu
„._^_.r
v-^-Jl^j
■^to ^^^
^^mM
BW^^^
^Phjj^i
i^
siflP^^T
^^"N
A^^^*T^
jH
D
A-^
r \
^^on
3^
^x-^^
/!/• \
\ Br
J*
%.
'\^ •'
W
M
iy
)te>
1
1
gon:"
.^
to.
JOURNAL OF
-149
PLATE 3
EXPLAXATIOX OF FIGURES
Deploying point of chain of Cyelosalpa affinis.
14 Dorsal view.
15 Side view, left side.
16 Ventral view.
Zooids on the right side are nuinbereil 1', 2', 3', etc.; those on the left, 1,2, 3, etc. A given zooid has the same number in all three views.
450
CHAIN OF CYCLOSALPA AFFIX IS
W. E. RITTER AND M. E. JOHXSON
~:'rrf
fmftH'i^l^:^'
(4 13 a II 10 4 6 H 10 9 8 "r
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
PLATE 4
EXPLANATIOX OF FIGURES
17-19 Chain of Cyclosalpa affinis
17 Ventral view of chain, showing unbroken part and four wheels. Somewhat diagramatic but drawn to scale. Natural size.
18 Dorsal view of same.
19 Peduncles of distal part of unbroken chain and of first two wheels.
20 Peneroplis arietinus Batsch, sp. Longitudinal section through the shell. Taken from Brady, Foraminifera, Challenger Expedition, vol. 9, plate 13, fig. 22.
21 Reophax membranaceus H. B. Brad}'. Taken from monograph of the Foraminifera of the North Pacific Ocean, Cushman, 1910, U. S. Nat. Museum Bulletin 71, p. 90, fig. 126.
22-25 Chain of Cyclosalpa affinis.
22 Enlarged view of the distal foot-pieces of the unbroken part of the chain.
23 Enlarged view of the foot-pieces of the first wheel.
24 Diagramatic representation of three stages in the development of the circulatory system of the chain.
25 Cross section through chain.
452
CHAIN' OF CYCLOSALl'A AFFINL \V. F.. RITTEK VXD M. E. JOHNSON'
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
453
ON THE FORMATION, SIGNIFICANCE AND CHEMISTRY OF THE WHITE AND YELLOW YOLK OF OVA
OSCAR RIDDLE
FrDiit Ihe Lnhoralories of Zoology and Experimental Therapeutics, rniversity of Chicago
THREE PLATES AND ONE TEXT FIGURE
Introduction -455
A method of measuring the rate of growth of rapidly growing ova 457
The rate of growth of the ovum of the common fowl 458
1. Large ova, more than 6.0 mm. in diameter 458
2. Small ova, less than 6.0 mm. in diameter 459
The thickness of the strata of white and yellow yolk in the egg of the common
fowl 461
The coincicience of the amount of yolk deposited in a day, with the amount
of yolk contained in a layer of white and yellow yolk 462
Yolk stratification in eggs of other animals as seen in the light of its causation
in birds 462
On the chemistry of white and yellow yolk 467
On the mechanism of yolk formation and de-formation 470
1. The part played by the reversible action of enzymes 471
2. The role of the partition coefficient of the elements of yolk 475
3. These two factors and the histological data 477
Summary 482
Literature cited 485
INTRODUCTION
Very many thousands of pages have been written concerning yolk — its presence, formation, varieties and distribution in eggs. Indeed, the task of recording such a series of facts has been repeated on nearly every egg that has come under the closer observation of the biologist; while some eggs, notably those of the frog and the fowl, -have submitted their yolks to the observation and description of dozens of different investigators. Not JOURNAL OF MORPHOLOGY, VOI,. 22, NO. 2
455
456 , OSCAR RIDDLE
withstanding this great amount of study and description, the literature fails to give satisfactory answer to any of the following questions: (1) Precisely how and where does yolk originate? (2) Why, or how is it that there are two kinds of yolk, (a) smaller spherules (often with enclosures) of white yolk, and (b) larger spherules of finely granular (often pigmented) yellow yolk? and, what is the relation between these? (3) What is the meaning of the stratified condition of the yolk of some eggs, eggs in which layers of white yolk alternate with layers of yellow yolk? (4) What are the chief chemical differences between these two kinds of yolk?
Thinking that we are now able positively to answer questions 3 and 4, and that these solutions bring some light upon the first and second questions, we submit the following data and considerations. These are presented with a minimum of reference to the enormous literature; otherwise this communication must have been increased to several times its actual size.
In carrying out this work, and now in the presentation of it, the author would say that he has not forgotten that 'yolk' is 'non-living substance' and therefore from a certain standpoint has but a minor interest to biologists. But, standpoints change. Until Johannes Miiller declared, and Van Beneden clinched the point, that the yolk of eggs is not living matter, and that it contrasts absolutely with the other part of the egg — the protoplasm — yolk had an all-absorbing interest to naturalists as a substance per se. In the years that followed, yolk has been studied largely with a view to cataloguing its diverse occurrences, forms, origins, distribution, tingibility, etc.; its interest to most students has flagged; though its often overweening bulk in the most studied of all cells has frequently won for it unwilling and tedious description. Perhaps one day we shall have a new standpoint. At any rate, we are only now beginning to realize that, though yolk is non-living substance, it is nevertheless organized substance and a very refined product of the vital laboratory; that it is a product laid down in the meshes of protoplasmic elements; and that the very act of its laying down is a signal of important metabolic states and capabilities of these living elements. More of the im
WHITE AND YELLOW YOLK OF OVA 457
port of the relation between this 'organized' and this 'living' material we shall know later; in the meantime, each bit of information is doubly welcome because it concerns the most interesting form of protoplasm — the egg — at what is probably its most interesting period. Perhaps, then, the substance that has seemed to have but blundered in where it could blind us most, may itself prove to be a mirror for many a secret that we have elsewhere sought in vain.
A METHOD OF MEASURING THE RATE OF GROWTH OF RAPIDLY GROWING OVA
The present studies began with an attempt to learn the cause of the stratified condition of the yolk of the hen's egg. It was suggested to me by results of an earlier study ('08) that the alternate layers of white and yellow yolk in the egg may be the result of the daily rhythm of nutrition — connected with high and low blood-pressure — which I had discovered in birds, and which I had shown to be the cause of the alternate fault-bars and fundamental bars of birds' feathers; it being there found that the daily variation in nutritive conditions in birds is sufficient to produce structurally perfect, and structurally imperfect parts in their rapidly growing feather germs. To test the suggestion, then, one might need only to learn the rate of growth of a bird's egg. What is the rate of growth in the eggs of the common fowl? This had not been determined, and no way of determining it was known.
It occurred to me that Sudan III might be used for this purpose. Knowing that Sudan was not destroyed in passing through the intestinal wall, (Daddi) that it circulated tied to the fatty acids of the food, and that the fatty acids of the food were laid down unchanged in the egg (Henriques and Hansen, '03), I inferred that Sudan given with fatty food would be laid down in the egg.
Moreover, it seemed possible by regulating the dosage and using proper intervals between feedings, to get laying hens to put this bright pigment down as definite bands in their growing ova, and thus enable one to determine the rate of growth.
The first experiment was as successful as the last. When such Sudan-containing eggs* were hard-boiled and sectioned under
458 OSCAR RIDDLE
water it was easy to measure the distance between the innermost borders of two such rings of Sudan, and thus to identify this amount of growth with the time which was known to have intervened between the two feedings.
Having thus discovered a method^ (described in detail by me elsewhere, '10) of measuring the rate and time of growth of ova, many data were collected on this point ; the distance between the normal strata (layers of white and yellow yolk) of the egg was carefully measured; later the problems and considerations growing out of the results were further followed up. We give here the following short statement of the observations and conclusions:
The radius of the hen's egg increases during the last few days of its growth by about 2.0 mm. per twenty-four hours. The thickness of a layer of white yolk and a layer of yellow yolk taken together is usually about 2.0 mm. Our conclusion is that in the fowl a layer of white and another of yellow yolk are laid down each twenty-four hours. Other facts at hand indicate that the yellow yolk is laid down under the best nutritive conditions, while the white yolk is a sort of growth-mark left by poorer nutritive conditions.
THE RATE OF GROWTH OF THE OVUM OF THE CO-MMON FOWL 1. Ova of more than 6.0 mm. in diameter
Table 1, section A, and plates 1 and 2 have been prepared to show the rate of growth of the larger ova as this is indicated by the Sudan method. The reader is referred to the table and plates in order to learn the kind of evidence on which the first conclusion is based. The amount of this evidence could be increased several times. It will be seen that the radius of the larger ova contained in the ovary of a fowl may increase by rather more than 2. mm. during twenty-four hours; also that this rate of growth is quite variable and may often fall to one-half the above amount.
Other data in our possession show that this rate not only varies for the eggs of different ovaries, but for different eggs of the same
' First announcement of the method, and of some of the present results, Riddle ('07).
WHITE AND YELLOW YOLK OF OVA
459
ovary which were grown at different periods. It has been shown moreover by the Sudan-method that the rate of growth may be reduced not only to one-half that given above, but to absolute zero; this, however, is only a confirmation of what inference has long declared must be so, since ova may even decrease in size while in the ovary, i. e., they may be resorbed.
TABLE 1
Showing under section A the rate of growth of hens' eggs as this was measured in central and peripheral parts of the yolk by means of Sudan. The numbers in the first column refer to the number which this egg bears in the plate. In section B are recorded measurements of the thickness of pairs of white and yellow yolk strata in central and peripheral regions of the ovum as this could be seen with unaided eye or with addition of iodine solution. The seven measurements here chosen arbitrarily from nearly forty in the records, are consecutive measurements of eggs from different hens.
A
B
NO. OF E
GG
24 HOURS RADIAL GROWTH
IN MM.
NO. OF OVARY
THICKNESS OF A PAIR OF
YOLK STRATA
Central
Peripheral
Central
Peripheral
5 PI.
1
2.2
1^ PI. 1
!
1.67
3='
1
1.41
1
2.16 2.16
4a
1
1.3
1.7
2
1.54 1.54
5=^
1
1.7
3
1.75
6=^
1
1.64
4
2.5 2.5
1
2
1.8
1.8
5
1.4 1.4
2
2
1.5
1.5
6
2.0 2.0
5
2
1.3
7
1.47 1.47
1
Average .
1.53
1.67
Average
1.85 1.81
2. Ova of less that 6.0 mm. in diameter
It hafi not been possible to obtain a deposit of Sudan in eggs smaller than 6.0 mm. in diameter. This failure is explained by the fact that these ova are growing very slowly, as compared with the more advanced ova, and the intake of the stained food is here not rapid enough to give a perceptible effect. We shall see, moreover, that this white yolk — for ova of this size are com
460 OSCAR RIDDLE
posed entirely of white yolk — is much poorer in fat than is yellow yolk. Since fat is the only food that can carry Sudan this is another reason for the failure of Sudan to appear in them. The Sudan method is therefore not available for the determination of the rate of growth in these eggs.
One bit of evidence of another sort concerning this rate of growth was obtained and may be recorded. In fig. 3, pi. 2, is shown the striated appearance which the peripheral white yolk of one of the small eggs showed after having lain in a quantity of Mann's formalin-alcohol mixture for a few weeks. Here the noteworthy facts are, that a striation exists, and that the lamellae are not thicker than 0.25 mm. Whether these lamellae are made up of still smaller strata which really represent days of growth I am quite unable to say. I doubt somewhat that the radius of these small eggs is increased by as much as 0.25 mm. in twentyfour hours; any^^ay these strata offer some evidence — ^in the light of what we know of succeeding yolk strata — that these small eggs do not grow faster than 0.25 mm. per day.
One must ask what is the meaning of the extraordinary difference in growth-rate of eggs under, and over, 6.0 mm. in diameter? Wliat old mechanism is inhibited or what new one brought into action, that accounts for this procession of cells — each with months of slow and constant growth behind it — coming to a point from which each jumps in a day from its accustomed rate of increase, to a rate that is probably from eight to twenty times higher? Do the folHcular cells now become more permeaable than formerly to the ingredients of yolk? Is the increased vascularity of the follicular envelopes, that certainly occurs at this time, a cause or a result of the new activity? To these questions there comes no answer. But to us there are few events in the history of the primary oocyte of the fowl more interesting than this one. All the more interesting it is, too, because of its glaring apparent teleology. Here is an ovum within five to eight days of extrusion^ a ad containing less than the hundredth part
^ It is true, however, that if the yolk grow less rapidly than normally the egg remains longer in the follicle; showing that the time of ovulation is not controlled by heredity but is governed quite completely by conditions.
WHITE AND YELLOW YOLK OF OVA 461
of the yolk necessary to make it capable of producing an animal. Nevertheless five to eight days suffice to supply the missing ninety-nine parts.
THE THICKNESS 0;F THE STRATA OF WHITE AND YELLOW YOLK OF THE COMMON FOWL
The measurement of the thickness of a layer of yolk offers some difficulties and- can rarely be done directly on a single layer; the reasons being that one stratum merges very gradually into another and that the strata are often very indistinct. More frequently, though by no means in every egg, a series of wellmarked layers can be found and a measurement made over all; the number of strata — or rather of pairs of strata — may be easily counted. When the total measurement is divided by this number one obtains the thickness of a combined layer of white and yellow yolk.
The result of eight such measurements is recorded in section B of table 1. These are typical of nearly forty reliable measurements, and indicate a thickness of about 2.0 mm. (1.4 — 2.5) for a layer of white and yellow yolk combined.
The layers of yolk can sometimes be seen in the fresh eggs, proving that they are not artifacts; but for the purpose of measurement it is usually best to hard-boil them, and section (under water) from one side until the exact center of the egg is reached. Sometimes it will be found advantageous to put the egg thus prepared in weak iodine solution for a time. This treatment seems occasionally, though not always, to strengthen the contrast between the layers of white yolk and those of the yellow variety.
For reasons stated above it is impossible satisfactorily to measure the thickness of a layer of white yolk. It can be said with confidence, however, that this so-called layer has but a fraction of the thickness of the adjoined yellow layer. Perhaps one errs but little in saying that the former usually has from one-fourth to one-eighth the thickness of the latter.
462 OSCAR RIDDLE
THE COINCIDENCE OF THE AMOUNT OF YOLK DEPOSITED IN
A DAY, WITH THE AMOUNT OF YOLK CONTAINED IN A
STRATUM OF WHITE AND YELLOW YOLK
A comparison of the two sections of table 1 shows quite convincingly, I think, that the figures, which in the one column indicate the amount of a day's growth, are of the same order of magnitude as those which in the other column indicate the thickness of a stratum of yolk. This fact, and another one, namely, that we know thai there exists in birds a daily nutrition rhythm capable of producing daily growth-marks in their rapidly growing feathers, convince us that a layer of white yolk and another of yellow yolk is laid down during each twenty -four hours.
The well-developed appearance of the yellow yolk, its large yolk-spherules audits much greater thickness than that of the white layer, all indicate, moreover, that this layer, like the broad fundamental bar of the feather, is grown under the best nutritive conditions; while the narrow layer of white yolk with its small spherules gives indication that it, like the fault-bar of the feather, is grown under poor nutritive conditions.
Since I have shown that the poor nutritive conditions which produce the fault-bar occur in the later hours of the night — 1 :00-5:00 A.M. — / consider it as practically certain that the white yolk of the ovum is produced at the same time, and that the yellow yolk is produced during all other hours of the day.
The layer of white yolk of the hen's egg is then a growth-mark left at the ever-changing boundary of the ovum; it represents the results of yolk formation under sub-optimal conditions. It is indeed incomplete, unfinished yolk, as is apparently indicated by the histological data already known, and by the chemical evidence which I shall present in another section.
YOLK STRATIFICATION IN OTHER ANIMALS AS SEEN IN THE LIGHT OF ITS CAUSATION IN BIRDS
With the story of the white and yellow yolk of the bird in mind it becomes most instructive to reexamine many of the peculiar types of yolk distribution which from time to time have been re
WHITE AND YELLOW YOLK OF OVA 463
ported and figured by embryologists and cytologists; for now we can feel fairly sure that wherever we meet alternate layers of white and yellow yolk, such layers indicate just so many alternations of better and poorer nutritive conditions during the time these layers were being formed. The better and poorer nutritive conditions doubtless applying to the organism as a whole. ^
A zonal arrangement of yolk similar to that of the bird has been reported in at least four other groups, viz., turtles, lizards, skates, and m3rxinoids. Some yolk patterns are known which are not distinctly zonal but intermediate to it and the type of yolk arrangement which is usual in small eggs; these help to bring all yolk distribution under a single principle or set of principles.
In order to avoid much tedious description in the text, and also to present more clearly and accurately this part of the subject, I have prepared plate 3, which is to a large extent a reproduction of figures which are not new. To what is shown in the plate, and in the explanation which accompanies it, I here add the following:
In all ripe ova, as in all the growth stages during which yolk is being deposited in the ovum, a layer of yolk composed of very small spherules (white yolk) is to be found at the extreme periphery of the egg. If larger yolk spherules (yellow yolk) also occur, they occupy more central portions of the egg. There is, moreover, scarcely an exception to the rule that the germinal vesicle or egg-pronucleus is immediately surrounded by similar small spherules and not by large ones.
It seems also to be very generally true that in those ova in which considerable yolk is developed, and in which the germinal vesicle makes its way from the center to the periphery of the egg (or remains near one side of the cell) it leaves in its wake a cyhnder of white yolk to which in some cases has been given the name of Pander's nucleus.
All of these features are shown in eggs of such widely separated forms as the skate (fig. 6) the amphibian (fig. 5), the lizard (fig.
^ On the other hand, some eggs, e. g., those of the salmon, may undergo their chief growth at the expense of the somatic tissues and while no food whatever is being ingested. The conditions here, however, are essentially constant and therefore produce no stratification of the yolk.
464 OSCAR RIDDLE
8), the birds and at least in some mammals (fig. 3). These are the forms, too, which — with the exception of the mammal — ^in addition show a stratification of the main body of the yolk. Two other forms are known, the turtle and the cyclostome (Bdellostoma) in which the stratification and other features occur, as in the above mentioned eggs, except that no Pander's nucleus has been found.
How may we explain at one and the same time the essential similarity of the yolk distribution in eggs of widely separated forms, and the often essential dissimilarity of its distribution in the eggs of closely related species? There seems now no doubt that all can be accounted for when one knows two things : first, the length of the growth period; and, second the chief fluctuations in the nutrition of the animal during the growth period of the eggs.
Most ova have no stratification, then, because the yolk is grown in a short season — -the animal not being subjected to such severe alternations as winter and summer, while the process is going on; or, because the eggs remain very small and develop little yolk'; or, again, because some ova have the extraordinary capacity of growing at the expense of somatic tissues. In such cases fluctuations in the nutrition of the animal are of little moment to the egg; the latter being able to feed well at the expense of the organism as long as it continues to live.
When stratification is present, however, I believe this to be a positive declaration that nutritive fluctuations did occur in the organism, and the number of the strata to be a reliable index to the number of such fluctuations. The presence of yolk stratification in the eggs of an animal then is an invitation to the naturalist and physiologist to look for important nutritional variations in that animal.
Thus far definite causal and time relations between such stratification and nutritional fluctuation has been determined only for the bird. What this time period is in BdeUostoma we can now only conjecture; but the fact that in a mature specimen eggs of a wide range of size exist possibly argues that these eggs are several years in forming. The further fact, that the animals lose much blood and become much weakened at each yearly spawning
WHITE AND YELLOW YOLK OP" OVA 465
period, is significant in that here may be found the means of a nutritional depression which produces a layer of white yolk in all of the remaining eggs of the ovary. If this be the true explanation one can readily understand the lack of stratification in the eggs of the related Petromyzon (fig. 2) since this form spawns but once in a lifetime.
In the skate the main growth period of the oocyte is probably completed in less than a year. The nine or ten pairs of strata figured by Riickert (fig. 6) are probably produced at the rate of about one per month. Whether this refers merely to the number of times the animal has fed during this time, or otherwise, nothing seems to be known.
The amphibian egg has a short growth period, and derives its growth material too from substances stored in the body, and is thus independent of external food supply. Doubtless these facts — together with its usually moderate size — will account for the actual configuration of its yolk.
The eggs of two reptiles — turtle and lizard — show very evident, but dissimilar, yolk strata. What the time, or the nature of the nutritive fluctuations are, that may produce these strata in Lacerta, I can make no suggestion.
In the egg of the tortoise Munson ('04) seems not to have identified (fig. 1) the so-called inner and outer cytocoel as layers of white yolk. A study of my own preparations, however, convinces me that such is their nature and the term cytocoel therefore is unnecessary. The turtle's egg has then alternate layers of white and yellow yolk somewhat comparable to those of the bird.
1 have found indications of four pairs of such zones in some eggs; or rather, by comparing the strata of different eggs from the same animal I have found such indications. But I am not now sure that four such pairs exist, nor that only four exist. Certainly several very thin strata can sometimes be found within
2 mm. of the periphery of some ova.
One wonders much whether the well-marked innermost layers of the turtle's egg can be the indications of years of growth. Agassiz ('57) showed that these ova undergo their greatest growth in four interrupted stages extending over four years. Our predic
466 OSCAR RIDDLE
tion is that further examination of these yolks, by proper methods for differentiating the strata, will show four pairs of white and yellow zones, to correspond to four yearly periods; each year supplying a period of growth and of rest, or at least of more rapid, and of less rapid growth.
Of the mammal's egg shown in fig. 3 it can be said that the several conditions of its growth seem to be closely similar to those of the amphibian egg which it so much resembles. To be sure, this egg may not, like the amphibian, develop at the expense of substances stored in the body; but, so few eggs are here developing at one time that an adequate food supply is always assured.
We believe then that these data practically give answer to the very important question which has been so well put by Riickert ('99, p. 585):
Diese Uebereinstimmung des Selachier— speciell des Torpedo-Eies, mit dem Vogelei ist, wenn der Vergleich sich zunachst auch nur fur die grobere Structur durchfiihren lasst, immerhin eine auffallende Thatsache.
Es wiirde die Mtihe wert sein, bei einer erneuten Untersuchung der ohnedies seit vielen Jahren vernachlassigten Dotterentwickelung nach Anhaltspunkten zu suchen, ob die Aehnlichkeit nur dadurch hervorgerufen wird, dass die beiderlei Eier unter gleichen Bedingungen sich entwickeln, oder ob es sich um einen durch Vererbung auf das Vogelei iibertragenen Vorgang handelt; mit einem Wort, ob eine Analogie oder Homologie vorliegt. Im letzteren Falle wiirde sich der Schluss Ziehen lassen, dass das meroblastische Ei des Vogels resp. der Sauropsiden ein primar meroblastisches ist wie das Selachierei und das Saugetierei kein tertiar sondern ein sekundar holoblastisches wie das Amphibienei.
The similarity noted above of the amphibian and marsupial eggs is another case in point. My results indicate that the likeness of yolk distribution in these two eggs, and in those of selachian and bird cited by Riickert, does not rest on heredity in any narrow sense of the word, hut on the fact thai they develop under like conditions.
ON THE CHEMISTRY OF WHITE AND YELLOW YOLK
The conception of white yolk which arose from the preceding work was that such yolk is a halted, or intermediate stage, in the
WHITE AND YELLOW YOLK OF OVA 467
development of yellow yolk. This same conception had been urged on histological grounds by several workers, though opposed by others. The chemistry of the two substances was then appealed to for further evidence of a sort which it alone could give.
An examination of the rather abundant literature on the chemistry of yolk showed that it contained none of the data which our problem required. Analyses of yellow yolk have indeed been made by Prout, by Gobley and bj^ Parke; but it was believed that the extraction methods of their time did not effect a complete separation of the fat from the other constituents of the yolk. These determinations have therefore been made anew. That such was really necessary may be indicated b}^ the fact that Parke ('67) extracted only 66.7 per cent of fat and phosphatids, whereas my analyses always yielded more than 70 per cent of these constituents. It was also imperative of course that results of analyses which were to be compared should be obtained by identical methods. Apparently no analysis of white yolk had been made, so that this had to be done.
Since, moreover, the metabolism of yellow yolk includes not only its formation but also its de-formation into absorbable constituents, it was considered necessary to take account of yolk in a late stage of such modification. Such yolk is met with in two rather different situations: Normally, the whole yolk of the egg (yellow yolk) is subjected during the incubation period to the digestive, i.e., disintegrative action of the embryonic tissues — entoderm and yolk sac. Under such modifying action does yellow yolk become more like white yolk, or does it become less like it? A similar digestive action occasionally overtakes an ovum in situ, i.e., while still in the ovary and surrounded by follicular cells. These are the so-called 'resorbed ova.' How does the yolk of such an ovum in an advanced stage of resorption compare with the yellow yolk which it was before the beginning of resorption? Has it become more like, or less like white yolk?
The complete results of my analyses with a consideration of their points of chemical interest, and an account of the preparation of materials, and of methods used, will be pubhshed elsewhere.
468 OSCAR RIDDLE
I may say here that the fat and phosphatid extractions were made with the methods recently discussed and described by Prof. Waldemar Koch, in whose laboratory these analyses have been made. At this time it seems most desirable to present only the amount and sort of data which is necessary to give a clear picture of the major differences between the two forms of yolk under consideration, and to answer the two questions just stated above.
TABLE
2
MOISTURE
IN PER CENT or SOLIDS
Fat
Phosphatids
Extractives
Protein
1
47.8
63.2 '
49.2
88.1
49.2
45.7
40.7
36.8
20.9
15.3
15.9
11.1
0.6
2.0
2.4
3.4
28.8
2
3
4
35.2
38.7 43.5
1 = analysis of fresh egg-j-olk (yellow yolk) (17.670 gr.)
2 = analysis of a resorbed ovum (1.834 gr.)
3 = Average of three analyses of contents of (9) yolk sacs (18 da. inc.), (78.821 gr-)
4 = analysis of white yolk (6.019 gr.)
Table 2 has been so arranged as quickly and accurately to tell the story. Nos. 1, 2 and 4 are single and quite typical analyses. The several analyses of the yolk sac contents varied considerably, and therefore an average of three separate analyses of yolksacs of eighteen days incubation is here given in preference to a single analysis. The white yolk was taken from a great number of eggs under 6.0 mm. in diameter, the yolk being removed without carrying over any traces of the enveloping membranes.
The quantitative differences in each of these chief components of white and yellow yolk are remarkable. Quite as striking and conclusive, too, are the numbers which show that ivhen yellow yolk is subjected to digestive action, in either of the two situations named, each and every component approaches more nearly to the quantity characteristic ofiohite yolk.
It cannot be said, however, that these data conclusivelj^ answer the question we have raised as to whether white yolk is an inter
WHITE AND YELLOW YOLK OP OVA 469
mediate stage in the formation and indeed of the de-formation (digestion), of yellow yolk; although they do strongly support that view. There seems to be an alternative, namely, that the figures under nos. 2 and 3 approach the composition of white yolk more and more, only because the amount of that sort of yolk originally present in the egg is not diminishing, or is diminishing but slowly, whereas the yellow yolk is here being digested very rapidly. For, it must be remembered that, although we are considering a mature hen's egg as our type of yellow yolk, it still contains white yolk in quantities not easy to estimate; though we are accustomed to think of this amount as small, probably between 5 and 15 per cent of the total.
Parallel to the chemical data are the histological conclusions that it is always white yolk and never yellow yolk that is found applied to a surface into which yolk is being ingested. This is true for the germinal disc of pre-embryonic stages, and for the advancing entoderm and yolk-sac of the embryo (Balfour, Agassiz). Virchow ('91, p. 105) however, questions the correctness of this statement. It is certainly almost always true for the nucleus, or germinal vesicle of the primary oocyte, a seemingly significant fact upon which I shall publish observations elsewhere. Our chemical data themselves show, however, that the alternative cannot be true unless there is several times as much white yolk in an egg as we have reason to believe exists there. In any event the certain and interesting fact remains that when the yolk complex of the hen's egg is subjected to digestive and absorptive processes, the fat and phosphatids digest and disappear much more rapidly than does the protein.
ON THE MECHANISM OF YOLK FORMATION AND DE-FORMATION
Having presented data to answer questions three and four of the introductory statement, we may now consider the first and second questions in the light of these results, and with the help of other facts. Precisel}^ how and where does yolk originate? Why or how is it that there are two forms of yolk; or, what is the relation between these?
470 OSCAR RIDDLE
I purpose to preface this inquiry with a statement of my two main conclusions, or theses. (1.) The formation and the de-formation of yolk are one and the same subject. The processes of building are also the processes of tearing-down; only an equilibrium changes. These two sister-subjects have, however, long paraded as independents. The formation of yolk has been considered a subject the investigation of which was connected with a wide variety of study such as the migration of fully-formed yolk granules from follicular cells into the ovum; the origin of yolk granules from migrated particles of the chromatin, or the nucleolus; or again their formation by the yolk nucleus, or by mitochondria, etc. On the other hand, when the other phase of yolk metabolism — its de-formation — was concerned, observers have been pretty generally satisfied to speak only of 'a digestion and ingestion of yolk.'
(2.) Given a region into which the elements of yolk — ivith their vast amount of potential energy — can go and can exist without undergoing oxidation, and yolk {or some of its elements) will there be increased or decreased in amount subject to an equilibrium which is a function of two 'factors; (a) the reversible action of enzymes and, (b) the partition coefficient of the elements of yolk. We do not state that all desirable proof of this thesis is at hand, but we do insist that a very considerable body of evidence supports it. Having been led to the formulation of this view, and to the acceptance of it to the fullest extent ourselves, we shall here outline the evidence which we believe will likewise commend it to others.
It is not necessary to discuss separately what we have called theses one and two. Both rest upon the question of the presence, the effectiveness, and the modus operandi of the two factors which we have proposed as the immediate agentsof yolk transformations ; whether such transformations be of growth or of 'digestion,' whether they be progressive or regressive in character. The discussion therefore hangs upon these factors and we shall consider them separately.
Before proceeding in this direction, however, it is well to be reminded that these theses are the physiological and explanatory counterpart of an histological dictum which in certain of its aspects has been for many years ably maintained by several noted
WHITE AND YELLOW YOLK OF OVA 471
histologists ; but which has apparently not gained universal acceptance : A spherule of yellow yolk may arise from a spherule of white yolk; in the normal destruction and utilisation of the yellow spherule, a white spherule may be again produced.
1. The part played by the reversible action of enzymes
Kastle and Loevenhart ('00) proved the reversibility of the action of lipase — the enzyme concerned in the analysis and synthesis of fat ; and we have seen that fat is the chief constituent of yolk. Wohlgemuth ('05) demonstrated the presence of lipase in the yolk of the fowl's egg. It was shown by Henriques and Hansen ('03) that the fatty acids of the food, i.e., of foreign fat, are laid down as such in the hen's egg. Since we know that this fat did not originate within the egg; and, since we are assured that fat as such does not pass through living cells, but that it is previously split into alcohol and constituent fatty acids, we must believe that the foreign fat found by Henriques and Hansen was synthesized within the egg cell; or, that it was synthesized in the neighboring follicular cells and thrown from their inner margins into the egg. This last alternative is not true as will be pointed out later.
Thus we come by means of the above series of facts directly to the proof of the existence within the fowl's egg of the synthesis — one side of the enzyme action — of the most voluminous constituent of the yolk.
Has the existence of the splitting action of lipase in the egg also been demonstrated? I believe it has practically been so demstrated by Liebermann's ('88) determination that only the merest traces of free fatty acids are present in the fresh egg, whereas large amounts are present at seven and fourteen days of incubation. The existence of a splitting activity of lipase in the hen's egg is moreover a matter that probably no one will question. From these facts then I think it must be said that the reversible action of lipase within the hen's egg has been indirectly demonstrated.
In fact, one familiar with the picture presented by the deposit and absorption of the yolk of eggs, can but wonder that this picture has not been before specifically pointed out as an example^
JOURNAL OF MORPHOLOGT, VOL. 22, VO. 2
472 OSCAR RIDDLE
a typical example — of the reversibility of lipase effecting speedy and rhythmic transformations. The example too, becomes of considerable zoological interest, since certainly nowhere else does this simple physiological principle have such a relation to interesting features of morphology as just here. For, not only does it in these cases often completely change the features of the egg-cell, but it results in a condition (telolecithal) which later gives direction to a host of events of early development — cleavage, gastrulation, etc. — which proceed from the egg.
When we have spoken above of proof of the synthetic and of the analytic action of lipase we mean, of course, proof that each of these reactions may predominate in the egg. The burden of our whole statement is that both sorts of reaction are going on simultaneously (since the reaction is a reversible one), but that the conditions in the egg are, as a matter of observed fact, shown to be such that during the growth period the synthetic reaction normally exceeds the analytic; and that during incubation the reverse is true.
We say nothing in this connection of the origin and disintegration of the proteins of the egg. This group does not furnish, at present, examples specially proved for conditions in the egg, as do the fats. The reversibility of proteolytic enzyme action has however been demonstrated.
With yolk-forming enzymes (lipase, etc.) accelerating a series of reversible reactions in an egg-cell in which traces of yolk have been deposited, what are the factors which favor each side of the reaction, and thus induce either an increase in the amount of yolk, or a decrease in the traces that already exist? We believe that for the ovum of the fowl which we shall more specifically consider, some of the factors effective elsewhere may be ignored.^ The daily temperature fluctuations, for example, are relatively slight, etc. There seems good reason to believe that the amount and pro ■* There are several other factors or conditions which possibly, even probably, play parts in the storage of facts, i. e., building of yolk, in the egg; most of these however are factors supplementary to those of distribution coefficient and enzyme reversibility; though some are not. Some such factors known to be effective in fat-storage elsewhere are: (1) quantity of lipase (Kastle and Loevenhart) ; (2) different species (?) of lipase (Hanriot) ; (3) alkalinity
WHITE AND YELLOW YOLK OF OVA 473
portion of the reacting substances present is here, as elsewhere under these conditions, the factor that determines whether the amount of yolk shall from time to time increase, remain constant, or diminish. What then are the conditions in the fowl that would tend to modify the amount of these reacting substances in the egg?
In answer to this we revert to the facts forecasted at the beginning of this paper in regard to our own earlier demonstration ('08) of a daily rhythm of better and poorer nutrition in birds; which rhythms coincide with periods of higher and lower blood-pressure. It was there made certain that very rapidly growing organs (feather-germs) were usually unable to pass over the period of the nightly (1:00-5:00 a.m.) reduction in blood-pressure without showing defects; which defects were proved to be due to insufficient nutrition.
Now we think there is no doubt that these facts lead to an answer to the above question. The egg (like the feather germs) doubtless derives fewer nutritive particles from the blood at this time than during the rest of the day. Possibly, even probably, the low blood-pressure induces at this time feeble but effective currents of fluid from these cells towards the blood and lymph; for it is probable that under low-blood-pressure the volume of the blood tends to increase at the expense of the fluid of the tissues. At any rate it seems certain that at this time the intake of the food substances from the blood is reduced, with the result that the equilibrium of the reaction is shifted. Thus the morphological picture becomes changed. Now growth will proceed more slowly. It is now that the granules must remain small, and poor in fat. It is now that some of the larger yolk spherules (yellow yolk) may possibly suffer reduction to smaller spherules (white yolk) ;
(Hanriotj ; '4) presence of other bodies eg. lecithin (Hewlett); (5) reducing conditions, i. e., conditions favorable to the formation of fat from carbohydrate and protein by reduction. A further reason for only a mere mention of these factors here is that the data for the egg are at present too meagre.
The factors which have to do with the formation and storage of the protein constituents of yolk, and of their union with lipoids and fats to form yolk, are less known than those factors which involve fat metabolism only; therefore the latter only are treated here. Macallum ('91) has some interesting statements on related subjects, and further points out that similar processes occur ill the formation of yolk and in the production of pancreatic zymogen.
474
OSCAR RIDDLE
the former being robbed more rapidly of their fat than of their protein. Now a layer of white yolk is produced in the egg.
In fig. A is shown a diagrammatic representation of how these fluctuations in the quantity of food-products of a fatty nature in the blood-stream would effect changes in size in oil drops, if these latter were separated from the blood by thin and semi-permeable membranes — the conditions existing at the surface of an egg. Section A represents growing conditions — predominance of fat synthesis due to rapid ingress of the constitutents of fat. Section B
C^5po-V
Text Figure A=1
Idealized representation of the relations of the periphery of a mature sauropidan egg to the blood and lymph. Follicular cells not shown; these considered pervious as vitelline membrane, or by their intercellular spaces offering free access of lymph to that membrane. A = optimum growth conditions. B = metabolism of an oil drop in equilibrium. C = impoverished blood bearing away elements of yolk, with extension of white yolk area at the expense of yellow yolk. b. cap. = blood capillar}'; v.m. = vitelline membrane; iv.y. = granules of white yolk; y.y. = granules of yellow yolk. See text .
stationary conditions; as much of fatty ingredients is being given off into the blood, as is being taken from it. Section C droplets reduced in size as a result of continued contact with a blood stream poor in fat.
WHITE AND YELLOW YOLK OF OVA 475
2. The role of the partition coefficient of the elements of yolk
Of less importance than the reversible action of enzymes, but following upon it, is the distribution between the yolk'" and the blood of the soluble substances concerned in these reactions according to their relative solubilities in these two solvents. There can be no doubt that this distribution, or partition coefficient is a factor in determining the amount of soluble substance which comes from the blood, lymph, follicular cells, or vitelline membrane, to the periphery of the yolk, and vice versa. Such is a physical necessity. The constitutents of fat for example, quite certainly enter the egg in soluble form and must there be subject to the laws of solubility.
The chief thing incumbent upon us in this connection is to point out how this partition coefficient may act selectively in modifying the amount of the reacting substances in the egg; i. e., how this principle may contrive at one time to increase, and at another to decrease the quantity of yolk contained in the egg. Remembering that it is the amount of reacting substances present that decides whether yolk formation or yolk de-formation may occur, the answer can perhaps be more easily given in reference to fig. 1. Let this figure now represent the periphery of the ovum of a turtle, in contact with the lymph and blood streams. During the summer, when the constituents of fat are probably most abundant in the blood, some of these must, on account of their solubility, pass into the egg and there later be built into yolk; their former places being continually taken by new particles from the constant supply of the blood stream. Under these conditions yolk spherules grow, as is represented by section A. In winter, conditions become as in C. The turtle cannot now digest food (Riddle, '09). Its heart-beat and other activities, hoVever, require food for their continuance, and the blood becomes depleted of food. The reversible action of its yolk enzymes is not likewise suppressed, but these now as before set free soluble yolk
^ The word 'yolk' is here made to stand for the whole body of the egg cell. Perhaps egg-protoplasm, follicular cell, vitelline membrane, and yolk, should all be mentioned instead of 'yolk' alone.
476 OSCAR RIDDLE
constituents. For such to be set free now, however, is to leave the egg entirely; for now the distribution coefficient of each of such substances brings a portion of it into the blood or lymph; and here it is not allowed to accumulate — to saturate this solution and then cease to act, — but is taken up by other organs ; while the blood thus freed from traces of it continues to pick up more of such particles as it passes the ovum. Because of this principle then an ovum may not be able to hold all the yolk that it has once acquired. Apparently we can explain the broad zones of white yolk in the turtles in this way, and the known facts seem to require the mechanisms we have described.
Of course we do not mean to infer that no other factor than the two we are describing have to do with certain aspects of yolk metabolism. For example, these two m^y have little causative influence in deciding that very important matter as to when the rapid growth of the hen's yolk is to begin. Here lie mysteries perhaps of the folhcular cells, or something else, perhaps more distant from the point of actual yolk formation. We are dealing only with the immediate mechanism of yolk formation and deformation.
The possible role of lecithin in increasing the solubility of fatty acids and soap in the follicular cells and in the yolk is an attractive subject. Moore and Parker ('01) have shown how enormously the solubilities of these substances are increased by the addition of small amounts of lecithin and bile salts. I have ascertained the presence of lecithin in the follicular membranes, but as yet have not enough analyses for comparison to draw conclusions. I have determined also, as is indicated in Table II, that the lecithin content of the white yolk — i. e., the layer just beneath follicular membrane, and usually between it and the yellow yolk — is smaller in amount than that of the yellow yolk.
As a concluding word on the role of the partition coefficient we record our belief that it alone accounts for the presence of the yolk coloring matters — vitello-lutein and vitello-rubin — in the yellow yolk, and not in the white. These are lipochrome pigments, soluble only in fat and fat solvents, and are abundant in the large yolk spherules, probably because, as we have shown by compara
WHITE Ax\D YELLOW YOLK OF OVA 477
tive analyses, these spherules abound in fat. Of similar intei-est is the discovery of Miescher ('97) that at the time of the development of the eggs of the salmon the blood of thfese animals is unusually rich in lecithin, fat and globulins.
3. These two factors and the histological data
One hardl}^ has a right to mention the words 'histology of yolk' without entering upon the consideration of an enormous literature. Since my own contribution is not primarily of histological nature, and for reasons stated at the outset, I refrain from doing so, although by my results I am seeking to put some rather new and additional interpretations upon histological conclusions, and to answer some questions in which histological, and to a less extent microchemical methods have before been largely used.
The view that intermediate forms of spherules exist, connecting white yolk with the yellow yolk spheres, has been maintained by Riickert, Sarasin, Disse, Kolliker and others. The region under the germinal disc of avian, reptilian and selachian ova have furnished the most and the clearest pictures of the transition forms. Previous authors have, however, generally considered only the formation of the yellow from the white spheres during growth, and have not considered the reverse of this process as it occurs during the destruction of the yolk. The engulfing of whole granules of (white) yolk by the entodermal cells has been recorded by His ('00) and others. This I would observe is, if true, not a real contradiction of my thesis, since these granules doubtless later undergo the ordinary processes of digestion in the entodermal cell. Similarly I would note that the presence of yolk granules in follicular cells — demonstrated by many observers — only illustrates the mechanism we have described at work in another cell; the classic example of this sort of formation being the fat globule in the cell of the intestinal mucosa. On the other hand the finding of such granules in a follicular cell is no guarantee whatever that the granule is thrown as such from that cell into the egg. The granule may here, as in the mucosa cell, again undergo digestion and pass from it in solution.
478 OSCAR RIDDLE
As regards yolk formation in insects, conditions are peculiar; the nurse cell seems here largely to carry out the work of yolk formation; while certainly the de-formation process normally is carried out by the egg only.
All observers agree that the outermost layer of yolk in any egg or growing oocyte consists of finely granular yolk. If this were otherwise our general theory of yolk formation would be untenable. Sarasin ('83) was led to the odd idea that the zones of the Lacerta egg were developed outermost first, and the central ones last. I think our demonstration of the nature of this growth in the bird, and the considerations that have followed, will convince that Sarasin's view is untenable.
Yolk spherules have been seen to grow after the egg leaves the ovary by Agassiz ('59), Van der Stricht ('07) and others. This growth is quite surely due to the spherule taking up by osmosis water-particles from the albumen or other fluid encountered by the egg; such growth is not of the nature we have described, though neither of the above mentioned authors has made the distinction.
In regard to the conclusion of many cytologists that yolk arises from the egg-nucleus, and of still others that it arises from the follicular cell nuclei, or from these cells in toto, I may append the following to show that we can exclude all of these as inadequate in the case of the yellow yolk of the fowl's egg. I have calculated that during the last day that a hen's ovum remains in the ovary it may deposit more than 5000 cubic mm. of yolk! Evidently too much work for an egg-nucleus. Again, since the radius of such an egg is increased by 2.0 mm. per day, this means that if yolk formation be a function of the follicular cell, each such cell must here produce daily a column of yolk 2.0 mm. long and of the diameter of the cell ; that is to say each such cell must form more than 50 times its volume of yolk per day, or more than twice its volume per hour ! Evidently too much vicarious labor for a cell.
It appears then that an exclusive origin of yolk from the nucleus, or within the follicular cells is impossible in the birds. The quantities of yolk laid down daily are amounts compatible with substances undergoing physical translocation by osmosis, solution.
WHITE AND YELLOW YOLK OF OVA 479
etc., but not compatible with the probable rate of organic synthesis in the restricted regions of either the nucleus or follicular cell. It is of course necessary for all of the material entering into yolkformation to pass through or between the follicular cells; but each particle of this material may have, by undergoing the synthesis in situ in the egg, twenty-four hours or longer to accomplish this; whereas we have seen that if it originated within the follicle each cell would there have to organize completely its own volume of yolk material and empty itself of this more or less solid material at least once in each twenty or thirty minutes of the day.
Since such theories of yolk formation as have been proposed are now shown to be inadequate in a case where a test can be applied, and since it seems clear that the mechanism of yolk building which we have here outlined and described is necessarily present wherever and whenever yolk is formed, there is at present no valid reason for believing that any dissimilar method of yolk formation exists.
In a certain sense, no general theory of yolk formation has as yet been stated. That is to say, no outline of the processes involved in yolk-building and of the conditions affecting these processes has been attempted, and bur own effort leaves at least important chemical phases of the problem quite untouched. Previous efforts have been largely devoted to features of the histogenesis of yolk granules, and to the identification of some cell organ as the directive agent of yolk formation. Thus such cell structures as centrosome, nucleus, chromatin, nucleolus, mitochondria, yolk-nucleus, etc., have each been several times proposed as the seat or source of yolk. Whilst for some eggs, particularly those in which all of the yolk plainly could not have so circumscribed an origin, the seat or source of such yolk was centered upon a similar structure of the follicular cell; yolk particles have sometimes been described as arising in such cell and later making their way through the follicle cell membrane, vitelline, or other egg membranes, into the periphery of the egg. But theory usually has extended only to the matter of the source of yolk, to the relation between the white and yellow granules, or to the designation of one or another cell-organ as the directive agent of yolk forma
480 OSCAR RIDDLE
tion. There has been no theory to cover the long series of points involved, some of which are the following: WTiat are the conditions which permit yolk to form? In what situations and about what structures does it form (this point much studied and discussed)? What are the processes involved, — what is the mechanism of yolk formation? How are the different forms of yolk genetically and chemically related? How account for the variable amount, distribution, and stratification of yolk?
The statements concerning cell-organs as directive agents of yolk-building have often been quite misleading. This could hardly be otherwise since we have had here attempts to 'explain' a process, not in terms of other processes, but in terms of structure — an error not uncommon even in modern biology. One gets the idea from some descriptions of yolk formation that the nucleus is the absolute, immediate and ultimate source of yolk; and this in spite of the fact that yolk is never present within the nucleus, but onh' outside of it. Just how a vanishingly small fragment of chromatin, thrust from nucleus into cytoplasm — i.e., into an environment so new as to imperil its own existence, — may guide and direct the very rapid production of a thousand times its own volume of yolk (a new and very different substance from itself) we have not been told. Much apparently has been left to the imagination of the reader who is evidently expected to bridge for himself the gap that exists between the chromatin particle in situ and the yolk building process in operatio. But, the high regard which some adherents of this theory have for the kingly chromatin evidently persuades them that chromatin particles — which certainly are thrown from the nucleus into the cytoplasm, and about which traces of yolk certainly are sometimes found — comprise material of such superior quality that the base and foreign matter which meets their Midas-like touch must turn at once into golden yolk! By other workers mitochondria, and still other structures, have been similarly endowed with what would seem to be wonderful and transforming power. The writer would not undervalue the great amount of very valuable work that has led to the determination of the cell-elements about which yolk forms. But it seems to him that much less valuable than this painstaking work
WHITE AND YELLOW YOLK OF OVA 481
are the inferences that have too often accompanied it to the effect that the structure about which yolk is found to form, is itself the active agent in the yolk formation. From the facts already brought forward we see that whatever the out-wandering chromatin particle — the invisible id or biophor — may be able to accomphsh in directing the course of differentiation in the highly complex Uving cytoplasm, the building of a single inert yolk granule by a plainly visible amount of chromatin is a task which clearly quite surpasses it! At any rate a task which it does not accomplish.
Yolk formation as it is indicated by the facts presented in this paper may be connectedly outlined as follows: Yolk will be formed (1) when conditions are such in the egg, follicular cell, food-supply, or organism that excess of food may enter the egg; but (2) in those regions only where some excess of food fat (and protein) can exist without undergoing oxidation; (3) the maintenance of such excess of food is dependent upon the amount of food, or upon marked fluctuations in the amount of food outside the egg, and (4) upon the distribution coefficient of the elements of yolk in the substance inside and immediately outside of the egg, and doubtless by other undetermined conditions within the egg; (5) the actual and active processes of yolk increase or decrease are essentially identified with the partially known synthetic and analytic, — i.e., reversible-action of the enzymes which act upon the constituents of yolk; (6) in the first stages of the growth of a (white yolk) yolk spherule the proportion of fats and phosphatids in its composition is small; (7) in later and more complete (yellow yolk) stages the proportion of these constituents is large.
In my opinion what we now most need to know is how those conditions arise which permit yolk-building to begin. We need further knowledge on*points (1) and (2) of the above. That is, we need to know why an unusual amount of food enters the egg at this particular time in its history. At present we do not know whether such cause lies inside or outside the egg. Again, what is the source of those reduction centers where foods which yield energy so easily as do the fats may not undergo oxidation but be built unchanged into yolk? It is possible, of course that nucleus,
482 OSCAR RIDDLE
chromatin, mitochondria, or centrosome, etc., of the egg, maj later be shown to have special causal significance with regard to such changes in amount of food-intake, or with the production of reducing centers, which we now recognize as basic and unknown features of the conditions which primarily initiate yolk formation. If so, then such cell-structure will have been shown to bear indirect causal relation to a result which it was formerly credited with 'causing' directly ; the test of this lies with the future. But, some at least, of the more direct and immediate features of yolkbuilding are quite certainly those which have been described in these pages.
SUMMARY
1. A method of measuring the rate of growth of large, rapidly growing ova has been found. It consists in feeding, at known intervals, the fat stain Sudan III to animals developing such ova.
2. Ova of the common fowl smaller than 6 mm. in diameter grow extremely slowly as compared with ova of larger size.
3. The time interval between the beginning of rapid growth of the 6 mm. egg, and the breaking of the egg from the ovarian follicle (ovulation) is normally between five and eight days. In most cases it is either six or seven davs.
4. The radii of ova which are larger than 6 mm. usually increase nearly 2 mm. during each twenty-four hours.
5. The thickness of a layer of white yolk together with an adjacent layer of yellow yolk is nearly 2 mm.
6. A pair of such yolk layers is therefore produced during each twenty-four hours.
7. We conclude that the layer of white yolk in the hen's egg is laid down during poorer nutritive conditions obtaining in the later hours of the night (1-5 a.m.) and that the yellow yolk is deposited during the better growth conditions of the rest of the day.
8. Reasons are found for beheving that white yolk wherever found is but a stage in the formation, or the de-formation, of yellow yolk. That it remains as the final form of yolk, only where it is slowly grown or is halted by sub-optimal growth conditions.
WHITE AND YELLOW YOLK OF OVA 483
Yellow yolk, on the other hand, probably indicates, wherever it is found in ova, rapid growth under better nutritive conditions.
9. The presence of alternating layers or zones of yolk (Schichtung) in the ova of some animals thus receives an explanation. A period of poor nutrition corresponds to each of such zones of white yolk; a period of better nutrition to each layer of yellow yolk.
10. The time of formation of a pair of such zones is known in the birds to be one day; in the turtle and myxinoid perhaps a year; in the skate possibly nearly a month; in the lizard this is quite unknown.
11. This 'Schichtung' of the yolk, and other peculiarities of yolk distribution, have produced great similarity in the gross morphology of eggs of widely separated forms, e. g., selachian and bird ; amphibian and marsupial. We can be confident that such similarities do not depend upon heredity in a strict sense, but upon the fact that these eggs have developed under like conditions.
12. The gross chemical composition of white and yellow yolk, and of yellow yolk undergoing de-formation or digestion (a) by the embryo and (6) by the follicular cells, have been determined and comparisons made.
13. White yolk contains much more water, proteid, and extractives, and much less fat and phosphatid than does yellow yolk.
14. When yellow yolk is digested, in either of the two situations named, its constituent parts are not digested, utilized, and absorbed at a uniform rate; but in such a way that the composition of what remains approaches the gross normal composition of white yolk. In such digestion fat and phosphatid are broken down more rapidly than is protein.
15. The immediate mechanism of yolk formation and of yolkde-formation are the same. Chiefly involved are two factors — not previously applied here — which we recognize as (a) the reversible action of enzymes, and {h) the partition coefficients of the several constituents of yolk.
16. The presence of the native lipochrome coloring matter — vitello-lutein — in the large spherules of yellow yolk only, is
484 OSCAE RIDDLE
probably due to the fact that these spherules contain much fat, and the lipochrome pigment is soluble in fat and fat solvents only.
17. The origin of the yolk of the fowl's egg from the nucleus of this cell, or from the nuclei of the follicular cells, is shown to be impossible. It is not probable that the essential features of yolk synthesis in any egg resides in either of these alleged sources.
18. An attempt is made to outline the processes involved in yolk formation.
LITERATURE CITED
Agassiz, L. and Clark, H. J. 1857 Contributions to the natural history of the United States, vol. 2.
Caldwell, W. H. 1887 On the embryology of monotremata and nmrsupialia. Phil. Trans. Roy. Soc, vol. 178.
Dean, B. 1899 On the embryology of Bdellostoma stouti: Festschrift f. V. Kuppfer.
Henriques, V and Hansen, C. 1903 Ueber den Uebergang des Nahrungsfettes in das Hiihnerei, und uber die Fettsaure des Lecithins. Skand. Arch, f. Physiologic, vol. 14.
Herfort, K. V. 1900 Der Reifung und Befruchtung des Eies Petromyzon fiuviatilis. Arch. f. Anat. u. Entwick., vol. 57.
His, W. 1900 Lecithoblast und Angioblast der Wirbeltiere. Histogenetische Studien. Abhdl. der math-phys. Klasse d. konigl. sachs. Gesellsch. d. Wiss. Leipzig.
Kastle J. H. and Loevenhart, A. S. 1900 On lipase, the fat-splitting enzyme, and the reversibility of its action. Amer. Chem. Jour., vol. 24.
Liebermann, L. 1888 Embryochemische Untersuchungen. Pflugers Archiv, vol. 43.
Miescher, F. 1897 Histochemische, physiologische Arbeiten. vol. 1, Leipzig.
Moore, B and Parker, W. H. 1901 On the functions of bile as a solvent. Proc. Roy. Soc. Lond., vol. 68.
Munson, J. p. 1904 Researches on the oogenesis of the tortoise, Clemmys marmorata. Amer. Jour. Anat., vol. 3.
Parke, J. L. 1867 Ueber die chemische Constitution des Eidotters. Med. -chem. Untersuchungen f. Hoppe-Seyler, heft. 2.
WHITE AND YELLOW YOLK OF OVA 485
Riddle, O. 1907 The rate of growth of the egg-yolk of the chick, and the significance of white and yellow yolk in vertebrate ova. Paper before Amer. Soc. Zool., Chicago. Abstract in Science N. S. vol. 27, 1908, p. 945.
1908 The genesis of fault-bars in feathers and the cause of alternation of light and dark fundamental bars. Biol. Bull., vol. 14.
1909 The rate of digestion in cold-blooded vertebrates: the influence of season and temperature. Amer. Jour. Physiol., vol. 24.
1910 Studies with Sudan III in metabolism and inheritance. Jour. Exp. Zool., vol. 8.
RtJcKERT, J. 1899 Die erste Entwickelung des Eies der Elasmobranchier. Festsch. f. v. Kuppfer.
Sarasin, C. F. 1883 Reifung und Furchung der Reptilieneier. Arb. aus d. zool. Inst. Wiirzburg, vol. 6.
Sarasin, P. und C. F. 1887 Zur Entwick. und Anat. d. Ichthyophis glutinosa. Ergeb. naturw. Forsch. auf Ceylon, vol. 2, Wiesbaden.
Van der Stricht, O. 1907 La vitellogenese et la deutoplasma de I'oeuf de chauvesouris. Comptes rendus de 1' Assoc. Anat. Lille.
Virchow, H. 1891 Der Dottersack des Hiihnes. Festsch. R. Virchow., vol. 1.
Wohlgemuth, J. 1905 Ueber den Sitz der Ferment in HUhrerei. Zeitsch. f. physiol. Chem., vol. 44.
PLATE 1
EXPLANATION OF FIGURES
All figures natural size. 1-6 and la-6a represent series of eggs grown simultaneously in two Sudan-fed hens. About 20 milligrams of Sudan fed to each hen at 2 P.M., January 27, and at 10 a.m., January 30 (sixty-eight hours).
The bird bearing series la-6a killed February 2, 10 a.m. (70 hours after last Sudan began to be deposited in yolk).
1 Egg laid January 27, with pear-shaped, more solid 'waxy' interior; also two prominent circles of 'modified yolk' near periphery.
2 Egg laid January 29. The outer border line here represents Sudan deposited from feeding of January 27. This layer was 1 mm. in thickness. The two circles of 'modified yolk' showing here as in figs. 1, 3 and 4. The size of each of the yolks at the time of the modification is indicated by these circles.
3 Egg laid January 31; see above.
4 Egg laid February 2. Two layers of Sudan. The time between Sudan feedings was sixty-eight hours; the amount of yolk deposited in this egg during that time was 6.2 mm. = 2. 2. mm. in twenty-four hours. Section nearly in plane of germ.
5 Egg laid February 4. The two layers of Sudan here as in fig. 4, were 6.2 mm. apart = growth of 2.2 mm. in twenty-four hours. Section at right angles to plane of germ.
6 Egg laid February 7. Shows spreading, or dilation, effects in Sudan layers. Apparently the 'spreading' is mostly outwards, though this figure well represents neither the position nor the condition of each layer. This effect noted in eggs that have remained long in ovary, or, as in this case, in laboratory at high temperatures.
la Egg laid January 27. To unaided eye the outer 10 mm. of one side of this egg showed very plainly six pairs of yolk layers = 1.67 mm. each. Interior contained somewhat solid, waxy body 15 X 10 mm.
2a Egg laid January 29. Seven very distinct pairs of layers of white and yellow yolk. The white yolk represented by dotted lines; the yellow yolk by the spaces between these. Probably another layer central to those indicated in figure.
3a Egg 40 X 27 X 28 taken from oviduct (see first statement above). Inner borders of two Sudan layers are 4 mm. apart = 1.41 mm. growth per twentyfour hours. Section through plane of germ.
4a Egg 31 X 26 X 28 from ovary (February 2). Distance between inner borders ofSudanlines = 3.6mm. = 1.3 mm. in twenty-four hours. Section nearly in plane of germ. Figs. 3a and 4a show the spreading or diffusion of Sudan in the region of the germ. Distance between inner border of outer Sudan layer and periphery = 5. mm. ; this growth of seventy hours = growth of 1.7 mm. in twenty-four hours.
5a Egg 24 X 21 X 20, from ovary (February 2). The first feeding of Sudan (January 27) left but faint traces of the dye in this small egg (see small crescent). By mistake the inner border of the thick layer of Sudan of this figure was placed 7 mm. from periphery instead of 5 as it actually was. This 5 mm. = growth of seventy hours = 1.7 mm. growth in twenty-four hours.
6a Egg 15 X 16 X 19, from ovary (February 2). Smallest egg in which trace of Sudan was present; this somewhat diffuse and indicated by dotted circle. From inner border of circle to periphery = 4.8 mm. the growth of seventy hours = 1.64 mm. growth in twenty-four hours.
All eggs are drawn as perfect circles, although, as is indicated above, the boiled yolks quite constantly show three unequal diameters.
486
WHITE AND YELLOW YOLK OF OVA
OSCAR RIDDLl
la
2a
3a
4a
5a
487
6a
PLATE 2
EXPLANATION OF FIGURES
Figures all drawn twice natural size, and then reduced one-sixth.
1 Section at right angles to germ of an egg( 32 X 30 X 28) that showed with schematic clearness its rate of growth. The Sudan deposited in the first (inner) broken line was fed forty-eight hours before a following feeding. Thereafter the feedings of the dye were made at thirty-six hour intervals. A growth of 1.8 mm. per twenty-four hours is indicated for the three intervals of thirty-six hours.
2 Through gei'm of egg, 33 X 29 X 27. Shows well the manner of deposit of dye (and therefore of yolk) in immediate region of germ. The fan-shaped figure of dilute Sudan lying deeply beneath the germ is perhaps however not a correct picture of the original position of the dye, but a 'spreading' effect. The first two feedings of Sudan were here twenty-four hours apart; the next forty-eight hours and the last, twenty-four hours later. A growth of 1.5 mm. per twenty-four hours is here indicated for time between first and last feedings.
3 A series of small ova (4-7.5 mm. in diameter) from the ovary of a laying hen which had been once fed about 25 milligrams of Sudan and killed on following day. Only three of these ova (6.5 to 7.5 mm.) showed any trace of the dye. In the drawing ofthe egg of 7.5mm. the position of the layer of dye was placed two mm. too near center of egg. The number along side each ovum indicates its actual diameter in millimeters. The egg of 7 mm. was of special interest. After lying in a quantity of Mann's balanced formalin-alcohol solution for a few weeks the striated appearance of the outer portion of its white yolk was visible with binocular. The lower right hand figure is an attempt to represent the structure of the outer 1.75 mm. of the 7 mm. egg. Eight striae could be here distinguished. Apparently therefore the striae have a thickness of about 0.25 mm. The stain was found to be confined to the large, yellow yolk granules.
4 Egg, 34 X 29 X 31, through plane of germ. Central heavy irregular lines are from Sudan feeding three days apart. Bird was laying at long intervals and next feeding delayed ten days and following this two feedings two days apart. This egg therefore at least seventeen days in developing. But its abnormality is evidenced by the crumpled appearance of the innermost layers of Sudan (the corners of innermost layer are too sharp in drawing) and by a curious depression of the germ. Another peculiarity of this egg was its presentation of a brightly Sudancolored vegetative pole (the stratum of dye here near surface), and a normally colored animal pole (except for the small depressed region of germ).
5 Later egg, sister to no. 4. Much more rapid growth than in 4. First two feedings thirty-six hours apart; others twenty-four. A growth of 1.3 mm. per twentyfour hours is here indicated.
488
Vv'HlTE AND YELLOW YOLK OF OVA
OSCAR RIDDLE
6§
o
Gi
»? >
©
7h
.jargre
yolk gran. stain
-clefts
•white yolk
489
PLATE 3
EXPLANATION OF FIGURES
The distribution of white and yellow yolk in the ova of vertebrata with special reference to zonal or lamellar formation.
1 Egg of tortoise, Clemmys (from Munson, '04), c.c. = cytocenter (centrosphere); y.c. = inner cytocoel of Munson, and what I should call inner or first layer of white yolk; i.y.l. = inner yolk layer of Munson, and inner or first yellow yolk layer I have called it: o.cy.c = outer cytocoel or second layer of white yolk; 0. y. I. = outer yolk layer, or second layer of yellow yolk; s. c. I. = subcuticular layer.
2 Mature egg of Petromyzon (from Herfort, '00). Very small granules in the external ooplasm, gradually merging into the large granules and large vacuoles of internal ooplasm.
3 Nearly grown egg of Phascolarctus (marsupial) (from Caldwell, '87). The darker crescentic body is coarsely granular yellow yolk; the clear area around the nucleus, which is also continued around the periphery of the entire egg is of finely granular white yolk.
4 One end of large (3 cm. long) egg of Bdellostoma (cyclostome) to show stratification of its yolk (from Dean, '99). The fine curved lines represent points richest in minute yolk-spherules (white yolk).
5 Mature egg of Ichthyophis glutinosa (amphibia) (from the Sarasins, '87) 6X9 mm. with central 'latebra' of white yolk; this connects above with the germinal vesicle, forming a nucleus of Pander beneath the latter.
6 Mature egg of Torpedo (from Riickert '99) in meridional section. The lenslike germ above. A central 'latebra' without stratification (Riickert says this is composed of dark, not light, substance). The dark layers are composed of loosely bound, but larger yolk platelets (white yolk?) ; the wider lighter strata of more closely packed but somewhat larger platelets (yellow yolk?).
7 Hen's egg photographed to show something of the concentric deposition of Sudan III. Dark lines = Sudan; these bright orange-red in original. The appearance here is very similar to the always less evident stratification of white and yellow yolk; the^narrow lines of Sudan in the photograph simulating the faint and narrow lines of the white yolk.
8 Part of immature egg of Lacerta (from Sarasin, '83) showing well-marked layers of white and yellow yolk (I infer that the dark lines represent white yolk) ; about one-fourth of egg is here shown. The germinal vesicle lies just outside the figure, above and to the right; all layers are seen to converge toward it, and to become gradually modified in its vicinity.
490
WHITE AND YELLOW YOLK OF OVA
OSCAR UIDDI.K
mm
,('■1(1 '^
/ ■ \
^l^>:
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2
491
SOME PROBLEMS OF COELENTERATE ONTOGENY
CHARLES W. HARGITT
From the Zoological Laboratory, Syracuse University
THREE PLATES AND THREE TEXT FIGURES
CONTENTS
Introduction 494
Material and methods 495
Observations 497
A Pennaria 497
1 Cleavage 499
2 Nuclear aspects 499
3 Amitosis 501
B Hydractinia echinata Flem 502
1 Cleavage 503
2 Ectosarcal features 508
3 Early embryo, moi'ula 508
4 Organization of the embryo 509
5 Entoderm formation 509
6 The larva, planula 510
C Clava leptostyla Ag 511
1 Maturation 511
2 Nuclear behavior 512
3 The chromatin 513
4 Nucleolar behavior 516
5 Later development 517
6 The morula 518
7 The germ layers 518
a. Ectoderm 519
b. Entoderm 520
Review and discussion 523
1 Origin and growth of germ-cells 524
2 Doctrines of homology 529
a The germ layers 531
b The planula 531
c The morula 532
d The blastocoel 534
e Cleavage homology 535
3 Amitosis 537
Summary 541
Bibliography 542
JOURNAL OF MOHPHOLOGT, VOL. 22, NO. 3 SEPTEMBER, 1911
493
494 CHARLES W. HARGITT
INTRODUCTION
In the course of investigations carried on by the writer during several years, certain facts have come to light which seem to have important bearings upon several problems of general ontogeny. In various papers phases of these have been suggested, but only incidentally has any attempt been made to discuss their significance or their probable correlations as developmental phenomena. With further investigations still additional facts have been observed, and similar investigations by others have tended to convince me of their importance in a still larger degree. When the honor to cooperate in the preparation of this memorial volume was submitted, it seemed that no more appropriate subject came within the scope of the writer's researches than that involved or implied in the above caption.
My introduction to coelenterate morphology began many years ago with the problem of the origin of sex-cells, a subject at that time brilliantly exploited by Weismann, whose Entstehung der Sexualzellen bei den Hydromedusen, Zugleich ein Beitrag zur Kentniss des Baues und der Lebenserscheinungen" ('83), has long been a recognized classic in its line. It was ably supplemented by the hardly less brilliant researches of Metschnikoff ('86), " Embryologische Studien an Medusen. Ein Beitrag zur Genealogie der primitiv Organe."
The first contribution to the subject by the writer was a very brief and tentative paper before Section F, of the American Association for the Advancement of Science, in 1889. It was adversely commented upon by one who had accepted without question the then prevalent dogma that Hydrozoa were distinguished from all other Cnidaria by the origin of the sex-cells exclusively from the ectoderm. Under this adverse criticism no further utterance was made on the subject for several years, though there was no lapse of interest or investigation.
In the meantime, an observer here and there had dared to question the conclusiveness of the earlier dogma. Little by little facts were accumulating which cast further doubts upon the matter, and even compelled the conclusion that Weismann 's
SOME PROBLEMS OF COELENTERATE ONTOGENY 495
fundamental contention was inconclusive. Results to be cited from various sources will tend to show that the early attempt to formulate a general theory of embryogeny on the basis of the origin of sex-cells was no less defective and inadequate than it was hasty. For some time past phases of my researches have forced the impression, which has deepened as the investigations have extended, that not a few of the earlier views as to coelenterate ontogeny were seriously defective, or absolutely in error at many points. Certain of these T have taken occasion to point out from time to time, as occasion arose. The purpose of the present paper is two-fold: First, to submit accounts of the development of several species of Hydromedusae which have been under investigation for some time; and secondly, to point out certain errors as to the ontogeny of the groups which, from various reasons, had become associated therewith.
MATERIAL AND METHODS
1. Material. The material upon which the results herein described are based (with the exception of that of Pennaria australis, for which I am indebted to Mr. Edgar J. Bradley, of Australia, to whom my thanks are hereby acknowledged) was collected by the writer at various times within the past two years, and chiefly in the immediate vicinity of Woods Hole, though some of that of Clava was collected at Harpswell, Maine. It is a pleasure to express my thanks to the directors of these laboratories for various courtesies.
Attention will be given primarily to two species of Pennaria, and to a single species each of Clava and Hydractinia. Other species will be given attention in relation to the several problems with which the paper has to do.
2. Fixation. In my earlier work great difficulty was encountered in reference to killing and fixing reagents. For killing my first lots of eggs of Pennaria picro-nitric and picro-sulphuric solutions, then much in vogue, were used; but to my sorrow these were found to be almost worthless. This was particularly the case with picro-sulphuric. Almost the whole of one summer's
496 CHARLES W. HARGITT
collection was absolutely worthless by reason of the almost exclusive use of this reagent.
Hermann's and Flemming's solutions afforded fairly good fixation, but subsequent staining was very difficult. Perenyi's solution was absolutely worthless with both Eudendrium and Pennaria material and has since been discarded. The only solution which gave reasonably good and fairly constant results was a strong solution of corrosive sublimate to which had been added 5 per cent of glacial acetic acid.
In later work I made use of various solutions of formaldehyde, but with only fair results. A 10 per cent solution in sea-water gave a good general fixation for immediate use. Combination with corrosive did not seem materially to better it. There was found also to be great variability in different species as to this matter. This was particularly apparent in eggs heavily yolk laden as compared with those in which yolk was lacking, or present in only small quantities. There was also great difference in later differentiating other cytoplasmic elements. For example, in the peculiar proteid granules present in eggs of Clava the first, and only satisfactory reagent was picro-acetic acid (p. 217, Biol. Bull, vol. 10, '06).
In 1906 my attention was directed to Bouin's picro-aceticformol. It was thoroughlj^ tested upon eggs of Pennaria and Hydractinia, and was found to be far superior to any thus far employed. I have since used Zenker's fluid with good results in fixation of eggs of several species. It is worth while to emphasize the importance of this feature of fixation, especially as it relates to coelenterate material. I have called attention to this in several previous papers, but it is absolutely imperative in order to warrant trustworthy results that particular attention be given to this matter.
3. Imbedding. In another respect I have learned to my cost the importance of prompt working up of coelenterate material after fixation. Attention was directed to this point in my paper on Pennaria ('04b, p. 455). This precaution has been abundantly confirmed by later experience, and I take occasion here to emphasize its importance once more. The value of this has been
SOME PROBLEMS OF COELENTERATE ONTOGENY 497
vouched for by Smallwood ('09). My present method in this particular is to imbed the material in paraffine as early as possible after reasonable time has been given for proper hardening and dehydration. This imbedded preservation may apparently be indefinitely prolonged without detriment. But in my experience it is impossible to preserve material of this group for any considerable period in alcohol without having it suffer considerable deterioration. This is particularly the case with those cytologic factors of mitosis and allied features so important in modern problems of embryology.
4. Staining. This, like the matter of killing and preservation is one of much importance and of varying grades of difficulty, as it related to the problem under review. As in the preceding, I had long since called attention to the extreme difficulty in the staining reactions of coelenterate material. This was most marked, in my experience, in the eggs of Eudendrium and Pennaria. Others have also found similar difficulties with this phase of technique. G. T. Hargitt ('09, p. 163) has recently devoted some attention to the subject, and my own results have been confirmed by those described in his paper.
Difficulties experienced in my earlier work in Pennaria, and the later work on Clava, were such as to leave doubt, particularly in relation to the phenomena of maturation, leading me to conclusions, tentatively adduced, which subsequent work has not confirmed, as shown by G. T. Hargitt (op. cit.) and Smallwood ('09), and by facts herein described.
OBSERVATIONS
A . Pennaria
Except for additional facts which have come to light in relation to a species of Pennaria, the development of which has been hitherto unknown, no particular attention would be given to the subject in this connection. Since the issue of my detailed paper on the early development of Pennaria tiarella ('04), repeated observations on the living eggs have confirmed my previous
498 CHARLES W. HAEGITT
results in every detail, so far as the general facts are concerned. I think it may now be regarded as beyond doubt or cavil that these results, anomalous as they may appear, are absolutely normal and conclusive. Furthermore, when analogous cases to which I had directed attention, and others to be cited in a later connection, are taken into consideration, it seems rather strange that "early cleavage differing widely from what we have come to think as typical" should be given as adequate grounds for a reexamination of the case! However, when it is recalled that, with certain investigators, it is more important to reduce vital phenomena to a set of formulae, or to corral all development within a common law than to recognize facts as they are, the wonder is less strange than it might at first seem! But additional facts are now availaible from a most unexpected source, and of such character as to remove Siuy further grounds for question or doubt. Somewhat over a year ago I had the good fortune to receive from Mr. Edgar J. Bradley, of Adelaide, Australia, a collection of hydroids, and along with them several colonies of Pennaria australis Bale, together with the medusse and eggs, which had been taken in tow-nets just at the height of the breeding season. The only feature of regret as to the eggs is that they had not been preserved in other than weak formalin, in order to have made them available for cytological study. But, as it is, they show in surface study the external aspects of developmental behavior to such perfection as to leave little to be desired. Figures 5 to 8 are sketches of a few of these stages, which speak for themselves. As will be seen at a glance, they duplicate in a most striking way similar stages in the development of Pennaria tiarella. If one were to pass under review separate series of eggs of the two species, without pains to have critically determined them in advance, it would be practicalh' impossible to say which belonged to the one species and which to the other. There are the same ectosarcal features, — papillae, bridges, strands, etc., in both; the same bizarre, amoeboid characters, the same anomalous phases of cleavage, 'every egg a law unto itself, and finallj^ the same end resultant, a normal embryo. Later phases of development of the Australian species were not present, hence further compari
SOME PROBLEMS OF COELENTERATE ONTOGENY 499
son was impossible, though there is no reason for doubt as to its subsequent similitude and results. A comparison with figures 1 to 4, of Pennaria tiarella, will make this more evident.
The fact that these eggs had been taken with the tow-net in the open harbor, and had been preserved shortly after in formalin, leaves no grounds for serious question as to their normal condition, and confirms completely the results of my own precautions ('04b, p. 474), to guard against possible effects of artificial conditions of the laboratorj^ These additional facts, together with others of like character which have since come to our knowledge, especially those described by Brooks and Rittenhouse ('07) must suffice once for all to establish the perfectly natural and normal phenomena of extremely erratic and indeterminate modes of cleavage and consequent organogeny.
1. Cleavage. There is nothing new to add concerning the cleavage features of the eggs of Pennaria tiarella. Concerning this feature in Pennaria australis little attempt will be made to give detailed descriptions. The figures cited will afford all that is necessar}'^ as to the general surface aspects. As already stated, there is such essential conformity in every respect to the corresponding stages in Pennaria tiarella that there seems small occasion to do more than refer to the figures and descriptions of the former paper ('04b). While the fixation does not give material fit for cytologic details, it is fairly good for general comparisons. Eggs carefully stained and cleared show fairlj^ well the general internal conditions, and here, as in the surface features, there is essential likeness to corresponding stages in Pennaria tiarella.
2. Nuclear aspects. Brief reference may be made to a fesv points under this head.
Fragmentation. In several of my earlier papers ('04b, pp. 460-1), attention was called to certain nuclear phenomena of a rather peculiar character. Among these was what seems to be a rather promiscuous dissolution, or disintegration of the nucleus and the dispersion of the greater portion of it into the cj^toplasm. To designate this process I used the term fragmentation, long previously employed to designate phases associated with direct nuclear division, and apparently first employed bj^ Van Beneden (Wilson, the Cell. p. 64).
500 CHARLES W. HARGITT
In recent papers both Smallwood ('09) and G. T. Hargitt ('09, pp. 197-8), have expressed doubt as to the process in the eggs of Pennaria, the latter stating that "no sign of its fragmentation has ever been seen." But in the following sentence he adds, the supposed disappearance of the germinative vesicle at this time, I believe to be due simply to the usual dissolution of the nuclear membrane and the mingling of karyoplasm with cytoplasm."
Smallwood expresses similar doubt, saying:
If by fragmentation of the nucleus is meant that the entire nucleus disappears and its contents disperse throughout the cytoplasm, then I find no evidence of such a process in these hydroids. But what shall be said of the chromatin changes before maturation in Hydractinia and in Pennaria after maturation, where large quantities of chromatin migrate into the cytoplasm? (Op. cit., p. 228.)
It was chiefly in this latter sense that I had used the term, and observations of Coe, Lillie, and others were cited in support of facts found in Pennaria. It may also be admitted that there seemed to be evidences of the entire dispersal of nuclear substance through the cytoplasm and their subsequent reorganization into new nuclei. ('06, p. 227, etc.). Further reference to this will be made in another section.
Contention for fragmentation was based almost wholly on chromatin behavior. The facts which I urged in this connection were those involving, first, the enormous dissipation of chromatin and its absorption by the cytoplasm, during the phases of maturation; and secondly, the achromophilous condition of the chromatin at a slightly earlier time. These facts have not been disputed. Whether my inferences or interpretations are valid is quite another matter. As to that upon which I have laid most stress, viz., the disintegration and dispersal of a preponderating portion of the chromatin, certainly not less than 90 per cent in many cases, and that it has little or no subsequent function as chromatin, — I am still firmly convinced of its validity and of the vast significance it involves as to chromosome theory.
Concerning the achromophilous condition above referred to, I have little to add to my previous accounts. G. T. Hargitt
SOME PROBLEMS OF COELENTERATE ONTOGENY 501
('09, p. 165), whose detailed experiments on differential staining have surpassed my own, was perplexed as to this condition. At the end of the growth period, the nuclear reticulum shows so little affinity for basic stains that there appears to be, so far as this test shows, no chromatin present in the entire nucleus. I can suggest no explanation for this peculiar condition of the chromatin at this period, but it is normal and characteristic of this stage." I am now convinced that there is here a chief ground for my failure to distinguish certain phases of maturation, and my subsequent error in the assumption of their possible suppression or modification in certain cases.
3. Amitosis. Concerning a further problem, that of amitosis, I am in doubt so far as Pennaria is concerned, even as at the time of my previous work. My chief grounds for this view are the facts first cited, and those of the multivesiculate aspects of the nucleus during cell proliferation. And here again Smallwood and G. T. Hargitt ('09), and later Beckwith ('09), all confirm my basis of facts. They find in these vesiculate nuclear conditions essentially the same results which are described in my accounts. Without exception their interpretations differ from mine. To them these facts are believed to be obscure phases, chiefly telophases, of mitosis. While I freely admit the force of their contentions, there are still good reasons for maintaining the plausibility of my own views and interpretations. This is especially the case concerning Eudendrium. Here there seemed to be clear and positive examples of amitosis, as shown in fig. 23, a and h, plate 15 ('04a). It may not be amiss to state here that all these examples of amitosis occurred in association with those 'nuclear nests' so intimately involved in the syncytial phase of development concerned in entoderm formation. The conditions are somewhat different in Pennaria, yet sufficiently similar to lead one to anticipate similar processes, and these appeared probable in the vesiculate 'nuclear nests' mentioned above. But in no case were there found the specific and positive examples figured in the case of Eudendrium. The same must also be said of Clava. But further discussion of this will be reserved for a later section.
502 CHARLES W. HARGITT
B. Hydractinia echinata Flem.
During the summer of 1907 I was fortunate in securing large numbers of this hydroid in the height of its breeding season, and took occasion to study the development and life history of the species. Some account of the life history has already been given ('08) which obviates any call for emphasis here on this point. The early development was studied from living material during two summers and at the same time material was carefully preserved for cytological study. This latter was turned over to my colleague, Dr. Smallwood, and his results have already appeared ('09). It only remains for me to submit such accounts of my observations as seem important in order to afford a more or less complete and connected description of phases of development, especially when correlated with Smallwood's account referred to above.
There are numerous points of difference between my observations and those of Bunting ('94), some of which may be due to the fact that her studies were restricted to material obtained from the small colonies living upon shells occupied by hermit crabs, while my material was derived chiefly from colonies of enormous size, obtained from piles of docks or similar habitat, but with comparison from the former sorts. As pointed out in the paper referred to above ('08, p. 98), there is no adequate reason for regarding these hydroids as other than a single species, hence any differences to be cited must be incidental rather than fundamental.
One of the first points of difference to be noted is concerning the time at which the liberation of sexual products takes place. According to Bunting this is between the hours of 9:30 and 10:30 p. M. That it occurs during the night I have repeatedly demonstrated. Further, that it may occur in certain cases about the time stated by Miss Bunting, I have also found true. But that it may also occur at a much later hour, and also at varying hours, I have also found to be the case. Some of the best cleavage series obtained, especiallj' for the very early stages, were in the mornings from seven to nine o'clock. That is to say, the eggs had been deposited some time after midnight, and at the hours
SOME PROBLEMS OF COELENTERATE ONTOGENY 503
named were in early stages (two- to eight-cell) of cleavage. This would seem strongly to indicate their deposition at perhaps five or six o'clock in the morning or thereabout, as recorded in my notes of July 11th and 12th. In other cases development had reached the morula stage at nine a. m., which would lead to the conclusion that liberation of sexual products had occurred about midnight. While it is true that in many hydroids the liberation of eggs and sperm occurs at a fairly constant time, yet there are others in which this is not the case, and in which such ripening and discharge is a more or less continuous process during the breeding period.
The character of the egg is much like that of Pennaria, though it is much smaller. Both are alike in general texture of protoplasm, contain yolk, and similar inclusions. There is present a pigment similar to that in the eggs of Pennaria, though less marked in color. Like those of the latter, the eggs are devoid of a definite membrane. They are rather heavy and sink promptly when set free. By reason of this it was practicable to suspend colonies in shallow vessels within wire baskets under docks in freely circulating water and with little liability of their being lost. This was a matter of some importance; for, despite the best precautions, these hydroids soon deteriorate in vitality under the artificial conditions of the laboratory, while by suspending them in open waters about the docks they thrive almost as if in the natural habitat.
jf. Cleavage. So far as I am aware the only definite work on cleavage of Hydractinia is that of Bunting ('94). In this paper we have a characteristically symmetrical portrayal of the process. In general surface aspects it is represented as almost mathematical in its regularity and symmetry.
That the earlier cleavage phases in perhaps a majority of the eggs conform to this in greater or less degree is probably true. But that it represents with any degree of accuracy the average behavior of this phenomenon as a whole none who had carefully followed it could for a moment admit. It has been difficult to conceive how, except by a selective process, any such account could have been formulated. It is quite easy to see that by directing
504 CHARLES W. HARGITT
attention only to eggs which exhibited the regulation aspects of cleavage, and disregarding, as abnormal, those of differing aspects, just such an account might easily have been made up; and this in all probability may have been the method followed.
It is not strange that under prevailing conceptions as to formulated 'laws of cleavage' this method might naturally have been adopted. In the case of Pennaria the present writer deliberately disregarded an entire batch of eggs which were so erratic in behavior as to suggest the probability of pathological conditions. But, by whatever method one may explain the miatter, certain it is that there is a measure of irregularity in a large proportion of the eggs of Hydractinia, especially after the third or fourth cleavage furrows, which at once takes them out of the usual category of geometrical order or s>^llmetry and puts them, if not in the Pennaria class of chaotic irregularity, at least consigns them to the category of the indeterminate and unsymmetrical.
However, it is not my purpose, in thus discrediting an account which gives so inadequate and misleading an impression, to goto a similar extreme in the other direction and convey the impression of predominantly erratic cleavage. On the contrary, let it be noted that in perhaps a majority of the eggs of Hydractinia echinata the cleavage, while seldom exhibiting an approach to geometric order or synrnietrj^, is yet more or less regular and orderly. In such cases cleavage begins, as usual, at the animal pole, cutting vertically downward, and generally divides the egg into symmetrical halves, which adhere to each other by a narrow band, or connective of cytoplasm at the lower pole. The second cleavage likewise may begin at the upper pole and at right angles to the previous division, or may begin at the center and work outward, thus dividing each half into symmetrical fourths, giving a fairly typical four-cell stage. The third cleavage, which is usually equatorial, often begins at the center and extends toward the periphery, a process more or less common in eggs of hydroids. The subsequent phases may continue more or less orderl}^ as in earlier stages, but often grow increasing!}^ irregular and independent, though resulting in a sjTametrical embryo. On the other hand, figs. 14 to 22, which are camera sketches of living eggs,
SOME PROBLEMS OF COELENTEEATE ONTOGENY 505
show liow strikingly irregular and unsymmetrical cleavage may be in eggs of a given lot, developing under identical conditions. But in these cases the first two ot three cleavage furrows are more or less synnnetrical. In not a few, however, cleavage is^ distinctly erratic from the first, the first furrow dividing the egg into very unequal portions. In such cases the irregularity becomes usually increasingly more so as cleavage goes forward.
A very interesting case is that shown in figs. 9 to 13, which occurred at irregular intervals within a period of about forty minutes of constant observation during which the sketches were made. The egg was kept under observation for several hours, or till the morula w^as apparenth^ completely formed. Fig. 9 show^s what may be regarded as a four-cell stage, the central portion comprising the main body of the cell, while at opposite poles are three other blastomeres, in each of which the nuclei were distinctly visible. In fig. 10 the small blastomere at the upper pole has divided, so that now we have a five-cell stage. It remained in this condition about fifteen minutes, when a most curious thing happened, the small blastomeres, x and ?/, being the factors of most interest. At first the blastomere y became detached from its connection with the cell body as shown in the figure, and later the other blastomere x did the same thing, both thus becoming absolutely free, in which condition they continued about thirty minutes. At the end of this time they resumed division and went forward to complete segmentation and formed what seemed to be a perfect, though very small morula, shown in fig. 11, b. The other portion exhibited something of the same tendency. For example, the small blastomere z cut itself free as had y, but it later drew back and, fusing with the cell body, continued as an integral part of the egg in its later development. Figs. 12 and 13 show the general aspects of this portion, which went forward normally and became a perfect morula and later gave origin to a normal planula. The small segment exhibited the same aspects, but later in the day began to show signs of degenerative tendency and finally disintegrated entirely without assuming the larval condition. This can hardly be ascribed to its minute size, for other embryos, which were otherwise apparently similar in every way, suffered the same fate.
506 CHARLES W. HARGITT
Among the anomalous aspects of cleavage which I have encountered in the development of these and other hydroid eggs not the least singular or significant is the occurrence, now and then, of what may be designated as blastomeric autotomy. That is to say, occasionally one finds during the earlier stages of cleavage, most commonly at the first, the complete separation of the primary blastomeres, which continue to develop as independent eggs, and from which independent embrj^os arise, giving origin to two polyps. I have called attention to something of this in Pennaria. The same thing has been found in at least two cases in Hydractinia. In one case actually followed from beginning to end the sequence of events may be briefly described. At the first cleavage of an egg which was in a marked degree unequal, the two blastomeres separated entirely, each part developing quite apart, and in a perfectly independent fashion. One of these segmented in a fairly regular and symmetrical fashion, while the other portion was markedly irregular from the first. It should be observed that the rate of cleavage in the former was much slower than in the latter, which exhibited a marked tendency toward amoeboid aspects as shown in the figures already cited.
It seems perfectly clear, therefore, that we have in these aspects of development a perfectly normal, and . not particularly rare mode of segmentation, involving the origin of two, and probably even three or more embryos from a single egg, in a perfectly natural and spontaneous way.
Among these anomalous aspects, which were numerous as well as various, those shown in text figs. A, B and C will be interesting. In this case the first cleavage w^as about normal, beginning at the animal pole and extending downward to the lower, where the blastomeres remained attached by the connective shown in fig. A for some time. The second cleavage was the not unusual type shown in fig. B where it was directed centrifugally and in a horizontal instead of a vertical plane ; and as it continued the connective was resorbed, leaving the two blastomeres quite free for a time during which they moved into the position shown in the figure, when the vegetal blastomere of the one side came into contact with the animal blastomere of the opposite part, in which posi
SOME PROBLEMS OF COELENTERATE ONTOCxENY
5o;
tion they fused and remained for some time. Fig, C shows the condition when the four-cell stage -had been established. As will be seen, the blastomeres had rotated until they became, as it were, fitted into close contact with each other, and the development of another connective bound them in that relation for some time. It may be noted that later development went forward with average regularity.
2. Ectosarcal features. In my paper on Pennaria ('04b, p.
469) attention was directed to certain very conspicuous aspects
which were designated as 'ectosarcal phenomena,' and which
comprised various more or less superficial excrescences, such as
papillae, films or bridge-like connectives between blastomeres.
508 CHARLES W. HARGITT
etc. They were described in some detail and various suggestions and comparisons submitted as to their significance.
In the eggs of Hydractinia very similar structures were encountered, though less conspicuous and less constant than in Pennaria. Certain of these are shown in figs. 14 to 22. As to their significance or function I have nothing new to offer beyond that previously suggested. Their more obvious function would seem the two-fold one of connecting adjacent blastomeres, and affording coordinating bonds for the entire egg during development. These suggestions could hardly apply to the papillose structures of the surface, and their presence must be regarded as problematical.
3. The early embryo, morula. With the progress of cleavage toward completion the irregularities of surface, due to ectosarcal structures above mentioned and erratic cleavage, which were so conspicuous a feature, gradually disappear and the eaibryo tends to become more or less typically spherical and blastula-like. But in comparativel}^ few Hydrozoa does a typical blastula occur. In my earlier accounts of the development of Eudendrium and Pennaria attention was directed to the presence of a true morula as the embryonic stage resulting from cleavage, and also to the entire absence of a stage comparable to a coeloblastula. This is likewise the condition to be found in Hydractinia and Clava. Rittenhouse ('07) has shown the same to be the case in Turritopsis nutricula. It will be shown in a later connection that this is the dominant type of cleavage embryo throughout the entire phylum. At no time is there to be found in either Hydractinia or Clava a distinctive or permanent cleavage cavity, though there may often be found certain intercellular spaces which are designated as such, but the correctness of which may be seriously challenged. This, again, will be considered in more detail in a later section.
For some time following the apparent completion of cleavage and the estabhshment of the morula there seems to be a period of quiescence. This is such, however, only in appearance; for in reality there exists a condition of active cell proliferation, as may easily be demonstrated by means of sections of embryos at this time. This has been especially demonstrated and emphasized
SOME PROBLEMS OF COELENTEKATE ONTOGENY 509
in the cases of Eudendrium and Pennaria, but is no less true in the present instances.
4. Organization of the embryo. This has usually been assumed to consist fundamentallj^ in the formation of the tissues, ectoderm and entoderm. In part this assumption is correct, but only in part. For example, the morula may remain for some time entirely devoid of these tissues in any definitive sense, and even in the later larval stage the entoderm may not arise till .a late period. As has long been known, among the first evidences of organization is that associated with the formation of the ectoderm. Indeed, this is only what might naturally be expected as one comes to consider the primary function of such a tissue, or its analogue, throughout the animal kingdom. The embryo, no less than the adult organism, requires superficial protection against external conditions. And from protozoon to mammal provision is made to this end by ectosarc and epidermis, and in the embryo by the ectoderm, which may be regarded as the primary tissue of the embryo.
5. Entoderin formation. But up to this time there is no definite differentiation of entoderm. It is true, that one will find what has long been designated as entoderm, namely, an interior mass of embryonic matter more or less cellular, but without differentiation of any sort. By some students of hydroid development this condition has been described as the ' end of entoderm formation' (Ende der Entodermbildung). In reality one may better designate it as the beginning of entoderm formation, though even this might be open to question. What we have at this time is simply an interior embryonic mass, often a syncytium, within the enclosing ectoderm, if this be yet differentiated; and of this mass but a very small proportion ever participates directly in entoderm formation. For the sake of clearness it seems desirable formally to recognize this condition by giving to it such name as may express the fact, and at present no better term seems available than 'pro-entoderm.' This only implies the existence at this time of material, a primordium, out of which in varying ways will be developed the definitive entoderm of the larva.
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 3
510 CHARLES W. HARGITT
6. The larva, planula. The life history of the morula is comparatively brief, perhaps from six to eight or ten hours, the period varying considerably. During this time the definitive ectoderm has been established, cilia developed and the free-swimming larva, the planula, begins its career. Concerning the structure of this organism there is no occasion for special details. It differs little if at all from that characteristic of others whose structure has- been repeatedly portrayed, and is too well known to need further account. In the present instance, as in those of numerous others, at the time of the assumption of this condition the larva is still a solid mass, with little organization beyond the above mentioned ectodermal differentiation. A definitive entoderm may not become established till relatively late in larval life, as I have repeatedly pointed out in other cases, and only after a process of physiological differentiation, as shown in a later section. The first evidence of a coelenteron appears as a slit-like cavity in the larval axis, which later enlarges as the reduction and absorption of the pro-entodermic mass proceeds. Finally, by such graduated method does the entoderm become established. At no time is there a mouth or other means of communication with the outside during phases of embryonic or larval history. Planulae of Hj^dractinia have been frequently reared under artificial conditions, and readil}^ transform into the final, or polyp state. Soon after the larva attaches itself the mouth is established by a terminal opening which arises by a rupture and rearrangement of the adjacent cells. Tentacles arise in the usual manner, first three in number, followed shortly by three others at intermediate points, and slightly below the first series. At the base of the polyp there arise root-like stolons, two or more in number, which mark the origin of the hydrorhizal network characteristic of the species.
C. Clava leptostyla Ag.
In connection with the work on Hydractinia I have taken occasion to re-examine the material upon which was based the work embodied in m}'- previous paper on Clava ('06), and have also carefully studied sections of new material which had been fixed in
SOME PROBLEMS OF COELENTERATE ONTOGENY 511
Bouin's piero-formol, and in Zenker's solution, and carefully stained by several of the most recent methods. So far as it relates to the organization of the egg or its cleavage no occasion has been found for modifying in any essential the earlier conclusions. These I believe to be confirmed in every detail, and lead me to reafhrm the former account. Concerning some few points in relation to the phenomena of maturation and nuclear behavior, including phases of germ-layer formation not touched upon in the previous paper, it is necessary to reconsider and add to the former results.
1. Maturation. Concerning the phenomena associated with maturation my observations will be very brief. In the former paper the general facts were explicitly stated and no occasion has been found to call for essential modification. Both in living eggs and in sections of stained material polar bodies were found and described. In connection with earlier accounts of this feature in other eggs of hydroids one may find such expressions as "About this time the nucleus becomes indistinct and finally disappears;" the nucleus "fades from view when the ovum is deposited." These accounts relate almost wholly to observations upon living eggs, and 1 have repeatedly verified them both in the living, and in sections of fixed eggs. WTiile in themselves such accounts ma}'seem to have little of distinctive value, in a morphologic sense, yet, as expressive of physiological conditions they seem to me to have ver}' large significance. In the first place, these observations described what was actually seen and its fidelity to fact can not be questioned. In the next place, cytological study of fixed material confirms just these accounts. As eggs grow toward maturity the germinal vesicle is large and conspicuous. But as they approach the phase of metabohsm involved in maturation a marked change occurs, as is well known. The chromatin network, which has been conspicuous, gradually disappears, and in many cases loses absolutely its affinity for stains. With dissolution of the nuclear membrane, a still further change occurs, which is exactly what these accounts describe, namely, the mingling and fusion of nucleus and cytoplasm to such a degree that it is often difficult to differentiate them by any of the usual methods.
512 CHARLES W. HARGITT
It is just at this time that the maturation phenomena are in process of development. In my previous account some doubt was expressed as to the presence of mitosis. Critical study of fresh material has enabled me definitely to confirm the facts of maturation mitoses attested by Smallwood ('09) in Hydractinia, and Beckwith ('09), in Clava. A critical review of my earlier material only went to confirm the previous doubt; all of which but tends to resolve the case to one of technique. In the newer material both maturation mitoses were distinguishable without serious difficulty.
2. Nuclear behavior. In addition to the foregoing discussion some f\irther reference to points of nuclear behavior seems desirable. In several of my earlier accounts attention was directed to the migration of the nucleus to the distal periphery of the egg as it approached maturity. As is well known, many earlier students of nuclear physiology have sought to correlate directly the nucleus with nutritive' functions during the growth period, and its location has been said to conform to this conception, and numerous citations made to facts recorded in phyla above protozoa. So far as the Hydrozoa are concerned I am not able to confirm this view. In the growing oocyte of Clava the germinal vesicle is rarely if ever directly contiguous to the nutritive surface of the spadix, and in the period of later growth invariably migrates to the distal surface and comes to lie in immediate contact with the outer wall of the gonophore. While I have not made any attempt at this time to take up the problem for critical inquiry and investigation, yet my general observations tend to render extremely doubtful the view above suggested, at any rate in any very explicit and causal sense. That the nucleus may function in this matter in a general way as in many other vital functions may be probable, yet that its primary or fundamental and direct part in the oocj^te has to do with nutritive more than other functions of cell life seems more than doubtful. It may easily be shown that processes of nutrition, along with other phases of metabolism, are functions of the entire cell working as a whole. In the earlier paper attention was directed to the phenomena of metabolism as related to the origin and development of the pig
SOME PROBLEMS OF COELENTERATE ONTOGENY 513
ineiit granules of the egg, and it was pointed out that they appeared first in the region of the nucleus, and from this extended as a peripheral zone over the entire egg, the process continuing up to, and even beyond, the phenomena of maturation, which would seem clearly to imply that at most the nucleus was concerned only in the origin of this process, since it early became involved in other functions of very different character.
S. The chromatin. In addition to what has been said in this connection as relates to Pennaria and Hydractinia a few facts may be mentioned as directly bearing on the matter of nuclear fragmentation. In figs. 31 and 32 are shown phases of nuclear behavior associated with maturation. Mg. 31 is a careful drawing of a condition not at all unusual in these eggs. Here one finds undoubted evidences of chromatin fragmentation and dispersal prior to the dissolution of the nuclear membrane. As will be noted, there is as yet no definite disintegration of the nucleolus, which is quite intact, though with a large vacuole. Chromatin granules are variously distributed through the nuclear network, chiefly at nodal points as shown. But the same sort of granules are to be seen just outside the nucleus, and are indistinguishable from those shown in the next figure, in which the nucleus is in process of disintegration, the membrane being entirely dissolved, and the network also surely disappearing. Here also the nucleolus is about to collapse, being flattened on one side, as if ready to go to pieces. Numerous cases of this sort occur in these eggs and seem to confirm what has been said above, that a degree of fragmentation both of nucleus and chromatin is apparently a constant feature. In a few cases I have found these features actively associated with maturation, the first polar body having been already formed.
From this it will be noted that fragmentation of the nucleolus may not occur until that of the nucleus is well under way, as shown in the figures already cited.
I have called attention to the problem of nuclear fragmentation in several of my earlier papers, ('04a, '04b, '06,), and in a paper now in press on the development of Cyanea ('10), attention is directed to very similar conditions associated with maturation.
514 CHARLES W. HARGITT
It is not a new problem; many investigators having directed attention to the matter. Strangely., however, there has been no very serious attempt to explain its probable significance, further than to suggest, as Wilson had done long ago ('00), its possible relation to the extremely active metabolism involved during the grow^th period of the egg. Unfortunately, there seem to be serious difficulties involved in such an interpretation ; for example, the fact that in many ova growth seems to be almost nil. Furthermore, it is not certain that just this type of fragmentation occurs in all eggs just at this period. But whatever may be its significance certain it is that in large numbers of organisms there seems to occur at this time this very interesting fact, that a largely preponderating proportion of chromatin is lost, or at least takes no part in the formation of the chromosomes, and so is a negative factor so far as relates to chromosome function or theory. The bearing of this on the question of chromosome individuality is not without great interest and importance; but no attempt can be made to discuss the problem here. It may be suggested, however, that defenders of the extreme views of chromosome individuality in its morphological sense, — and only in such sense has it any essential significance, — are confronted with a problem, the complexity of which is beyond estimate, and the difficulty of which is hardly less so. Let him who finds difficult the intricacies of 'germ-plasm' hypothesis beware in essaying to unravel the no less intricate mystery, or miracle, of preserving individuality in the metabolic maze through which chromatin must pass in every cytogenic cycle!
In the previous paper (op. cit. '00, , p. 227), attention was directed to the very anomalous aspects of nuclei during early cleavage, features of which were shown in several sketches. Certain of these presented rather strong indications of amitosis, though the condition of the chromatin was such as greatly to obscure the finer details of nuclear structure, and the suggestion was made that "certain phases of the mitotic mechanism may be disguised or actually lacking." As shown above, mitosis has been demonstrated during maturation, yet something of the original doubt remains as to mitosis during early cleavage, the newer material
SOME PROBLEMS OF COELENTERATE ONTOGENY 515
affording no appreciable advantage over that used in the former instance. At this time the chromatin appears only in the form of extremely irregular, flocculent patches, scattered here and there through the cytoplasmic cell-like aggregates. The same elongated, clavicular, or dumb-bell shaped nuclei previously described are found in the newer material treated by latest methods. Under ordinary treatment the chromatin stains so intensely as seriously to obscure details of structure. Only by prolonged destaining and clearing, and by more delicate staining with picro-hematoxylin has it been possible to reduce somewhat the flocculent aspects. When this is done one may distinguish a granular chromatin constitution, but the chromosomes have defied all attempts to render them distinctive either in form or in number. This relates to conditions in early cleavage, as was pointed out in the previous paper, aspects of mitosis become fairly clear in later cleavage. Beckwith describes mitoses in early cleavage but makes no reference to the anomalous conditions here described.
With all the pains taken in preparation of my material it must be allow^ed that these conditions are not artifacts, but facts entitled to the same consideration as others of similar treatment. It must be admitted however, that, even at best, our latest refinem.ents as to staining technique must be accepted as only tentatively trustworthy. In other words, it becomes more evident every day that in protoplasmic and nuclear metabolism there are such incessant and intricate variations of chemical conditions that one may not assert that a given stain or fixing agent affords any certain test of a given state at a given time. On the contrary, it will not be denied that a given stain may act in one manner on one cell and on another very differently; or indeed, that it may in another case fail utterly to yield any results whatsoever in differentiation. Under the aspects of chromatin organization, or perhaps better, lack of organization, as here portrayed, it has not been possible to obtain any definite information as to the number or character of the chromosomes. But it may be said, as before mentioned, that in Clava, as in Pennaria and Hydractinia, there is an enormous fragmentation and dispersal of chromatin at the time of maturation, most of which must be utterly
516 CHARLES W. HARGITT
lost as chromatin, unless some may be recovered during phases of cleavage, as suggested in the previous paper. Some further reference to this feature will be made in a later section of this paper, in connection with the discussion of theoretical problems involved in the general subject.
Incidentally it may be worthy of mention that in at least one case two germinal vesicles have been demonstrated in a given egg. So far as the writer is aware this is a rather rare occurrence, though probably not more so than that of hermaphrodite gonophores, as described in the previous paper (p. 211). Fig. 23 is a careful camera sketch of these nuclei. There was not the slightest evidence to show that there might have been a fusion of two oocytes in this case, as sometimes happens in other hydroids, the egg being only of average size.
4. Nucleolar behavior. In the previous paper (p. 221), attention was directed to some aspects of nucleolar changes associated with maturation. Among these that of vacuolation was particularh^ mentioned, as was also that of the migration of the nucleolus from the nucleus into the cytoplasm. The latter feature is rather unusual, and is not probably of any large significance. More important is the fragmentation which is a rather common feature. Prior to maturation the nucleolus exhibits various phases of vacuolation. In some cases several vacuoles of slightly differing sizes appear, some of which may later fuse into a larger vacuole. In other cases one finds a single large vacuole which finally occupies almost the whole of the nucleolus, at which time it may happen that the body collapses upon itself, or gradually goes to pieces, i.e., fragments. In other cases there may be in a given nucleus two nucleoli, one highly vacuolated and evidently degenerating, the other having all the appearance of a new organ, without signs of vacuoles.^ These changes usually occur while the nucleus is
^ In this case the larger, vacuolated nucleus exhibited a most interesting phase of apparent fragmentation. Almost the whole organ comprised a single large vacuole, and adhering to the outer surface were numerous deeply staining spherical granules borne upon delicate pedicels, the whole resembling a sort of pin-cushion aspect. Just what significance such a condition may have in relation to nucleolar metamorphosis, or its bearing on the problem of chromosome formation, as
f
SOME PROBLEMS OF COELENTERATE ONTOGENY 517
yet intact ; and in some cases, at any rate, the entire fragmentation, or dissolution of the nucleolus may occur before the nuclear membrane disappears. In other cases nucleolar dissolution and disappearance take place coincident with the nuclear dissolution and maturation, as stated in a preceding section. In some instances the nucleolar degeneration seems to involve a gradual process of shrinkage, probably by solution or absorption by the nucleoplasm. It has been no part of my present purpose to study the matter of nucleolar genesis, or the possible relation of nucleolar metabolism to the genesis and differentiation of chromosomes. An interesting and varied literature on this subject has grown up with recent times, some of which seems to have a profound significance in relation to chromosome theory. But to enter upon this phase would involve time and details far beyond the scope of the present paper.
5. Later development. It is not the purpose to enter into any considerable details as to later aspects of development save as they are found to be more or less exceptional as compared with other species concerned. As to cleavage no occasion has been found to modify the account already given in the former paper (pp. 223 to 227). There is much here in common with that known in Tubularia, Hydractinia and Pennaria. While in a certain proportion of the eggs cleavage is more or less regular; on the other hand, in a large proportion, irregularity and lack of order or symmetry is the rule. This is particularly the case with those ova which are flattened laterally against the sides of the spadix of the gonophore. In the case of ova terminally placed in the gonophore the shape is more nearly spherical, and in such cases the tendency is toward regularity. This is what one might naturally expect; yet there are notable exceptions, and one will do well to remember the extremely erratic cleavage of such ova
suggested in a following sentence, I am not prepared to suggest. The nucleus of this egg was in rather typical resting condition, and its chromatin of the usual granular spireme aspect. As stated in another connection, different modes of fixation and staining have appreciably different effects on the egg cytoplasm and nucleoplasm, so that much more elaborate observations would be necessary ere one might venture any very positive opinions on so complex a problem.
518 CHARLES W. HARGITT
as those of Pennaria, Hydractinia and Turritopsis, where the freedom of the egg from all influence of gonophore walls, etc., ought to afford perfect conditions as to regularity of cleavage. It may not be improbable that external conditions of pressure, etc., have an appreciable influence on cleavage, but such facts as those just cited clearly indicate that there are other factors concerned which are probably more potent than the merely physical ones of pressure, gravity, etc.
6. The morula. As in Hydractinia and Pennaria, the early embryo in Clava is a morula. Cleavage results in a solid mass of cells, with only incidentally a sign here and there of an intercellular space, and in only rare instances anything comparable with a segmentation cavity. Indeed, one might venture to aver that such cavity is conspicuously absent throughout the ontogeny of this hydroid. As already pointed out, this is not peculiar, but rather the general fact in hydroid development. Something further will be said on this point in a later connection. There is nothing in the morula stage in Clava which differs appreciably from that of the other species already referred to. As the embrj'-o reaches the morula condition it assumes the usual spherical shape, whatever may have been the shape of the egg during cleavage or growth. Evidently the walls of the gonophore do not afford any very serious barrier to this change, for one finds all conditions of shape from the flattened lateral pocket of the growing oocyte to the spherical terminal capsule, and the oval capsule of the planula, all derivable in turn from the first as the embryo grows and finally emerges as a pear-shaped planula.
7. The germ layers. What has been said on this subject in connection with Hydractinia may be affirmed of Clava. Granted the assumption of a morula as the primitive embryo, and there is no occasion for question or discussion concerning the segmentation cavity, delamination, multipolar immigration, etc. Absolutely nothing of these is involved in the case under consideration. At the time of the completion of cleavage, — indeed before this, when the morular aspect first begins to take shape, — we have only a spherical cell mass, with syncytial tendencies, and as yet without sign of tissue differentiation. In fig. 29 is shown such
SOME PROBLEMS OF COELENTERATE ONTOGENY 519
an earl}' morula-like embryo. In this is shown an oval embryo without definitive ectoderm, or sign of entoderm. This condition persists for some time, the only changes distinguishable being that of cell proliferation, or perhaps more precisely, nuclear proliferation; for in most cases it is not possible to distinguish the presence of cell boundaries of any definitive sort. But a most remarkable thing becomes apparent under careful staining, — namely, the fact that the internal mass shows a differential staining reaction, represented by the shaded interior. This I take to be indicative of an important physiological change, namely, an incipient entodermal differentiation directly related to the primary purpose of entoderm development, that of digestion. While the results do not as yet warrant a dogmatic pronouncement on this matter, they do tend to confirm a view I have already proposed (cf. Science, March 25, 1910). It has generally been assumed that the ectoderm is the primary germ layer, and morphologically this is undoubtedly true. But if the suggestion just made be confirmed by later experiments one will have to aver that, physiologically speaking, the entodermal function is the first to express itself. Further consideration of this point will be deferred to a later section.
a. Ectoderm. The development of this tissue is a graduated process. With the establishment of a surface layer of cells of more or less similar character one is not warranted in designating it as a definitive ectoderm. For, as Rittenhouse has 'pointed out ('07, p. 453) :
Even those cells which are at the surface at the completion of segmentation cannot be regarded as primitive ectoderm, for in the breaking down of the cell boundaries, the formation of the syncytium, and the recasting of the cells, it is quite impossible to say what changes of protoplasm may take place.
Furthermore, it must not be overlooked that, with a primary layer of cells established, there are yet other ectodermal elements to be taken into account, such as interstitial cells, cilia, nettling cells, etc. Only with the formation of the supporting lamella can it be claimed that the definitive ectoderm is really established.
520 CHARLES W. HARGITT
b. Entoderm. As in the case of the ectoderm, what has been said as to entoderm formation in Hydractinia will apply for the most part to Clava. What has been said above in reference to the morula as the primordial embryo applies in this connection. Entoderm formation is a graduated process, and in its morphology a much slower process than that of the ectoderm. In its physiological genesis it may be said to outrun the ectoderm; for its functions begin almost immediately after the completion of cleavage. As was pointed out in an earlier section, the internal cell-mass included within the primordial ectoderm is not in any sense a tissue, but rather a primordium,^ — pro-entoderm. For some time following the nuclear proliferation of this mass continues. But at the same time another, and extremely different process is under wa}^, namely, that of cellular and nuclear disintegration and destruction. Out of this interior mass relatively few cells will survive to constitute the definitive entoderm of the polyp. What is taking place is in reality a struggle among these cells for nutrition, reminding one of the ingenious theory of Roux ('81), 'Der Kampf der Theile im Organismus,' though, so far as I am aware,. this author never applied his theory in this direction. It is not until after the planula has become free that a definitive entoderm is finally established; indeed, this does not become established until near the metamorphosis of the planula into a polyp, though one may trace stages in the process much earlier. Wliat really happens is that the same sort of vicarious process of nutrition occurs as that by which, inmanyhydroids, the oocyte grows; that is, the devouring of sister cells or primordial ova; in the later stage occurs the similar process of digesting the pro-entoderm cells and making their substance available as nutrition to the embryo. As is well known, these pro-entoderm cells are richly laden with yolk granules, as were also originally the cells of the ectoderm. But long after the ectoderm has exhausted this primitive supply the entoderm is reducing its surplus cell mass for similar ends.
With the gradual advance of this process the coelenteron of the larva grows larger, appearing in sections as an axial slit of irregular outline, and later assuming a more regular aspect and
SOME PROBLEMS OF COELENTERATE ONTOGENY 521
becoming more capacious. As the entoderm cells become definitely organized and adjusted in contact with the supporting lamella the entoderm may be said to be established as a tissue. But this does not become complete until a large proportion of the proentoderm mass has been reduced and appropriated by the embryo. There yet remains masses of cells in the cavity along with yolk fragments and other debris variously distributed.
Earlier accounts of the differentiation of the entoderm differ in several particulars from that here given. In the first place, it has been generallj^ assumed that the entoderm is early established, an error which I have taken occasion in several accounts to correct. In the next place, the exact mode of its differentiation has not been very critically studied, nor the fate of those parts of the pro-entoderm not directly concerned in entoderm formation. For example, Korschelt and Heider, following the older accounts, have asserted that following the establishment of the entoderm the remaining cell-mass undergoes fatty degeneration, serving in part as food matter, with a residual mass of debris, the fate of which is not formally stated. I have not found in my preparations any evidence of such fatty degeneration, though, as stated above, I have found direct evidence of the operation of digestive ferments. According to Wilson ('83) something akin to amoeboid engulfment of these cells and their intracellular digestion is tentatively suggested:
These appearances suggest, though they do not prove, that the yolk granules and spheroids pass bodily into the cells. I have never seen them in the act of passing into the cells, but the technical difficulties are great, and the other considerations seem sufficient weight to warrant the provisional acceptance of the view advanced.
That something of such amoeboid engulfment may occur is not altogether improbable; though I have found shght evidence of it. We know that in the growth of the oocyte in certain species just such a process does take place, and its occurrence in the slightly later history of the embryo would be what might naturally be expected. Indeed, I have, in an earlier part of this paper, suggested such a process in the behavior of the cells of the pro
522 CHARLES W. HARGITT
entoderm during differentiation. But associated with the process there are strong evidences of the action of digestive ferments which are set free by these cells in which this process is first set up and carried forward. This likewise takes place in the case of the oocyte during growth, as has been shown by many recent observers. The suggestion of Metschnikoff long ago, that intracellular digestion forms the dominant, if not the only digestive process in coelenterates, is not borne out by recent investigations. For example, it is well known that medusae, actinians, etc., capture highly organized prey, such as Crustacea, fish, etc., and digest it quite after the fashion of the higher Metazoa. The same thing is easily demonstrated in hydroids, in which small organisms, like worms, copepods, etc., form an important food supply. Gland cells are well known in the entoderm of Hydrozoa and are evidently associated with digestive functions. Hence it seems more than probable that enzyme digestion is no less a feature in this than in other animal groups; and that it more than any other, is the mode involved in the reduction of this inner cell mass of the planula is almost certain. This in no wise contravenes the fact of the presence of yolk granules in the entoderm cells, for they were original constituents of these cells, just as in the case of the primordial ectoderm cells. Whether such yolk granules are ever taken in entire by the larval entoderm may be open to doubt, at least till better sustained by direct evidence than at present.
So far as I am aware, the general conception herein outlined as to the physiological processes involved in this phase of larval development has not been hitherto proposed. Of its fundamental validity there seems no serious objection and much direct evidence. In brief, it involves these facts: First, that of the pro-entoderm mass of cells relatively few go to constitute the definitive entoderm of the planula. In the second place, the primary process involved in the necessary differentiation must be one of selection. So far as one can distinguish these pro-entoderm cells are alike in form and potency. The primary demand in embryogeny is growth, which involves nutritive material in some form. And the onlv source of such is that associated with these cells
SOME PROBLEMS OF COELENTERATE ONTOGENY 523
as yolk granules, which can only become available by the dissolution of the cells which contain them. Supposing that originally it was equally distributed, it could only remain so by the further assumption that cell division was likewise equal and continuous throughout. This we know is seldom the case, being in general inversely as the amount of yolk varies. Hence those cells whose growth and metabolism became more rapid would first exhaust their own deutoplasm and demand supplies from outside. And here must originate the struggle among cells which has been emphasized above.
Assuming the substantial truth of the conception we must face the implication that the older views as to the ontogenic and phylogenic significance of the germ layers are discredited by these further facts, as they have also been in theory. I believe we may, therefore, conclude that fundamentally the phenomena involved in germ-layer formation are primarily physiological processes, and relate to protective, motor, and nutritive ends; and that only secondarily, if at all, can they be supposed to have any significance in ontogeny or phylogeny.
REVIEW AND DISCUSSION
As stated in an early section of this paper one of the purposes in view was to review certain phases of current and earlier theory and doctrine concerning problems of ontogeny, in the light of recent knowledge, and to seek to point out and correct such errors as may easily come within the scope of pertinent discussion. This seems to the writer particularly important and desirable just at this time of virile criticism and readjustment.
For some time the conviction has grown that not a few of the earlier views and theories touching ontogenetic problems had outlived their days of service, and that new facts were demanding new methods of interpretation. For example, who today pretends to invoke, in its original content, the Recapitulation Theory in correlating ontogeny and phylogeny? Who would seriously defend, or use the so-called laws of cleavage in interpreting every phase of egg development? And so one might multiply examples.
524 CIIAKLE8 W. HAIKJITT
The fact remains, however, that just these outgrown syst(>nis or tlieories still cumber the literature, the available text-books and inniiuals for introducing students to tlu^se subjects of present day biology, nnich to the re])r()ach of its leaders and sponsors.
With th(^ desire^ to aid, in however small a degree, in correcting phases of error, or what appear such, the writer will aim under this section of the paper briefly to pass in review the chief aspects of the problems involved, and, so far as may be practicable, will endeavor to show distinctive examples of inadequate theor}^ and erroneous implications and d(>(hu'tions.
1.
Or if/in, viuUiplicalion, and growth of germ-cells
It seems woi't h whik^ briefly to sununarize results which observations, more particularly my own, have brought to light on these several aspects of ontogeny. Many of the facts have already been made known in previous papers, but care will be taken to avoid, as far as may be, any unnecessary duplication, giving attention chiefly to those features relating to ])hases which seem to call for consideration. Concerning the earlier controversy as to the mere place or tissue in which the germ-cells arise, it is no longer necessary to nmltiply words. Recent work from various sources, and especially that of Goette ('07), has, I believe, placed the subject beyond further dispute. That there is any such region as may be designated a 'Keimzone' or 'Keimstiitte' may be at once dismissed jis jibsohddy without warrant as a general proposition. Furthermore, that the germ cells have their origin in the ectoderm alone in hydromcdusae may be similarly denied and dismissed as unworthy of further inquiry or doubt. And still further, I am thoroughly convinced that the still more recent controversy as to the hypothesis of the 'germ-plasm,' if not as clearly a, delusion as the preceding, is yet without the slightest support from the ontogeny of the grouj:) imder review.
It is a matter of easy demonstration that in many species of hydroids the egg may be followed in every detail from its origin as an ectoderm or an entoderm or interstitial cell through its gradual differentiation and growth to maturation, as a distinct individual
SOME 1M?<)BLEMS OF (XIELENTEH ATE ONTOCiENY 525
cell, widioul the slightost toiidoncy to luiiltiplicjition. That is to yay, in species of Eudciidrium, Hydractinia, Campamilaiin, Pachycordyle, and others, there is at no time any or^;;in wWwh is ovarian in character, within which masses of primordial ova arise and pass throiij;!! o()j2;onial and oocytic phases familiar in other species to be mentioned later; but a }i;iven cell of tlu^ entoderm which is to give rise to an egg begins to grow, and either in situ or after migration into the gonophore, develops directly into atypical egg, and iatei-, aft (M- fertilization, gives ris(^ directly to an embryo and linally to an individual polyp. On the olhcMhand, in many cases, e.g., Tennaria, Tubularia, Syncoryne, Hydra, large numbers of primordial ova, arise in what may be regarded as an ovaiy where, by a series of cytological changes, they exhibit the oogonial and oocytic phases i-eferred to above. These somewhat strikingly different modes of oogenesis may, foi- convenience be designated as the 'direct' or 'individualized' and the 'indirect' or 'oogonial' modes. That they are sharply distinct, or qualitativ(^ly diff(n-(nitiated types of oogenesis is not claimed. In this, as in other phases of development, there are all shades of intergradation and relation to be found in tlu^se and other species of Cnidaria.
Correlated with these apparently widely divergent modes of origin are those of nutrition and growth. In the 'direct' or 'individual' ova nutrition is almost invariably likewise through the direct medium of the adjacent, tissue cells, which supply by diffusion the appropriate nuti'itive plasnui. On the other hand, in ovarian eggs, which involve oogonial and oocytic generations, there arise indefinite masses of primordial ova; and the growth of certain of these as ovarian eggs, is largely through the active ap})ropriation of the excess primordial ova, which are literally devoured whole, or predigested to a liquid plasma, which is then absorbed. Illustrations of both these processes are too familiar to call for special emphasis. While the two processes of nutrition are thus apparently different, intermediate cases are not unknown, e.g., Eudendrium hargitti, recently described by Congdon ('06, p. 39) has been found to comprise something of both modes. And, though it belongs to a genus in which oogcMiesis
JOURNAL OK MORPHOLOGY, VOL. 22, NO. 3
526 CHARLES W. HARGITT
is associated in its nutritive relations with the direct activities of the tissue cells of the parent organism, yet in this particular species the egg certainly turns parasite, if not cannibal, and devours bodily the cells of either ectoderm or entoderm as may happen to afford it particular support at a given time. And one finds in these growing ova of E. hargitti eggs literally loaded during most active growth with the engorged nuclei of tissue cells, the exact counterpart of those conditions found in Pennaria and Tubularia in which the growing eggs are similarly packed with the primordial ova of the ovarian tissues.
In his earlier studies on heredity Weismann admits that germcells may be derived from somatic cells, e.g., (Essays on Heredity 1, p. 209) :
It is quite impossible to maintain that the germ-cells of Hydroids or of the higher plants exist from the time of embryonic development, as indifferent cells, which cannot be distinguished from others, and which are only differentiated at a later period. Such a view is contradicted by the simplest mathematical consideration; for it is obvious that none of the relatively few cells of the embryo can be excluded from the enormous increase by division, which must take place in order to produce the large number of daughter individuals which form a colony of polyps. It is, therefore, clear that all the cells of the embryo must for a long time act as somatic cells, and none of them can be reserved as germ-cells and nothing else; this conclusion is moreover confirmed by direct observation.
In later discussing ths feature, while still contending that in most cases germ-cells arise early in ontogeny, Weismann is yet compelled to admit that in Hydrozoa these do not arise till very late, and indeed in individuals of a later generation, (Evolution Theory, vol.1, p. 410). Notwithstanding this admission he still contends for his dogma of germinal continuity:
Here the primordial germ cell is separated from the ovum by a long series of cell-generations, and the sole possibility of explaining the presence of germ-plasm in this primordial cell is to be found in the assumption that in the divisions of the ovum the whole of the germ i^lasm originally contained in it was not broken up into determinant groups, but that a part, perhaps the greater part, was handed on in a latent state from cell to cell, till sooner or later it reached a cell which it stamped as a primordial germ-cell. Theoretically it makes no difference whether these germ-tracks, that is, the cell generations which lead from the ovum to
SOME PROBLEMS OF COELENTERATE ONTOGENY 527
the primordial germ-cells, are short or very long, whether they consist of three or six or sixteen cells, or of hundreds and thousands of cells. That all the cells of the germ-tracks do not take on the character of germ-cells must, in accordance with our conception of the maturing of determinants, be referred to the internal conditions of the cells and of the germ-plasm, perhaps in part also to an associated quantum of somatic idioplasm which is only overpowered in the course of the cell divisions. This splitting up of the substance of the ovum into a somatic half, which directs the development of the individual, and a propagative half, which reaches the germ-cells and there remains inactive, and later gives rise to the succeeding generation, constitutes the theory of the continuity of the germ-plasm (p. 411).
Theoretically, the hypothesis is interesting and developed with much plausible argument. Yet its demonstration is far from evident, indeed quite beyond demonstration, as has been frequently pointed out by many of his critics. How'ever, Weismann insists that there are evidential facts :
The hypothesis does not depend for support merely on a recognition of its theoretical necessity, on the contrary there is a whole series of facts which may be adduced as strongly in its favor. Thus, even the familiar fact that excision of the reproductive organs in all animals produces sterility proves that no other cells of the body are able to give rise to germ-cells; germ-plasm cannot be produced de novo.
It is passing strange that he should ignore the body of facts concerned in regeneration, and among them the reproductive organs. And it is still more strange that in support of this he should cite in detail the Hydrozoa as illustrating and supporting the hypothesis, ignoring the well-known facts that among these are abounding evidences which afford insuperable objections to just these assumptions. The present author has, in many cases, shown that gonads may be as readily regenerated by hydroids and medusae as any other organs ; and that not for once or twice, but repeatedly in the same specimen, and that de novo and in situ; not the slightest evidence being distinguishable that any migration through pre-existing 'germ-tracks' occurred. The assumption that in these animals the gonads have been shifted backwards in the course of phylogenetic evolution, that is, have been moved nearer to the starting point of development" seems so at variance with knowai facts as to be difficult to appreciate
528 CHARLES W. HARGITT
or respect. That ' ' the adherence of the sexual gonophore to the hydroid colony has made a more rapid ripening of the germ-cells possible," or that * 'nature has taken advantage of this possibility in all cases," as claimed by Weismann, is but another example of subservience to theory ; for I cannot believe he can be ignorant of the general fact that there is not the slightest evidence that in hydroids with fixed gonophores the germ-cells ripen more rapidly or more frequently.
It is in vain to attempt to bolster up these speculations by cleverly designed diagrams; for such devices are too often mere products of a vivid imagination. Furthermore, it is difficult to account for the dogmatic persistence with which this writer seeks to sustain the view that the germ-cells originate exclusively in the ectoderm. In the earlier work, which makes up his splendid monograph already referred to, he has admitted again and again the probable origin of the cells in the entoderm (pp. 215-217). But in his 'Evolution theory' (p. 415), it is asserted, in no single case is the birthplace of the germ-cells to be found in the entoderm, but always in the ectoderm, no matter how far back it may have been shunted." And in citing cases in support of the point he refers to Hydractinia and Podocoryne, both of which are known to prove the exact opposite, as shown by Bunting ('94) and Smallwood ('09), as well as by the writer in numerous similar cases.
The following critique by Lloyd Morgan ('91) is pertinent in this connection:
This germ-plasm resides in the nucleus of the cell; and it would seem that by a little skillful manipulation it can be made to account for anything that has ever been observed or is likely to be observed. It is one of those convenient invisibles that will do anything you desire. The regrowth of a limb shows that the cells contained some of the original germ-plasm. A little sampled fragment of Hydra has it in abundance. It lurks in the body-wall of the building polyp. It is ever ready at call
Now, although I value highly Professor Weismann's
luminous researches, and read with interest his ingenious speculations, I cannot but regard his doctrine of the germ-plasm as a distinctly retrograde step. His germ-plasm is an unknowable, invisible, hypothetical entity. Material though it be, it is of no more practical value than a mythical germinal principle. By a little skillful manipulation,
SOME PROBLEMS OF COELENTERATE ONTOGENY 529
it may be made to account for anything and everything. The fundamental assumption that whereas germ-plasm can give rise to body-plasm to any extent, body-plasm can under no circumstances give rise to germplasm, introduces an unnecessary mystery .... The fiction of two protoplasms, distinct and yet commingled, is in my opinion, little calculated to advance our knowledge of organic processes.
It has been assumed, as the foregoing citations clearly show, that there is some predetermined order of sequence and relation as to the origin, nutrition, growth, etc., of germ-cells, not only in such a group as the hydrozoa, but throughout the animal kingdom. And with this as a postulate assiduous search has been directed to its support. It is not necessary that one should, a priori discredit the method, for it is perfectly scientific, — within limits. The fault which must be emphasized is that it has been so conspicuously partial and dogmatic. Facts quite as accessible, quite as convincing, have been silently ignored ; and it is thus that such work or method becomes both unscientific and untrue. I believe the foregoing facts must suffice to show that, both as to origin, differentiation and growth, the germ-cells of the Hydrozoa, so far from sustaining the doctrine of the germ-plasm, afford the strongest and most direct evidence to the contrary.
2. Doctrines of homology
[f one were asked to indicate the dominant conception which characterized the biological activity of the greater part of the nineteenth century he could hardly go far amiss in phrasing it somewhat as follows : The perennial and irrepressible search for homologies! This would be confessedly the case with so much of the period as comprised the Darwinian epoch of biology. But the conception belongs quite as properly to the seething period of the biological renaissance of the early half of the century, and finds expression in the researches of von Baer and Cuvier, Lamarck and St. Hilaire, and a long roll of hardly less distinguished names. But strangely enough the doctrine had antipodal significance under the early, as contrasted with the later epochs of thought. To the first homology embodied the postulate of types of creation according to the conception of ' archetypes' of plan and
530 CHARLES W. HARGITT
structure, details of which have been elaborately developed by Owen ('48) whose well-known 'Homologies of the verfebrate skeleton' is its best expression. But to these naturalists homology meant likeness of structure merely, with the implications of ideals and design. To naturalists of the later period the conception took on an infinitely larger scope and significance. Like the former, they were free to accept likeness of structure as an index of homology; but following the blazed trail of Lamarck and St. Hilaire, they conceived in the doctrine the key to lineage. To them homology involved kinship ; and ' archetj^pes' as such had no vital meaning. It is not strange that, under the masterful hand of Darwin, the newer doctrine gave to biologists a working hypothesis comparable with that of gravitation, and at once placed biology on the foundation of scientific certitude.
To naturalists of both periods must be ascribed well deserved praise. Both sought in the most conscientious and critical manner to discern the facts of homology. Among both were those of divergent and conflicting views, ' von Baer and Cuvier versus Lamarck and Hilaire; Agassiz versus Darwin. In both were elements of important truth; in both were extremes of mischievous error. It is not the purpose to undertake any critical review of the phases of confhct involved in these antithetic aspects of one of the most profound of biological doctrines; but rather, ignoring extremes of the earlier period, v/hose errors have largely gone into oblivion, to point out in briefest way wherein, under the ardent impulse of the newer view, something of extravagant over-valuation has come to have a retarding and mischievous influence upon biological thought and progress. It hardly need be said that in this matter attention will be directed to those points in particular which have come under my own lines of research. A similar duty has been ably performed upon a larger scale by several brilliant authorities, among them Wilson ('94), Morgan ('03), Montgomery ('06).
a. The germ-layers. No occasion exists for a review of the origin of the conception of germ layers developed through the work of Wolff, Pander, von Baer, Remak and Kolliker. It is sufficient to my purpose to cite the astute observation of Huxley
SOME PROBLEMS OF COELENTERATE ONTOGENY 531
as to the likeness of the diblastic tissues of coelenterates and the mucous and serous layers of the embryo ('49). Let it be noted, however, that Huxley does not designate these as homologous, but rather as analogous. Ten years after his first utterance he remarks It by no means justifies the assumption that the Hydrozoa are in any sense arrested developments of higher organisms. All that can be justly affirmed is, that the Hydrozoon travels for a certain distance along the same great highway of development as the higher animals." (Oceanic Hydrozoa, p. 2.)
Interestingly enough, the embryological researches of the time, led by Kowalevsky, Gegenbaur, Haeckel and others, centered about this pregnant conception of Huxley and led Haeckel to formulate his famous Gastraea Theory, with all its far-reaching implications as to the homology of the germ layers of all embryos, "from the lowest sponge to man." Of course, the gastrula at once sprang into a position of commanding importance in embryology, and as the prototype of Haeckel's hj^pothetical gastraea became a focal factor in embryological thought for a whole generation. It is not strange, therefore, that the Coelenterata, as the distinctively diploblastic phylum of the animal world, should earl}^ come in for a more than usual measure of interest and concern ; and as the theoretical ancestral phylum from which all higher metazoa must have arisen, should have at once assumed a unique and dominant phylogenetic importance. When, however, it is clearly known that in only a single class of coelenterates does gastrulation occur, and that in no case is the gastrula, as an embryo, known, it seems remarkable to the point of surprise that the theoretical postulate should still be cherished by not a few students of phylogeny. Current literature, however, furnishes abundant evidence of just such adherence to tradition.
b. The planula. As is well known the planula is the distinctive larva of the entire phylum, including also the sponges. It has generally been assumed that the planula is a specialized gastrula, and that in some early species its enteron must have been formed by gastrulation. In this again there is involved the further inference and implication of the dominance of the biogenetic
532 CHARLES W. HARGITT
law. Granted the diploblastic character of coelenterate and sponge; granted further, the gastrula stage in ontogeny throughout a large proportion of higher Metazoa, who could well resist the conclusion jumped at by Haeckel as to the necessary homology of gastrula and planula, facts to the contrary notwithstanding!
c. The morula. It has just been stated that the planula is the distinctive larva of coelenterates. Another ontogenetic stage, however, must not be overlooked, namely, the morula. Of this one hears little now-a-days, though formerly it was a name fairly common in the literature of embryology. Even so recent a text-book as that of Korschelt and Heider devotes to it a single brief paragraph or so. They remark, we shall see that examples of such a mode of origin of the two primary germ-layers are still ascribed to many Hydroids and Anthozoa, though probably the greater part of the cases referred to this method can be reduced to epibolic gastrulation, in which events the morula stage, as being a schema founded on erroneous assumptions, would have to be omitted." As an illustration of subserviency to dominant theory this sentence is a brilliant example! As a matter of fact epibolic gastrulation is absolutely unknown in coelenterate development, cases given of its occurrence having, without exception been proven egregious errors.
It might be questioned whether the morula, as a stage, should be given recognition. But when it is taken as the counterpart of the blastula, a stage everywhere recognized, but comparatively rare in the phylum under review, the objection vanished. The morula is far and away the dominant cleavage embryo in Hydrozoa and common in other classes, the Scyphozoa alone excepted. Accounts of its structure and origin given in an earlier section obviate any call for details here. Suffice it to say, that the comphcated methods described by Metschnikoff ('86, p. 70), while interesting and ingenious, are but of small value. That delamination and immigration (polar or multipolar), may occur need not be questioned; but that they occur in any such degree of frequency or constancy as to constitute laws of entoderm formation none who has had to do with the problem would hesitate to deny.
While less importance is attached to this problem of germ layers
SOME PROBLEMS OP COELENTERATE ONTOGENY 533
than formerly, one still finds it more or less dominant in embryology. In his book, ' The Development of the Frog/ Morgan ('97) gives the subject usual attention; and in the still more recent book, 'The Development of the Chick,' Lillie ('08) devotes several pages to the subject, and it crops out repeatedly in the earlier chapters. The germ layer theory came to have a larger place than might otherwise have been the case in the attempt to discover some ultimate embryological basis for homology, and similar warrant for the so-called Biogenetic Law, or Doctrine of Recapitulation.
It has long passed as a cardinal doctrine in embryology that the primary germ layers form a constant, and more or less infallible basis for homology, — a sort of court of last appeal where other criteria fail. But not a few recent results have tended to force the concession that even here there have been hasty generalizations. Not only in modes of formation and development have the germ-layers been found to differ widely, but in their function and fate in ontogeny there has likewise been obvious variation and discrepancy at many points. Balfour long ago called attention to discrepant modes of mesoderm formation, and recent experimental results have shown that organs of usual ectodermic origin are far from dependent on such mode of derivation. In coelenterate ontogeny the most radical divergences as to modes of origin are too well known to call for extended review. From the extreme mode of delamination exclusively in entoderm formation as pointed out by Metschnikoff in Geryonia, and since confirmed in substance by Brooks ('86), who calls it very peculiar, and without any exact parallel," to that of gastrulation in Scyphozoa, with its confusing variations and exceptions, which involved those rancorous discussions of Claus, Goette, and others, and the more usual mode through the morula, the entire gamut of germ-layer formation might seem to be epitomized. But despite the misguided and essentially mischievous (however well meant), efforts to derive all these phases from a mythical gastraea, now long a discredited and discarded phylogenetic monstrosity, the fact remains that there is probably no genetic relation whatsoever between them.
534 CHARLES W. HARGITT
d. The hlastocoel. As another phase of the germ-layer problem, the cleavage-cavity calls for some passing notice. Formerly it had large attention at the hands of embryologists, and, though less emphasized at present, it has not passed out of consideration. One can hardly consult a current paper dealing with early development without meeting the problem of the origin of the cleavage cavity and its later fate in ontogeny. It is not necessary that one should assume to discredit entirely any possible morphogenic significance to this cavity in any group of organisms; but one does not need to study any considerable series of ontogenies to have forced upon him the conviction that its importance has been greatly exaggerated and correspondingly misinterpreted. One of the first impressions to be gathered from any considerable comparison of coelenterate embryology is that of the conspicuous absence of any definite blastocoel. It is only necessary to cite such figures as those given on plates I to III, illustrating these phases in Pennaria, Tubularia, Clava, Hydractinia to make this point very evident. It is true that here and there at certain stages of cleavage may be found irregular intercellular spaces which have been designated in general as segmentation cavities by those describing them. Spaces they undoubtedly are; but they are not cavities which have any permanence, either of form or position, but shift, or disappear under the erratic adjustments of the blastomeres ; and one might about as well speak of the morphologic significance of the interstices in a box of oranges or bag of potatoes as of these promiscuous intercellular spaces.
Another feature may also be mentioned in this connection. That is the rather significant fact that in many species, such as Hydractinia, where during very early cleavage a cavity may appear incidental^, it almost immediately disappears, becoming totally and permanently obliterated by encroaching cells. And even in certain species, where a more or less characteristic cavity arises and persists for a time, as in certain Geryonids, I am constrained to interpret it as having a physiological rather than morphological significance, — a sort of embryonic receptacle for the deposition of cytolymph, or other substances developed during cleavage, or possibly for food matters derived from the water
SOME PROBLEMS OF COELENTERATE ONTOGENY 535
during early cleavage, or even later the retention of infusoria, as claimed by Merejkowsky ('83), though this may be doubtful. ^
Hence the facts herein adduced, together with the further fact of its extreme variation as to size, shape, position, or, still more significantly, its absolute absence in a large proportion of the species of the entire phylum, afford ample warrant for the conclusion that, so far from having any necessar} morphogenic or ontogenic significance, the blastocoel may be said to be absolutely devoid of anything of the sort, least of all of any relation to phylogeny.
e. Cleavage homology. With the later development of the doctrine of homology there came to be involved varying phases of embryology, as shown above. One of its latest aspects is that concerned with cleavage, which has assumed a place of commanding influence within recent years, as expressed in the flood of literature which sprang into existence dealing with the subject from every point of view, — normal, artificial, experimental. Conklin ('97), has stated well the subject as follows:
In the whole history of the germ-layer theories I see an attempt to trace homologies back to their earliest begimiings. This problem is as important today as it ever was, and whether one find these earliest homologies in layers or regions of blastomeres or the unsegmented ovum itself, the quest is essentially the same. Within this question of the earliest homologies is included another of great and present interest, viz., the significance of cleavage.
With the broader implications and relations of this subject there is neither the time nor occasion for extended review in
^ I can but express the strong conviction that those who contend for the presence in such cases of a definitive segmentation cavity and blastula are in serious error. It seems not at all adequate to aver that the absence of any true blastocoel is due to the 'abbreviation of this stage of development,' as G. T. Hargitt ('09) has designated it. As suggested above, but for the earlier theoretical significance involved in the matter, it may be doubtful whether any such contention would be made as that under review. To the writer it seems a pity to waste words over the subject in the form of argumentation. The facts are their own best exponent, and with these clearly apprehended there ought to be small occasion for controversy. The presence or absence of syncytial conditions has nothing whatever to do with the problem. Long before a syncytium has developed the morula has arisen as shown above, a fact as incompatible with the blastula as the planula is independent of the gastrula.
536 CHARLES W. HARGITT
this connection. With certain hmited aspects of the problem as they relate to coelenterate ontogeny facts have come to knowledge which demand consideration. In a general way it may be said of the problem of cleavage homology that two rather divergent schools of biologic thought have grown up. One of these, ably represented by Driesch and 0. Hertwig, maintain that cleavage is a more or less general and quantitive process, the resulting blastomeres being largely equipotent in later development, their individual values depending largely upon relations of position, etc. The other wing of thought would hold that cleavage is fundamentally a qualitative process, involving a nicely predetermined and 'orderly sifting of materials,' resulting in a splendid 'mosaic work,' each cell fitted into its predetermined place with mathematical precision. Under the latter conception 'cell lineage' became the dominant problem of embryological research.
As a corollary to this, not only were blastomeres factors of supreme concern, but the natural and almost necessary implication followed that there must of necessity be predetermining factors in the unsegmented egg even more fundamental than those in the blastomeres. Hence came into prominence the search for evidences of 'formative stuffs', 'prelocalized germinal areas', etc. Waiving all further consideration of this particular aspect of the problem in its theoretical implications, I may very briefly cite facts concerned with coelenterate cytology, and attempt to show their bearing in the case.
In earlier contributions on the subject. of cleavage, particularly in Pennaria and Clava, and further facts given in previous parts of the present paper, attention has been directed to facts which must be their own exponents. As to any blastomere homology in any of these cases it is difficult to conceive. Furthermore, both under normal conditions and through experiment, it has been demonstrated over and over that one or many blastomeres may be detached without in the least modifying the course of development in any particular. With such pictures as those in figures 1 to 30 before one, he would need be possessed of a measure of imagination beyond compare who could discern any sign of a
SOME PKOBLEMS OF COELENTERATE ONTOCxENY 537
'mosaic work'! And what shall be said as to the existence of prelocalized germinal areas in such ova? I have searched throughout the phylum for ova having any semblance of such, but without evidence of its existence. It was thought for a time that Clava might be a case, but the most painstaking efforts to detect it were only negative. For a time Conklin believed he had found such in the ova of Linerges, and so pronounced; but his final utterance ('08, p. 166), recalls this: The view expressed in my preliminary note on the development of Linerges, that the three layers of the egg give rise to the ectoderm, the entoderm, and the mesogloea is not confirmed by further study."
That there may be certain special distribution of egg material I have already shown in the case of several hydromedusae, but this is by no means implies that it is germinal in character or definitely prelocalized.
3. Amitosis
In several earlier papers I found occasion to call attention to what seemed to be amitosis in cells during early cleavage. In several of these the evidence seemed direct and positive; in others the indications were somewhat general and indirect. The fact that several later students of cleavage in eggs of hydroids failed to confirm my results, while in the case of several others there has been very explicit confirmation, leads me to briefly review the matter as it appears at the present time. As I have elsewhere stated, the question of amitosis is purely one of fact. Whatever may be the implications of amitosis in its theoretical bearings on problems of heredity or otherwise, it must be evident that to attempt to discredit it on such grounds, or others of like nature, can only result in confusion worse confounded. One fact is just as sacred as another, and just as much entitled to respect and consideration, and is bound sooner or later to be taken account of. The extreme attitude of Ziegler, Vom Rath, and certain other cytologists who would have us believe that amitosis is to be found only in senile or pathologic tissues, will have to be abandoned as altogether unwarranted. Cytologists no less capable and conscientious, in growing numbers, accept amitosis as a normal
538 CHARLES W. HARGITT
and not rare mode of cell division. The following recent utterance of one of the avowed conservatives will show how just is this claim: ' 'Accepting the idioplasm hypothesis . . . what do we know of its transmission? We may answer with assurance that it is transmitted from cell to cell by division; and we may safely presume, I think, in most cases by mitosis, though the direct or amitotic process may play a larger role than was formerly supposed." (Wilson, '09.)
My first suggestion concerning the problem was made in connection with my early account of Pennaria ('00) ; and the same year, Allen, one of my graduate students, made a similar statement in connection with the development of Tubularia. In a paper on regeneration, my son, G. T. Hargitt ('03), described abundant amitoses in the regenerating hydranths of Tubularia, and suggested the probable relations of the process to rapid growth and metabolism. In several contributions Child has also described amitosis, and in one in particular ('07), gave a brief account of the process in a series of organisms from coelenterates to birds. In one of these he made bold to predict that future investigations will probably show that amitosis is at least as important in the life of the cell as mitosis." How timely was this prediction may be inferred by an examination of several recent papers on the subject, particularly by Patterson ('08), on 'Amitosis in the pigeon's egg,' and Glaser ('08), 'A statistical study of mitosis and amitosis in the entoderm of Fasciolaria.' In both these studies it is interesting to find so striking a vindication of Child's forecast. Patterson finds that at certain stages amitosis is quite as common as mitosis; and suggests "it seems very probable that amitosis is the result of special physiological conditions, which create a stimulus to cell division, .... whatever factors are involved in bringing about the rapid growth of any region would seem to be concerned in causing amitosis." This affords an interesting agreement with the suggestion made by G. T. Hargitt, as quoted above. Glaser also concludes "that amitosis plays in this instance "(Fasciolaria) "an important, if not the chief part in the differentiation of a definitive tissue."
SOME PROBLEMS OF COELENTERATE ONTOGENY 539
These several series of facts afford, as I believe, very strong confirmation of my own results as related to Eudendrium, Pennaria and Clava. They are also further supported by interesting results described by Young ('08) in connection with the ' Histogenesis of Cysticercus pisiformis.' In this paper is found the somewhat radical suggestion that cells may arise de novo, from a 'cytoblastema,' much as held by Schwann long ago.
It is not necessary to review this phase further than to point out its relation to a similar suggestion made by the writer concerning a somewhat similar origin of cells after nuclear fragmentation in both Eudendrium and Clava. Several of Young's figures are strikingly similar to those given in connection with Eudendrium.
Another most interesting confirmation of my results is to be found in the account of Brooks and Rittenhouse ('07), of the development of Turritopsis. In this case both mitosis and amitosis are found occurring 'simultaneously in the different cells of the segmenting egg.' The varying size of the amitotic nuclei and their reticular structure, confirm with utmost exactness my earlier accounts. Furthermore, the association of amitosis with an approach toward syncytial conditions also resembles the condition in Eudendrium, as does also the metabolism associated with yolk digestion. I cannot agree, however, with the authors that there is any such relations involved in any of these processes as would bring them into conformity with the theory of Flemming, Ziegler, and others, that they presage degenerative ends. On the contrary, they seem to me to be most intimately associated with the intense metabolism and rapid growth of histogenesis.
It remains briefly to refer to phases of nuclear behavior so characteristic in the cleavage of Pennaria, and to a less extent in Clava. Among these are such features as the highly vesiculate aspects of the nuclei during early cleavage, and the equally anomalous features of clavicular, reniform and dumb-bell shaped nuclei. These facts have been very abundantly confirmed by the several researches of Hargitt, Smallwood, and Beckwith, already cited. Their interpretations, however, differ very widely from my own, and with plausibility and force. At the same time I fail to per
540 CHARLES W. HARGITT
ceive that the facts are not quite as clearly within the amitotic mode, and with apparently quite as strong evidence in support of the latter view. Granted the facts of amitosis as a normal process in cytogeny, and this is no longer open to denial, its occurrence along side of mitosis must be allowed. And even where one investigator may find mitosis, another may find both mitosis and amitosis; and this I have shown in the cases already cited, and my results have been confirmed in similar cases by others.
But there is still a further word in this connection. The fact that the nuclear vesicles differ so markedly in size, shape and number is rather difficult to interpret on the basis of mitosis alone. Are these several vesicles derived from single chromosomes, or from several which have fused? If the latter, how shall we correlate the fact with the further fact, admitted by both Small wood and Beckwith, that it is not essential that these vesicles should fuse between successive mitoses? But how then shall we attempt to explain the assumed exact nuclear equivalents of every mitotic division? But if, on the other hand, it be held that these nuclear vesicles are originally derived from single chromosomes, as seems more likely, how are we to account for the marked diversity of size, number and shape? These queries are not suggested out of any captious spirit, nor on the other hand, as affording an insuperable objection to the interpretations given by these authors, but as more or less clearly pertinent questions which warrant consideration in connection with the problem concerned.
The further assertion of Smallwood in this connection that ' ' the mere shape of the nucleus in Pennaria is no indication of amitosis," may be looked on as somewhat of an evasion of the real issue. I have nowhere made such a claim; but if such were the case it might with pertinence be replied that mere shape is not the point at issue. On the other hand, we are here concerned with particular and anomalous shape, a very different matter. Whether shape has any significance in this relation depends to a marked degree upon the kind of shape. That a reniform, or dumb-bell shaped nucleus 'is no indication of amitosis' may be flatly denied, where it is more or less prevalent. Given such shapes, while
SOME PROBLEMS OF COELENTERATE ONTOGENY 541
spheres, spindles, etc., are held to be types of normal nuclei, I think it must be allowed that the burden of proof that the former are but phases of the latter is upon the champions of exclusive mitosis, and thus far the evidence which they submit has not been convincing.
With the facts herein presented, and the cumulative evidence available from a wide range of observation and authority clearly appreciated, it seems difficult to evade the force of the conclusion which is implied. The writer believes, therefore, that his earlier tentative suggestions concerning amitosis as a mode of nuclear activity is not only not discredited nor disproved by later researches, but is rendered both credible and probable.
SUMMARY
The main points embodied in the paper may be summarised as follows:
1. Later observations on the development of Pennaria, and including a new species, go clearly to confirm the earlier results, and to show that it is not peculiar to a single species or to a given locality.
2. Observations on the development of Hydractinia echinata also confirm much of that found to occur in Pennaria, including cleavage, ectosarcal features, formation of germ layers, etc.
3. The same may be said in general as to Clava leptostyla. New facts as to certain histogenic aspects seem established, and the significance of the early embryo, — the morula, — is emphasized.
4. Concerning the origin and growth of germ cells it becomes more and more certain that the theoretical contentions of Weismann find no warrant in the ontogeny of Coelenterates, and particularly in that of Hydrozoa, the group especially claimed by him.
5. A review of earlier doctrines of homology goes to show that they have been greatly overestimated as criteria of phylogeny. This includes especially the features involved in germ layers, the early hydroid larva, cleavage homology, prelocalization of ger JOURNAL OF MORPHOLOGY, VOL. 22, NO. 3
542 CHARLES W. HARGITT
minal areas, etc. The facts of homology and ontogeny as related to phylogeny leave much to be desired ere it will be possible to sustain the earlier conceptions of recapitulation and its enormous implications as to biological philosophy,
6. Amitosis as a factor in cytogenesis is a question of fact. Cumulative evidence from almost every field of cytology goes to show that it is neither rare, nor limited to senile or pathologic conditions of cells or tissues. Its significance in cytogeny is difficult to overestimate. It is not unknown as a factor in embryogeny in any of the great phyla of nature. As a fact it is no less sacred than any other, and must be reckoned with in any final doctrine of development.
BIBLIOGRAPHY
Allen, Carrie M. 1900 A contribution to the development of Parypha crocea;
Biol. Bull., vol. 1, p. 291. Balfour, F. M. 1885 A treatise on comparative embryology; London, 2nd
Edition. Beckwith, Cora J. 1909 Preliminary report on the early history of the egg and
development of certain hydroids; Biol. Bull., vol. 16, p. 183. Brooks, W. K. 1886 Life history of North American Hydromedusae; Mem.
Bost. Soc. Nat. Hist., vol. 3. Brooks, W. K. and Rittenhouse, S. 1907 On Turritopsis nutricula; Proc.
Bost. Soc. Nat. Hist., vol. 33, p. 429. Bunting, Martha 1894 The origin of sex-cells in Hydractinia and Podocoryne,
and the development of Hydractinia; Jour. Morph., vol. 9, p. 203. Child, C. M. 1904 Amitosis in Moniezia; Anat. Anz., Bd. 25, p. 545.
1907a Amitosis as a factor in normal and regulatory growth; Anat.
Anz., Bd. 30, p. 271.
1907b Studies on the relation between amitosis and mitosis; Biol.
Bull., vol. 12, pp. 89, 175; vol. 13, pp. 138, 165. Conklin, E. G. 1897 The embryology of Crepidula; Jour. Morph., vol. 13.
1908 The habits and early development of Linerges mercurius;
Carnegie Inst., Washington, Pub. No. 103, p. 153. Glaser, O. C. 1908 A statistical study of mitosis and amitosis in the enteron
of Fasciolaria; Biol. Bull., vol. 14, p. 219. Hargitt, C. W. 1900 The natural history and early development of Pennaria
tiarella (McCr.); Am. Nat., vol. 34, p. 387.
1904a The early development of Eudendrium; Zool. Jahrb., Bl. 20,
p. 257.
SOME PROBLEMS OF COELENTERATE ONTOGENY 543
Hargitt, C. W. 1904b The early development of Pennaria tiarella (McCr.); Archiv f. Ent-Mech., Bd. 18, p. 453.
1906 The organization and early development of Clava leptostyla Ag.; Biol. Bull., vol. 10, p. 207.
1908 Notes on the Coelenterates of Woods Hole; Biol. Bull., vol. 14, p. 97 et seq.
Hargitt, C. W. and G. T. 1910 Studies in the development of Scyphomedusae;
Jour. Morph., vol. 21, p. 217-262. Hargitt, G. T. 1903 Regeneration in Hydromedusae; Arch. f. Ent-Mech. d.
Organismen, Bd. 17, p. 64.
1909 Maturation, fertilization and cleavage of Pennaria tiarella and Tubularia crocea; Bulletin Mus. Comp. Zool. Harvard, vol. 53, no. 3.
Harm, K. 1902 Die Entwickelungsgeschichte von Clava squamata; Zeits. f.
wiss. Zool., Bd. 73, p. 115. Hertwig, O. 1892 Text book of embryology; English trans., Huxley, T. 1849 On the anatomy and affinities of medusae; Phil. Trans. Royal
Soc. London, part 2, pi 413.
1859 Oceanic Hydrozoa; Ray Society, London. KoRSCHELT AND Heider 1895 Text book of the embryology of invertebrates;
English translation. LiLLiE, F. R. 1908 Development of the chick; Henry Holt, New York. Merejkowsky 1883 Development de la meduse Obelia; BuJl.de la Soc. de France. Metschnikoff, E. 1886 Embryologische Studien an Medusen; Wien. Montgomery, T. H. 1906 The analysis of racial descent in animals; Henry
Holt, New York. Morgan, Lloyd 1891 Animal life and intelligence; Ginn and Company, Boston. Morgan, T. H. 1897 Development of the frog; Macmillan, New York.
1903 Evolution and adaption; Macmillan, New York. Owen, R. 1848 Homologies of the vertebrate skeleton, London. Patterson, J. T. 1908 Amitosis in the pigeon's egg; Anat. Anz., Bd. 32, p. 117. Rittenhouse., S. See Brooks and Rittenhouse.
Roux, W. 1881 Der Kampf der Theile im Organismus.
Smallwood, W. M. 1909 A reexamination of the cytology of Hydractinia and
Pennaria; Biol. Bull., vol. 17. Weismann, a. 1883 Entstehung der Sexualzellen bei den Hydromedusen.
1889 Essays on heredity; English translation, Macmillan, vol. 1.
1904 Vortrage iiber Descendzentheorie; English translation, 2 vols. London.
Wilson, E. B. 1884 The development of RenilLa; Phil. Trans. Roy. Soc. London, vol. 174.
1894 The embryological criterion of homology; Biological Lectures, Woods Hole, Boston.
Young, R. T. 1908 The histogenesis of Cysticercus pisiformis; Zool. Jahrb., Bd. 36, p. 183.
EXPLANATION OF PLATES
All figures made with the aid of Abbe camera lupida. Those of living eggs in outline only. Details supplied free hand. No attempt has been made to give exact magnification of living eggs, the erratic shapes making this extremely difficult.
PLATE I
EXPLANATION OF FIGURES
1 to 4 Pennaria tiarella. Varying aspects of cleavage in early phases, as a basis for comparing that of Pennaria australis. p. protoplasmic connective or strand; a very common feature in these eggs, x, a blastomere of second cleavage. In fig. 3 it will be noted that this blastomere segments more rapidly than the lower. This is very common, and continues in fig. 4.
5 to 8 Pennaria australis. Cleavage here resembles in a marked degree that of the preceding species.
9 to lib Hydractinia echinata. An extremely erratic cleavage. In fig. 10 are shown several interesting features, viz. the blastomeres at x, y, z. At fig. 11 they are shown in a later stage, in which x and y are just becoming detached. Their later history is shown in figs. 11, o and b.
PLATE 2
EXPLANATION OF FIGURES
12 to 22 Various phases of cleavage of similar eggs. Figs. 14 to 18 phases of cleavage in a single egg at intervals of ten minutes. The irregular spaces shown are interesting as so-called segmentation cavities. As a matter of fact, they are but aspects of the peculiar ectosarcal and amoeboid activities, and hence absolutely devoid of any blastocoel relations. Such is likewise the case with other similar features in other cases.
19 to 22 Varying phases in the cleavage of another egg, sketched at intervals of fifteen minutes. In these are shown in typical aspects the ectosarcal features more or less common in these eggs. The various strands, papillae, etc. are conspicuous.
12 to 13 Aspects of later cleavage of the egg body shown in figs. 9-11.
23 Gonophore of Clava showing the unusual feature of two perfectly formed and typical germinal vesicles in a single egg. X about 100.
546
SOME PROBLEMS OF COELENTERATE ONTOGENY
CHARLES W. HARGITT
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 3
547
PLATE 3
EXPLANATION OF FIGURES
24 to 27 Cleavage aspects of Tubularia crocea. (Reproduced from drawings made by G. T. Hargitt illustrating his paper on early development of Pennaria tiarella and Tubularia crocea. Bull. Mus. Comp. Zool., vol. 53, no. 3, by permission.)
24 Cleavage planes which are complete, are more or less vertical, but the equatorial furrows are shown in several of the blastomeres.
25 to 26 Two sections of an egg showing extremely elongated and erratic aspects. The several spaces shown are designated as cleavage cavities. This viejw I have taken occasion to question in the text of the present paper.
27 So-called blastiila stage. This point I have also shown to be a mistaken view. In fact it may be questioned if in any case the term blastula should be applied to early stages of cleavage such as this.
28 Section of an egg of Pennaria tiarella in early cleavage. This is an egg which has shown an unusual regularity in cleavage behavior. N shows a typical resting nucleus, of which several others are shown. At A'^' is shown a nucleus in what seems amitotic cleavage. In this egg are seen also several inner spaces, but which are extremely transient phases.
29 Morula of Clava. The pro-entoderm has been tinted to show the early physiological differentiation of these cells. A discussion of this may be found in the text.
30 Morula of Hybocodon prolifer. As compared with preceding figures of Tubularia, Pennaria, et al., it shows the same indefinite intercellular spaces, but no distinctive blastocoel. Several nuclei here shown resemble much those of Clava, and appear in some cases in amitotic division.
31 Nucleusof Clava just prior to maturation. The nucleolus is conspicuously vacuolated. Chromatin is in process of fragmentation and dispersal. Stained by picro-hematoxylin which differentiates the yolk beyond mistake, and makes certain the chromatin nature of these granules. X 800.
32 Nucleus of Clava in process of fragmentation and dissolution. First polar body already discharged. Nucleolus in process of collapse; chromatin fragmentation well advanced. Stain as in previous egg. X 800.
548
SOME PROBLEMS OF COELENTERATE ONTOGENY
CHARLES W. HARGITT
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 3
549
PHYSIOLOGICAL ANIMAL GEOGRAPHY
VICTOR E. SHELFORD
Department of Zoology, The University of Chicago
NINETEEN FIGUKES
CONTENTS
I Introduction 552
1 Faunistic animal geography 553
2 Physiological animal geography 554
II The physiological characters and distribution of particular species of
tiger beetles 556
A Material : general habits 556
1 Reproductive processes 557
2 Larva and pupa 557
3 Food 558
B Habitat relations of soil-inhabiting tiger beetles 559
a C. purpurea limbalis 559
1 General behavior of adults 559
2 Ecological relations of adults 560
3 Ecological relations of larvae 561
4 Experimental studies of habitat selection 566
5 Geographic distribution 567
6 Geographic variation in habits 568
b Cicindela tranquebarica 568
c C. sexguttata 575
d Other species . . .'. 584
C General considerations 586
1 Importance of breeding instincts and breeding place 587
2 Relation of behavior to the habitat and associated forms. . . 588
3 Meaning of variation in habits 589
4 The relation of geographic to local distribution-governing
factors 590
III- The physiological characters and distribution of groups of species (formations) 591
A Zoological opinions and difBculties 591
B Nature of the environment 592
C Environmental relations of animals 592
1 Comparison of the environmental phenomena of plants and
animals 592
2 The most important relations of animals 595
a The method of investigation 595
551
552 VICTOR E. SHELFORD
3 The relation of physiological characters to geographic range. . 596
a Laws governing the reactions of animals 597
b Law of minimum 597
c Law of toleration of physical factors 598
4 Tentative laws of distribution 600
D Classification of environments 600
1 Elementary principles of classification 601
2 The best index of geographic complexes 601
E The animal formations 602
1 Classification of formations 603
a Principles of classification 603
b The geographic formations of the world 604
IV The problems, methods and relations of physiological animal geography . 607
A Problems 607
1 Behavior 607
2 Physiology 608
B Methods 609
C Relations to other subjects 609
D The future biology 611
V General summary 612
Acknowledgments 613
Bibliography 615
I. INTRODUCTION
Only a working knowledge of the facts of animal geography is necessaiy for the recognition of at least two or three lines for the development of investigation, and for organization into a science. Likewise, a casual inspection of the existing literature, indicates clearly that only one or at most two of the possible lines of investigation have received attention; facts have been accumulated very largely from the point of view of animal structure, and organization has been based on evolution. Physiological lines have been proportionately neglected. It is our purpose to point out some of the possibilities of investigation and organization along physiological lines. ^
1 When my paper on the life-histories and larval habits of the tiger beetles ('08) was prepared, it was my intention to follow it a year later with one on their ecology and distribution. While attempting to prepare this, it became evident that many of our so-called principles of distribution and some of the methods employed in its study, were not wholly trustworthy as a basis for generalization. Before interpreting the beetle data, further observation and the examination of the existing literature seemed advisable. We present here a point of view which developed in connection with this uncompleted task.
PHYSIOLOGICAL ANIMAL GEOGRAPHY 553
The materials of animal geography may be roughly classified into fact and interpretation. Interpretations have been related to genetic or historical geography, fields in which speculation has been common. Since the methods employed and the conclusions reached are quite familiar, we will take up the discussion so far as possible, from the point of view of facts.
The facts of animal geography fall under tw^o main heads: (a) facts concerning the structural and the taxonomic differences and resemblances of the animals of different parts of the world, and (6) facts concerning the physiological and ecological characters of animals which enable them to live under the geographic conditions in which they are found, and the effect of geographic environments upon their behavior, physiology, and mode of life. The former is what is commonly known as faunistic animal geography, the field in which nearly all the investigation has been concentrated.
1. FAUNISTIC ANIMAL GEOGRAPHY
a. Point of view. The point of view is essentially that of speculative evolution, of the evolution of animal groups and of the evolution of barriers and land masses as related to the distribution and dispersal of animals from the supposed centers of origin (Wallace, 76; Osborn, '02). The subject is even more strongly committed to speculative evolution than any other phase of biological science.
The study of so many phases of biology from the point of view of an inadequate conception of evolution, which has been so prevalent during the past forty or fifty years, has probably materially retarded the unification and progress of biological science as a whole. In the case of the geographic aspect, the damage done is quite beyond repair, for the great mass of ecological data accumulated during that period has not been preserved and the conditions which make such observations possible have passed away, too frequently forever, before the hand of civilization (Haddon, '03; Webb, '03).
554 VICTOR E. SHELFORD
b. History. The history (Ortmann, '96) of faunistic animal geography is largely that of the ideas of regions, centers of dispersal, barriers, etc. The only recent writer who has advanced other ideas is Seitz ('91), who called attention to the resemblances of the forms of similar habitats in different parts of the world (Beddard, '95).
2. PHYSIOLOGICAL ANIMAL GEOGRAPHY
a. Point of view. There are two distinct points of view for biological investigation. One is that of evolution; the other, that of physiology, or the explanation of the organism in terms of physics and chemist^y^ One may make a physiological explanation of the behavior or structure of an organism and in no wise explain its evolution. On the other hand one may make an evolutionary explanation of an organism without making any contribution to its physiology. The study of physiological animal geography may be conducted independently of the problems of evolution. It does not need to be concerned with centers of origin, or paths of dispersal, or with other problems of faunistic animal geography. In this paper we are concerned solely with the physiological relations of animals to natural environments.
b. History. Crude expression of some of the ideas which should be included in this subject is no doubt as old as biology itself. From the standpoint of particular taxonomic groups, various writers on natural history, ecology, behavior and physi-' ology have from time to time touched upon the relation of habits, ecology, or physiology to geographic distribution of particular species (Semper, '81).- Ortmann ('07) especially emphasizes the importance of a knowledge of the habits as a means of interpreting distribution. Indeed, this is an important principle of ecology, and lacks definite formulation rather than recognition.
From the point of view of all the animals of a given set of environmental conditions, Thomson gives the early history in his introduction to Brehm ('96). The early observations were made by men whom Thomson designates as naturalist travelers
- Semper' brought a large number of these facts together as they existed in 1879.
PHYSIOLOGICAL ANIMAL GEOGRAPHY 555
of the biological type, the most noted of whom were contemporaries of Darwin, such as Bates, Belt, Wallace and Brehm. Though anthropomorphic, at least in his wording, Brehm stands as one of the foremost writers of the time in this field of animal behavior.
He had unusual power as an observer of the habits of animals. His particular excellence is his power of observing and picturing animal life as it is lived in nature, without taking account of which biology is a mockery and any theory of evolution a one-sided dogma. The success of the pictures which Brehm has given us of bird-bergs and tundras, of steppes and desert, of river fauna and tropical forest, raises the wish that they had been complete enough to embrace the whole world. Thomson.
An excellent discussion by Craig ('08) who compares the behavior and adaptation of the birds and mammals of the steppe of North America with those of the forest,^ is the only recent paper of this kind, by a zoologist, which has come to my attention. There has been, so far as I have found, no comparison of the behavior of the animals of the different deserts or different steppes, etc.
While physiological animal geography is a subject for experimental study, experimental methods can hardly be said to have been used in the study of geographic distribution. Experimental researches which have involved distribution are limited chiefly to investigations of the reactions and local distribution of aquatic and cave animals (Banta, '10).
In the field of plant geography, Schimper's ('03) work indicates the first step in the development of the world-wide aspect along physiological lines (Cowles, '09). This work opened a new and fruitful field for experimental work and field observation. Here Warming ('09), Cowles ('01), Whitford ('01), Transeau ('03, '05) and others contributed much from the observational side, while others have done important experimental work.
In the presentation of data and in the discussions here we illustrate two points of view for investigation by classifying the materials roughly into (a) those related primarily to the par ' The work of Adams ('05, '09), and Ruthven ('06), was conducted with reference to all the animals but from a genetic rather than a physiological point of
556
VICTOR E. SHELFORD
ticular species of tiger beetles, and (h) those related to the entire group of animals inhabiting a given environmental complex.
II.
THE PHYSIOLOGICAL CHARACTERS AND DISTRIBUTION OF PARTICULAR SPECIES OF TIGER BEETLES
A. MATERIAL: GENERAL HABITS
The tiger beetles are graceful, predatory, swift^flying insects, whose bright colors and great variability have long been familiar.
The following general account of habits applies to all the species especially considered here. The life-histories consist of the egg, three larval stages, the pupa and the adult. When the beetles
Fig. 1 From left to right — the ventral, side, and dorsal view of the ovipositor
of Cicindela purpurea with segments numbered. Three times natural size.
Fig. 2 The egg of C. purpurea in position in the hole in the ground made by the ovipositor. One and one-half times natural size.
Fig. 3 The egg. Three and one-half times natural size.
emerge from the pupal stage in summer, they are not sexually mature. Many species hibernate during the winter following emergence. Hibernating species (Shelf ord, '08), reach sexual maturity after several warm days of spring. Previous to sexual maturity, the animals are in a different physiological state than when sexually mature, and they accordingly behave differently, congregate in different places, and never attempt to use the ovipositor.
1. Copulation arid egg-laying
The beetles copulate on warm days, especially when the atmospheric humidity is high. The eggs are laid in small vertical holes, 7 to 10 mm. deep, made by the ovipositor (figs. 1, 2 and 3).
2. The larva and pupa
The larva, which on hatching excavates a vertical, cylindrical burrow in the position of the ovipositor hole, is elongated, yellowish, and grub-like, with a n-umber of brown spots on each abdominal segment, and with a dark-colored, strongly chitinized head and prothorax of unusual form. The head bears two pairs of large ocelli on the outer border of the upper surface, two pairs of small ones on the lower surface immediately below them (figs. 4 and 5). The mandibles, instead of extending downward or for
Fig. 4 The larva, side view; h, hooks. Three times natural size. Fig. 5 The anterior half of the larva; an, antennae; trip, maxillary palp; m, mandible; o, ocelli. Three times natural size.
ward as is usual in insects, are curved upward, amd when closed, overlap above the anterior end of the clypeus. The lower side of the head is somewhat hemispherical, the upper side flattened, and, with the appendages, almost semicircular in outline. The prothorax is semicircular, flattened above, and projects at the sides. Taken together, the head and prothorax form nearly a circle. The meso- and metathorax and abdomen are soft and fleshy. On the dorsal side of the fifth abdominal segment is a
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 3
558 VICTOR E. SHELFORD
hump-like outgrowth which bears a pair of long, curved, anteriorly directed hooks {h, fig. 4), a pair of short vertical spines, and manystrong bristles. The last two abdominal segments are also armed with strong bristles.
In moving up and down in the burrow the larva uses the dorsal hump, the legs, and the last abdominal segments. The animal turns around in the burrow by bending the anterior part of the body dorsally, and forcing the head past the dorsal side of the abdomen which is held in position while the anterior part is moved by means of the feet. When at rest in the burrow, the animal assumes a zigzag (Enoch, '03) position as shown in fig. 4. When waiting for prey at the mouth of the burrow, the same general position is maintained, but the head and prothorax are bent at right angles to the longitudinal axis of the meso-metathorax. The legs, the vertical spines of the dorsal hump, and the strong bristles of the last two segments hold thfe animal in position. The head and prothorax just close the round opening and the mandibles are extended. If a small or medium sized insect pass near, the larva strikes at it with its head, by suddenly straightening the body in the region of the meso- and metathorax (Geoffroy, 1762), at the same instant closing the mandibles with a snap that can be distinctly heard, if the prey escapes them. If the insect caught be of small size, the larva darts backward to the bottom of the burrow with its prey which is devoured at leisure, the inedible parts being brought to the surface and cast out. If the prey be large (for example, a cabbage butterfly, as was observed by Weed, '97), it is held at the entrance of the burrow. The forward projecting hooks of the dorsal hump serve to prevent the butterfly from dragging the larva out of its hole, while its blood is being withdrawn. The pupa is of the usual beetle type (fig. 6). Pupation takes place in the ground.
3. Food
The food of both larvae and adults consists of sow-bugs, centipedes, spiders, dragon-flies, butterflies, beetles, flies, and larvae of all sorts, in fact, any small animals that come within
PHYSIOLOGICAL ANIMAL GEOGRAPHY 559
reach. If larvae are not fed, they will not die for a week or two, or even longer, but the lengths of their periods of growth are greatly increased.
b. habitat relations of the soil-inhabiting tiger beetles (characteristic data)
The environmental relations will be illustrated by the relations of three species of Cicindela purpurea Oliv. subspecies limbalis Klg., tranquebarica Herbst, and sexguttata Fabr.
Fig. 6 The pupa. Three times natural size.
a. Cicindela purpurea limbalis
The adults are beautiful red and green, though not strikingly conspicuous forms. Eggs are laid in June; the larvae hibernate usually in the second instar and pupate in the second summer. The imagoes emerge about a month after pupation, hibernate, and become sexually mature late in the third June. The larval life lasts twelve to thirteen months; adult life, ten months; two years between generations.
1 . General behavior of adults. They are not strong fliers, but are very alert and start to fly whenever one approaches them. The form of the moving object is not important; ^ize and movement produce the reaction apparelitly without reference to form and color. I have not been able to ascertain that they turn and face an approaching person with any degree of uniformity,
as is asserted by Comstock ('04), and have never seen them fly into vegetation, or crawl into crevices.
2. Ecological relations of adults, a. General conditions at the point of study. My studies have been conducted along the west shore of Lake Michigan between Lake Bluff and Winnetka,
Fig. 7 'Diagram showing Lake Michigan bluff as seen from the zenith. U, level
surface of upland; BL, bluff; SB, sandy beach; M, water, L. Mich.; /, piers; toward the left is north; sand has lodged on the north side of the piers. AB and
CD indicate positions of cross-sections below.
Fig. 8 Cross-section AB. Slumping bluff stage. The adults of C. limbalis are distributed from .4— B ; the larvae, sparingly, from E to F. Other letters as in fig. 7.
Fig. 9 Cross-section CD; stage of some bluff stability and bare clay exposure. Adults of limbalis between C and D; larvae plentiful between G and H. Other letters as in fig. 7.
Illinois, but my attention has been concentrated on the habitats near Glencoe, Illinois.
Between the points mentioned, the lake is eroding its morainic shores. Steep banks have been formed by this action which are from 1L4 meters (38 feet) to 20.4 meters (68 feet) in height.
PHYSIOLOGICAL ANIMAL GEOGKAPHY 561
The steepness of the slope makes conditions severe for plant life. It is only where inactivitj^ of forces of erosion has decreased the steepness of the slope, that scattered plants are present. Where the slope is still less steep, the bluff is sometimes covered with forest.
On the upland adjoining the bluff are stretches of meadow, woods, and sometimes pastures, all intersected by paths, roads, and ravines. All these furnish bare ground which apparently is essential to these tiger beetles. At the base of the clifT is frequently found a narrow stretch of sandy beach, which varies in width from 1 to 25 meters (figs. 7, 8 and 9).
h. Local distribution. The adult beetles of C. limbalis are found on the upland near the bluff in all of the bare places just
Fig. 10 The burrow of C. purpurea limbalis; p.c, pupal cell. One-third natural size.
described, and on the steep clay bank and the sandy beachabout equally distributed in proportion to the area of the bare soil exposed (figs. 7, 8, 9). If the number be grekter in any one of the situations, it is on the sandy beach. If the adults be about equally distributed on the different areas, which of these are we to consider the habitat of the species? Let us inquire into the habits of the larvae.
3. Ecological relations of the larvae, a. Local distribution. I have carefully watched the larvae of this species (fig. 10) in their external environmental relations for five years in the vicinity of Glencoe. They are found almost exclusively on the clay bank (fig. 11). Occasionally larvae are found in bare places on the steep banks of the ravines. Three or four individuals were once found on the top of the bluff in a bare place on level ground,
EXPLANATION OF FIGURES
11 The habitat of C. purpurea limbalis near Glencoe, Illinois, showing several stages in the development of the forest on the bluff. The area to the right of the imaginary line between a and 6 is stable enough to support some sweet clover. Here the tiger beetle larvae are most abundant. The area between lines joining a and b and a and c is in the early shrub stage. To the left of a c the shrubs are denser, and larger and some trees are present.
12 Habitat of C. tranquebarica in the pine zone of the ridges at the south end of Lake Michigan. The dark portion in the foreground is the shadow of a tree. At the left is the cattail zone of the depression; between a and h, the sedge zone; between b and c the zone of high depression plants. The white blossoms here are those of Parnassia caroliniana; their distribution, September, 1906, corresponds approximately to the distribution of the larvae of C. tranquebarica which arose from eggs laid in May and June, 1905. The portion above and to the right of c represents the higher portion of the ridge and the habitat of C. scutellaris.
but these are the only ones found in such a situation, as compared with over six hundred actually dug from the clay bank. They are entirely absent from the sandy stretch at the base of the bluff. I have dug very many larvae of Cicindela lepida from this situation, and have never found a single larva of C. limbalis. However, larvae of C, limbalis are not equally abundant on all parts of the clay bluff. The portions which are very steep, subject to land slides in the spring, and very dry in summer, are essentially without larvae. The forest covered portions are without larvae. The shrubby parts are inhabited only in the open places. The bare places with a few herbaceous plants have the greatest number of larvae.
b. Migration of larvae. As I pointed out in 1908, the larvae of this species rarely migrate, but remain at the point where the egg was laid. Only fifteen per cent of them left their burrows during a period of two or three weeks after they had been dug from their normal habitat and placed in holes made with a wire in moist sand. In eighty-five per cent of the cases the larvae smoothed off the sides of these burrows, and remained in this very unnatural situation — one in which all of the physical conditions had been changed. The steep, sloping clay had been replaced by level sand, resistantly packed particles of clay, by coarse sand grains, and the sohd edge of the burrow (fig. 10) by the crumbling sand. In the field, I have never seen larvae crawling on the ground. Burrows have been found empty in a few cases in digging about six hundred larvae. In one or two cases the dead larvae w^ere found in the burrows and as these would soon disintegrate and leave the burrow apparently empty, vacant holes may have been left in this way. Then again, ants may overcome a larva, and after chewing off" its antennae and tarsal joints, drag it from the burrow. While larvae may occasionally migrate, the empty holes are not so numerous but that their occurrence may be due to other causes.
c. Local distribution of larvae dependent upon adjustment in egg-laying. The larvae vary in position from year to year apparently with the weather conditions at the time of egg-laying. They live for a little more than a year. In 1906 the full grown larvae
PHYSIOLOGICAL ANIMAL GEOGRAPHY 565
were found on the higher and drier parts of the clay bank. The eggs from which these larvae were hatched were laid in June of 1905. The total rainfall at Chicago, from January to June inclusive, was 42.5 cm. (17.1 inches), for April, May and June 29.0 cm. (11.5 inches), and for May and June 21.0 cm. (8.4 inches) and for June 8.0 cm. (3.2 inches).
In 1907, on the other hand, they were on the low places near the springy situations and in small gullies, the eggs from which these larvae hatched having been laid in June 1906. The rainfall from January to June inclusive in 1906 was 29.0 cm. (11.6 inches), from April to June inclusive 14.5 cm. (5.8 inches), for May and June 10.0 cm. (4.0 inches), and for June 4.7 cm. (1.9 inches).
The failure of the larvae to migrate stands out clearly even a year after the egg-laying took place. The larvae of this species usually adjust the depth of their burrows to the temperature conditions of the sand in which they were placed under experimental conditions. In nature, however, I doubt that these larvae can dig their holes deeper when the soil becomes dry and the temperature high, because at such a time the clay is very hard.
d. Relation of larval habitat and distribution of food to the distribution of adults. In a natural indentation of the coast at Lake Bluff, Illinois, a beach of considerable w4dth has been deposited and the bluff bears a very dense forest. No larval habitat is present, and the adults of C\ limbalis are not present on the beach. Their food is at least as abundant here as where the clay bank is bare.
As we have stated, the tiger beetles feed over an area much more diversified and much greater in extent than the breeding place or larval habitat.
The adult beetles feed on any available animals. The feeding areas which are adjacent to like breeding places differently located, are frequently very different, and are occupied by very different food species. The food is then of necessity different for forms living in different places.
In captivity the adults have been fed with lean meat of various sorts (beef, pork and mutton) which they eat readily when fresh. They also pick up the ants, Thysanurans, etc., in the cages, and when not fed, devour their own species. The food relations are. then, highly regulatory, the animal feeding on available food.
4. Experimental studies of habitat selection, a. Methods of experimentation.^ Do the adults select their egg-la3dng place? To answer this question, adults were placed in cages containing soil of several kinds. Each kind was so arranged into steep and level parts, that about one square foot of each type was exposed. The adults placed in the cage were taken when the species was copulating freely. The soil was kept very moist up to the time the first ovipositor holes were made because this species lays only in moist soil. After this the wetting of the soil must be done very cautiously, so as to prevent washing eggs from the ground in steep parts. Accordingly, the holes were not obliterated from day to day and the counts are not accurate for the soil in which a large number were made because eggs are sometimes laid very close together and adjoining holes destroyed. Some eggs are deposited in irregular cracks and crevices where they are likely to be overlooked. The greatest care was taken to discover every hole made in the soils in which larvae do not occur in nature.
h. Results. The following table shows the approximate number of holes made in the clay and probably the actual number
The distribution of ovipositor holes and larvae of C. purpurea limbalis under experimental conditions
Lot I.
Lot II.
Lot III.
Holes. . Larvae . Holes . . Larvae Holes. . Larvae.
CLAY,
9 ETS.
FOREST
HUMUS, 1 PT.
CLEAN
CL^x
HUMUS, 1 PT.
HUMUS
SAND,
9pts.
SAND
s
L
S
L
S
1
s
L
S
L
9
21
5
12
1
17+
7+
1
24
10
»
S = steep; L = level.
- Each experiment requires daily attention for from one to two months, as well
as considerable greenhouse space.
PHYSIOLOGICAL ANIMAL GEOGRAPHY 567
made in the other soils, together with the number of larvae which appeared; 80 per cent on the steep slope, 98 per cent in clay.
The count of holes includes some in the first stages of digging, mere scratches on the ground, and others which had been excavated to the usual depth with or without eggs being laid.
c. Factors controlling egg-laying. Pairs taken in coitus were placed in cages containing sand only and level clay only. No larvae appeared in either case. The experiment with the level clay has not been repeated. Females placed in cages containing rough, steep clay, deposited eggs. Eggs are also absent from dry soils, whether steep or level.
Slope, kind of soil and soil moisture are factors governing the deficiency or absence of eggs. A deficiency or excess in any one of these respects decreases the number of eggs laid, or causes them not to be laid at all. The animals are in the condition for egg laying for a short period.
d. Method of selection. It has been determined by opening holes that eggs are not laid in all, and in one case the first holes made bj^ a female were empty. This would tend to show that they try the soil before laying the eggs, but I have not been able in other cases to determine whether the first holes contained eggs or not. To determine this, it would be necessary to watch R female all of the time during several days.
d. Geographic distribution, a. Distribution of the species. This species occurs from the Island of Mount Desert on the coast of Maine, northward along the coast of Nova Scotia and New Brunswick, up the St. Lawrence River, through the region of the Great Lakes, and westward across the northern part of the great plains to Alberta, and south along the eastern slope of the mountains; southward in the upper Mississippi Valley to St. Louis. Its place in the great plains and southern prairie region is taken by other forms recognized as other color varieties, but so far as is known, similar in habits. These forms are the varieties or subspecies transversa, splendida, amoena, denverensis, and ludoviciana. Splendida has recently been recorded by Sherman from western North Carolina. Occasional specimens are recorded from Kentucky, Tennessee, and northern New Jersey. It is evident,
568 VICTOR E. SHELFORD
since none of these states were mentioned in the old state records of early collectors, that the beetles have dispersed into this region with the cutting of the timber and the building of roads and railroads, which have exposed large areas of clay bank.
There are also taxonomic difficulties and lack of knowledge of larval habits. No map of the distribution of the species will be published until we have investigated the subject further. It is clear, however, that the distribution area of limbalis is one in which moist clay banks are common. It represents the margin of the ice sheet, the region of extensive clay deposits which are being eroded rapidly, and the slope of the mountains where erosion is also rapid. It is closely correlated with the behavior of the species. Its geographic distribution appears to be determined by the same factors as its local distribution.
6. Geographic variation in habits. The relations to soil and topography do not vary greatly geographically. The various races mentioned as occurring in the southern part of the range of the series differ sufficiently in structure and color to constitute subspecies in the opinion of good taxonomists. Still a number of observers, Messrs. Lantz,Wickham,Wolcott, Smyth, and Cloverdale, tell me that the adults of all are associated with clay banks.
Near Chicago the larval life is a little more than a year, thirteen to fourteen months, and the adult life ten to eleven months. Criddle ('10) has confirmed his statement ('07) that the larval life lasts two years in Manitqba. The depth of larval burrows in Manitoba is 15 cm., near Chicago, Illinois, 5-10 cm.; the adult burrows at Aweme are 15 cm.; at Chicago, in captivity, 5-8 cm.
b. Cicindela tranquebarica Herbst
The usual color of the adults in eastern North America is brown. The life-history differs from that of C. limbalis in the following points: (1) Eggs are laid in May; (2) larvae pass the winter in third stage.
1 . General behavior of adults. They are a little shyer than C. limbalis and more difficult to capture. They start when approached by a moving object, and when alighting, frequently
PHYSIOLOGICAL ANIMAL GEOGRAPHY 569
turn toward the observer. They almost never alight on vegetation. When caused to fly up from a narrow path, they frequently fly in a circle and return to a point behind a person moving forward. I have never seen them crawl under objects when pursued. They excavate burrows for the night and cloudy days.
2. Ecological relations, a. Area of special study. They have been studied specially at the south end of Lake Michigan. Here the species is found only on the ridges with pines. These ridges were originally thrown up under water near the shore. By the falling of the surface of the lake, which has amounted to a total of 18 meters since glacial times, ridges have been left out of water perhaps about as fast as they were formed. We have, then, a series of them of different ages, arranged in order of age. The youngest are nearest to the shore. Their width varies from five to thirty meters. Long, narrow ponds of corresponding age occur between the ridges. As a given ridge came above the surface of the water, it often received wind-blown sand; there is little or no vegetation on the youngest ridges.
h. Local distribution. C. tranquebarica is absent from the ridges with sparse vegetation. On the ridges on which young conifers are found together with various herbaceous plants along the pond margins, C. tranquebarica is present. Adults are numerous along the margins of the ponds and all over the ridges, particularly on the sandy 'blowouts,' or points where the wind has removed some of the sand and keeps the vegetation from growing up. The beetles frequently burrow into the sand for the night and for hibernation. Food is abundant on the white sand areas and the beetles find advantage in its conspicuousness, which no doubt causes them to congregate on these places to feed.
When an area of denuded sand, in which ponds or depressions are present, is deposited or exposed, vegetation appears first nearest the water. Humus accumulates, blackening the soil and making conditions favorable for more plants, so that a turf is soon formed near the water. Similar processes are going on higher up on the side of the pond margin and it is soon captured by the plants. It is on the ridges in which the soil just above the very moist or sedge zone is blackened by humus, but still not completely occupied by the roots of plants, that we find C. tranquebarica.
570
VICTOR E. SHELFORD
The succession of plants does not end here, and we find shrubs coming in and the turf migrating farther and farther up the slope of the pond margin. Shrubs shade the pond margin. The pines on the ridges are displaced by oaks and the undergrowth of herbaceous plants becomes denser; the pond margins are densely covered with turf or shaded by shrubs and trees. Though the higher portions of the ridges, namely, the feeding grounds, are still bare, C. tranquebarica is not to be found. The species must then have some vital relations to the pond margin.
Fig. 13 The upper part of the burrow of C. tranquebarica, pupal cell shown by dotted line. One-third natural size.
3. Ecological relations of the larvae, a. Local distribution. The general behavior of the larvae is similar to that of C. limbalis The holes are, however, deeper and straight (fig. 13). The larvae of C. tranquebarica are found in clay, alluvium, or sand, and have been reared or identified from all of the kinds of soil mentioned in the discussion of the adults. In our sandy area of special study, they are found near the pond margins only. In all the locahties referred to in connection with the adults, the larvae have
PHYSIOLOGICAL ANIMAL GEOGRAPHY
571
been found in soils with a moisture content similar to that near Chicago.
b. Migration of the larvae. The larvae of this species rarely migrate. I have watched the larvae that appeared in the experimental cages after the soil had been permitted to become very dry, but none of them moved during several weeks.
c. Variation in local distribution. The distribution of larvae in Gary in 1906 corresponded to that of the white blossoms of Parnassia caroliniana which are shown in fig. 12. Their position varies from year to year, according to the rainfall as in the case of C. limbalis.
d. Relation of the larval distribution to the distribution of the adults. On the pond margins where herbaceous plants have been displaced by shrubs, C. tranquebarica is not present, although the higher parts of the ridges are bare and much like they are where C. tranquebarica is present, indicating that the adults disappear with the larvae.
4-. Experimental studies of habitat selection a. Method. Adults of C. tranquebarica were placed in cages which were much like those which were used in the work on C. limbalis, but the soil was all essentially level.
b. Results. The results were as follows :
The distribution of ovipositor holes and larvae of C. tranquebarica under experimental
conditions
CLAT
HUMUS
SAND, 9 PTS.
HUMUS, IpT.
CLEAN SAND
SAND, I PT.
HUMUS, 1 PT.
Holes
7
13+
19
wanting
1907 I
Larvae
4
25
1
wanting
Holes
?
3
25+
18
wanting
I
Larvae
11
3
31
1
wanting
1908/
Holes
16
wanting
29+
11
46
Larvae
5
wanting
41
7
24
c. Factors controlling egg-laying. One striking difference is that the females did not lay with the same precision as did the
572 VICTOR E. SHELPORD
females of C. limbalis. Very many holes were made in the fresh, clean sand, but eggs were laid in only a few of them. These holes in the fresh sand have frequently been opened and found to be without eggs. Why fresh, clean sand should be so attractive to the females and fail to satisfy the final act of egg-laying is strange. Pure humus appears to be avoided when either moist or dry.
During the experiments, the different kinds of soil were kept as nearly equally moistened as possible, but a slight depression was provided in each. These were wetter and were especially selected by the females when standing water was not present. Eggs are not laid in dry or very wet soil. Moisture is evidently an important factor in controlling the egg-laying. I have found the beetles copulating and depositing eggs in my cages, on damp, cloudy days. This has not been observed in the case of most other species. It would appear, then, that light is not very important. However, as in the case of C. limbalis, deficiency or excess in one factor is sufficient to cause the soil to be avoided or only little used.
5. Geographic distribution. The habitat relations of C. tranquebarica are less definite than those of C. limbalis. We have found it on the bare clay of the overflow flats of the Arkansas River at Dodge City, Kansas, depending on stream moisture; on a path at the top of a terminal moraine at Waverly, New York, depending on climatic moisture ; on alluvium along the Des Plaines River at Lyons, Illinois; and on the residual and alluvial soils of various parts of Colorado, New Mexico, Nevada and Idaho. In nearly all these localities, the soils examined were similar in their moisture content. The species is always nearer water courses in the more arid climates. The only place in which the soil moisture was deficient about the burrows was at Las Vegas, Nevada, at the height of the dry season. This is a region of winter rain, where the soil would be much moister in spring, the egg-laying season of the species. The larvae were much nearer the water (Las Vegas Wash) than I have found them in the moister climates. The bottoms of the burrows were nearly as moist as we commonl^^ find them near Chicago.
PHYSIOLOGICAL ANIMAL GEOGRAPHY
573
This species iiicliules several races which seem, according to the accounts of entomohjgists and my own observation, to be very similar in habits. It stretches across the middle region of North America, and ranges from the sea level to 7536 feet and throughout four of the zones of Merriam without regard to vegetation, efficient temperature or other climatic condition (table 3). A consideration of the races involved is necessary (Horn, '05; Wickham, '06). The records represented by dots (fig. 14) adjoining the Pacific Coast are for well recognized races. All others have been cast into synonomy by good taxonomists. The remaining records including 1-9 of fig. 14 are then for a single race. Furthermore two of the races sometimes recognized, horiconensis and Wickham's southern race, have been produced b}^ suitable conditions during the late larval and pupal life. Table 3 shows the relation of a single race to climatic conditions.
The relation of the distribution of C. Iranquebarlcd to climate. Vegelalid/i a ml rainfall are approximated, especially for Alberta axd B. ( '. Life zones an a piimxi niattd tvhere detailed maps are not available. The vegetation at Kalso is in (ju'stian l)iit the species has been taken in the mountain.'^, near, where there can be U I th iloutdthat it is coniferous forest. The numbers at the left refer to fig. I4.
PLACE
STATE
ALTITUDE
IN FEET
.MEAX:
RAIN FALL
I.N INCHES
VEGETATION
ZONE
Cmerriam)
COLLECTOR
1
Woods Hole
Mass.
5
45.0
deciduous
forest
Transition
Author
2
Meridian
Miss.
358
58.0
deciduous
forest
Lower Austral
U.S.X.M.
3
Alamosa
Col.
7536
15.0
steppe
Upper
Author
4
Aweme
Man.
1180
17 45
steppe
Boreal
Criddle
5
Innisfail
Alb.
3600
15.0
steppe
Boreal
T. X. Willin ji
6
Calientc
Nev.
4407
7.0
desert
Lower Austral
-Author
7
Hagernian
Id.
2600
10
semi-desert
Upper Austral
Author
8
Kalso
B. C.
1870
25
conifer forest
Boreal
L W. Cockle
9
Bridge])ort
Cal.
64
3.7
desert
Lower Aus
Wickham
tral
JOURNAL OF .MORPHOLOGY, VOL.
574
VICTOR E. SHELFORD
Fig. 14 The distribution of C. tranquebarica as shown b^- locality records.
The map indicates general topograph}-. The numbered localities are selected to
show relations of the distribution of a single race to topography, climate, vegetation, and Merriam's zone (p. 573).
PHYSIOLOGICAL ANIMAL GEOGRAPHY 575
Such distribution is characteristic of species which occupy environments made by streams, lakes, soil, or other local conditions. Such species are local in their distribution.
6. Geographic variation in habits. The life-history at Chicago is similar to that of C. limbalis. Criddle expresses the opinion that the larval life is two years in Manitoba, but has not yet confirmed the statement. The depth of larva burrows at Chicago is 22-50 cm. ; at Aweme, Manitoba, 43-50 cm.
c. Cicindela sexguttata
This is a brilliant green form. Its life-history difiers from that of C. limbalis in the following points : (1) Egg laying occurs about one week later, (2) larvae pass the winter in both second and third stages, (3) the adults emerge in August, but usually remain in the pupal cells until spring.
1 . General behavior of adults. The adults of this species are less alert than those of the other species just discussed. They frequently fl,y and alight on leaves of bushes. When frightened in the woods they frequently crawl under a leaf or other object on the ground. Sometimes they remain very quiet for a time when the body is not all covered and the bright green wing covers stand out in contrast to the brown leaf under which they are hiding.
They crawl under the bark of trees at night and in cool or cloudy weather, both in nature and in cages, and rarely dig into the soil. But one individual moved soil when in captivity. This one was in a cage in which a piece of bark lay on the sand present. It was found to have removed a small amount of sand to make room for its body under the bark.
2. Ecological relations of adults. C. sexguttata has been studied in Massachusetts, New York, Illinois, Indiana, and Tennessee. It lives only in or about forests of a certain particular type. It is entirely absent from those that have developed on low% wet ground, such as marsh forests and humid climate flood-plain forests; it is not found in the early stages of the oak forest nor in the beech and maple (the climax forest of eastern
576 VICTOR E. SHELFORD
North America) nor in the cotton wood or true coniferous forests. It is abundant in and about the white-oak, red-oak hickory forest (figs. 15 and 16).
CUmatic conditions influence the relations of this species to different types of forests, e.g., in eastern Tennessee they are found in much more xerophytic forests than in the vicinity of Chicago where the rainfall is appreciably less.
EXPLANATION OF FIGURES
15 General view in east Tennessee.
16 An open place in the oak and hickory forest of the mountain side, a typical C. sexguttata habitat. The individuals were seen copulating on the log in the foreground.
578 VICTOR E. SHELFORD
The Cumberland mountain district was originally completely forested. The forest of the valleys was chiefly beech and maple; of the mountain slopes, oak and hickory; of the mountain tops conifer. The soils are various, resulting from many different kinds of rock. We were unable to find this species in the strictly red cedar, pine or beech forest. However it occurs in the more mesophytic oak containing ravines of strictly conifer forests and in forest of mixed oak and conifer. Sherman records it from such situations also. This is not true near Chicago.
The beetles come out into the little streaks of sunshine on fallen trees and bare ground in the early forenoon to feed. The writer has seen them picking up insects from the logs in such locations. From my observations in the field I am confident that bare spots of mineral soil, fallen trees, etc., are essential to this species.
Fig. 17 The burrow of sexguttata. One-third natural size.
It is only in such places in virgin or little disturbed forests that I have found them copulating. However it is not a particular type of forest that is essential to this or any other species of tiger beetle, but a certain environmental complex in which a certain consistency and moisture of soil and a certain amount of sunlight and bare ground are the essential things.
3. Ecological relalions of larvae, a. General behavior. The burrow resembles that of C. limbalis and is shown in fig. 17. This larva is less active than those of the other species, but otherwise is similar in habits.
6. Local distribution. The larvae of this species are very difficult to find because they are for the most part under leaves.
PHYSIOLOGICAL ANIMAL GEOGRAPHY
579
In eastern Tennessee I found them in bare spots on the steep mountain slopes where steepness of slope had prevented the accumulation of leaves, and in parts of the forest that had been fired recently and the leaves accordingly removed. They occur in clays resulting from the weathering of the following rocks: Briceville shale, Newman limestone, Knox dolom'te, Chicamauga limestone, and Conasauga shale. Near Chicago they have been found on clay till, and on sandy till mixed with humus.
4. Experimental studies of habitat selection, a. Method. This species has been placed in cages containing various kinds of soil as have the others. The light relations were the same as in all of the other experiments, although the light is of more importance.
b. Result: Tables 4-8. Distribution of larvae of C. sexguttata under experimental conditions.
TABLE 4 Relation to slope; sand dry at surface
1907
CLEAN
SAND
SAND, 9 PT.
HUMUS 1 PT.
CLAY, 9 PT. FOREST
' HUMUS, 1 PT. HUMUS
s
L
S 1 L
s
L S j L i S
L
Holes
Larvae...,
0+ 0+
12 4
0+' 0+ 1
7 : 5 1 jo
S= steep L = level
Relation to shade; sand dry at surface. Sunlight in cages is reduced to one-third glass roof and cage screen
CLEAN SAND
SAND, 9 PT.
HUMUS, 1 PT.
CLAY
POBE8T
HUMUS
REMARKS
s
u
^
u
s
u
Holes
?
?
8 8
8
1
S = shaded
Larvae .'
U = unshaded
580
VICTOR E. SHELFORD
TABLE 6 Clean sand, moisl
1908
CLEAN
SAND
SAND, 9 PT.
HUMUS, 1 PT.
CLAY
sand.Ipt.
hcmus.Ipt.
ONE LEAP PLACED IN
EACH CAGE
Lotl
Holes
5 1 5+
4 1 6
1
2+ 0+
18
4
04 1
■10412
54 13
1
1
5
1
36
29
Lot 2
Holes
Clay very dry
Larvae ....
4
246
3
10 3
11437
Lot 3
Holes
Two under leaf
Lot 4
Holes
Lai'vae
Three under leaf
TABLE 7 Relation to thick covering of leaves
Larvae .
SAND, 9 FT. I humus, 1 PT. I
34
28
SAND, 1 PT. HUMUS, 1 PT.
f All soil covered i with leaves ex[ cept clean sand
TABLE 8
Total larvae shown in tables 6 and 7
— Hru"s.Vjx. -- nir^^l^T. I «-—
Larvae .
90
56
43
Grand total 210
Total under leaves: 76.
PHYSIOLOGICAL ANIMAL GEOGRAPHY 581
c. Factors controlling egg-laying. The relations to light and leaves are interesting. The larvae were frequently at the edges of the piles of leaves in such positions as eggs might be placed by females, the posterior half of whose bodies were concealed by the leaves (fig. 18). Females have been found making ovipositor holes, when the posterior two-thirds of the body was under a leaf.
The depositing of eggs in the unshaded portions of the cages may be due to the reduced intensity of light, or the shadow of the cage frames, which falls upon some parts of the soil at any
Fig. 18 Showing the position of the larvae of C. sexguttata in a cage. The
black spots represent the larval holes. The stippled portion represents the
approximate area in shadow during the middle of the day.
given time. Nearly all parts pass into shadow for a time during each day. All of the eggs may then have been deposited when all or a part of the animal's body was in the shade. The reduction of the intensity of the light to one third that of normal out-ofdoor light may act as a partial shadow to this species. The experiments will be repeated in the full sunlight.
As will be seen by an inspection of the tables most of the eggs were deposited in the sand with a little humus. None were laid in pure forest humus. The fresh sand and clay were ignored when they were allowed to remain dry. Muddy places were avoided. It is evident that egg-laying is governed by (a) kind of soil, (b) soil moisture, (c) slope, (d) light, (e) temperature (activity only
582 VICTOR E. SHELFOBD
on warm days). Under conditions unsuitable in any one factor, few or no eggs are laid.
5. Geographic distribution. The geographic distribution of C. sexguttata is exactly what the general habitat relations would lead one to expect. Within fifty miles of Chicago, I have found it alwaA^s associated with the white-oak and red-oak and with a single exception also the shag bark hickory. The same is true in east Tennessee. Comparing the distribution of the trees, we find that the combined extent of the white-oak and hickory represent almost exactly the distribution of this species (fig. 19). That is, the geographic distribution is the exact function of the local distribution (Ruthven, '07).
Fig. 19 A combination of the maps of Schimper '03, and Transeau, 03, showing the geographic plant formations of North and northern South America and the distribution of Cicindela sexguttata.
la, c, d are forests with broad thin leaves.
la. Dense tropical evergreen forest, rain-forest.
Ic. Dense temperate evergreen forest, temperate rain-forest.
Id. Deciduous forest. The large black dots in this area represent locality records of C. sexguttata; the heaviest dots combined with crosses are placed over the centers of states from which it is recorded. The lines .r and y show the relation of its distribution to that of two characteristic trees of the deciduous forest. The continuous line {x) represents the distribution limits (except along the Atlantic Coast of the white-oak (Quercus alba) ; the broken line {y) represents the distribution limits of the shag bark hickory (Hicoria alba), except where its limits are coincident with those of the white-oak. The distribution of these is not the same as that of the deciduous forest because the map is based on area with more than twenty per cent of woodland. In the savanna region (36) these trees occur along the streams as does C. sexguttata.
2. Coniferous forest (with narrow thick leaves).
This is mixed with the deciduous forest in the region of the Great Lakes. In southern Unites States it does not properly belong to this map because it is dependent upon soil rather than climate (p. 600).
Sa. Tropical steppe and savanna.
Sh. Temperate savanna.
3c. Temperate grassland or steppe.
4. Evergreen forest with broad thick leaves.
5a. Scrub or thorny forest which merges into desert.
5b. Desert; 3-5 is very arid desert-like steppe.
Unshaded area in the north is tundra.
PHYSIOLOGICAL ANIMAL GEOGRAPHY
583
19
584 VICTOR E. SHELFORD
It is not to be understood that these forms are in any way directly related to the trees, but the trees represent the general conditions in which the beetles will live and reproduce. The species is an inhabitant of one of the 'climatic' realms and will be found continuously distributed where the forests are continuous.
6. Geograjjhic variation in life-history. C. sexguttata rarely appears in northern localities in the autumn and it is probable that it remains in the pupal burrows until spring. The species is reported as appearing both autumn and spring in some southern localities. At Chicago, the adults appear during April and May, while in the western part of the geographic range of the species they do not appear until late in June, after the heavy rains which soften the soil, so that the imagoes can dig to the surface.
d. Other species
1. Experhnental studies of habitat selection. By similar methods, I have determined the breeding place of the following species: C. scutellaris, high, dry sand with a little humus, or sand which is not shifting; C. formosa generosa, slightly shifting sand; C. lepida, shifting white sand; C. duodecimguttata, very moist dark soil ; C. punctulata, soils with some humus and moist at egg-laying time; C. purpurea, same as punctulata, but in moister places, not repelled by considerable grassy vegetation, bare spots necessary as breeding places. In every case the range of the adults is far wider than the breeding grounds.
2. Geographic variation in habits. In captivity the larvae of all the species studied at Chicago close the burrows near the mouth and go to the bottom when the soil is dry. Here they remain inactive until water is applied. No such closures have been noted in the field, except C. lepida, which lives on the dry sand dunes. Criddle ('10) says:
In Manitoba, there are often long intervals of inactivity of the larvae of manitoba, venusta, limbata, lecontei, and probably others, during the summer months. At such times the larvae close their burrows at
PHYSIOLOGICAL ANIMAL CEOGKAPHY
585
the to]), and renuiiii apparentlN' williout food, and do not grow appreciably. In 1907, larvae of venusta and limbata closetl their holes on June 12, and some did not appear again until August 25, nearly two-anda-half months. A few, however, would open up at night, throw out a lot of earth, and then retire again. These larvae were always active when dug out. This strange habit may be due to the dryness of the soil to some extent, though it is not altogether so, as holes have remained closed during wet weather, and they are always opened in autumn or late summer, and deepened before winter, no matter what the condition of the ground is. The extreme heat of the sun may also be a factor of some importance. The beetles are unquestionably influenced by temperature, and will go into winter quarters earlier on a dry, hot fall than they do during a cold one, and hot summer days are much ]ireferred for commencing winter homes.
The following table shows the depth of burrow'S at Chicago and at Aweme, Manitoba. It includes all available data.
TABLE 9
Showing the relative depths of hibernation burrows of adults and burrows of larvae, of the same species, in the same soil. Manitoba, Criddle {'07, '10). Compare with table 10
i
■ o z
Larvae
i
Adult B
CHICAGO
IBERJ.ATION
CHICAGO
AWEME
AWEME
SPECIES
Depth of
Burrow
Depth of Burrow
Depth of
Adult in
Hibernation
Depth of
Adult in
Hibernation
cm.
cm.
cm.
C7n.
C. linibalis
clay
5-10
15-20
clay
5-8
7-15
C. trauquebiirioa
sand?
22-50
43-50
clay
15
15-30
C. formosii
sand
30-50
125-200
sand
62-117
C. scutellaris. .
lecontei
sand
25-45
70
sand
25-64
C. lepida
sand
60-90
145-175
C. 12-guttata . .
allu
vial
5-10
15-37
clay
10-15
5-25
C. repanda
sand
10
clay
5-10
15-20
586 VICTOR E. SHELFORD
TABLE lU
Shvin'rig the coniparadve mctcorulngical ronditiotis in Chicagu and Brainlui {25 miles northwest from Aweme) during active period of tiger beetles, April to September, and during the time of digging hibernation burrows {September)
2
a
w to
S O
il
S
is
1. 1
"^ o
E
is [
^g
5
<
a
■ ■ 05
•< B
K a.
s
■^ s
D f
a
< s :
ss
5
S j.
S a
S a
tt.
< ^
2 K
O '^
2 «
zS
5?^
^
Z B
r
3 '^
-. ^
K '^
< ^
a f
a sc
a 'f
&
PS
H
s
K
S
S
S
H
in.
Ars.
<ie5.
per cent ,
about
deff.
<?e9.
deff.
Chicago
19.3
1695
70
100
4S
71
57
33.4
Brandon
12.45
1310
6S
SO
33
66.4
38
17 45
On comparison of the data for the two points in question in table 10, we see that the amount of rainfall, the extremes of temperature and the length of season as well as the amount of sunshine differs widely at Chicago and in the vicinity of Aweme, Manitoba. Comparing these data with those found in table 9, we note that the larval burrows are deeper in the climate which is most arid and coldest in winter. Likewise the depth of the hibernation burrows is greatest where the temperature is lowest during the period of digging.
The shorter season, fewer hours of sunshine, and drought accompanied by the periods of inact vity described by Criddle may be the cause of the longer ife-histories referred to in the case of the three species especially considered here.
C. GENERAL COISTSIDERATIONS
We have noted various features of tiger beetle behavior. X discussion of the general bearing of this will now be presented under the following heads: (1) Importance of the breeding instincts and breeding place, (2) the relation of behavior characters to habitat and associated forms, (3) the meaning of variation in behavior, (4) relation of local and geographic distribution; importance of various factors.
PHYSIOLOIGICAL ANIMAL GEOGRAPHY 587
/. The unportance of the breeding instincts and the breeding
place
We have shown that the adults range over an area much greater than that which the larvae inhabit and that a species is entirely absent where feeding habitats .of the adults is represented and the egg-laying place or larval habitat absent.
Those tiger beetles which hibernate in situations different from the one in which the larvae are found, always return to the breeding place to deposit eggs. When the breeding place disappears, the species also disappears. The larval habitat or egg-lajdng place is much narrower and more definitely circumscribed than any other part of the habitat. The breeding place and the breeding instincts are, then, the center about which all other activities of the organism rotate. They are the axis of the environmental relations of these organisms.
a. Comparison with other activities and relations. The breeding place and breeding instincts must usually be considered in connection with the feeding ground, and feeding instincts as well as other factors. The tiger beetles will not breed where there is not sufficient nourishment for considerable periods. The feeding place is often the second consideration after breeding. In the tiger beetles, however, the feeding structures and habits are so generalized that their food is plentiful everywhere, and the food relations need only be mentioned. A third important environmental relation is that to means and place of escape from those environmental factors which tend to destroy the organism, such as its enemies, extremes of weather or climate, etc., but all these are of secondary importance.
b. Fixity of breeding instincts. The determination of their degree of modifiability or fixity would require experimental work which I have as yet been unable to accomplish. There is, however, good evidence from field study that the breeding instincts are most fixed of all the instincts. Since such behavior characters in the tiger beetles are usually specific or racial, they are probably modified only by the gradual processes of evolution.
588 VICTOR E. SHELFORD
2. Relation of the hehavior to habitat and associated forms
There is the greatest difference in the behavior of the different species. I have as yet been unable to study this critically, but it is at least a very promising field.
0. Behavior and habitat. As we have noted, C. sexguttata has for example, various peculiarities of behavior which are related to the forest conditions in which it lives, which are not possessed by other forms. As we noted, when it is frightened from a rock or bare place, it frequently alights on the leaves of a low tree or bush and crawls under the bark of trees for the night, or even to hide when pursued. None of the other species which I have studied behave in such a manner. I have never seen C. tranquebarica crawl under objects when pursued. It does not alight on the green leaves of trees or shrubs when they are present. It excavates burrows instead of crawling under objects. The behavior of these two species is correlated with the general environmental conditions.
b. Inter-physiology , and inter-psychology . Tarde ('03) has recently written an article on inter-psychology — the psychology of the relations of individuals of the same species (man). To this should be added the behavior between different species, while acting or living together as one. He suggests that the social psychology of man may be traced to the inter-psychology and physiology of the lower animals. If this is true, then we can be more certain that the inter-psychology of the higher forms has developed from the inter-phj^siologv of the lower forms (Craig, '08 [2]).
I have looked for the inter-physiological manifestations in these beetles, but have found none striking except the mating instincts. There is little or no social life. I have found animals belonging to totally unrelated species attempting to copulate in some cases where the two are dissimilar.
It seems quite evident from my observations that the more marked phases of the behavior of the tiger beetles arise not from inter-physiology, but from relations to the species which are quite
PHYSIOLOGICAL ANIMAL GEOGRAPHY 589
different in behavior and habit from the beetles themselves. This I propose to call intermores-physiology or psychology.'^
c. Intermores-physiology. We have seen the behavior of these beetles when pursued, their flight, alighting only to wait for the moving object to come near when they start up again, the hiding under leaves of C. sexguttata, etc. All this is illustrative of the behavior which is related to forms antagonistic in behavior and habits.
The study of the behavior of forms which live together in the same situation from the point of view of the relations of the behavior of the different species is a promising field of investigation. It will throw much light on the problems of psychology as well as ecology.
3. The meaimig of varMtion in habits
We have noted geographic difference in the length of the lifehistory and the depth of the burrows. In 1908 we pointed out that severe conditions increase the length of the various stages. Criddle has noted that the larvae do not feed for a considerable period in the summer. This accords fully with my experimental results on the larvae of the Chicago species They stop ceding and close their burrows when the soil becomes too dry, or the condition otherwise severe. The lengthened life-history of Manitoba forms may be due to the shorter seasons and the failure of the larvae to feed for a considerable period.
We pointed out also ('08) that the larvae respond to stimuli by deepening their burrows. The soil conditions in Manitoba have not been studied, but the different depths of the burrow under different experimental conditions is suggestive.
The correspondence between experimental results and the differences in the so-called habits in the different localities suggests that the apparent variation in habits may be only a regulatory behavior response that probably would be found common to most individuals of the species. This could be settled by experimental study.
^ Mores (Latin), 'behavior,' 'customs,' 'habits.'
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 3
590 . VICTOR E. SHELFORD
4. The relation of local and geographic distribution
We have suggested in the case of the three species here considered, that the geographic distribution of each is the geographic distribution of the conditions in which it lives and breeds. We have visited several of the different climatic realms in which each of these three and many other species occur. So far as ordinary observation can go, the breeding and general living conditions are similar in the different localities, even though the climate, as in the case of Cicindela tranquebarica, is very different. The same conditions are found by the species through its moving near to water in the arid climates, as compared with the more moist climate.
It is customary to conclude that condi ions are the same because the species is the same. Here we have tried in a general way to determine whether the species is the same throughout its range, by the study of the condition, and experimentation on the animals. This is the only mode of attack that can yield definite results. It is highly desirable, however, to carry the observations on soil and other environmental factors further with the use of physical factor instruments. It is equally desirable to conduct actual experiments on each species of beetle at a number of points, especially those near the outskirts of its geographic range. This would, no doubt, reveal differences of detail which we have overlooked, but which cannot, so far as present observation goes, be of great importance.
5. Factors governing geographic distribution
Our data show clearly that the breeding pe iod is crucial as determining the local distribution, and that an excess or deficiency in any one factor is sufficient to decrease the number of individuals present, or cause them to be absent entirely. Any factor differing sufficiently from the optimum for a given species is sufficient to limit its distribution. There can be no important difference whether a deficiency in moisture is due to insufficient rainfall or to distance from or height above water, or whether an excess of temperature is due to latitude or exposure and accordingly the same factors must govern both ocal and geographic distribution of the tiger beetles.
PHYSIOLOGICAL ANIMAL GEOGRAPHY 591
III. THE PHYSIOLOGICAL CHARACTERS AND DISTRIBUTION OF GROUPS OF SPECIES (FORMATIONS)
A. ZOOLOGICAL OPINIONS AND DIFFICULTIES
There is, I believ^e, a general opinion among laboratory zoologists to the effect that no important generalizations can be made from data concerning the environmental relations of animals. In other words, the data of natural history cannot be organized into a science.
There are at least three good reasons for the prevalence of such views. The first of these is that such zoologists are often familiar with only a few of the very common species of animals, common because their habitat relations are such that they can flourish in the conditions which civilization produces or because they do not have definite habitat relations, being in this respect an exception to the rule. The lack of attention to the taxonomy of common forms is also a factor. Animals which belong to different species, genera, or even families, are often quite similar in appearance and so are sometimes regarded as single species. Articles regarding American species have occasionally been published under the names of European species not found in this country, or at least rare and confined to northern latitudes.
The second reason results from the fact that relations of animals to their environment are not understood. Often we do not discriminate between the important and unimportant periods of relation to environment in a life-history. The third reason lies in the fact that the environment of animals is also not understood and the various stages and phases have not been classified so that habitat relations can be readily described.
The lack of knowledge of taxonomy and the simpler facts of natural history requires no discussion. On the other hand, our knowledge of animal behavior and animal physiology has been but little applied in the study of animals in nature, and the knowledge of environments, which is in the hands of the plant ecologists and geographers is not at all well known among zoologists.
592 VICTOR E. SHELFORD
B, THE NATURE OF THE ENVIRONMENT
The animal environment is a complex of many factors. Each is dependent upon another or several others, in such a way that any change in one factor affects several. Some of the most important environmental factors are water (atmospheric moisture), oxygen, carbon-dioxide, nitrogen, temperature, pressure, currents, excretory products, food, enemies, and materials for abode (soil, vegetation, etc.) . In nature, such combinations of these and other factors, in the proportion requisite for the maintenance of the life of a considerable number of animal species, are cal ed environmental complexes (Davenport '03).
C. ENVIRONMENTAL RELATIONS OF ANIMALS
The only features which space will permit us to discuss are the physiological and ecological relations. In this field we must confine ourselves to a comparison of plants and animals and the bearing of the important environmental relations on geographic distribution.
1. Comparison of the environmental phenomena of plants and
animals
An organism is a system of inter-dependent and definitely related processes (i.e., a system in dynamic equil brium). The definite relations of the inter-dependent processes of the organism (dynamic equilibrium) may be disturbed by changes in the external forces which surround the organism and to which the processes are adjusted (Jennings, '06). Such a disturbance is what we ordinarily call stimulation.
With this idea as a background, we give in parallel columns a comparison of the more obvious relations of plants and animals to their environments, as shown by experimental work. The column on the right is written by Dr. H. C. Cowles, Associate Professor of Plant Ecology in the University of Chicago.
PHYSIOLOGICAL ANIMAL GEOGRAPHY
593
a Comparison of the responses 'of (motile) animals and (sessile) plants
ANIMALS (motile)
I. Animal behavior is evident because of motility;
II. When an external stimulus is applied to an animal, it responds mainly by movements which are usually more or less random, and which bring the organism into various conditions, one of which relieves the disturbance and the organism resumes normal activity, in conditions which brought the relief. These conditions are not necessarily advantageous. . . . (Jennings.)
III. Animal behavior is usually plas tic, i.e., may be modified by external stimuli, but sometimes appears rigid.
IV. (Animal structure is only slightly
plastic. The plasticity usually occurs in the early stages). Structural modifications are rarely of importance in the life of the animal.
PLANTS (sessile)
I. Plant behavior is inevident because of lack of motility.
II. Plants respond more evidently through changes in form and
III. Plant structures are usually plastic but frequently appear rigid.
IV. Structural modification of plants is often of importance in the life of the plant.
V. The motile organisms of a given
V. The plants of a given habitat
habitat usually have common
usually have common structure
behavior characters. Com
and functions, or those that
bined structural and behavior
are ecologically equivalent. "
characters of comparable forms
of a given habitat, or of similar
habitats are ecologically equiv
alent."
VI. The breeding activities of ani
VI. The reproductive organs and em
mals are probably least modifi
bryonic stages of plants are less
able and least regulatory, but
modifiable than the vegetative
are governed by the same laws
stages of adults.
as the other activities. (Shel
ford, '07, '10).
' The meeting of the same general conditions in a different way constitutes ecological equivalence. The term was first used by Cowles.
594 VICTOK E. SHELFORD
b. Discussion of the parallel statements. (II) Animal (ojmotile organism) distribution at any given time is a better index of the condition at that time than the distribution of plants, because when the conditions at a given point become unfavorable, the animals (or motile organism) move to another situation, while the plants (or sessile organisms) remain or die.
(V) The fifth is not well established. However, a preliminary testing, for example, of the rheotaxis of a arge number of brook animals has shown them to be strongly positive, strong positive rheotaxis being a common behavior character. Many of them have special means of attachment which may be brought into play with great speed.
The darters are strong swimmers and are able to live in rapids by virtue of their swimming powers and positive reaction, while the snails (Goniobasis) which occupy similar situations, are able to maintain themselves because of the strength of the foot and positive reaction. The two are ecologically equivalent. The sixth statement appears to be generally true, but needs experimental confirmation.
The proposition may be summarized as follows : The behavior and general mode of life of animals are the superficial equivalent of the structural phenomena in the vegetative parts of plants. Behavior and vegetative structure are convenient indices of physiological conditions within the organism.
To illustrate this still further, let us consider the plants of the sand areas at Chicago and in Manitoba. As compared with Chicago plants, the plants of Manitoba differ in size and structure under the more arid conditions found at the point where Criddle's studies were made. The Manitoba tiger beetles do not, so far as I can find, differ from the Chicago forms in any of the structural characters which have to do with their meeting those conditions, hut they dig their holes deeper and require longer time for transformation. The tiger beetles of the desert and semi-desert and the tropical sand areas (Bates and Westwood, '52; Snow '77; Lucas, '83) are usually nocturnal or crepuscular; those of moister and cooler areas are diurnal^ — a difference in behavior. Desert plants are structurally adapted to withstand the desert conditions
PHYSIOLOGICAL ANIMAL GEOGRAPHY 595
(Schimper '03) and differ in this respect from plants of cooler, moister situations. Again, the difference between the tiger beetles which deposit their eggs in different soils is not structural difference in ovipositor, but a different physiological response of the organism.
The activities mentioned are general and may be separated, into feeding, breeding, etc. Probably all are governed by the same general laws. In the study of all the animals of a given environment we are confronted with the question of what activities are most important, just as in the study of particular species.
2. The most important environmental relations of animals
The strength of a chain is the strength of the weakest link. The activity which determines the range of conditions under which a species will be successful is the activity which takes place within narrowest limits. Failure to recognize this principle is in part responsible for the general chaotic state of our knowledge of natural history. Men have often failed to interpret the relations of animals to their environments because they have regarded the records of the occurrence of all stages of the life-history as equally important. They have considered the occurrence of the most motile stage in the life-history important, as for example, the occurrence of an adult butterfly. Plant ecologists would have met with equal success if they had studied only the environmental relations and distribution of wind-disseminated seeds. While we have indicated above that the breeding activities are most limited (Merriam, '90; Allen and Verrill fide Merriam, '90; Adams, '08; Shelford, '07, '10; Reighard, '10; Herrick, '02; Clark, '10), there are no doubt exceptions to this, and at the present stage of our knowledge the subject is one for investigation. Whatever this activity may prove to be in a given case, it is of great importance and deserves comment, both as to method of investigation and bearing on distribution.
a. Method of determining the most important activities. The first step is field observation — the continuous watching of animals throughout a number of life cycles. Experimentation is almost
596 VICTOR E. SHELFORD
always necessary also. It is only under unusually favorable conditions that the relative importance of the various periods of the life-history of an animal can be ascertained, without experimentation. On the other hand, experimentation must be correlated with field observation. Simple experimentation on the behavior of animals in the laboratory does not illuminate this matter to any appreciable extent.
3. The relation of physiological characters to geographic range
Our studies of animal distribution usually consist of a list of names of species with a statement of the distribution of each, followed by such interpretation as suits our particular purposes. Attempts actually to study the environment in any detail, or the reactions of animals to the conditions of environment are rare indeed. Furthermore, the groups most studied (higher vertebrates) are probably least dependent upon their environmental complexes; they are often decidedly migratory and because of their size least adapted to experimental study.
Some quite extensive attempts to correlate geographic range with meteorological conditions have been made, but always with only implied reference to the physiological character of the organisms themselves, and usually with the use of species as an index of conditions. A few factors have been emphasized, and these usually in the sense of barriers. Merriam ('90, naming also Allen and Verrill but not citing their papers) emphasizes temperature; Walker ('03) atmospheric moisture. Heilprin ('87, p. 39), like most paleontologists, emphasizes food. There appears to be no adequate basis for the idea that the same single factor governs the distribution of most animals. Such a conclusion probably results from leaving the organism out of consideration.
Since the environment is a complex of many factors, every animal lives surrounded by and responds to a complex of factors, at least in its normal life activities within its normal complex (Jennings, '06, p. 180). Can a single factor control distribution?
A large amount of physiological study of organisms has been conducted with particular reference to the analysis of the organism itself, but with little reference to natural environments.
PHYSIOLOGICAL ANIMAL GEOGRAPHY 597
Many of the factors and conditions employed in such experiments are of such a nature that the animal never or rarely encounters them in its regular normal life. Other experiments are, however, attempts to keep the environment normal, except for one factor (Jennings, '06, p. 180). These have demonstrated that in ordinary reactions an animal responds to the action of a single stimulus. Certain general laws govern the reaction of animals to different intensities of the same stimulus.
a. Laws governing the reactions of animals. The laws governing the stimulation of animals in the experiments of the laboratory are familiar subjects in the textbooks of physiology (VerwornLee '99). With respect to a given factor used in the experiment, it has been found that there is a range of conditions within which the activities of the animal proceed without marked stimulative features. These are called optimal conditions. Take, for example, temperature. There is in most animals which have been subjected to experimentation with temperature, a range of several degrees in which the animal is not markedly stimulated (optimum). As the temperature is raised or lowered from such a condition, the animal is stimulated. If the temperature be continuously raised, a point is reached at which the animal dies. The temperature condition just before death occurs is called the maximum. The lowering of temperature produces results comparable in a general way to those of high temperature. The condition just before the death point is reached is called the minimum. With various limitations, unimportant in this connection, the same is true with respect to each of the various factors which an animal encounters in nature. Which factor determines the limitations of occurrence of an animal on the earth's surface? The answer to this is suggested in Liebig's Law of Minimum.
6. Law of minimum. Liebig's law of minimum is summarized by Johnstone ('09, p. 234):
A plant requires a certain number of food stuffs if it is to continue to live and grow. Each of these food substances must be present in a certain proportion. If it is absent the plant will die; if present in a minimal proportion the growth will also be minimal. This is true no matter how abundant the other food stuffs may be. The growth is then dependent upon the amount of food stuff present in minimal quantity.
In nature this law applies both geographically and locally. As
apphed to animals it includes both food and material for abode.
The presence, absence and success of a species is determined hy
the necessary material which is absent or present in minimal
quantity.
c. Law of toleration of physi-M factors. We have noted (p. 581) in the case of the tiger beetles, that for the egg-laying to take place the surrounding temperature and light must both be suitable, the soil must be moist, probably also warm, and must satisfy the ovipositor tests with respect to several factors. Egglaying, the positive reaction, is then probably a response to several factors. Furthermore, after the eggs are laid, the conditions favorable for egg-laying must continue for about two weeks if the eggs are to hatch and the larvae reach the surface of the ground. The success of reproduction depends, then, upon the qualitative and quantitative completeness of the complex of conditions. The negative reaction, on the other hand, appears to be different. The absence of eggs, the number of failures to lay and therefore the number of eggs laid in any situation can be controlled by qualitative or quantitative deficiency or excess with respect to any one of several factors. The presence, absence, or number of eggs laid is, then, determinable by a single factor, according as it is near the optimum or near either the maximum or minimum tolerated by the species. It is, however, not necessary that a single factor deviate ; the effect is similar or more pronounced if several deviate.
In nature the presence or absence, or success of a species or group of species, its numbers and sometimes its size, etc., are largely determined by the degree of deviation of a factor or factors from the range of optimum of the species or group of species. The cause of the deviation in the factor or factors is not of importance. For example, in the case of a soil inhabiting species such as Cicindela tranquebarica, to which considerable moisture is necessary, the cause of the deficiency in one case may be climatic deficiency in rainfall, in another a rapid drainage due to steep slope and porosity of soil. The former is what we have called a climatic (geographic) condition and the latter a
local condition. The evidence for the law of toleration as applying to distribution is good so far as the local distribution is concerned and, since the same factors are involved in the geographic,
there is no difficulty in the application of the law to geographic
distribution also. The fact that in so far as our observation can
go at present, the tiger beetles are found in similar conditions
throughout their ranges, is also good evidence for the application
of both the laws of minimum and toleration to geographic distribution. In fact the law of minimum is but a special case of the
law of toleration. Combinations of the factors which fall under
the law of minimum may be made, which makes the law of toleration apply quite generally; for example: food and excretory
products may be taken together as constituting a single factor.
From this point of view the law of toleration applies, the food
acting on the minimum side, excretory products on the maximum.
d. Application of the law of toleration to geographic distribution. The so-called centers of distribution are often only areas
in which conditions are optimum for a considerable number of
species (Transeau, '05; Adams, '02 and '05). The relation of the
law to centers of distribution is shown in the diagram below;
above the line is the scale of stimulation with the limits of toleration shown and below the parallel relation of the distribution and
relative abundance.
Minimum limit Range of Maximum limit
of toleration optimum of toleration
I \ \ I
I II I
< Center of distribution - — ■ — ^
Absent Decreasing Greatest abundance Decreasing Absent
On account of the nature and distribution of climatic and vegetation conditions, it follows that as we pass in one direction from a center, one factor may fluctuate beyond the range of toleration of a species under consideration ; but as we pass in another direction it is very likely to be a different factor. The. divisions of Merriam's zones into arid and humid portions is an illustration of this, and seems to constitute a begging of the temperature question.
600 VICTOR E. SHELFORD
4. Tentative laws of distribution
On this general basis tentative laws of distribution may be formulated.
a. Governing the limit of geographic range. The geographic range of any species is limited by the fluctuation of a single factor (or factors) beyond the limit tolerated by that species. In nonmigratory species the limitations are with reference to the activity which takes place within the narrowest limits. In migratory species this activity limits the range only during a part of the life cycle.
b. Governing distribution area. The distribution area of a species is the distribution of the complete environmental complex ivithin which it can live as determined (1) by the activity which takes place within the narrowest limits and {2) by the animal's power of migration. Barriers in which some one factor of the complex fluctuates beyond the limits of toleration of the species at all periods of its life-history may prevent the animal from reaching all the suitable habitats, but this is the result of the working of the laws rather than an exception, and faunistic animal geography begins where physiological animal geography ends.
D. CLASSIFICATION OF ANIMAL ENVIRONMENTS
While this is a necessary subject for discussion, it is with much hesitancy that I undertake it here, where brevity is necessary. Obviously, since our subject is physiological animal geography, we shall confine our attention mainly to those aspects which are geographic in extent in the sense that they are nearly uniform over a considerable area of the earth's surface.
If one is to understand the most elementary principle of synecology,^ he must first recognize the distinction between local (edaphic, Schimper, '03; minor and secondary, Adams, '08), and climatic or geographic (extensive) environmental complexes (major,
~ Synecology is the ecology of formations. In the classification of formations and environments, no nomenclature has been established for the larger or climatic units. Dr. Cowles tells me that plant formations do not represent climate and therefore 'climatic' should not be used. However, every ecologist and geographer knows the significance of 'climatic' and 'local.' The geographers object
PHYSIOLOGICAL ANIMAL GEOGRAPHY 601
Adams, '08). The climate of a region and all that goes with the climate are a climatic or geographic complex. Opposed to these are local complexes, such as water (streams or lakes), soil, exposure, or lack of exposure, etc. For example, in the Mohave Desert the climatic conditions are characterizable as hot, arid, etc., but within the desert are streams fed by mountain rainfall. These streams are local conditions in themselves, and also produce other local conditions such as moist soil, etc. These are not dependent upon the dominant conditions within the desert.
The relation of local and geographic condition has been the subject of much careful consideration by Cowles ('01), Schimper, ('03), Shelford ('07), Adams ('08) and ('09).
We will turn our attention, then, first to an inquir}^ as to the best index of climatic or geographic condition or major environmental complexes.
1. The index of climatic or geographic conditions or of major environmental complexes
The vegetation from the standpoint of whether it is forest steppe, or desert, etc., does not involve animals, and represents climatic complexes in a general way. It is the most important factor in the control of temperature, moisture, light, food and material for abode^ and is a good index of the conditions which surround animals. Tentatively it may be used as a basis of classification of the animal environments. A knowledge of these environmental complexes may be acquired from the data of physiography, meteorology, plant ecology and physiological plant geography (Schimper, '03).
to the use of the term 'geographic' for the climatic enviromnents because, to them, the local environments are equally geographic. Every zoologist understands the meaning of geographic' and 'local.' Adams' terms, 'major' (climatic) and 'minor' (local) are to be preferred but one must continually explain their meaning. The writer uses 'climatic' and 'geographic' here because their meaning is clear.
- Material surroundings have been regarded as of great importance in the case
of mammals. Hagenbeck states that he always supplies an environment which resembles as far as possible the natural environment. He has imitation icebergs for polar bears, etc., and finds that this adds greatly to the success of keeping his animals in captivity.
E. THE ANIMAL FORMATION
Animals select their habitats, probably by trial and error, as is indicated by the making of additional holes and parts of holes by the tiger beetles only to reject them without laying eggs. The simple fact of selection is, we believe, very familiar to all naturalists.
A given environmental complex is selected by a number of species. All of the animals of a given habitat constitute what is known as an animal formation (Warming, ('09) ; Clements, ('05) ; Schimper, ('03); Adams, ('08); Grisebach, ('48, fide Clements).
It follows that there is a certain physiological or ecological similarity and ecological equivalence in the forms that thus select the same or similar complexes. It follows^ that the animals of different deserts, different deciduous forests, different steppes, etc., are ecologically and physiologically similar or ecologically equivalent if the deserts, the forests, and the steppes, etc., are similar (Adams '05).
Tentatively, formations may be characterized in general physiological or ecological terms (mores). The characterizations of plant formations have thus far been largely based on growth-form. Attempts to find structural similarity among animals of similar habitats, while not failing in particular cases, have led to no good results or generalizations (Ritter, '09). The great difficulty with this point of view is that it must, because of the great difficulty of investigation, remain for a long time largely a matter of speculation. The attempts which have been made are based on natural selection speculations or Lamarckian speculations.
It should be noted further that the relations of a given group of animals to their habitat and to each other is more complex than that of the plants which are commonly treated in this manner.
The conspicuous plants of a given environmental complex, except in the tropical forests, are usually rooted in a single plane which greatly simplifies the relations of plants to their environments.
^ A term is needed to cover such characters. The term mores (Latin), 'customs,' 'behavior,' 'habits' is suggested as best covering the need. It stands opposed to form and forms; thus steppe mores meaning the behavior of characteristic steppe animals or an animal or animals with characteristic steppe behavior.
PHYSIOLOGICAL ANIMAL GEOGRAPHY 603
Animals, on the other hand, have different habitats which are not related to one plane, and so must be separated into similar groups for purposes of the comparison of one formation with another. For example, the animals which burrow into the ground in a given environmental complex cannot be compared with those that live in trees in another, but must be compared with subterranean forms. Accordingly, for comparison, animals must be separated into : (a) burrowing forms, (b) ground forms, (c) arboreal forms, etc.
1. Classification of animal formations based on environmental
relations
a. Principles of classification. We have noted that all of the animals of a given environmental unit constitute a formation, and that environmental units are classified into climatic or geographic (extensive) and local. The groups of animals which occupy the climatic or geographic environments may be called ' climatic or geographic animal formations.' The groups of animals which occupy the local environments are called local formations (societies, or associations).
If one is to study the relation of animal physiology and behavior to the environmental conditions, in so far as this can be done by field study, these distinctions must be kept clearly in mind. For example, in dealing with animals of the great North American steppe area, to treat together all forms found here (as is common practice) would lead to endless confusion from our point of view. The forms which belong to the water (aquatic), those that live in the timber along the ravines, in the sand areas, are forms belonging to local formations. Those that occupy the plains proper belong to the steppe formation. Some forms may belong to both, in which case the facts should be taken into account.^"
1" An animal should be associated: first, with the breeding conditions; second, with the feeding conditions; third, with the conditions affording shelter. Calvert_ ('08) attempted to find correlation between the distribution of Odonata and vegetation zones with negative results. Aside from the reasons given by the author, it should be noted that Odonata breed in the water and, excepting forms breeding in water holding plants, belong to local conditions, and no correlation was to have been expected. Correlation of the distribution and species is, however, not essential to our point of view.
604 VICTOR E. SHELFORD
b. Climatic or geographic animal formations of the world based upon physiological similarity and ecological equivalence under similar conditions'^ The distribution of the similar environments is given by Schimper ('03) and Transeau ('03, '05) and in fig. 19. Only the environments and distribution of the formations is given here; much concerning the mores of the different format ons may be obtained from the existing literature but we do not have it well enough organized to present here.
1 Formations of forests with broad, thin leaves.
a Tropical rain-forest formations (fig. 19, la). Environment: Dense forest with broad thin leaves, two or three heights
of trees, uniformly distributed rainfall and nearly uniform temperature. Distribution: Large areas Mexico and Central America (Belt, '88), ^^
and South America (Bates, and [Clodd, '93J), southern Asia and East
Indies (Wallace, '94), and several small areas in Africa (Garner, '01). b Monsoon-forest formations. Environment: Similar to the rain-forest but with a dry season in which
the leaves fall. Distribution: Adjoins areas of rain-forest, c Temperate rain-forest formations (fig. 19, Ic). Environment: Similar to the tropical rain-forest, but much less luxuriant
and in different climatic conditions. Distribution: East coast of northern Mexico, southern U. S., western
Chile, southern Japan (Kobelt, '02), New Zealand. d Temperate deciduous forest formations (fig. 19, Id). Environment: Similar to the temperate rain-forest, but much less dense
and deciduous. Distribution: Eastern North America, north to the Great Lakes; Chile,
north to 35° (Darwin, '45, p. 242); Europe, north of the Alps (Mosley
and Brown, '63, p.) and south of 60° (Kobelt, '02; Brehm, '06); Japan
and vicinity of Okhotsk.
2 Formations of forests with narrow, thick leaves (coniferous forest formations;
further study will probably subdivide these) (fig. 19, 2).
Environment: dense evergreen forests with little undergrowth.
Distribution: North America, north of the Great Lakes and Columbia River extending southward in the mountains (Seton, '09) ; Eurasia, north of 60° and southward in the high mountains (Brehm, '96).
" This outline is essentially that arranged for a committee of the Geographic Society of Chicago on the Classification of Geographic Materials, and is parallel to one for plants by Dr. H. C. Cowles and to one for Human Geography by Dr. J. P. Goode and Miss J. B. Obenchain.
12 Some characteristic literature on the natural history of the various formations is cited where possible. See Thomson's introduction to Brehm ('96 j.
PHYSIOLOGICAL ANIMAL GEOGRAPHY 605
F'ormations of savannas and grasslands.
a Warm savanna and steppe formations (fig. 19, 3a). Environment: Dry season in spring; scant rainfall; grassland with scattered thorny trees, occasionally thickets, and dense forests along larger streams. Distribution: The great plains of Africa (Roosevelt, '09-' 10), and South America, b Cool savanna formations (fig. 19, 36). Environment: Similar to the warm in aspect, but more often with trees
in groves. Distribution: A narrow belt nearly surrounding the Great Plains, Uruguay, Australia, and eastern Siberia (Brehm, '96). c Cool steppe formations (fig. 19, 3c). Environment: Cool, dry, winters cold, grassland with trees only along
the principal streams. Distribution: The great plains of North America (Craig, '08;Seton, '09), south central Asia (Brehm, '96), De La Plata southward to Patagonia (Hudson, '92). Formations of forests with broad, thick leaves (fig. 19, 4).
Environment: Subtropical conditions with winter rain and hot, dry
summers. Distribution: California, the Mediterranean region, Chile (near Valparaiso, Darwin '45), South Africa, southwest Australia. Formations of deserts and scrub areas (semi-desert).
Environment: Various types of arid condition with thorny vegetation. a Scrub or semi-desert formations (fig. 19, 5a). Distribution: Mexico, Texas and Central America (Belt, '88; Bailey, '05), eastern Brazil, southern South America, arid Australia (in part), northeastern Africa (Plowden, '68), India, and China, b Desert formations. Distribution: Southwestern North America (Merriam, '90), South America, Sahara and Arabia (Brehm, '96), central Australia and south Africa. Tundra formations.
a Arctic tundra formations.
Environment: Cold, treeless, with short cold summers. Distribution: Circumpolar (fig. 19). b Alpiuie tundra formations. Environment: Similar to a. Distribution: Mountains above the tree line. Formations of fresh water.
a Still water formations (lakes, ponds and sluggish streams). b Turbulent water formations (swift streams and eroding lake shores). Formations of the sea and its shores (amphibious formations, principally breeding on shore, feeding in sea).
a Ice-bound shore formations (Arctic) (Brehm, '96; Shackleton, '10). b Tropical and temperate shore formations.
c Oceanic Islands formations (the island fauna, representing the migration of land animals by sea) (Wallace, '92).
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 3
606 VICTOR E. SHELFORD
9 Formations of the waters.
a Formations of the sea (marine) (M'Intosh, '04).
1 Limestone bank formations (littoral) (Brooks, '93).
2 Rocky (eroding) shore formations (Littoral) (Verrill, '72; King and
Russell, '09).
3 Sandy (depositing) shore formations (littoral).
4 Open sea formations (pelagic) (Heilprin, '81).
5 Deep sea formation (mudline and abysmal).
It should be noted that the various formations of the Ust are to be found duplicated or essentially so in various parts of the world. This point of view emphasizes the resemblances in the behavior and ecology of forms living under sirnilar conditions. In the case of the great zoogeographic regions, there is no duplication, and differences are emphasized. '-*
" The distribution of aquatic animals is governed by : a. Kind of bottom (Sumner, '09). b. Depth, current, temperature and all other factors which are modified by depth, etc.
" There are no doubt several valid objections to such a classification, when thus statically stated and as mapped by some workers, such as Schimper. We present it thus because the recognition of the existence and general features of a phenomenon must -precede its analysis. However, one of the most important of these objections arises when one inspects a number of maps of the distribution of species. Such an inspection shows that the distribution areas of some species are bounded by the limits of the deserts, steppes, forests, etc., while those of others bear no relation to these regions. The former afford no difficulties while the latter deserve further comment. Species that, apparently, do not fit our classification fall under three heads : 1 . Species whose range is far greater than that of any realm or plant formation, covering perhaps several realms. 2. Species that occupy only a part of the plant formation in which they belong. 3. Species whose range lies within a region intermediate between two realms or plant formations.
The first group is made up of species dependent wholly or in part upon local conditions. Some species are always associated with local conditions, e.g., C. tranquebarica, p. 574, fig. 14. Such forms are relatively independent of climate, geographic plant formations, etc., and are dependent upon such conditions as are afforded by streams, sand areas, lakes, etc.
The species which are in part dependent upon local conditions usually belong properly to the climatic or geographic conditions of one formation, and invade another formation in local conditions which happen to be like the geographic of the one, in respects essential to that species. For example, some of the species of Orthoptera belonging to the great plains, or North American steppe region, invade the sand areas in northern Indiana where the climate is suitable for forests. Such phenomena are common and have been discussed by Adams ('02, '09).
PHYSIOLOGICAL ANIMAL GEOGRAPHY 607
IV. THE PROBLEMS, METHODS, AND RELATIONS OF PHYSIOLOGICAL ANIMAL GEOGRAPHY
A. SOME PROBLEMS OF PHYSIOLOGICAL-ANIMAL-GEOGRAPHY
1. Behavior problems. That the behavior of animals reflects their general conditions of existence, I think will not be seriously doubted. Some of the geographic problems may be stated as follows :
a. Behavior and geographic conditions. How much, and what features of the geographic conditions, for example, such as the steppe, the tundra, or the tropical forest, are reflected by the behavior of animals? Are these characteristics acquired by the individual or are they hereditary? In connection with the first question, I quote Brehm on the Arctic fox:
His whole character and conduct are quite different from those of our reynard and his near relatives. One scarcely does him injustice in describing him as a degenerate member of a distinguished family, unusually gifted, intelligent, and ingenious. Of the slyness and ingenuity, the calculating craft, of his congeners he evinces hardly any trace. His disposition is forward, his manner officious, his behavior, foolish. He may be a bold beggar, an impudent vagabond, but he is never a cunning thief or robber. He follows his worst enemy; without fear he approaches a man sleeping in the open, to snap at a naked limb.
The behavior of the penguins of Antarctica as described by Shackleton is equally interesting. Is it, or is it not, a picture of the hard struggle, intense cold, and monotony of the tundra?
" — Continued.
Our second group (or species which occupy only a part of the formation to which they belong) is important. Maps of the distribution of trees, by Transeau ('05) illustrate this. An inspection of these shows that there is a central area in the formation, in which species are most numerous, and in which we may conclude the conditions for the majority of the forms are best (optimum). Suitable investigation would no doubt .show that species thus narrowly distributed are limited by the termination of their necessary conditions, and that relative numbers are dependent upon the law of toleration.
Our third type, or species which occupy intermediate ground between the realms, are few so far as observation has been recorded (Ruthven, '07).
The above discussion is, however, based on the distribution of morphological species. If, however, there are physiological differences, behavior differences, or even regulatory responses in the different formations, morphological species and their distribution, are unimportant matters.
bUo VICTOR E. SHELFORD
h. Inter-psychology and inter-physiology (between ecologically similar forms). The problems of the inter-psychology (Tarde, '03) and inter-physiology (p. 588) are equally important in connection with the relations suggested above. Some aspects of inter-psychology are not inter-specific, but concern forms with similar habits. In the steppes ecologically similar animals frequently act as one species. Mr. Roosevelt has said: One of the most interesting features of African wild life is close association and companionship so often seen between totally different species of game" (Roosevelt, '09). Mr. Roosevelt shows the zebra and hartebeest herding together.
c. Intermores-psychology and physiology (between ecologically dissimilar forms, or antagonistic forms). The relations of animals of different size, habits, etc., to one another involves the most striking features of behavior. Much of the behavior which tends to protect the species from enemies falls under this head. This aspect of behavior has its geographic as well as its local significance. For example, the problem of the effect of the presence or absence of large carnivores on the behavior of other animals present in a climatic formation would deal with the broader geographic side.
d. Geographic variation of mores. The phenomenon of geographic variation in behavior and physiology probably usually belongs to wide ranging species. The best available data are probably on the nesting habits of birds (Knowlton '09).
2. The 7nore purely physiological problems. Let us illustrate by the desert. The dominance of the reptiles in the desert is well known, and Dr. A. P. Mathews has called my attention to the fact that the excreta of reptiles is uric acid which is a substance of low osmotic pressure passing out with the feces in a dry state; little water is used in the disposal of the excreta. This, together with the thick skins, enables reptiles to meet the conditions of the desert. Desert mammals must meet the same conditions. In these, water is required to wash the urine out of the tubules. Mammals are few in the desert ; their physiological relations there are not well known; Swain ('03) has pointed out the high specific
PHYSIOLOGICAL ANIMAL GEOGRAPHY 609
gravity of the urine of the Cahfornia coyote. The fact that many mammals do not drink for long periods in the steppe and desert regions is well known. Livingstone ('58) noted it in the Kalahari Desert, Roosevelt ('09) in east Africa, and Craig ('08) in the case of the prairie dogs and birds of Dakota. (Verworn, '99 p. 280.)
B. METHODS
The methods of physiological animal geography have been indicated from time to time throughout the paper. The method may be characterized as combined experimentation and field observation, each conducted with reference to the other, and both conducted with reference to animal formations. A typical study with reference to the steppe would consist of {a) a field study of a number of carefully chosen steppe species, accompanied by experimental study of their behavior and of their physiology; (b) a comparison with a similar study in a steppe in another part of the world; (c) a study of steppe species ranging outside the steppe with a view to ascertaining variation in behavior or behavior differences, etc.; {d) a comparative study of steppes and other formations.
C. RELATION OF PHYSIOLOGICAL ANIMAL GEOGRAPHY TO OTHER
SUBJECTS
The problems of physiological animal geography lie close to those of human geography, sociology and psychology, and offer a field of observation which may be accompanied by experimentation.
1. Human geography. Its relation to human geography is especially intimate. Indeed, geographers have been disappointed with the data which zoology has furnished them. It is almost exclusively data concerning the taxonomy and morphology of animals. The parallelism between the geographic phenomena in animals and the relation of culture to environment lies not in the color and structural adaptations of animals, but in their behavior
610 VICTOR E. SHELFORD
characters which enable them to Uve under a given set of conditions and the behavior which those conditions produce. ^^
It is to be hoped that geographic studies such as we haVe outHned may be conducted on wild animals in connection with the geographic and psychological problems in man (Waxwieler, '06).
2. General biology and evolution. From the biological side alone, the more purely physiological problems present an interesting field which is sure to jdeld results of far-reaching importance. The day is, I believe, rapidly approaching when the physiologist will find it necessary to give more attention to the study of animals unacclimated to the gases and artificial surroundings of the laboratory. Indeed, the failure of students of behavior to study their animals in nature is probabl}^ constantly leading to misinterpretation.
It has not been m}^ purpose to point out the relations of this subject to the evolution of species. However, to the question of the evolution of behavior characters, of instincts, etc., this point
1^ While attempting to make comparisons between human society and man on the one hand, and plants and animals on the other, geographers, sociologists, and psychologists — in so far as I have been able to read their writings along this line — have compared structure in plants and animals with what is obviously not structure in man, namely, his culture and mental make up. AVaxwieler compares human society with the whole animal kingdom as constituting another society. McGee ('96) takes a similar position. In discussing the relation of cultures to environment, he says: "When the law of biotic development is extended to mankind, it appears to fail; for the men of the desert and shore land, mountain and plain, arctic and tropic, are ceaselessly occupied in strife against environmental conditions which transform their subhuman associates, yet men remain essentially unchanged, some taller, some stouter, some swifter of foot, some longer of life than others, yet all essentially Homo sapiens in every characteristic.
More careful examination indicates that the failure of the law when extended to man is apparent only. The desert monads retain certain common physical characteristics, but develop arts of obtaining water and food, and these arts are adjusted to the local environment. ..." He continues with the citation of other cases. In the light of our present knowledge, such adjustment of arts is comparable only to the adjustment of wide ranging species of animals in food, nest building, materials used in nest building, and other features of ecology and behavior (see also Hubbard '96; Mason '96). Goode ('04) called attention to the fact that physical changes in man are slow as compared with the changes in culture (see also W. S. Tower, '10).
PHYSIOLOGICAL ANIMAL GEOGRAPHY 611
of view is very important. Wliile many aspects of these problems are not geographic, many others are, and the study of physiological animal geography bears the same important relations to the study of evolution of behavior as did faunistic animal geography to evolution of species in the beginning of its study. The relations of animal behavior to the evolution of species has never been appreciated. It is obvious that the behavior of all animals is regulatory and tends on the whole to preserve the species and to retain it in the environmental complex to which it is adjusted; still only slight changes in the physiological characters of an animal will cause it to select a slightly different complex, open entirely new avenues of migration and change the distribution of the group of species to which it belongs. Such a change in physiological character would bring a group of species into an entirely different relation to all the so-called factors of evolution (McDougal, '08 ; Tower, '07, ' 10) . Students of experimental evolution have, in no case that has come to my attention, made any study of the behavior characters of their new races, while the morphological features have been pursued with vigor. Is it not time that students of evolution began to study the effects of behavior on evolution?
D. THE FUTURE BIOLOGY
In this paper we have sharply separated evolution and structure on the one hand, from physiology and behavior on the other. Space, clearness, and the condition of the subjects have forbidden that we attempt to unite them here. While it may be expedient to continue in this manner until our knowledge of physiology and behavior is commensurate with that of the other subjects, the following of such a course indefinitely, with respect to either morphological or physiological aspects of biology cannot, if it be general, bring about the best development or unification of biological science. Indeed, its present lack of unity is traceable to such a course followed until recently by zoologists generally.
If our understanding of the data of physiological cytology be correct, we may expect to find so-called structures of some sort within or among the cells concerned in function, which stand for
612 VICTOR E SHELFORD
or are correlated with each physiological state and physiological condition to which we have referred. Our methods may not at present be sufficiently delicate to detect such structure, or the processes which lie back of it, but we may, it is believed, confidently expect the necessary methods for the detection of such structures and processes, and especially their correlation with and relation to the more permanent and more easily recognizable morphological conditions.
We classify the responses and changes in animals as evolution, modification by the environment, behavior and physiological response. Are not all these, after all, but different expressions of the same or similar processes? Future investigations must answer this question and it is around this question that the future of much that is known as biology hinges.
V. GENERAL CONCLUSIONS AND SUMMARY
1. Distribution and dispersal
a. Every animal selects an environmental complex as its general habitat (pp. 566, 571, 579).
6. The breeding grounds are usually the most important index of the true habitat (p. 587).
c. Each species is usually distributed as far as its environmental complex extends, unless barriers are encountered; faunistic animal geography begins where physiological animal geography leaves off (pp. 573, 582, 598).
d. The success of a species within a territory and its limitations to that territory are determined by fluctuation of one or more environmental factors, toward or bej^ond the limit tolerated by the species (p. 599).
e. Species which select those environmental complexes which are determined by streams, soil, or other situations which occur only locally, are local in their distribution (pp. 574, 5).
/. Animals which select a habitat which is geographic in extent and which represents the dominant conditions of an area, are distributed throughout their area and are usually not so wide ranging as the species which select the local conditions (p. 582).
PHYSIOLOGICAL ANIMAL GEOGRAPHY 613
g. The dominant vegetation of a given area which possesses some degree of uniformity of cUmate (as, for example, the deciduous forest of the United States), is the best index of general conditions, as it not only presents the results of the conditions, but makes certain types of eiivironmental complex for the animals (pp. 582, 601).
h. The field of plant ecology and of ecological plant geography present the best data on the distribution of animal environmental complexes (p. 601).
2. The physiology and behavior of animals
a. In animals, behavior characters take the place of growthform in plants. Animal formations may be characterized by the behavior, physiological, and habitudinal relations (mores) of the constituent animals, while plant formations are superficially characterized by structural characters which indicate the physiological conditions of the constituent plant's (593).
b. Animal behavior, physiology and general mode of life (mores) probably reflect the geographic conditions such as climate, general surroundings (vegetation) and other animals present (pp. 588, 607).
c. Physiological animal geography offers a field for experimentation and observation which will have important bearing on human geography, sociology and psychology, and the general problems of biology and evolution (p. 609).
ACKNOWLEDGMENTS
The author wishes to express his indebtedness to the staff of zoology of the University of Chicago, especially to Prof. C. O. Whitman for encouraging the study of natural history; to Dr. C. M. Child, who suggested our type of experimental study of the tiger beetles several years ago: he is indebted also to Dr. H. C. Cowles for much advice and information in the field of plant ecology; to Dr. Wallace Craig and Prof. H. H. Lane for criticising the entire manuscript ; to Professor William Ritter and Mr. Ellis
614
VICTOR E. SHELFORD
L. Michael for suggestions; and to Mrs. Mabel Brown Shelf ord for tabulating the data furnished by the gentlemen whose names appear below.
I am especially indebted to the following, who very kindly sent me locality records included on the maps. Many of the localities are in remote parts.
Mr. C. C. Deam Mr. J. D. Evans Prof. J. S. Hine Mr. James Johnston Mr. A. W. Andrews Mr. W. Knaus Prof. H. F. Wickham Prof. A. L. Melander Prof. S. A. Forbes Mr. H. R. Hill Mr. G. P. Mackenzie Dr. E. C. VanDyke Dr. R. H. Wolcott Prof. D. E. Lantz . Prof. E. P. Felt Prof. Wm. Mcintosh Mr. Germain Beauleiu Mr. Chas. Stevenson Mr. B. H. Walden Mr. E. D. Harris
Mr. Albert L. Borrows Prof. G. W. Herrick Mr. Chas. W. Leng Prof. R. S. Woglum Mr. G. M. Dodge Mr. C. N. Ainslie Mr. Norman Griddle Mr. Tom Spalding Prof. S. A. Johnson Mr. Wm. Beutenmliller Mr. C. S. Brimley Mr. E. P. Venables Mr. I. W. Cockle Mr. E. M. Anderson Mr. T. N. Willing Dr. Henry Skinner Mr. F. F. Crevecoeur Mr. James Hunaen Prof. F. H. Snow Mr. H. P. Loding
I am also indebted to Dr. Swartz, Dr. Henry Skinner, Mr. William Beutenmliller and Dr. Samuel Henshaw for the privilege of examining the collections in their charge, from which data were obtained.
PHYSIOLOGICAL ANIMAL GEOGRAPHY 615
BIBLIOGRAPHY
Adams, C. C. 1902 Southeastern United States, a center of distribution of flora and fauna. Biol. Bull., vol. 3, pp. 115-131.
1905 Post-glacial dispersal of the North American biota. Biol. Bull., vol. 9, p. 53. ,
1908 Ecological succession of birds. 'The Auk, vol. 15, pp. 109-153. Bibliograph^^
1909 Isle Royale, Biological survey of Michigan. Rep. Mus. Univ. of Mich. Geol. Surv. (Lansing). Bibliography.
Bailey, V. 1905 Biological survey of Texas. U. S. Dept. Agr., N. A. fauna, no.
25. Banta, a. M. 1910 A comparison of the reactions of a species of surface Isopod
with those of a subterranean species. Jour. Exp. Zool., vol. 8, nos. 3
pp. 243-310 and, 4, pp. 439-488. Bates, H. W. and Westwood, J. O. 1852 On the habits of the Coleopterous
genus M^gacephala. Trans. Ent. Soc. London, 2d Ser., vol. 2,
pp. 49-58. Bates, H. W. (E. C. Clodd) 1893 The naturalist on the river Amazons, fith
Edition, London. Beddard, F. E. 1895 Zoogeography. Cambridge University Press. Belt, Thomas 1888 The naturalist in Nicaragua. London. Brehm, a. E. 1890 Vom Nordpol bis zum Equator. Stuttgart.
1896 From north pole to equator. London. Translation by M. R.
Thomson, introduction with bibliography by J. A. Thomson. Brehm, A. E. and Rossmassler, E. A. 1864-67 Die Tiere des Waldes, 2 Bd.
Leipzig (not seen). Brooks, W. K. 1893 Salpa in its relation to the evolution of life. Johns Hopkins Univ. Studies, Biol. Lab., vol. 5, p. 129. Calvert, P. P. 1908 The composition and ecological relations of the odonate
fauna of Mexico and Central America. Acad. Nat. Sci. Phila. Proc,
vol. 60. Bibliography, pp. 486-88. Clark, F. N. 1910 Plan for promoting the whitefish production of the Great
Lalles. Bull. Bur. of Fish, vol. 28, 1908 pp. 637-642. Clements, F. E. 1905 Research methods in ecology. Lincoln, Nebr. CoMSTOCK, J. H. 1904 Manual for the study of insects. Ithaca, N. Y. Cowles, H. C. 1901 The physiographic ecology of Chicago and vicinity. A
study of the origin, development and classification of plant societies.
Bot. Gaz., vol. 31, pp. 73-108, and 145-182.
1901 Plant societies of Chicago and vicinity. Bull. Geog. Soc. , Chicago,
no. 2.
1909 Present problems in plant ecology. Am. Nat., vol. 43, p. 356.
1911 A textbook of botany. Part on ecology. American Book Co.
In press. Craig, Wallace 1908 North Dakota life: plant, animal, and human. Am.
Geog. Soc, Bull. vol. 40, pp. 321-415. Bibliography.
1908 The voices of pigeons regarded as a means of social control. Am.
Jr. Sociol., vol. 14, pp. 86-100.
616 VICTOR E. SHELFORD
Griddle, N. 1907-1910 Habits of some Manitoba tiger-beetles. Can. Ent., vol.
39, April, pp. 10.5-114; 2d paper, vol. 42. Jan., 1910, pp. 9-16. Darwin, Charles 184.5 A naturalist's voyage around the world. London. Davenport, C. B. 190.3 The animal ecology of the Cold Spring sand spit.
Univ. of Chicago, Dec. Pub. 1st series, pp. 157-76. Enoch, F. 1903 The life history of Cicindela campestris, Proc. Ent. Soc,
London, pp. 10-19. Garner, R. L. 1901 Apes and monkeys, their life and language. Boston and
London. Geoffroy, E. L. 1762 Histoire des insects. Paris, 1762, T. i, p. 139. GooDE, J. Paxil 1904 Human response to the physical environment. Journal
of Geog., pp. 333-343. Haddon, a. C. 1903 The saving of vanishing data. Pop. Sci. Monthl}', vol.
62, pp. 222-229. Hagenbeck, K. 1909 Beasts and men. London. Heilprin, a. 1887 The distribution of animals. Appleton. Herrick, F. H. 1902 The home life of wild birds, a new method of the study
of the photography of birds. New York. Horn, Walther 1905 Systematischer Index der Cicindeliden. Deutsch Ent.
Zeit. Suppl., Feb. 1.
1908-10 Genera Insectorum, Fasc. 82. Carabidae-Cicindelinae. Hudson, W. H. 1892 The naturalist in La Plata. London. Hough, R. B. 1907 Hand book of trees of the northern States and Canada.
Lowville, New York. Hubbard, G. G. 1896 The Japanese nation: A typical product of environment.
Smithson. Rep. 1895, p. 667-88. Jennings, H. S. 1906 Behavior of the lower organisms. Alacmillan. Bibliography. Johnstone, James 1908 Conditions of life in the sea. Cambridge. King, L. A. L. and Russell, E. S. 1909 A method for the study of the animal
ecology of the shore. Proc. Roy. Phys. Soc, Edinburgh, vol. 17, no. 6,
pp. 225-253. Knowlton Z. H. 1909 The birds of the world. R. Ridgway ed. Henry Holt. KoBELT, W. 1902 Die Verbreitung der Tierwelt. Leipzig. Leng, C. W. 1902 Notes on Cicindelidae of Louisiana. Journ. N. Y. Ent. Soc,
vol. 10, pp. 131-136.
1902 Revision of the Cicindelidae of boreal America. Trans. Am.
Ent. Soc, vol. 28, pp. 9.3-186. Livingstone, David 1858 Missionary travels and researches in South Africa.
Harpers. Lucas, H. 1883 (On the habits of Mantichora.) Bull. Soc. Ent. France (6),
T. 3, pp. Ixxi-lxxii. MacDougal, D. T. 1908 Heredity and environic forces. Science, Jan. 24. M'Kintosh, Prof. 1904 On the distribution of marine animals. Ann. and
Mag. of N. H., vol. 13, ser. 7, pp. 117-130. Mason, O. T. 1896 Influence of environment upon human industries or arts.
Smith. Rep. 1895, pp. 639-665.
PHYSIOLOGICAL ANIMAL GEOGRAPHY 617
McGee, W. J. 1896 The relation of institution to environment. Smith. Rep.
1895. pp. 701-11. Merriam, C. H. 1890 Result of a biological survey of the San Francisco mountain region and the desert of the little Colorado. Arizona. U. S. Dept.
Agr., N. A. Fauna, no. 3.
1898 Life zones and crop zones. U. S. Dept. Agr., Biol. Survey Bull.
10. MosLEY, O. AND Brown, E. 1863 The natural history of Tutbury. London, Ortmann, a. E. 1896 Grundztige der marinen Tiergeographie, Jena.
1907 The crayfishes of the State of Pennsylvania. Mem. Carnegie
Mus. Pittsburg, vol. 2, pp. 343-523. OsBORN, H. F. 1902 Law of adaptive radiation. Amer. Nat., vol. 36, pp. 353 363. Plowden, W. C. 1868 Travels in Abyssinia. London. ,
Reighard, Jacob. 1910. Methods of studying the habits of fishes with an account of the breeding habits of the horned dace. Bull. Bur. of Fish.,
vol. 28, 1908 pp. 1112-1136. RiTTER, W. E. 1909 Halocynthia johnsoni, a comprehensive inquiry as to the
extent of law and order that prevails in a single animal species. Univ.
of Cal. Pub., Zool., vol. 6, pp. 65-114. Roosevelt, T. R. 1909-1910 African game trails. Scribner's. RuTHVEN, A. G. 1906 An ecological survey in northern Michigan. Rep. Mus.,
Univ. of Mich. (Rep. Geol. Surv., 1905).
1907 Reptiles and Amphibians from southern New ^Mexico and Arizona. Am. Mus. of Nat. Hist. Bull. vol. xxiii, pp. 483-604.
1908 The faunal affinities of the prairie region of central North America Am. Nat., vol. 42, pp. 388-393.
1908 Variations and genetic relationships of the garter snakes. U. S. Nat. Mus., Bull. 61.
Sargent, C. S. 1880 The forests of North America. 10th Census, vol. 9. 1905 Manualof trees of North America. Boston and New York.
Schimper, a. F. W. 1903 Plant geography upon a physiological basis. Translated by W. R. Fisher, Oxford.
Seits, a. 1901 Zoogeographische Betrachtungen. Zool. Gart., 193 and 232.
Selous, F. C. 1893 Travel and adventure in southeast Africa. London.
Semper, Karl. 1881 Animal life as affected by the natural conditions of existence. Appleton, N. Y.
Seton, E. Thompson 1909 Life-histories of northern animals, New York.
Shackleton, Ernest 1909 The heart of the Antarctic, Philadelphia.
Shelford, V. E. 1907 Preliminary note on the distribution of the tiger beetles and its relation to plant succession. Biol. Bull., vol. 14, no. 1. pp. 9-14.
1908 Lif e-histones and larval habits of the tiger beetles (Cicindelidae) . Linn. Soc. Journ. Zool., vol. 30, pp. 157-184.
1910 Ecological succession of fish. Trans. 111. State Acad., vol. 3. pp. 108-110.
618 VICTOR E. SHELFORD
Sherman, F. 1904 List of Cicindelidae of North Carolina with notes on species. Ent. News., vol. 15, no. 1, pp. 26.
1908 Notes on tiger beetle elevations. Ent. News, vol. 18, p. 360. October.
Snow% H. F. 1877 Amblychila. Trans. Ivans. Acad., vol. 6, p. 29.
Sumner, F. B. 1910 An intensive study of a small area of sea bottom. Bull. Bur. Fish. vol. 28, pp. 1225-1263.
Swain, R. E. 1905 Some notable constituents of the urine of the coyote. Am. Journ. Phys., vol. 13, no. 1.
Tarde, Gabriel 1903 Inter-psychology. International Quarterly, vol. 7, pp. 59-84.
Tow'ER, W. L. 1907 An investigation of evolution in Chrysomelid beetles of the genus Leptinotarsa. Carnegie Inst., no. 48. Washington, D. C. 1910 The determination of dominance and the modification of behavior in alternative inheritance. Biol. Bull., vol. 18, no. 6; vol. 20, no. 1.
Tower, W. S. 1910 Scientific geography; the relation of its content to its subdivisions. Bull. Am. Geog. Soc, vol. 42, Nov., p. 801.
Transeau, E. N. 1903 On the geographic distribution and ecological relation of bog plant societies. Bot. Gaz. vol. 36, pp. 401.
1905 Forest centers of eastern America. Am. Nat., vol. 39, no. 468, pp. 875-88.
Verrill, a. E. 1873 Invertebrate animals of Vineyard Sound. Rep. U. S.
Comm. Fish and Fisheries. 1871-72 pp. 295-538. Verworn, Max 1899 General physiology. London. Translation by F. S. Lee. Walker, A. O. 1903 Atmospheric moisture as a factor in distribution. S. E.
Nat., vol. 8, pp. 43-i7. Wallace, A. R. 1869 The Malay Archipelago. London.
1876 Geographical distribution of animals. New York.
1892 Island life. London. Warming, E. 1909 Oecology of plants. Oxford. Waxweiler, E. 1906 Equisse d'une Sociologie. Inst. Solvay Inst, de Soc. Notes
et Mem., Fasc. 2, 306, Bruxelles. Webb, Sidney 1903 The diminution and disappearance of the south east
fauna and flora. South East Naturalist, vol. 8, p. 48. Weed, C. M. 1897 Life-histories of American insects. New York. Whitford, H. N. 1901 The genetic development of the forests of northern
Michigan. Bot. Gaz., vol. 31, p. 315. WiCKHAM, H. F. 1902 The habits of American Cicindelidae. Davenport,
la. Acad. Nat. Sci., vol. 7, pp. 206-228.
1906 The races of Cicindela tranquebarica Hbst., Ent. News., Feb., pp. 43^8.
ON THE OLFACTORY ORGANS AND THE SENSE OF SMELL IN BIRDS
R. M. STRONG
From the Hull Zoological Laboratory, University of Chicago
TWO PLATES AND FOUR TEXT FIGURES
CONTENTS
1. Introduction 619
2. Literature 620
A. General 620
B. Peripheral olfactory apparatus 621
1. The nasal chambers 621
2. The olfactory epithelium 622
3. The olfactory nerves 623
C. Central olfactory apparatus 623
1. The olfactory lobes 623
2. The olfactory fiber tracts 625
D. Observations of behavior 625
E. Reports of experimental studies 626
3. Methods and material 629
A. Morphological 629
B. Experimental 632
4. Morphological results 641
5. Results of experimental studies of the sense of smell in ring doves 646
6. Control experiments with white rats 650
7. Results of other experiments and observations 650
8. Conclusions 652
Bibliography 655
1. INTRODUCTION
The work described in this paper was done principally at the University of Chicago. About six years ago the writer became interested in the question of whether birds possess a sense of smell or not. The subject was assigned to a student for some preliminary study in the spring of 1905. During this time it became apparent that something more than simple direct tests would be
619
620 . R. M. STRONG
necessary, and the writer decided to study the problem by means of a labyrinth in which a demonstration of olfactory ability would require the association of an odor with the location of food. Such work was done mostly in the year 1907-8.
During the autumn of 1909, the writer enjoyed the privilege of studying the unique collection of bird brain material in the Senckenbergisches Neurologisches Institut at Frankfurt am Main, Germany. Though a large number of bird brain sections representing a good many species of birds were studied, no new facts of importance concerning central olfactory relationships were discovered from them. On careful examination of the literature it became apparent that there was need of a comparative study of the lobes and nerves, which could be done to advantage with the fine series of partly dissected heads in Professor Edinger's collection. These were placed at my disposal for further dissection and study. The nasal chambers also were studied.
The writer wishes to express his hearty thanks to Prof. Dr. Ludwig Edinger, the director of the Institute, for the opportunities afforded and for his helpful interest. Thanks are also due to Dr. W. M. Cooper of Frankfurt, to Dr. Priemal, the director of the Frankfurt Zoologischer Garten, to Prof. John B. Watson of Johns Hopkins University, and to Mr. W. H. Osgood of the Field Columbian Museum of Chicago for additional material and for courtesies received. Through the kindness of Mr. Seth Smith and Mr. R. I. Pocock, the privilege of making a test of the olfactory sense in Apteryx was enjoyed at the London Zoological Gardens. Assistance in the preparation of the drawings was received from Mrs. Strong.
2. LITERATURE
A. Ge7ieral
During the early part of the last century, a spirited controversy was waged by a number of naturalists over the question of the existence of an olfactory sense in birds. So much evidence on the negative side was brought forth as to put the general occurence of a sense of smell in birds in doubt ever since.
THK SENSE OF SMELL IN BIRDS 621
The presence of normal olfactory apparatus in birds has been recognized by a number of writers. In Apteryx, according to Parker ('91) and Owen (71), the olfactory organs are relatively large for a bird. In other groups of birds the olfactory apparatus is generally recognized as small when compared with mammals, but varying in size. Thus Scarpa ('89), Schultze ('62), and others speak of well developed organs in the swimming birds, in the wading birds, and in the birds of prey. Bumm t'83), recognizes a relatively large olfactory apparatus in the swimming birds, but not in the birds of prey studied by him. The gallinaceous birds and the singing birds are described as having very much reduced olfactory organs.
B. Peripheral olfactory apparatus
1. The nasal chambers. The nasal chambers of birds have been studied from various standpoints by a number of writers, including the following especially:
Beeker ('03) Common fowl (Gallus) ogeranus, Psittacus, Picus, Capri and duck (Anas) mulgus, Podargus, Sturnus, Corvus,
Born ('79) Chick (Gallus) and other forms
Cohn ('02) Chick Giebel ('76 Seventeen species of
Dieulafe ('04 and '05) Paroquet, birds
duck, turkey, dove, and vulture Mihalkovics ('98) Gallus
Exner ('72) Fowl, duck, dove, and Owen ('72) Apteryx
some finches Parker ('91) Apteryx
Ganin ('90) Eighteen genera of birds Schultze ('62) Falco, Strix, Gallus, Gegenbaur ('73) Columba, Gallus, Columba, Anas and other birds
Meleagris, Anser, Buteo, Strix, GypIn general, two or three turbinals or conchae are recognized as occurring in the nasal chambers of birds. According to Gegenbaur the so-called superior or posterior concha is better named a 'Riech-hiigel,' as, in the material he studied, it was found b}^ him to be only an elevation or projection which did not possess the characteristic rolling of a true turbinal. Beeker ('03) supported him in this position. In some species Gegenbaur found even a 'Riech-hiigel' lacking. The other turbinals are regularly designated as median or middle and inferior or anterior.
JOURNAL OP MORPHOLOGY, VOL. 22, NO. 3
622 R. M. STRONG
The turbinals of Apteryx are described by Parker ('91) as having an 'extreme complexity' (p. 49). In addition to the three turbinals already mentioned, he found 'anterior and ventral accessory turbinals.'
The absence of a posterior concha in the smaller species of birds was noted by Schultze ('62) and also by Giebel (76). The latter considered this structure also lacking in Corvus and Garrulus. He found three turbinals in Lanius excubitor, however.
Jacobson's organ occurs in rudimentary or vestigial form according to Mihalcovics ('98), Ganin ('90), and Cohn ('02), in the embryo bird. It is lost during embryonic life, though the median portions of the ducts of the nasal glands are regarded as modified Jacobson's organs by Ganin and Mihalcovics.
2. The olfactory epithelium. In Apteryx, according to Parker, all of the turbinals, except the so-called ventral accessory, are covered with 'Schneiderian membrane' (p. 51). Owen, using a different terminology for the turbinals of Apteryx, also described an extensive distribution of olfactory nerve fibers in this bird both on all of the turbinals, excepting the 'anterior,' and on the septum narium.
.In other birds studied, a much more limited distribution of the olfactory epithelium has been found. In the common fowl, according to Mihalcovics, the olfactory epithelium is limited to the posterior turbinal and to the adjacent wall of the nasal cavity up to the roof. A similar location was noted by Dieulafe ('04, p. 439). Preobraschensky ('92) found the olfactory epithelium limited to the posterior turbinal only, in the chick.
According to Schultze, the surface of the posterior turbinal may not always be entirely covered by olfactory nerve terminations. Doves were found to have the lower border free from olfactory epithelium. Gegenbaur believed that Schultze had the middle turbinal in mind, and the former writer says that doves do not have a posterior turbinal. Schultze states that in those singing birds which lack a posterior turbinal a very small part of the structure which corresponds to the middle turbinal receives olfactory nerve fibers. He also found the septum narium receiving olfactory nerve fibers, at least in those birds which have a posterior
THE SENSE OF SMELL IN BIRDS 623
turbinal. The structure of the olfactory epithelium has been mo^t fully described by Schultze, though it has been studied by Exner and Disse.
3. The olfactory rierves. The olfactory nerv^es of birds have been given very little attention except from the standpoint of their development in the chick. They have been figured in drawings of the internal anatomy of the head by Scarpa for a few species of birds. There are also a few figures by Gadow ('91), in Bronn's Thier-Reich.
A single pair of so-called olfactory nerves are generally recognized as connecting the olfactory epithelium with the brain. (A more appropriate term would be ' olfactory root.' See Edinger, '08, p. 252.) However, in Apteryx, according to Owen, there are more than two. He says that "the olfactory nerves perforate the anterior and inferior wall of the rhinencephalic chamber by se\eral foramina, but are closely invested and united by the neurilemma, especially along their upper surfaces, so as to appear for an extent of eight or nine hues, each as one large olfactory nerve." Parker ('91, p. 107), says that "the numerous olfactory nerves are given off from the ventral and anterior surfaces of the rhinencephal." The olfactory nerves of the vulture are stated by Owen ('66), to be larger relatively than those of the common turkey. They were found by Gage ('96) to be minute in the English or house sparrow.
The writer unfortunatel.y has not had access to some of the old works on bird anatomy which may have had accounts of olfactory nerves.
The early development of these structures in birds has been studied by Cohn ('02), Disse ('96-'98) in the chick, goose, and duck, by Kolliker ('90), Marshall ('78), Preobraschensky ('92), in the chick, and Belogowy ('09), in the chick.
C. Central olfactory apparatus
1. The olfactory lobes. The olfactory bulbs of many vertebrates are represented in birds, according to Edinger, by a Formatio bulbaris which covers the olfactory lobes except on their caudal dorsal borders. In the literature of bird brains, this com
624 R. M. STRONG
hination is usually designated as an olfactory lobe and will be so termed in this paper.
In Apteryx, the olfactory lobes are of considerable size according to the figures of Owen (72, pi. xlv, fig. 2). This writer says (p. 383), that the Rhinencephalon is as remarkable for its large size as is the mesencephalon for the smallness of its principal elements," in Apteryx. The optic lobes and nerves are small in Owen's figures. The only other reference to very large olfactory lobes in a bird is that of Klinckowstrom ('90), who figured and described large olfactory lobes for Fulmarus glacialis.
The olfactory lobes of a number of birds were studied by Bumm ('83) who noted that these structures are smaller in birds than in mammals. He states that the olfactorj^ lobes are well developed in the swimming birds, moderately large in the marsh birds ('sumpf Vogeln'), and much less developed in other orders. The ratios in weight of the olfactory lobes to the cerebrum are given for the 'Gans,' 'Schnepfe,' and 'Bussard.'
Some statistics are given for the olfactory lobes in forty- two species of American birds, the majority being Passerine forms, by Turner ('91). Turner concluded that '4here has been a gradual retrograde evolution of the avian rhinencephalon" ('91, p. 57). He observed that as we ascend the scale, the olfactory lobes move caudad and become smaller," and he also noted that they are fused and almost completely imbedded in the 'prosencephalon' in the higher groups. In another article, Turner ('91b), stated that the high development of the sense of vision in birds has been accomplished at the expense of the olfactorj' sense.
The histological structure of the bird olfactory lobes has been described by Turner ('91), and by Pedro Ramon y Cajal. The latter's account was not accessible to the writer, and it has been necessary to depend upon the statement of Edinger ('03, p. 403) of Cajal's conclusions. The structure is stated to be of the same type as in other vertebrates except for being simpler.
References to the olfactory lobes have also been made by Carus ('14), Elliot-Smith ('95), Stieda ('69), Herrick ('93) Schupbach ('94), and Kappers und Theunissen ('08).
THE SENSE OF SMELL IN BIRDS
625
2. Olfactory fiber tracts. Very little is known about the central relationships of the olfactory organs in birds. References are made to olfactory fiber tracts by Stieda ('69), Bumm ('83) ElliotSmith ('95), Munzer und Wiener ('98), and Kappers und Theunissen ('08).
Edinger('03) states that a basal bundle (tractusbulbo corticalis), consisting of a few fibers, passes into the brain base and is lost after a short course. Amputation of the olfactory lobes leads to degeneration of medullated fibers in the lobes only. According to the accounts of Edinger and Kappers, only a small number of olfactory fibers of the second order have been seen and these have not been satisfactorily traced. In the work of Kappers and Theunissen it is stated that olfactory fibers of the third order which connect the olfactory lobes with caudal non-cortical portions are more numerous. An olfactory cortex or hippocampus has not been demonstrated clearly.
D. Observations of behavior
A large number of miscellaneous field observations have been reported. As an illustration, an article by Rhoads may be cited. This observer states that while digging sweet potatoes in New Jersey, he noticed a luxurious growth of vines over a small mound in the field, and the potatoes dug at this place were unusually large. On inquiry, he found that a horse and a cow had been buried there during the previous winter. In the afternoon and during the following day, vultures came 'in scores', swooping to the ground about the mound. These birds continued to come 'for long after,' though not so numerous as at the time when the crop was plowed out. Rhoads could detect no 'taint' in the atmosphere, yet hundreds of vultures assembled 'from far and near.' He concluded that these birds were attracted by an olfactory stimulus.
According to Reeker ('99), a number of birds which were observed feeding on table scraps in a back yard declined to eat a potato which had been bitten by a cat. The author concluded that the potato was neglected because of an odor left by the cat.
JOURNAL OP MORPHOLOQT, VOL. 22, NO. 3
626 R. M. STRONG
Raspail ('99 and '01), credits birds with a very keen sense of smell. He gives a number of observations of occurrences in the field, and he reports various simple and unconti'olled experiments.
A caged condor was observed by Gill ('04) to become very much excited when, during the dissection of a rabbit, the strong odor of the abdomen escaped into the room in which the cage was placed. The operation is stated to have been carried on 'quite out of sight of the condor.'
The chances for error in the interpretation of this kind of evidence are so great that it has little value.
E. Re-ports of experimental studies
Here are included accounts only of experiments conducted with critical care. The following studies were made by Audubon C35):
1. An entire deer skin, including the hoofs, and provided with artificial eyes, was stuffed with dried grass, the whole being allowed to become 'perfectly dry.' The stuffed skin was exposed in a large field, and the observer concealed himself not far away. In a few minutes a vulture, soaring about, saw the deer skin and sailed down to it. The hide was torn open, and much grass was pulled out.
2. A large dead hog was hauled to a ravine and concealed by a covering of cane. As the weather was w^arm, the body became 'extremely fetid' in a couple of days. Dogs found the carcass and fed heartily upon it, but vultures sailing over from time to time did not find it.
3. A young pig was killed, and its blood was scattered about on the ground. The body was concealed by a covering of leaves. Vultures found the blood and followed it down the ravine to the body, which was then discovered and devoured.
4. Two young vultures were kept for some weeks in a cage where they became accustomed to receiving food. The birds were in the habit of hissing and gesticulating when the^^ saw food approaching. However, when food, either fresh or putrid, was brought up to the immediate rear of the cage where the vultures could not see it, no excitement was shown.
THE SENSE OF SMELL IN BIRDS 627
Audubon quotes some experiments by Bachman from Loudon's Magazine of Natural History, 1838, as follows:
1. A dead hare, two dead birds, and a wheelbarrow full of offal from a slaughter house were deposited on the ground at the foot of Bachman's garden in South Carohna. A frame was raised above the pile at a distance of twelve inches from the ground, and this was covered with brush, allowing air to pass under freely. Though hundreds of vultures passed over in the next twentyfive days, none noticed the meat.
2. A coarse painting on canvas of a sheep skinned and cut open was placed on the ground, where it was noticed by vultures. They walked over the painting and tugged at it with their beaks. The painting was then placed within fifteen feet of the offal mentioned above, but the offal was not touched.
3. The most offensive portions of the offal were next placed on the ground, and these were covered by a thin canvas cloth on which were strewn several pieces of fresh beef. Vultures came and ate the beef, but they did not discover the offal beneath the canvas. A rent was then made in the canvas, whereupon the offal below was seen and eaten.
Negative results were also obtained by Darwin ('34), pp. 184186), who experimented with a number of condors in a garden at Valparaiso, Chili. In his account Darwin says:
The condors were tied, each by a rope, in a long row at the bottom of a wall and having folded up a piece of meat in a white paper, I walked backwards and forwards, carrying it in my hand at the distance of about three yards from them, but no notice was taken. I then threw it on the ground, within one yard of an old bird; he looked at it for a moment with attention but then regarded it no more. With a stick I pushed it closer and closer, until at last he touched it with his beak; the paper was then instantly torn off with fury, and at the same moment, every bird in the long row began struggling and flapping its wings.
Darwin did not state whether the meat used in this experiment was entirely concealed by the wrapping paper until the package was torn open, nor did he indicate the opportunity for an odor to escape from the package. One is warranted in inferring from the account that the meat was detected before it was exposed to
628 R. M. STRONG
view. The description of this experiment would suggest that the meat was finally smelled, though one may also infer that the olfactory sense was probably not extremely keen in these birds. Darwin himself concluded that "the evidence in favor of, and against, the acute smelling powers of carrion-vultures is singularly balanced," but he evidentl}'^ believed that these birds find their food by sight.
Experiments of a different nature were conducted by Hill ('05) with a pair of turkeys. He employed a number of odors including many which should produce strong stimulation of nerves of general sensation, under the conditions of the experiments. Such substances as asafoetida, essence of anise, oil of lavender, valerianate of zinc, powdered camphor, chloroform, etc., were placed very near, or upon, one of two piles of food located in an enclosure into which the turkeys were admitted when they were to be fed. A number of trials were made to see whether the birds would make their choice of a pile of food because of the presence or absence of one of the odorous materials, but with negative results. No evidence of discrimination appeared, and even the fumes of prussic acid, though causing the bird to stagger, did not drive it from the first pile of food it happened to choose. Laboratory experiments were conducted by Rouse ('05), who observed the respiratory movements of pigeons when in the presence of oil of bergamot and lily of the valley. No 'appreciable reactions' to these materials were noticed, but a ' slight sensitiveness was shown to asafoetida,' and 'marked reaction' was produced by turpentine and ammonia. The writer recognized, however, that the reactions obtained may have been due to other than olfactory stimuli.
On the other hand, evidence has been obtained by Beebe ('09) at Bronx Zoological Park, New York, that the turkey vulture has a sense of smell. The following is taken from his account (pp. 467-8) of experiments with turkey and black vultures:
Three boxes were placed on the ground some distance apart, and the birds fed for a few days in various parts of the cage. Then after several days of fasting, a piece of tainted meat was placed under the central box. Care was taken to go through the farce of placing something
THE SENSE OF SMELL IN BIRDS 629
under each box so that no visual hints of the location of the meat was conveyed. The vultures were very hungry, yet they did not leave their perches and come to the ground, although they had watched their keeper intently. He now re-entered and threw down one or two small bits of meat. Within a second or two, almost as the meat left the hand of the keeper, every vulture swooped to the ground and was hissing and struggling for a portion of the food. Twice the black vultures walked close about the meat box without appearing to notice the odor which was clearly perceptible, even to persons outside of the cage. A turkey vulture walked to leeward, instantly turned and made his way to the box, which he examined on all sides. He was soon joined by two others of the same species, and all three took up their stations close to the source of the odor. Soon two black vultures came up, apparently impelled more by imitation than by actual discovery of the smell. All five birds remained fcr a long time grouped close to the box, going to it now and then, and examining it carefully. Thus even in the turkey vulture the sense of smell is certainly not highly developed, and compared with the sense of sight, is defective indeed.
The experiments of Benham ('06), gave evidence that Apteryx has an acute sense of smell.
A preliminary statement of conclusions which were made from the experimental work described in this paper was published by the writer (Strong '08).
3. METHODS AND MATERIAL A. Morphological
The olfactory lobes and nerves and the nasal chambers were studied in dissections of the bird head. As the olfactory lobes and nerves are always more or less imbedded in bone or cartilage and there is often considerable tough connective tissue directly over the lobes, dissection is not easy. A pair of small, pointed bone-forceps were found especially useful for the larger heads. For very small birds, a pair of scissors with curved points, dissecting needles, and a lens were employed. Measurements were made with the aid of a pair of dividers of the breadth of the cerebrum and of its length in the median line. The length of the olfactory nerve from the apex of the olfactory lobe to the nasal chamber and the diameter of the olfactory nerve were also determined. Of course all of these measurements were approx
630 R. M. STRONG
imate only, because it was impossible to select points for measuring with any precision. The olfactory nerve diameter was the most important of the measurements, and it could be determined with somewhat greater precision than the others. Nevertheless, the olfactory nerve cross section varies in form and size at different points, and a mean diameter for the median portion was consequently sought. The olfactory nerves in some species are flatter laterally than in others. Where they were very much flattened, lateral and dorso-ventral diameters were measured.
Photographs varying from one-half diameter to nearly natural size were made of the dissections in dorsal view. The heads were arranged so that the olfactory nerves were approximately at right angles to the axis of the camera. Some of the dissections were also pTiotographed from the side to give a lateral view of the olfactory lobes and nerves. Prints enlarged two or three diameters were then made from the negatives of subjects which were selected for illustration in this paper. A magnification of three diameters was used only for a few small birds. Most of this work was done with the Edinger drawing apparatus. When this apparatus was properly adjusted for an exposure, and before a print holder was inserted, a sketch was made of the image projected on drawing paper. This sketch gave the general outlines for the making of a drawing. References were made frequently to measurements and to the enlarged prints in the preparation of the drawings. Three methods w^ere thus available for securing accuracy in the drawing of the illustrations. Most of the brain mass has been included in order to show the relative size of the olfactory lobes.
Heads of the birds in the following list were dissected and studied: The classification used in Sharpe's Catalogue, ('74) and Handlist, ('99) is employed here and the orders are given so that the taxonomic distribution of the forms studied may be seen at a glance. The Sharpe Catalogue nomenclature has been followed consistently because at present there is no other work so representative of the birds of the world. As many of the specimens in the Edinger collection bore other names, they also are mentioned after the Sharpe Catalogue name in parentheses.
THE SENSE OF SMELL IN BIRDS
631
This material represents twenty-seven out of thirty-five orders of existing birds. The species which were studied in Professor Edinger's laboratory are indicated by an asterisk.
Order 1. Rheiformes
- Rhea americana,
rhea.
Order 2. Struthioni formes
SUB-CLASS RATITAE
- Struthio camelus, Linn. Ostrich.
Linn. Common Order 3. Casuarii formes:
- Dromaeus novae-hollandiae, Lath.
Emeu.
SUB-CLASS CARINATAE
Order 1. Tinamiformes:
- Nothura maculosa, Temm. (Rhychotus). Spotted tinamou.
Order 2. Galliformes:
- Crax carunculata, Temm. Yarrell's curassow.
- Lyrurus tetrix, Linn. Black grouse
(Tetrao).
- Caccabis petrosa, Gm. Barbary
partridge.
- Perdix perdix, Linn. Common partridge.
- Gennaeus nycthemerus.i (Euplocamus argentatus). Silver pheasant.
- Gallus gallus, Linn. Common fowl.
- Polyplectrum chinquis, P. L. S.
Mull. Gray peacock-pheasant. Order 5. Columhi formes:
- Columba livia, Bonn. Domestic
dove.
Turtur risorius, Linn. Ring dove.
- Goura coronata, Linn. Crowned
pigeon. Order 7. Ralliformes:
Rallus virginianus, Linn. Virginia rail.
- Fulica atra, Linn. European coot.
Fulica americana, Gm, American coot.
Order 11. Sphenisciformes:
- Spheniscus demersus, Linn. Cape
penguin. Order 12. Procellarii formes:
- Fulmarus glacialis, Linn. (Procellaria). Fulmar.
Order 14. Lariformes:
Hydrochelidon surinamensis, Gm. Black tern.
Sterna fuliginosa, Gm. Sooty tern.
Anous stolidus, Linn. Noddy tern.
- Larus ridibundus, Linn. Black
headed gull. Order 15. Charadriiformes:
- Belanopterus cayennensis, Gm.
(Vanellus). Cayenne lapwing.
- Pavoncella pugnax, Linn. Ruff.
(Machetes).
- Scolopax rusticola, Linn. European woodcock.
Order 16. Gruiformes:
- Anthropoides virgo, Linn. (Grus).
Demoiselle crane.
♦Eurypyga helias. Pall. Sun bittern. Order 18. Ardeiformes:
- Theristicus melanopsis, Gm. (Ibis).
Black-faced ibis.
1 The identification of 'Euplocamus argentatus' with any name in the Sharpe nomenclature was difficult. On consultation of other works it seemed probable that the bird is Gennaeus nycthemerus in the Sharpe system.
632
R. M. STRONG
- Platalea leucorodia, Linn. White
spoon-bill.
- Pseudotantalus leucocephalus, Gm.
(Tantalus). White headed ibis.
- Leptotilus crumeniferus, Less. African adjutant.
- Tigrisoma lineatum, Bodd. (T. brasiliense).
- Nycticorax cyanocephalus, Mol.
Night heron. Order 19. Phoenicopteriformes:
- Phoenicopterus roseus, Pall. (P.
antiquorum.) European flamingo. Order 20. Anseriformes:
- Plectropterus niger, Scl. Black
spur-winged goose.
- Anser anser, Linn. Domestic goose.
- Anas boscas, Linn. Domestic duck.
Order 23. Pelecani formes:
- Phlacrocorax carbo, Linn. Cormorant.
- Plotus anhinga, Linn. Snake bird.
- Sula bassana, Linn. Gannet.
- Pelecanus rufescens, Gm. Redbacked pelican.
Order 24- Cathartidiformes:
- Catharistes urubu, Vieill. (C. atrata). Black vulture.
- Cathartes aura, Linn. Turkey
vulture. Order 25. Accipitriformes:
- Circus aeruginosus, Linn. The Moor
buzzard.
- Accipiter nisus, Linn. European
sparrow hawk.
- Buteo buteo, Linn. European buzzard.
- Circaetes gallicus, Gm. Serpent
eagle. Order 26. Strigiformes :
- Bubo ignavus, Forst. Eagle owl.
Order 28. Psitlaciformes:
- Chrysotis auripalliata, Less. Golden-naped Amazon parrot.
- Chrysotis brasiliensis, Linn. Redtailed Amazon parrot.
Order 31. Coccyges:
Coccyges erythrophthalmus, Wils. Black -billed cuckoo. Order 32. Scansores:
- Rhamphastos cuvieri, Wagl. Cuvier's toucan.
Order 33. Piciformes:
- Gecinus viridis, Linn. Green woodpecker.
Order 36. Passeriformes:
- Sylvia atricapilla, Linn. Blackcap
Lanius migrans, Linn. Migrant shrike.
- Motacilla alba, Linn. White wagtail.
- Coccothraustes coccothraustes,
Linn. Hawfinch.
- Serinus serinus, Linn. Serin finch.
- Sturnus vulgaris, Linn. Starling.
- Corvus corax, Linn. Raven.
- Corvus corone, Linn. European
carrion crow.
Corvus brachyrhynchus, Brehm. American crow.
- Garrulus glandarius, Linn. European jay.
Cyanocitta cristata, Linn. American blue jay.
B. Experimental
The jfirst attempts by the writer to determine whether an olfactory sense exists in birds or not were simple experiments with such substances as cedar oil, asafoetida, oil of bergamot, clove oil, and hydrogen sulphide. The odorous material, placed upon rags and filter paper or in bottles, was usually held within a few feet or inches of the bird's head. Several nestling and adult ring
THE SENSE OF SMELL IN BIRDS
633
doves, two young crows, some common fowls, turkey vultures, and a paroquet were used as subjects for the tests. As these experiments proved to be worthless, it appeared to be necessary to develop more elaborate methods for distinguishing possible responses to olfactory stimuli, and a labyrinth in which the bird
t.t.
Text fig. A Labyrinth without cover; '/, entrance to chamber .1 ,■ h, liter washbottle; 6', small bottle which contained the odorous material and was located
here when the food box was in the position which is indicated in fig. C; i. glass tube
which connected odor bottle with the wash bottle; tt, tube leading to two small
bottle connections; d, entrance to main enclosure from cage.
would have the opportunity of finding its food was devised. This apparatus, shown in figs. A and B, included a central area five feet wide and ten inches high, v/ith four accessory chambers, each of which opened into the central enclosure at the middle of a side. The chambers were ten inches square and so arranged that a bird in the central enclosure could not see food placed at f.b., fig. C.
634
E. M. STRONG
The dimensions given above were chosen mainly in order that the apparatus might be accommodated to the size of the room available for the experiments, and they seemed to be suitable for the work with ring doves as the experiments proceeded.
R f.
Text fig. B Labyrinth with cover (c) in phice;/, funnel of exhaust; p, Richard's
air pump (two air pumps were used at this point in the oil of bergamot series).
The control of air currents naturally demanded an indoor location for the apparatus.
In order to reduce the danger of general diffusion of the odor, an air exhaust was arranged. A funnel with a diameter of 6| inches at the larger end was mounted with the small end up at the center of the cover to the main enclosure, and this was connected with a Richards- air pump which was attached to a water
- During most of the experiments with oil of bergamot, two of these pumps were used with a T connection in order to accelerate the removal of diffused odor from the central enclosure.
THE SENSE OF SMELL IN BIRDS
635
tap (fig. B). Gentle air currents were forced into each compartment through glass tubing arranged as seen in fig. B. The air currents were produced by a machine originally constructed for injecting blood vessels. In this apparatus (fig. D), water was passed from one closed tank into another by elevating the filled tank above the empty one. The two tanks were connected
f.b.
Text fig. C Diagram of labyrinth; A. B, C, D, chambers; d, position of entrance to main enclosure (m. e.) from cage;/, b., food box.
by rubber tubing (r.t., fig. D). The air displaced by the entering water in the lower tank was allowed to escape through stop cocks into two glass tubes. Each of these tubes communicated with a wash bottle (one-litre size) which was about three-fourths full of water. From the wash bottles, the air emerged through T tubes (fig. A, t.t.) and was conveyed to 125 cc. bottles (fig. A, b') opposite the four chambers of the labyrinth. These smaller
636
R. M. STRONG
bottles were mounted in dishes which contained wax in order to prevent them from being upset. The glass tubing used was of uniform diameter, and the ends which projected inside the chambers were carefully rounded so that they might present no discernible differences in appearance. This tubing had an inside
Text fig. D ; Air pressure apparatus (.not designed by the writer) ; t', upper tank;
t", lower tank; r. t., rubber tube which connected the tanks; s. c, stop cock.
diameter of about 5.5 mm. One only of the small bottles was used to contain the odorous material, and it was disconnected after each set of experiments to be put away until the next series of experiments were begun. Food (a mixture of canary seed, millet, and wheat with a small admixture of ground charcoal and oyster shells) was placed for each series of four experiments in
THE SENSE OF SMELL IN BIRDS 637
one of the four chambers (fig. C, f.b.), and the odor bottle was connected with the chamber which contained the food, where it replaced the bottle located there in the previous series of experiments (fig. A, b'.). Newspapers were laid underneath the entire labyrinth to prevent the birds from soiling the floor.
After the apparatus had been inspected to insure its being ready, the air pumps were started by turning on water at P, fig. B. The cage containing the bird was placed with the door against the entrance to the enclosure (fig. A, d); the cage door was then raised, allowing the subject to enter. It was not found practicable, though desirable, to place this entrance at the center because there was greater need of room for the exhaust funnel at that point. Immediately after the dove was admitted, air pressure was started by turning a cock at s.c, fig. D. This order of procedure insured the entrance of the odor into the enclosure by the time the bird started to look for food, and the postponement of turning on the air pressure until just after the bird had entered decreased the danger of a general diffusion of the odor in the central enclosure during the time which w^as ordinarily occupied by the subject in hunting for the food.
When the bird reached the seed, the air pressure was cut off to avoid unnecessary diffusion of the odor, but the exhaust was kept going until the last of the birds used in each set of experiments had found its food.
When the last bird had finished feeding and had been removed to its cage, the entire labyrinth was elevated several inches above the floor for ventilation ; a window and a door to the room containing the apparatus were kept open for a short time at least during the ventilation.
In order to determine so far as possible, the extent of the space in the main enclosure where the odors which were employed in the experiments might be expected to be perceptible, various tests were made. Strips of litmus paper about eight inches long were hung from standards placed at thirty-six approximately equal points inside the main enclosure. These strips were all dipped in distilled water just before air pressure was turned on. Ammonia water of several dilutions was placed in a bottle which was used
JOURXAL OF MORPHOLOGY, VOL. 22, NO. 3
638 R. M. STRONG
in place of the odor containing bottle. When a relatively strong solution of ammonia was used, all of the litmus paper strips turned blue in the course of a few minutes, those in the corners changing last. When solutions produced a diffusion of ammonia gas which seemed to be at all comparable in strength with the odor of oil of bergamot, so far as such a comparison could be made, a semicircular area of diffusion was indicated by the litmus paper. The radii of this area converged at the entrance of that chamber where the ammonia gas emerged, and its front extended out to the region of the exhaust funnel. The writer stretched himself on the floor inside the enclosure and attempted tests of the odor diffusion with his own olfactory organs. Musk and violet sachet powder were smelled with difficulty and only at the entrance where these odors were emerging. The odor of oil of bergamot was detected as far as eighteen inches from the point of emergence. The odor was not strong beyond a foot or less from the entrance of the chamber from which the odor was emerging.
Of course these experiments afford only indirect evidence concerning the size and form of the area in which the odor might be expected to be effective. However, the behavior of the birds and of some rats in the enclosure, when considered in connection with the tests mentioned above, have convinced me that the odor localization was sufficient to enable an animal which would be capable of odor discrimination to determine the compartment from which the odor emerged in the experiments. White rats appeared to have no difficulty in locating the source of the odor.
Choice of subjects. It was obviously necessary in using such apparatus that only tame and tractable birds be employed. Of various species which were considered, ring-doves seemed to be most suitable. The writer had on hand a number of hybrids between the white and blonde ring-doves Turtur alba and T. risorius which were unusually tame; four vigorous males w^ere chosen. Unfortunately, the habits of ring-doves do not suggest that they have any use for a sense of smell. Their food consists mostly of seeds which have practically no odor for the human olfactory sense. Nevertheless, their great docility, convenient size, and adaptiveness to cage life made them preferable, in the
THE SENSE OF SMELL IN BIRDS 639
writer's judgment, to less tractable birds whose habits might suggest greater olfactory possibilities. During the series of experiments there were many occasions when even these birds were none too manageable.
Choice of odor. An ideal odorous material for these experiments should be associated with the natural food-seeking habits of the subject. It should also stimulate only the olfactory sense organs. Unfortunateh", the first condition could not be realized, and the second could not be established with absolute certainty. A careful consideration of food materials did not suggest anything under practicable conditions which would supply a strong odor as measured by the human olfactorj^ organs. No odorous material of any kind could be found where the possibility of other than olfactory stimulation could be known to be absolutel}^ eliminated. It w^as therefore necessary to consider odors which do not have any known relationship to the experience of ring-doves.
The following were selected as being the least objectionable: animal musk, violet-sachet powder, and oil of bergamot. Eau de cologne was used for the first series of experiments, but it was finally abandoned as the WTiter became impressed with the possibiUty of alcohol stimulation by the alcohol contained in this compound. Animal-musk and violet-sachet powder when they were not dissolved in alcohol produced odors so weak that they were employed only a short time. A strong odor seemed desirable because these birds could not be assumed to have so keen a sense of smell as even man possesses. Oil of bergamot was finally chosen for the larger portion of the work because it seemed to combine strength with freedom from at least apparent danger of there being other than olfactory stimulation. In the absence of knowledge concerning the effects of odorous materials upon the dove's sense organs, the sensations of the experimenter could be the only guide in choosing an odor. It may be objected that w^hen held very near the nose, oil of bergamot produces almost painful sensations, which may involve tactile endings. Under the conditions of the experiments, however, the stimulation of olfactory endings was mild for the wTiter and there were no suggestions of any other stimulation.
640 R. M. STRONG
Management of birds. Before the taking of records was begun, the doves were trained to look for food without the use of any odor. They were induced to enter a chamber by a trail of seed which led to the food box. In the course of a week, thej^ lost their fear of entering the chambers, and it was found practicable to work with them twice a day.
The doves were kept in cages which stood in the room containing the apparatus. Their view of the apparatus was cut off by a screen, though this precaution was probably unnecessary. Even when the bird had been placed inside the main enclosure, on several occasions, before the seed box had been put in place, it gave no evidence of having profited by the opportunity it had had to see the location of the food, but went through the usual search for the proper chamber to enter. Thus, on one occasion, a dove after such an opportunity went to all three of the empty chambers before it entered the one containing the food.
Each bird was allowed enough time to eat all the food it wished. It then almost invariably returned to the main enclosure, and was driven out through the entrance d, fig. A, into a cage which had been placed alongside. After a few weeks, the removal was easily accomplished, for the doves became accustomed to passing out when a stick was waved over the apparatus.
Records. The significant movements of the birds were recorded usually by means of symbols. The sign = was used to indicate that the dove entered a chaiijiber so far that it should have been able to see whether food was present or not. The simple approach of a bird to or within a few inches of the opening of a chamber without an entrance w^hich would be complete enough to enable it to see whether food was present or not, was designated by the sign — . Thus the form SO— B = C = A indicated that dove No. 30 in the writer's breeding register w^ent to the opening of chamber B but did not enter. It then turned and entered chambers C and A, the latter containing the food. Notes were made of interesting or unusual variations in the behavior of the birds.
THE SENSE OF SMELL IN BIRDS 641
4. MORPHOLOGICAL RESULTS
The anatomical studies made by the writer have given evidence that (1), the olfactory organs of birds are of too great size to be set aside as non-functional, but that (2) there is a tendency in the bird series towards retrogression of these organs. Various types of olfactory lobes and nerves have been figured in plates 1 and 2, and measurements for some of the birds studied have been given in table 1.
In most of the orders, the olfactory lobes usually have essentially the form found in the doves and in the gallinaceous birds. They are so similar in the representatives of these two groups which were studied, that only one form has been chosen for illustration, and that is the species used by the writer in his experimental work (fig. 4). The so-called olfactory nerves usually leave the olfactory lobes in more or less close contact with each other, and they diverge gradually at some distance from their proximal ends as in fig. 4 ; but in some cases they are widely separated (figs. 1, 8, 9, 13, 16, and 17). In table 1, some variations have been noted under the head of remarks. The length of the olfactory nerves and their degree of separation seem to be adapted to the form and arrangement of other structures in the head, and with no functional significance that could be discovered. Olfactory nerves, so called, were found in all of the material studied except for Dromaeus, Spheniscus, and Fulmarus. In these three birds, the olfactory lobes have their anterior ends in contact with the nasal capsules, though Dromaeus may possibly be said to have very short olfactory nerves.
In no case was more than one pair of olfactory nerves found except possibly in Rhea. Unfortunately, the available material was not in a condition to give a satisfactory dissection for this species.
The most interesting olfactory organs in some respects were those found in Dromaeus, Fulmarus, Catharistes, Cathartes, and the Corvidae. In Dromaeus and in Fulmarus, the olfactory lobes are notable for their relatively great size (figs. 2, 3, 5, and 6). It will also be observed that there is a constricted connection of
642
R. M. STRONG
TABLE 1
Measurements are in millimeters. Unless otherwise stated in the column headed remarks, the olfactory lobes and nerves do not vary significanth/ in their size, form and arrangement from those of the doves and the gallinaceous birds. The species which appear in plates I and ^ have not been included here.
OLFACTORY NERVE
Rhea amerirana.
Crax c^runculata. . .
Ljn-urus tetrix
Perdix perdix
Gallus gallus
Columbia livia
Goura coronata
Fulica atra
Anous stolidus . . . Larus ribidundus . . Belanopterus cay ennensis
Scolopax fusticola. Anthropoides virgo Eurypyga helias.. .
9.0+ 0.7
11.5 1 15
12.0 0.5
8.0 0.4
12.0 0.7X1 10.0 I 0.6X.75
14.0 0.8
5.0 1.0
8.5 0.5X0
8.0 0.7
11.0 0.4
12.2 0.8
0.8 0.1
9.0 0.4X0.8
Theristicus melan-j
opsis j 9.0 \ 1.0
Platalealeucorodia.i 8.0 \ 12.0
Tigrisoma brasil iense
Nycticorax cyano
cephalus
Anser anser
Anas boscas
Phalacrocorax
carbo
Plotus anhinga
Sula bassana
15.5
14.0
7.0
12.5
1.0 1.5 12
z 1.0
17.0
0.7X1.2
CEREBRUM
19.5
20.0 14.0 12.5 14.0 12.0 16.5 17.0 18.5 12
12.5 15.0 19.0 12.5
22.5
27.5
34.0 Lobes rather large and resembling those of Dromaeus slightly. Turbinals elaborate
29.0 Olfactory lobes long
23.0 I
22.5 20.0 28.0 22.5 11.5
24. a
Olfactory lobes long as in duck
20.0
20.0 See next page. 30.5
19.2 I Nerves very close I throughout
togeth(
35.0 \ Lobes relatively small 34.0 j Lobes small and between anterior ends of hemispheres; ophthalmic branch of trigeminus very large
18.0 29.0
19.0 29.0 j
33.0 56.0 I Lobes long
20.0 26.0 Lobes long
23.5 29.5 Nerves relatively small
Lobes relatively small 24.5 i 37.2 Lobes have peculiar ventral position
THE SENSE OF SMELL IN BIRDS
643
TABLE 1 (continued)
OLFACTORY NERVE
CEREBRUM
Species
a
.2
REMARKS
Rhamphastos cuvieri
6.0
5
19.0
29.0
Lobes small; nerves wide
Gecinus viridis
Pelecanus rufescens
9.5
0.3
19.0
23.0
apart; no external nares;
nasal chamber very short Lobes very small Lobes small
Circus aeruginosus .
Accipiternisus
10.0
9.5
0.5X0.8
0.6
18.5 14.0
32.0
24.0
Buteobuteo
12.0
0.5X1.1
19.0
35.0
Lobes small and short
Bubo ignavus
0.8
21.0
41.5
Lobes nof, studied successfully
Sylvia atricapilla. .
Sturnus vulgaris. . . .
5.5
5.5
0.2
0.3
9.0
14
12.5
20.
Lobes and nerves very small
Lobes small; nerves separated
Garrulus glandarius
9.0
0.2X0.4
27.5
25.5
at post. ends.
Lobes minute and fused; olf.
nerves very slender
the olfactory lobes with the brain in Dromaeus. Such a condition was found in less conspicuous form in specimens of the order. Charadriiformes which were examined by the writer, and it also appears in a figure of the brain of the American woodcock, another member of this order, (Herrick, '93, pi. 26). The writer regrets that he has not had opportunity to study the Dromaeus material histologically to note the relations of the formatio bulbaris to the olfactory lobe proper. There was evidence obtained by dissection that there is a comparatively rich inervation of the relatively complicated posterior turbinals, by olfactory fibers. The surface of the middle concha is also elaborate, as can be seen in the posterior portion which has been included in fig. 2, and it seems quite possible that the middle concha in Dromaeus also receives olfactory fibers.
Though the relatively huge olfactory lobes of the fulmar have been described by Klinckowstrom, they were not adequately illustrated in the figure which accompanied his account, and the nasal chambers were not included. It has therefore seemed desir
644 . R. M. STRONG
able to the writer that a more complete illustration be published as has been done in fig. 5. The turns of the posterior turbinal may be seen at (d) where openings have been made in the preparation. The extensive rolling of this turbinal and the large size of the olfactory lobes in this species seem very significant, and the writer regrets that he has not had time to study the sense of smell in the living fulmar.
As the fulmar is one of a group of birds which are characteristically seagoing animals and which are in the habit of taking long trips over the sea, it is natural to suspect that these large olfactory lobes might be used as orientation organs if not serving for the sense of smell, (Cyon, '08). With this idea in view, the writer wrote to Prof. John B. Watson, of Johns Hopkins University during the early summer of 1910, while Professor Watson was conducting studies of orientation in the noddy and sooty terns at the Dry Tortugas Islands. Dr. Watson had discovered that these birds are able to find their way home over a presumably unknown area of the sea when taken from their nests for a distance of at least 850 miles, (Watson '08). Material of both the noddy and sooty terns was kindly furnished by Professor Watson. The olfactory lobes and nerves were both found to be relatively small even when compared with those of birds with olfactory organs of moderate size. The diameter of the olfactory nerves is given in table 1. In this connection it is interesting to note that Professor Watson ('10) conducted experiments in which individuals of both species had their external nares closed with wax, but the ability of the birds to find their way at sea was apparently not disturbed.
Although the crows and ravens have often been credited with a keen sense of smell, the olfactory lobes and nerves of the specimens examined were found to be surprisingly minute (fig. 17). In all of the Corvidae material studied, this condition prevailed. These structures were found with some difficulty lying deep in the interorbital space. iVs the Corvidae have been considered by some writers as standing at the top of the bird series, it is unusually interesting that the olfactory organs are smallest in this group. The minuteness of the olfactorj^ lobes and nerves
THE SENSE OF SMELL IN BIRDS 645
argues for the greatest reduction of the sense of smell in this faaiih'.
The olfactory lobes of the two parrot heads which were dissected were found to be so merged with the fore brain lobes as to be undefinable, (fig. 13). Apparently the organs of smell are not well developed in these birds. The representatives of the Passerine birds studied were all found to have extremely small olfactory lobes and nerves, as has been the observation of otherwriters. The condition shown in fig. 16, is characteristic for the finches.
The writer agrees with Scarpa ('89), Schultze ('62), and Bumm ('83) in finding olfactory" organs of good size in the swimming birds. These organs are also fairly large in the shore birds (Charadrii formes). The olfactory lobes and nerves of the birds of prey have, on the other hand, been found to be of relatively moderate size, thus confirming Bumtn's observation. In fig. 12, a peculiarly fused and elongated condition of the olfactory lobes has been illustrated. In the specimens of the other birds of prey which were studied, these organs w^ere more or less similar to those of the doves.
On comparison of all the available material, the olfactory organs are found to be, in general, largest in the so called low^er groups and progressively smaller in the higher orders. The sense of smell has evidently been disappearing in birds w^ith the great development of the sense of vision. It seems not at all improbable that the sense of smell may be practically lost in the Passerine birds.
The ophthalmic branches of the trigeminus after emerging from the orbits usually cross the olfactory nerves dorsally near their distal ends. Their course is often so close to that of the olfactory nerves at this place as to make it easy to confuse the trigeminus with the olfactory nerve. In the figures, the trigeminus nerve branches have been drawai slightly separated when desirable for the sake of clearness. In those forms where the trigeminus branches to the upper mandible were found to be especially large, the olfactory lobes and nerves were also of good size as a rule, and it seems not impossible that they may function in the feeding operations of birds like the ducks, the flamingos, etc., in more or less intimate relationship wdth the trigeminus.
646 R. M. STRONG
5. RESULTS OF EXPERIMENTAL STUDIES OF THE SENSE OF SMELL IN RING DOVES
The methods which were employed in this work have been described on pp. 632-9, where it will be noted that a labyrinth was used, and the birds were compelled to find their food in one of four chambers which were selected by the experimenter at random.
If the doves had entered the chambers entirely in hit or miss fashion with no clues or habit preferences which would influence their choice of a chamber, they w^ould be expected, according to the law of ei-ror, to distribute their first choices equally among the four chambers, and the compartment which contained the food would be entered first 25 per cent of the time. What actually happened is shown in the following tables:
TABLE 2
Cologne series
No. 62, 19 per cent; No. 30, 28 per cent; No. 92, 33 J- i)er cent ; No. 24, 24 + per cent
TABLE 3
Musk series
No. 62, 20 + per cent; No. 30, 28.9 + per cent; No. 92, 31 per cent; No. 24, 25
per cent
TABLE 4
Violet sachet powder series
No. 62, 50 per cent; No. 30, 40 j)er cent; No. 92, 28 per cent; No. 24, 75 per cent
TABLE 5
Oil of t'.ergamot series
No. 62, 37 + per cent; No. 30, 31 + per cent; No. 92, .36 + per cent; No. 24, 47
per cent
It will be seen in the table that when cologne and musk were used, the percentages of correct first entrances were not significantly different from what might be expected according to the law of error. With oil of bergamot, however, the percentage is
THE SENSE OF SMELL IN BIRDS 647
notably large for all four birds. This is especially true of No. 24, a bird which gave other suggestions of discrimination in its behavior than are indicated in the tables. The percentages are also large in the violet sachet powder series, but not much significance can be attached to this fact because of the small number of experiments in this series.
It is of course conceivable that the food itself might have had an odor for the doves, even though none was apparent to the writer. However, if there had been any olfactory stimulation b} the food, all four of the birds probably would have shown similar responses to the stimulus. That failure to find the food through a hypothetical odor from the food material was not due to disturbances produced by the odors which were employed in the experiments, was indicated in a short series of control experiments when no odors were used. In this series, the food was found entirely by chance, apparently.
There was an unfortunate tendency for the doves to enter chamber A first by habit. After this compartment had been entered, if it was empty, they would go through the usual hunt for the food and would ordinaril}'- show no further tendencies to enter a second chamber by habit. The extent to which the birds made their first entrance by habit is indicated in the following tables.
Tables showing the number of times each chamber was entered first:
TABLE 6
Cologne series
No. 62 A, 8; B, 14; C, (5; D, 5
No. 30 A, 13; B, 1; C, 2; D, 7
No. 92 A, 1S;B, 0; C, 4; D, 8
No. 24 A, 12; B, 4; C, 8; D, 5
TABLE 7
Musk series
No. 62 A, 11; B, 15; C, 14; D, 5
No. 30 A, 32; B, 4; C, 3; D, 2
No. 92 A, 38; B, 4; C, 3; D, 5
No. 24 A, 23; B, 8; C, 4; D, 5
648 R- M. STRONG
Violet sachet powder series
No. 62 A, 0; B, 2; C, 1; D, 1
No. 30 A, 3; B, 0; C, 0; D, 2
No. 92 A, 6; B, 0; C, 0; D, 1
No. 24 A, 2; B, 2; C, 0; D,
TABLE 9
Oil of hergamot series
No. 62 A, 35; B, 73; C. 54; D, 24
No. 30 A, 109; B, 17; C, 8; D, 12
No. 92 A, 94; B, 31; C, 12; D, 39
No. 24 A, 31; B, 42; C, 24; D, 44
In those experiments where A was not the first chamber entered, the percentage of correct first entrances made by No. 30 in the oil of bergamot series was 41, and 44 per cent for dove No. 92.
It will be seen that doves Nos. 30 and 92 both entered chamber A first a very large number of times. This habit became so confirmed in No. 30 that experiments with this bird were finally discontinued. Attempts were made to break up the habit, but no success was obtained, except when the food was placed regularly at one of the three other chambers. This could not be done much, of course, without seriously affecting the results of the experiments. It was desirable that there should be little difference in the number of times each chamber was used for food. It will be noticed that dove No. 62 did not develop such a habit and that No. 24 did not exhibit the tendency in the oil of bergamot series.
In order to test the possible odor discrimination of the doves after they had made one mistake, the means of errors made by the birds were calculated. If, for instance, the birds entered the chambers at random and did not go into any single chamber more than once, they would, in a sufficiently large series of experiments, be expected to have a mean of errors approximating 1.5. As a matter of fact, they often entered an empty compartment more than once before finding the food. Thus, on one occasion, No. 92 entered chambers B and C each three times before
THE SENSE OF SMELL IN BIRDS
649
finding the food at A. In this case D was entered once. The record of this result was written as follows: No. 92 =B = D =C = B=C=B=C=A. This repetition of errors increased the size of the mean appreciably, at first, in the cologne series. In the following tables it will be seen that the means were significantly small when oil of bergamot was used, in spite of the fact that errors were repeated occasionally.
TABLE 10
Cologne series
BIRD
MEAN
PROBABLE ERROR
OF MEAN
62
1.54831.58
1.7083+
1.9689+
0.179±
30
0.1578±
92
0.182±
24
0.189±
TABLE 11 Musk series
BIRD
MEAN
PROBABLE ERROR
OP MEAN
62
1.5
1.447
1,549
1.6
30
92
24
TABLE 12
Oil of bergamot series
BIRD
MEAN
PROBABLE ERROR
OF MEAN
62
1.18+
0.058±
30
1.369+
0.0703 ±
92
1.193+
0.0625±
24
0.922
0.0663 ±
The totals of trials given the doves for different odors wil found in the following table:
be
650
R. M. STRONG
Bird.
Cologne
Musk
Violet sachet powder
Oilofbergamot I 186
62
30
92
31
25
30
44
38
51
4
5
7
1 186
146
176
1
24
29
40
4
141
G. CONTROL EXPERIMENTS WITH WHITE RATS
A pair of rats were used as a test of the efficiency of the apparatus. They gave the following results when oil of bergamot was employed, the conditions being those which were furnished the ring doves.
PERCENTAGE
MEAN
TOTAL NUMBER
OF TRIALS
Male rat
0.62
0.406+ 59
Female rat
0.71
0.316+ 60
It is the writer's opinion that the rats found their food usually, if not always, when not by accident, through an association of the odor of oil of bergamot with the location of the food. In a short series of trials which were made without any odor, the rats appeared to find the food (.sunflower seed) by the method of trial. That these keen scented animals made so many mistakes is probabh' explained by their tendency to enter the first chamber they came to and sometimes the next in order before they made any attempt to localize the source of the odor of oil of bergamot which was all of the time entering the enclosure.
7. RESULTS OF OTHER EXPERIMENTS AND OBSERVATIONS
The writer spent about ten weeks of the winter of 1906 in Florida, where some observations were made on the habits of the turkey vulture. Some very simple experiments with meat wrapped in paper resulted negatively, but the conditions of the experiments did not warrant the conclusion that meat is not smelled by these birds. During a tramp through a pine forest,
THE SENSE OF SMELL IN BIRDS 651
a turkey vulture was flushed from the entrance of a gopher-turtle hole. The bird showed a great disinclination to leave the spot although other individuals which were seen by the writer outside of cities were disposed to be wild. A dead gopher-turtle was found inside the burrow. It was impossible to view the turtle except when in a position to look down the oblique burrow, and it did not seem probable that a bird when flying overhead could see the body. A very strong odor of carrion prevailed for some distance on the lee side of the burrow.
The writer could not rule out the possibility that the vulture had found the turtle outside of the hole through its sense of vision and had later pushed the body inside, but it seemed unlikely that this had happened. The circumstances all appeared to favor the conclusion that the carrion had been smelled, even though the e\idence was far from conclusive. The well known naturalist, Henry Nehrling, whose estate was the headquarters of the writer during his stay in Florida, has also conducted some experiments with turkey vultures. He placed meat under boxes, and vultures actually alighted on these covers.
On September 6, 1909, the writer attempted a simple experiment on the sense of smell in Apteryx, at the London Zoological Gardens. Through the kindness of Mr. Pocock, three specimens of Apteryx mantelli were available. These birds were kept in a pen next to one occupied by some young kangaroos. The study was made in the evening after dark at the usual feeding time for the birds.
At the advice of a keeper, earthworms were selected for the experiments, as much appreciated food. Some flower pots partly filled with soil were placed in a row at one end of the pen, and one of the pots contained the worms. A very small amount of light from a lantern was used in order to note the movements of the birds, and even the feeble light which was employed was enough to check the activity of two. The third individual approached the pots rather shyly and inserted its beak at random. On two occasions, it seemed as though contact with the worms must have been nearly or quite effected, but the bird did not discover the food at these times. Eventuallj^, the worms were located, appar
652 R. M. STRONG
entlj' by hit or miss probing with the bill. There was no evidence of the acute sense of smell described by Benham ('06), though, of course, its existence was not disproved. The danger of serious effects upon the kangaroos from the disturbance produced by these experiments was so great that no further attempts were made by the writer.
It seems highly desirable that opportunities should be furnished at our great zoological gardens for thorough-going studies of behavior, which need not cause any injury to the animals. Little can be accomplished, as a rule, under the conditions which prevail. A few animals might be withdrawn from exhibition for a period of some weeks or months without any serious loss to the public and with much gain to science. The zoological gardens are in a position to acquire and maintain animals, which makes them the logical places for the study of species that are rare or difficult to keep in captivity.
An uncorked bottle of oil of bergamot was held within a few inches of the nostrils of two emeus, but no results were obtained.
8. CONCLUSIONS
The author agrees with Edinger C08a), that a sense of smell should be expected to occur in birds. Thus on page 440, Edinger says :
That part of the brain which in man and other animals is undoubtedly concerned with the sense of smell exhibits a constant arrangement and microscopic structure, not only in them but in all vertebrates down to the cyclostomes. We are therefore justified in the conclusion that an animal which possesses this part smells even though from its behavior nothing may be safely inferred. Indeed we may judge of the importance of the sense of smell to the animal according as this organ is large or small in relation to the remainder of the brain.
Unfortunately very little is known concerning the sense of smell in the vertebrates outside of mammals, but it is highly significant in this connection, that the olfactory organs of fishes have been demonstrated to function for a sense of smell, by Parker TIO), and Sheldon ('11).
THE SENSE OF SMELL IN BIRDS 653
In birds with relatively large olfactory lobes, such as Dromaeus and Fulmarus, the sense of smell should of course be stronger than in the Corvidae. The writer knows, however, of no observations upon the sense of smell in the first two birds other than the unsatisfactory ones already mentioned in this paper.
Man is classed as a microsmatic animal, and the human olfactory organs are relatively very small ; yet no one with a normal olfactory sense would deny that the human olfactory organs give their owners a large number of more or less intense impressions. It has not been practicable to make an exact comparison of the relative sizes of the organs of smell in birds and man, but those portions of the olfactory apparatus which are known in birds seem more extensive, relatively, in some birds, at least, than those of man. From a morphological standpoint, then, such birds as the Fulmar should be expected to have a more acute sense of smell than man possesses.
It is quite possible that olfactory stimuli may simply reinforce reactions to other stimuli such as visual and tactile impressions. We may have mutual relations of stimuli such as were found by Yerkes ('05 and '06), for the sense of hearing in the frog. Such birds as the duck or the flamingo, for example, may possibly, in probing for food, have tactile sensations which are received through the huge trigeminus nerves, modified by olfactory stimuli. The negative results of most experiments with vultures may have been due to a mutual relation between the olfactory and visual senses which made it difficult for the bird to react to an olfactory stimulus only.
The author agrees with Turner ('91), that the great reduction of the olfactory organs which has occurred in the higher birds would seem to indicate that the development of keen vision in birds is being accompanied by a degeneration of the olfactory sense which may result in its total loss, eventually.
Though the doves in the experiments described on pp. 646-650, never learned to find their food with perfect accuracy during a series of studies which extended, twice a day, through the greater part of about nine months, it is evident that they were stimulated by at least one of the odorous materials used, i.e., oil of bergamot.
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 3
654 R. M, STRONG
Unfortunately, the writer did not find it practicable to eliminate the possibility of stimulation of the nerves of general sensation in the experiments. The operations which would be involved in cutting all of the nerves of general sensation which might possibly be concerned were, in the writer's judgment, too severe to be worth attempting. Not only the innervation of the nasal cavities, but also that of the mouth chamber, would be involved. It has been suggested by Professor Herrick that a series of tests be made with birds whose olfactory nerves had been cut, but it has not yet been possible to attempt this desirable experiment. However, the extreme tenuity of the odorous material where stimulation occurred would appear to require a sensitivity far more acute than that which is known to be possessed by general sensory endings. A quantity of onh^ about 5 cc. of oil of bergamot, for instance, was used and there was no significant loss in volume during the months which were occupied by the series of experiments.
In the author's judgment, the results of the ring dove experiments warrant the conclusion that the behavior of some birds at least may be affected by olfactory stimulation.
THE SENSE OF SMELL IN BIRDS 655
BIBLIOGRAPHY
Audubon, J. J. 1835 Ornithological bibliography; vol. 2, pp. 33^7.
Beebe, C. W. 1909 New World vultures. Part II. Zool. Soc. Bull. New York;
no. 32, pp. 465-470, 7 text figures. Beeker, a. 1903 Vergleichende Stilistik der Nasenregion bei den Sauriern,
Vogeln und Saugethieren. Morph. Jahrb. Bd. 31, S. 565-619, Taf. 22-24. Belogowy, J. 1909-10 Zur Entwickelung der Kopfnervon der Vogel. Ein
Beitrag fur Morphologie des Nervensystems der Wirbeltiere. Bull.
Soc. Natural. Moscou. S. 177-337, 9 Taf. Benham, W. B. 1906 The olfactory sense in Apteryx; Nature vol. 74, pp. 222 223. Born, G. 1879 Die Nasenhohlen und der Thranennasengang der amnioten
Wirbeltiere. II. Morph.Jahrb. Bd. 5, S. 401-429. Taf. 23-24, 3 text
figs. BuMM, A. 1883 Das Grosshirn der Vogel. Zeitschr. f. wiss. Zool. Bd. 38, S. 430 467. Taf. 24-25. Carus, C. B. 1814 Versuche einer Darstellung des Nervensystems und insbe sondere des Gehirns nach ihrer Bedeutung, Entwicklung und Vollen dung im thierischen Organsimus. 6 Tafeln. Leipzig. CoHN, F. 1902 Zur Entwicklungsgeschickte des Geruchsorgans des Hiihnchens.
Arch. mikr. Anat., Bd. 61. S. 133-150, 1 Taf. 5, figs. Cyon, E. v. 1908 Das Ohrlabyrinth als Organ der mathematischen Sinne fiir
Raum und Zeit. 432 pp., 45 text figs. 5 Taf., Berlin, Julius Springer. Darwin, C. 1834 Naturalist's voyage round the world. New ed. 1896. D.
Appleton and Co., New York. 519 pp. Dieulafe, L. 1904 Les fosses nasales des vertebras. (Morphologie et embry ologie) Jr. Anat.Phys., Paris. Annee 40, pp. 268-298, text figs. 1-10.
1904a Les fosses nasales des vertebras (Morphologie et embryologie).
Jr.Anat.Phys. Paris. Annee 40, pp. 414-444; text figs. 11-19.
1905 Les fosses nasales des vertebres. Jr.Anat.Phys. Paris. Annee
41, pp. 478-560, text figs. 24-52. Disse, J. 1896 Ueber die erste Entwickelung des Riechnerven. Sitzungsb.
Ges. Bef. ges. Naturwiss., Marburg. Jahrgang 1896, Nr. 7, S. 77 92. 3 text fig.
1897 Die erste Entwickelung des Riechnerven. Anat. Hefte, Abth. 1. Bd. 9, S. 255-300. Taf. 20-23.
1898 Early development of the olfactory nerve. Jr.Anat.Phys., London, vol. 32, (n. s. vol. 12) pp. 12-16; 4 text figs.
Edinger, L. (mit a. Wallenberg und G. M. Holmes) 1903 Untersuchungen liber die vergleichende Anatomie des Gehirnes. 5. Untersuchungen iiber das Vorderhirn der Vogel. Abh. Senckenb. nat. Ges. Frankfurt a M., Bd. 20, S. 343-426, 7 Taf. 11 fig.
Edinger, L. 1908 Vorlesungen iiber den Bau der nervosen Zentralorgane des Menschen und der Tiere. Bd. 2, Aufl. 7, 334 S. mit 283 Abbildungen. Leipzig, F. C..W. Vogel.
1908a The relations of comparative anatomy to comparative psychology. Jour.Comp. Neur. Psych., vol 18, no. 5, pp. 437-457.
656 R. M. STRONG
Elliot-Smith, G. 1895 Xotes upon the morphology of the cerebrum and its
commissures in the vertebrate series. Anat. Anz., Bd. 11, no. 3, S.
91-96 ExNER, S. 1872 Weitere Studien fiber die i^iechschleimhaut bei VVirbelthieren.
Sitzungsber. kais. Ak'ad. Wiss., Wien. Bd. 65. Abth. 3, S. 7-41, 3
-Taf. Gadow, H. 1891 Broun's Klassen und Ordnungen des Thier-reichs. Bd. 6,
Abth. 4. Vogel. Leipzig., C. F. Winter. Gage, S. P. 1896 Comparative morphology of the brain of the soft-shelled
turtle (Amyda mutica) and the English sparrow (Passer domestica).
Trans. Am. Micr. Soc. vol. 17. pp. 185-238, 5 pis. Ganin, M. 1890 Einige Thatsachen zur Frage uber das jacobson'sche Organ der
Vogel. Zool. Anz. Jahrg. 13, S. 285-287. Gegexbaur, C. 1873 Uber die Nasenmuscheln der Vogel. Jenaische Zeitsch.
Bd. 7, S. 1 21. 3Taf. GiEBEL, C. G. 1876 Die Muscheln im Geruchsorgan der Singvogel nach C. I .
Nitszch's Untersuchungen. Zeitschr. f. d. ges. Natirwiss., Bd.47
(n.f., Bd. 13), S. 485^91. Taf. 2 B. Gill, E. L. 1904 The condor's sense of smell. Trans. Nat. Hist. Soc. Northumberland und Durham; vol. 1, (n.s.) p. 40. Herrick, C. J. 1893 Illustrations of the surface anatomy of the brain of certain
birds. Jour. Comp. Neur. Psych, vol. 3, pp. 171-176, pi. 26. Hill, A. 1905 Can birds smell'^ Nature, vol. 71, no. 1840, pp. 318-319. Kappers, C. U. A. UND Theunissen, W. F. 1908 Die phylogenese des Rhinen cephalons, des Corpus striatum und der Vorderhirnkommissuren.
Folia Neurobiol. Leipzig. Bd. 1 Nr. 2, S. 173-288. 5 text fig.
Taf. 2-4. Klinckowstrom, a. 1890 Les lobes olfactoires du Fulmarus glacialis. (Biol.
For. Forh.) Verb. Biol. Ver. Stockholm. Bd. 3, p. 10-11; 1 fig. Kolliker, a. v. 1890 Uber die erste Entwickelung der Nervi olfatorii. Sitzungsber. physik.-med. Ges. zu Wtirzburg, 14, S. 127-133. Marshall, A. M. 1878 The development of the cranial nerves in the chick.
Quart. Jr. Mic. Sc. n.s. vol. 18, pp. 10-40; pis. 2-3. MiHALKoyics, V. V. 1898 Nasenhohle und Jacobson'sches Organ. Eine morph ologische Studie. Anat. Hefte. Bd. 11. Abth. 1. S. 1-107. Taf. 1-11. MuNZER, E. UND Wiener, H. 1898 Beitrage zur Anatomie und Physiologie
des Centralnervensystems der Taube. Monatschr. f. Psychiatrie u.
Neur. Bd. 3. H. 5, S. 379^06. Taf. 5-8. Owen, R. 1866 Comparative anatomy and physiology of vertebrates; vol. 2,
586 pp. 406 text figs., London, Longmans, Green and Company.
1872 On Dinornis (Part XVI) : Containing notices of the internal
organs of some species, with a description of the brain and some nerves
and muscles of the head of the Apteryx australis. Trans. Zool. Soc.
London; vol. 7, pp. 381-396; pis. 45^7. Parker, G. H. 1910 Olfactory reactions in fishes. Jr. Exper. Zool., vol. 8,
no. 4, pp. 535-542. Parker, T. J. 1891 Observations on the anatomy and development of Apteryx.
Phil. Trans. Roy. Soc, London, vol. 182. (B) pp. 25-134; pis. 3-19.
THE SENSE OF SMELL IN BIRDS 657
Preobraschionsky, S. S. 1892 Beitrage zur Lehre iiber die Entwicklung Jes
Geruchsorganes des Huhnes. Mittheilungen Embryol. Instit. Univ.
VVien. Heft. 5, S. -19. 1 Taf. Ramon \ Cajal, P. 1890 Origen y terminacion de las fibras nerviosasolfactorias.
Con 6 granados. Gaceta Sanitaria Municipal de 10 de Diciembre de
1890. Abstract by Edinger in Bericht iiber die Leistungen auf dem
Gebiete der Anatomie des Centralnervensystems. Jahre 1890, p. 15-16. Raspail, X. 1899 Les sens de I'odorat chez lesoiseauK. Bull. Soc Zool. France
T. 24, p. 92-102.
1901 On the sense of smell in birds. Ann. Rep. Smithson. Instit.
1899, p. 367-373. (Translated from Revue scient. (4) f. 12, p. 144-148. Reeker, H. 1899 Zum Geruchsinn der Vogel. 27. Jahresber. westfal. Prov. Ver. S. 44^5. Rhoads, S. N. 1883 The power of scent in the turkey vulture. Am. Xat., vol.
17, pp. 829-833. Rouse, J. E. 1905 Respiration and emotion in pigeons. Jour. Comp. Neur.
Psych., vol. 15, pp. 494-513, 7 text figs. Scarpa, A. 1789 Anatomicae disquisitiones de auditu et olfactu. Ticini.
(Examined in the Senckenbergisches Bibliothek at Frankfurt a. M.,
Germany. Notes including the bibliographical reference were lost
before this paper was written.) ScHULTZE, M. 1862 Untersuchungen iiber den Bau der Nasenschleimhaut,
namentlich die Structur und Endigungsweise der Geruchsnerven bei
dem Menschen und den Wirbelthieren. Abhandlungen der Naturf.
Ges. Halle. Bd. 7, S. 1-100, Taf. 1-5. ScHtiPBACH, P. 1904 Beitrage zur Anatomie und Physiologic der Ganglienzellen
im Zentralnervensystem der Taube. Centralbl. Physiologic. Bd. 17,
S. 750-754. Sh.\rpe, R. B. 1874-1895 Catalogue of the birds in the Briti.sh Museum. 27
vols.; published by Briti^sh Museum, London.
1899-1909 A hand-list of the genera and spefcies of birds; 5 vols. British Museum, London. Sheldok, R. E. 1911 Sens? of smell in selachians. Science, n. s., vol. 33, no.
845, pp. 389-390. Stieda, L. 1869 Studien iiber das centrale Nervensystem der Vogel und Sauge tiere. Zeitschr. f. wiss. Zool. Bd. 19, Heft 1, S. 1-94. Taf. 1-3. Strong, R. M. 1908 The sense of smell in birds. Science, n.s., vol. 27, p. 943. Turner, C. H. 1891 Morphology of the avian brain. Jour. Comp. Neur.
Psych., vol. 1, pp. 39-92, pis. 5-8.
1891a Morphology of the avian brain, pp. 265-286; pi. 18. Watson, J. B. 1908 The behavior of noddy and sooty terns. Carnegie Instit.
Washington. Pub. 103, pp. 187-225, 11 pis., 2 text figs.
1910 Further data on the homing sense of noddy and sooty terns.
Science, vol. 32, no. 823 pp. 470-473. Yerkes, R. M. 1905 The sense of hearing in frogs. Jour. Comp. Neu'r. Psych.,
vol. 15, no. 4, pp. 279-304, 7 text figs.
1906 The mutual relations of stimuli in the frog Rana clamata Daudin.
Harvard P.sychol. Studies, vol. 2, pp. 545-574, 10 text figs.
658 R. M. STRONG
PLATES
Original drawings made with the aid of photographs and the Edinger projection apparatus. The cerebrum and the cerebellum have been included to indicate the relative size of the olfactory lobes.
ABBREVIATIONS
c. a., Anterior concha or turbinal. c. p. Posterior concha or turbinal.
c. m.. Middle concha or turbinal. t., Ophthalmic branch of trigeminus nerve.
PLATE 1
EXPLANATION OF FIGURES
(All figures X l)
1 Dorsaiviewof olfactory lobes and nerves in Struthio camelus, (ostrich). Posterior ends of nasal chambers shown semi-diagrammatically.
2 Dromaeus novae-hollandiae, (emeu). Dorsal view showing olfactory lobes and a portion of the nasal chambers. A posterior portion of the roof of the left nasal chamber has been dissected away to show the contour of part of the posterior turbinal. The right posterior turbinal and a posterior portion of the right middle turbinal are also exposed. Contour lines for the eyes and bill have been sketched at the sides of the drawing for orientation.
3 Dromaeus. Lateral view of specimen shown in fig. 2, but without the nasal chambers.
4 Turturrisorius, (ring dove). Dorsal view to show olfactory lobes, olfactory nerves, and sections of the ophthalmic branch of the trigeminus nerve.
5 Fulmarus glacialis, (fulmar). Dorsal view of portion of dissected head with the brain case material which separates the eyes from the brain removed. The right nasal chamber and a posterior portion of the left nasal chamber are exposed. The posterior turbinal of the right side has been opened at d to show the turns or rolls of its structure. The middle and anterior turbinals have been mutilated slightly in the dissection.
6 Fulmarus. Lateral view of right olfactory lobe. A small posterior bit of the right nasal chamber is represented semi-diagrammatically in contact with the right olfacitory lobe.
7 Pavoncplla pugnax, (rufif). Dorsal view. Sections of the ophthalmic branches of the trigeminus nerve are figured pulled apart very slightly to avoid confusing their outlines with those of the olfactory nerves.
8 Pseudotan talus leucocephalus, (white-headed ibis). Dorsal view to show olfactorv lobes and nerves.
^KXSE OF SMELL L\ BIRDS
n. M. STRONG
JOURNAL OF .MORPHOLOOy VOL.
660 R. M. STRONG
PLATE 2 KXPLANATIOX OF FIGURES
Fi^ureH lo imd 16 are X I5. All other figures X 1.
9 Lei)totihi.s crumeniferu^, (African adjutant). Dorsal view to show olfactorj' lobes, olfactory nerves, and posterior portions of the nasal chambers. The ophthalmic branches of the trigeminus nerves are seen pulled slightly apart, and they enter the roof of the nasal chambers anteriorly. The main branches of the right olfactory nerve in the dorsal inner roof of the nasal chamber are shown as they appear in such a preparation. The corresponding area in the left nasal chamber has been opened.
10 Phoenicopterus roseus, (European flamingo). Dorsal view showing olfactory lobes, olfactory nerves, posterior portions of the nasal chambers, and ophthalmic branches of the trigeminus nerves. Contour lines for the eyes have been drawn.
1(^ Catharistes urubu, (black vulture). Dorsal view showing olfactory lobes and nerves, posterior portions of the nasal chambers, and trigeminus branches. The trigeminus nerves are seen ascending from the orbit and converging at some distance anterior to the junction of the olfactory nerves with the nasal chambers. Contour outlines for the eyes are sketched.
12 Circaetes gallicus, (serpent eagle). Dorsal view showing fused olfactory lobes, olfactory nerves, and posterior ends of the nasal chambers. The fused and lengthened condition of the olfactory lobes which has been shown here was not found in the other birds of prey which were examined.
13 Chrysotis auripalliata, (golden-naped Amazon parrot, j Dorsal view of dissected head. The olfactory lobes are seen to be fused with the diverging forebrain lobes, and they are practically non-definable by the method of dissection alone. The olfactory nerves are seen to diverge widely. The right nasal chamber has been exposed to show the turbinals. The tissue composing the orbits has been removed, and the eyes are consequently not separated in the figure from the cerebrum and the olfactory nerves.
14 Coccyges erythrophthalmus, (black-billed cuckoo). Dorsal view of olfactory lobes and nerves.
15 Motacillaalba, (white wagtail). X |. Dorsal view showing the very small andfusedolfactorylobesmergedwiththetaperingforebrainlobesso as to be undefinable by the method of dissection only.
16 Coccothraustes coccothraustes, (hawfinch). X 4- Dorsal view of portion of dissected head. The right nasal chamber has been partly exposed. The olfactory nerves are shown throughout most of their extent. Oibit tissue removed.
17 Corvus corax, (raven). Dorsal view showing the minute olfactory lobes and the slender olfactory nerves. The posterior ends of the nasal chamber are included.
SENSE OF SMELL IN BIRDS
R. M. STKONG
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 3 G61
ON THE REGULAR SEASONAL CHANGES IN THE RELATIVE WEIGHT OF THE CENTRAL NERVOUS SYSTEM OF THE LEOPARD FROG
HENRY H. DONALDSON
The Wistar Institute of Anatomy and Biology
FIVE CHARTS
The bearing of this investigation can best be understood by a short account of the steps leading up to it. In some earlier studies on the innervation of the muscles and skin of the leg of the bullfrog and of the leopard frog (Donaldson '98) (Donaldson and Schoemaker '00)^ the weights of the brain and of the spinal cord of the frogs were taken, and the percentage of water in these two portions of the central nervous system determined.
When the records thus obtained were assembled, the arrangement of them as they appeared on the chart suggested that the increase ifi the weight of the central nervous system might run parallel to a logarithmic curve, based on the weight of the entire body. The curve based on this datum alone was however found to fall away from the observed values as the body weight increased and hence a second factor, the value of which increased gradually but at a diminishing rate, was necessary to make the calculated values correspond to those observed. This second factor was found in the total length of the frog, the fourth root of which increased at such a rate that when the logarithm of the body weight is multiplied by the fourth root of the total length, the values obtained are an almost constant fraction of those observed. It remained then merely to multiply the number thus found by a constant to approximate the observed values.
' In the paper cited above and in several ottier publications from my laboratory, the leopard frog has been designated R. virescens brachycephala Cope. ►Since 1908 the name Rana pipiens has been used (see Donaldson, Science, vol. 26, p. 655, 1907).
663
664 HENRY H. DONALDSON
The formula for the weight. of the central nervous system was accordingly written
Weight C. N. S. = (Log. W X \'l) C
where 'weight C. N. S.' is the combined weight of the brain and spinal cord in milligrams ; W the body weight in grams ; L the total length of the frog in millimeters, and C, a constant empirically determined.
Corresponding results were obtained for both the bullfrog and leopard frog (Donaldson '02).
By this formula it is possible to calculate the approximate weight of the central nervous system of the frog from the data on body weight and total length, and also to show its growth.
The observations used in the foregoing stud}^, from which the formula was obtained, were taken from summer frogs {i.e., in the case of the bullfrogs, July and August, and in the case of the leopard frogs, June and July).
In commenting on these results, I pointed out at the time that it was necessary to avoid several sources of observational error. These are represented (1) By variations in the moisture of the frog, and therefore onh' frogs that have been kept moist for some hours at least, should be used. (2) By loss of weight during captivity, especially in frogs taken in the spring and early summer. Hence such frogs must be examined either as soon as caught or must be kept under special conditions or some correction must be made for the loss which they undergo. (3) By season ; as I noted that both in the few spring and autumn frogs which I had examined, the nervous system was apparentlj^ relatively lighter than in frogs killed during the midsummer.
In the course of this work, the first two sources of error were taken into account, and corrections made where they were deemed necessary. Also, as just stated, the third was escaped by using summer frogs only.
The difference thus found between the relative weight of the central nervous system in the summer and in the spring and autumn , appeared to me worth further examination, for unless it could
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM 665
be satisfactorily explained, the formula which had been suggested for determining the weight of the central nervous system seemed to have only limited apphcability.
For this reason, in 1901-02, I endeavored to get data which would show whether a regular seasonal variation took place. Since that time, I have examined for the same purpose other series of frogs in 1908 and also in 1909.
The general result of these observations is to show that the relative weight of the central nervous system of the leopard frog does change during the year, being constant during hibernation, low in the spring, high in the summer and low again in the autumn, when the frogs go into hibernation.
The discussion which follows is intended to present I. The evidence that a seasonal change occurs.
II. The biological interpretation of this change.
I. THE EVIDENCE THAT A SEASONAL CHANGE OCCURS. TECHNIQUE AND SOURCES OF ERROR
In all of the series about to be described, the technique for examination has been essentially uniform. Specimens of Rana pipiens, the leopard frog, alone were used. The frogs were kept moist for several hours before dissection. They were killed with chloroform and the body weight = (Bd. W.) taken to the nearest 0.1 gm. The frog was next either suspended or laid flat on its ventral surface, with the legs fully extended, and the distance from the tip of the nose to the tip of the longest toe taken with a jointed calipers and then read off on a scale to the nearest milUmeter = (total length). While in the ventral position, the long axis of the head was brought in line with that of the body by raising the head with a small wooden wedge, and with a vernier calipers the distance from the tip of the nose to the tip of the urostyle — the cartilaginous end of which was exposed by a slit through the skin — was measured and read to the nearest 0.1 mm. =(body length).
The frog was then placed on its back, opened and all the viscera removed.
666 HENRY H. DONALDSON
At this time any necessary correction in the body weight was made by subtracting from the initial weight, the weight of the ova or of undigested food distending the stomach. These were the only two corrections made to the body weight. The body weights of the females are always given without the ova.
After evisceration, the brain was exposed through its entire length and the spinal cord exposed as far down as the Til nerve. With spring compasses, the length of the brain from the tip of the olfactory bulbs to a point midway between the tip of the calamus scriptorius and the level of the III nerve = (origin of II nerve) was taken. This was recorded to the nearest 0.1 mm. = (brain length) The olfactory nerves were next cut through with a very fine scissors, and in the same manner a section was made between the tip of the calamus scriptorius and the III nerve, i.e., the level of the emergence of the II nerve. The choroid plexus over the fourth ventricle was removed and then the brain was raised from behind forwards on a narrow lifter, and the nerve roots severed as close to the brain as possible.
If the hypophysis was still attached to the brain, it was removed and the remaining mass at once placed in a closed weighing bottle and weighed to 0.1 milligram = (brain weight).
Similarly the spinal cord was exposed through its entire length and the conus just caudad to the XI nerve laid bare. With the spring compasses the length from this point to the level at which the cord had been severed from the brain was measured and recorded to the nearest 0.1 mm. = (cord length). The cord was then seized just below the conus with a fine forceps and raised so that the nerves could be clipped away close to the cord. The mass of the cord, thus deprived of nerves, was placed in a closed weighing bottle and at once weighed to 0.1 milligram = (cord weight.)
Both parts of the central nervous system were then dried at 90-95° C. for a week and re weighed. From these data, the percentage of water in the brain and in the spinal cord was determined to 0.1 per cent = (percentage water, brain) = (percentage of water, spinal cord).
For the present investigation, the foregoing determinations represent all that are necessary.
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM 667
Observations on the frogs examined at Chicago, 1901-1902
Beginning in the spring of 1901, I endeavored to examine four males and four females of R. pipiens each week from March 28, 1901, to the following March, 1902. As will be seen from table 1, this plan met with only approximate success.
The frogs examined in Chicago were obtained from a dealer who was in turn supplied from a wide range of country extending from southern Minnesota to Indiana.
It was assumed at the beginning of this work that season was the main cause modifying the relative weight of the central nervous system in these frogs, and that therefore the fact that the frogs were taken from different localities would not materially modify the results. Consideration of all the facts at present in hand now leads me to think, on the contrary, that the relative weight of the nervous system is modified by station, and that in the case in question, the frogs from different stations were mixed together. As a rule however, the lots which were obtained were rather uniformly mixed and so, except in a few cases, no serious discrepancies appeared.
As has been stated, it was planned to examine every week, four frogs of each sex or eight in all. From March 28, 1901, to October 2, this plan was followed with moderate success, although the number of complete records is less than of those incomplete. This has come about by reason of the fact that although the full number of frogs was examined almost every week, nevertheless when the percentage of water in the brain and in the spinal cord was determined later, it was found in some cases that it was beyond the normal limits. These I have set for the brain as 83.585.5 per cent and for the spinal cord as 79.5-81.5 per cent. When any record transgressed these limits for both the brain and spinal cord, such a record was excluded as it was assumed that deviation to this extent in both divisions of the central nervous system meant that the frog was in an abnormal state.
Between October 2nd and the end of the year, this deviation in the percentage of water made it necessary to exclude all the records. The frogs obtained by us at this season had evidently
668
HENRY H. DONALDSON
been caught much earlier and were suffering from the conditions under which they were kept.
It was then not until the end of January, 1902, that a few freshly captured frogs were brought in and these were used for series 37, the last one in table 1.
The frogs examined were selected from the dealer's tank less than 24 hours before they were to be dissected, and kept in proper
TABLE 1 Data on frogs from Chicago, 1901-02
VALUE OF C
3 I
4 i 4 ! 4 \ 4 I 4
4 4 3 4 3 4 2 4 2 4 4 4 4 4 4 3 4 3 2 3 3 4
1901
March
April
April
April
April
April
May
May
May
May
June
June
June
June
July
July
July
July
July
August
August
August
August
September 3
September 12
September 17
September 24
October 2
1902
January 30
3 10 16 23 30
8 15 21 28
4 11 19 26
3
9 17 23 30
6 13 20 27
Range
Brain 1 Cord
24.7,
24.8 I 25.4
26.9 j 27.0
25.0 I 28.6 j
27.3 I
28.1 I
27.2 I
29.4 ' 26.1 j 26.4*1 25.2*, 24.2* 25.1*1 27.9 I 30.0
29.3 I 27.0 j
29.4 I
28.0 j
28.5 I
27.1 :
29.2 I 24.9 i
28.6 j 28.1 I
24.8 i
(24.2-25.7) (22.5-26.7) (23.6-28.7) (22.9-29.4) (24.3-31.4) (23.0-26.3) (24.6-32.7) (24.1-29.8) (24.1-30.5) (24.4-30.9) (24.5-33.8) (23.5-29.1) (21.5-30.1) (23.5-27.5) (22.9-25.4) (23.7-26.3) (25.9-30.7) (27.3-31.4) (22.7-33.7) (25.3-30.9) (28.1-31.5) (24.8-31.1) (24.9-33.6) (23.6-31.2) (24.0-32.7) (22.8-25.9) (26.4-30.6) (26.4-32.3)
84.1 84.4 84.7 84.7 84.2 85.0 85.1 84.4 84.9 84.4 85.0 84.2 85.0 85.5 85.6 84.4 84.3 84.9 85.0 84.5 85.1 84.9 84.6 84.5 84.1 85.2 84.3 84.4
79.5 80.2 79.8 80.7 80.2 80.7 80.9 80.9 80.5 80.4 81.0 80.2 80.9
79.8 79.7 80.7 80.5 79.7 80.4 81.3 80.4 79.0 79.1 80.9 80.2 80.6
(22.9-27.8) 84.6 i 81.1
♦Seepage 671.
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM 669
jars in the laboratory until examined. Enough frogs were taken to give as a rule four of each sex, and the endeavor was made to have them of nearly the same size, i.e., 22-29 gms. in body weight, A glance at table 2 will show that this effort was largely successful. The measurements which were made included (1) the total length, (2) the body weight (3) the weight of the brain (4) the weight of the spinal cord, as well as the length of each of these portions of the central nervous system and the percentage of water in each.
These data, from 1-4, are necessary for determining the value of the constant C in the formula
Weight of C. N. S. = (Log W x \L) C.
It will be seen from the inspection of the formula that if there are two frogs of the same body weight and the same total length, but differing in the weights of their central nervous systems, then in the case of the frog with the lighter nervous system, the value of C will be less than in the other, for by the terms of the formula, the weight of the central nervous system is
C times (Log W x\l)
a value which in this instance is the same for both the frogs. It becomes therefore possible to express the changes in the relative weight of the central nervous system by the differences in the value of C — and it is this method which has been here employed.
This variation in the constant C not only enables us to express the variations in the relative weight of the central nervous system, but is by far the best method available for the purpose, as it is quite independent of the absolute size of the frog in any instance.
In table 1, the average value of C is given together with the range in this value for each series.
If these average values for successive weeks are plotted, they show great irregularity. By grouping the series however, so as to take the first one alone, then the next three series as the second group, and finally the remaining series up to the very last one in groups of four, and the last one again alone, we obtained the data
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 3
670
HENEY H. DONALDSON
TABLE 2
Data on frogs from Chicago, 1901-02. Fundamental table giving the mean values of all the data for each group
NUMBER OF SPECIMENS
BODY WEIGHT
M.
F.
3
12
16
14
12
16
14
12
2
3
8
14
14
12
15
14
14
2
gms.
21.2 22.6 21.3 22.3 24.4 22.4 23.3 23 19.4
165
2-4
170
5-8
168
9-12
13-16
168
174
17-20
166
21-24
166
25-28
165
37
166
Brain
SERIES
WEIGHT
LENGTH
PERCENTAGE OF WATER
1
gyns.
0.0793 0.0857 0.0870 0.0917 0.0859 0.0939 0.0944 0.0933 0.0793
13.7
13.7
13.6
13.9
14.1
14.2
14.2
14.1
13.4
84.1
2-4
84 6
5-8
84.7
9-12
84.6
13-16
85.1
17-20
84.7
21-24
84.8
25-28
84.5
37
84.6
Cord
1
0.0382
0.0410
0.0415
0.0423
0.0416
0.0443
0.0453
0.0429
0.0344
14.6
15.6
15.5
15.3
15.6
15.5
15.3
15.2
15.4
79.5
2-4
80.3
5-8
80.7
9-12
80.6
13-16
80.5
17-20 ....
80.1
21-24
80.4
25-28
80.2
37. .
81.1
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM 671
given in table 3. This table shows that the mean value of C rises from March 28th to a maximum in July and then begins to fall. In the case of one group, formed by the series 13, 14, 15 and 16, there appears an unexpectedly small relative weight of the central nervous system, while at the same time there is no ground to exclude these series as abnormal. I assume therefore that the series forming this group came from localities where the frogs had a proportionately small nervous system and that these were not mixed as in the case of the other series, with frogs from more northern stations, in which the nervous system happened to be larger. For reasons given earlier, the autumn fall is very incompletely shown, but the observation of the January series gives, as one would expect, a value of C similar to that found in March. As a control,
I have calculated the probable error of the mean ( ± .6745 ~' of C in all the groups of table 3.
As will be seen, in the groups that consist of four series, this is quite constantly about 0.3. This value is high but when it is considered that the ranges of the value of C in the several series
TABLE 3
Data on frogs from Chicago, 1901-02
'sPECTMENS^ VALUE OF C PERCENTAGE OF WATER
SERIES
M.
MEAN D.^TB
F.
Probable
Mean error of Brain
mean
i
Cord
1
3
12
16
14
12
16
14
12
2
1 1901
3 March 28 8 April 9 14 May 4
14 June 1 12 June 29*
15 July 27 14 • August 24 14 September 22
1902 2 January 30
■24.7
25.8 27.0 27.6
25.8 28.5 28.3 27.9
24.8
±0.16
±0.32
±0.29
±0.31
±0.27
±0.31
±0.30
±0.32
±0.67
84.1
84.6
84.7
84.6
85.1
84.7
84.8
84.5
84.6
79.5
2-4
5-8
9-12
13-16
17-20
21-24
25-28
37 ...
80.3
80.7
80.6
80.5
80.1
80.4
80.2
81.1
"See Comments on p. 671.
672 HENRY H. DONALDSON
are much the same (table 1) then it would follow that the size of the probable error of the mean would rapidly reduce as the number of cases was increased. It is by reason of this fact that I still consider the successive mean values of C significant, despite the large probable error.
When the data in table 3 are put in the form of a chart (1) where the ordinates represent the values of C on a base line of time in days, the relations aboA^e described are shown clearly. We conclude therefore from this series that the relative weight of the central nervous system of the leopard frog rises from the time that the frog appears in the spring until midsummer, arid
29(- VALUE OF c. ^ CHART I
• CHICAGO
FEB. MAR. APR. MAY. JUNE. JULY AUG. SEPT. OCT. NOV. DEC. JAN.
Chart 1 Based on data in table 3. Determination for (\ June 29, not entered.
then falls until the frog goes into hibernation, i.e., when the external temperature drops to 7°-10° C. =45°-50° F. (Torelle, '03). It should be added however that this critical temperature is probably modified by latitude and tends to become lower as we pass from south to north in the range of the frog.
In table 2 are given the necessary fundamental data relating to the several groups belonging to this lot of Chicago frogs.
Observations on the frogs from Minnesota, 1908
The exact locality of the frogs in this series is not known. They came however from Minnesota, probably from southern Minnesota and all from the same station. They were delivered in good condition in Philadelphia where they were examined. As
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM
673
table 4 shows, there were but three series and the numbers, except in the last series are small. The treatment of these Minnesota frogs was similar to that for the Chicago frogs, and need not be again described.
As shown in table 5, there is a spring series, a summer series and a late autumn series (entered in two parts) and in these three series the relative weight of the central nervous system as shown by the values of C, undergoes a seasonal variation corresponding to that found in the Chicago frogs — although the absolute values for C are much higher. This variation is exhibited in chart 2,
Data on frogs from Minnesota, 1908. Fundamental table giving the mean values for each series
NUMBER OF SPECIMENS
U. , F.
gms.
mm.
mm.
1
2
2
50.5
212
79.5
2
5
56.4
225
84.2
3
3
3
48.5
213
78.4
3'
3 9
64.9
230
85.4
Brain
SERIES
WEIGHT
LENGTH
PERCENTAGE OF WATER
gms.
mm .
1
0.1250
16.4
84.8
2
0.1475
17.5
85.6
3
0.1245
16.6
85.0
3'
0.1334
17.5
84.6
Cord
1
0.0577
18.5
79.9
2
0.0640
18.7
80.4
3
0.0582
18.1
80.1
3'
0.0665
19.3
79.9
674
HENRY H. DONALDSON
For the Chicago frogs we had no October observation, but in this case we do have one and it is seen that it occupies the position which we should expect, and thus supplements and extends the Chicago data.
When first computing the values of C for the Minnesota frogs, taken in October, it seemed important to use specimens having the same body weight as those used in the preceding series, so the first entry for October made in tables 4 and 5 is for the six specimens having an average body weight of 48.5 gms. This entry is designated 3. Later the value of C was determined for the re
TABLE 5
Data on frogs
from
Minnesota
1908
NUMBER OP
SPECIMENS
DATE
VALUE OF C
SERIES
M.
j
F.
Mean
Probable \
error of the
meant
Range
190S
1
1
1
1
2
2
March
26
28.1
±0.30
(27.0-29.4)
2
5
June
10
31 1
±0.65
(28.8-34.8)
- 3
3
3
October
19
28.2
±0,66
(24.5-32.6)
t3'
3
9
October
19
1
28.3
±0.36
(25.1-32.1)
- Series 3 (six cases) has an average body weight of 48.5 gms.
t Series 3' (twelve cases) has an average body weight of 64.9 gms.
X See comments on pp. 671-672.
maining twelve specimens having an average body weight of 64.9 gms. This latter record is entered in the tables as 3' . The value of C is the same in both series. This gives me the opportunity to correct a statement previously made, (Donaldson '10, p. 14) to the effect that the value of C is in a measure influenced by the absolute size of the frog. This conclusion I now think erroneous. It may be added that in the paper just cited the argument is not altered by the introduction of the data for the groups of frogs there excluded from comparison on account of their body weight. The most striking difference between the observations on the Minnesota frogs and those on the Chicago frogs is the high value
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM 675
of C in the former, but the discussion of this point will be deferred until the observations on the next lot of frogs have been presented.
The probable error of the mean values of C given in table 5 is open to the same interpretation as was given in the case of the Chicago frogs (see pp. 671-672).
We conclude from this study of the Minnesota frogs that the relative weight of the central nervous system is low in the spring, high in the midsummer and low again in the autumn, and these relations are shown on chart 2.
CHART 2
MINNEAPOLIS
3029
FEB. MAR. APR. MAY. JUNE. JULY AUG. SEPT. OCT. NOV. DEC. JAN.
Chart 2 Based on data in table 5.
Observations on the frogs from the Brandy wine, 1909
These frogs were brought from the Brandywine Creek near Philadelphia, and in some ways represent the best of the three lots. The plan was to examine about twelve specimens at intervals of a month or less between the appearance of the frogs in the spring and their disappearance in the autumn. The two February series, as well as those of October and November, were taken within a spring house, and those of the intervening series from the neighborhood. The frogs did not emerge until the end of March. The July series was overheated in transport and could not be used. In all there are eleven series recorded.
676
HENRY H. DONALDSON
As in the other eases, all the frogs examined in which the percentage of water was normal, have been included in the records, but those with abnormal percentages of water in the central nervous system have been omitted, hence many of the series contain less than twelve records (see table 6), which also gives the fundamental data for this lot.
TABLE 6
Data on frogs from the Brandywine, 1909. Fundamental table giving the mean values for each series
NUMBER OF SPECIMENS
M.
F.
gms.
mm.
mm.
1
5
5
30.3
176.
66.8
2
6
5
29.3
172
66.2
3
4
26.7
166
63.2
4
12
21.0
158
. 60.1
5
5
6
32.7
188
71.7
6
6
2
27.7
179
63 8
7
8
3
5
5
7
28.8
45.5
176
197
66.3
9
75.8
10
2
6
40.5
190
71.5
11
4
7
32.1
176
67.0
12
2
3
33.0
182
68.9
Brain
SERIES
WEIGHT
LENGTH
PERCENTAGE OP WATER
1
0.0881
0.0876
0.0867
0.0729
0.1074
0.0895
0.0962 0.1178 0.1058 0.0915 0.0927
14.7
14.5
14.6
14.4
15.4
14.2
15.0 15.9 15.3 14.7 15.0
83.9
2
83.6
3
84.2
4
5
6
85.0
85.3
84.6
7
8
85.0
9
84.8
10
85.2
11
84.8
12
84.8
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM G77
Cord
SERIES
WEIGHT
LENGTH
PERCENTAGE OF WATER
1
0.0394
16.1
79.9
2
0.0382
16.4
78.9
3
0.0354
14,8
79.4
4
0.0334
15.0
80.3
5
0.0429
16.5
80.4
6
0.0361
15.8
79.1
8
0.0417
15.7
SO. 9
9..
0.0497
17.6
80.0
10
0.0446
17.2
80.4
11
0.0392
16.3
80.7
12
0.0404
16.6
80.6
When the values of C for these frogs are plotted as in chart 3, we obtain a series of records which, although irregular for the midsummer season, fit very fairly with the entries in the two preceding charts. As in the Chicago lot the average value of the probable error (table 7) is 0.3. The size of the error in this case is open to the same explanation as was given before in the cases of the Chicago and Minnesota lots. The diminution on the relative weight of the central nervous system in the autumn is shown better in this series than in either of the two preceding.
The data as they stand justify in this case the conclusion which has been drawn from the other two sets of data; namely, that the relative weight of the central nervous system is low in the spring, high in the midsummer and low again the autumn. It will be noticed that the absolute values for C are again different from those in the preceding cases, being the lowest of all.
It is hard to compare the ranges of C on the charts 1, 2 and 3 because of the difficulty of constructing satisfactory curves by which to make the comparison. I made for my own use however a series of curves, employing each to control and correct the others and obtained the following relations of C, always taking the late March value as the initial one.
The Chicago frogs were found to range in the value of C from 24.7-28.0 or 3.8 points, the Minnesota frogs from 28.1-31.3 or 3.2 points, and the Brandywine frogs from 24.0 to 27.0 or 3.0
678
HENRY H. DONALDSON
points. Thus the percentage gains are 15.4 per cent, 11.4 per cent and 12.4 per cent respectively. This indicates that the frogs from the several localities change not only in the same manner, but also to about the same extent.
VALUE OF C.
CHART 3
BRANDYWINE
\y. JUNE. JULY. AUa SEPT. OCT. NOV.
Chart 3 Based on data in table 7.
When we take the mean of the percentages given above, we find it to be 13.1 per cent, the true maximum in all cases falling in July. This result can be controlled by another treatment of the data.
TABLE 7 Data on frogs from the Brandywine, 1909
NUMBER OF SPECIMENS
DATE
VALUE OF C
SERIES
M.
F.
Probable
Mean error of Range
mean*
1
5
12
5
5
4
n
1909
1
23.7 ' ±0.20 : (21.9-25.1)
2
February 22 23.8 ±0.36 (21.7-26.8)
3
March 23 24 ±0 29 (20.9-26.0)
4
April 21 24.8 ±0.20 1 (23.2-26.3)
May 20 26.8 ±0.61 ' (22.5-31.7)
June 19 26.3 ±0.30 (24.4-28.9)
5 . .
5 6
6
6
3 5 2
4 2
2
5
7 6 7 3
7
- 1
8
August 13 ' 26 1 i ±0 24 (24.1-27.1)
9
September 8 27.2 ±0.31
September 24 25.2 ±0.57
October 19 24.5 ±0.43
November 16 2S 9 ±0 Ifi
(25.2-30.6)
10
(21.6-29.6)
11
(20.2-28.7)
12
(22.7-26.0)
i
- See comments on pp. 671-672.
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM 679
The percentage chang
es in the value of C in each
of the three lots of frogs.
The March
value used as the standard
is underlined
Chicago lot
Minnesota lot
Brandywinb lot
o
^
»
o
H
b
■< w
h
■< H
B.
°H
go
o
H S
H
DATE
p ij —
<
§2
date
<
P
>
p.
>
0.
>
23.7
E
February 4
—1.0
February 22
23.8
-0.8
March 28
24.7
March 26
28.1
March 23
24.0
April 9
25.8
4.8
April 21
24.8
3.3
May 4
27.0
9.3
May 20
26.8
11.6
June 1
27.6
11.7
June 10
31.1
10.7
June 19
26.3
9.6
June 29
July 27
28.5 15.4
August 13
26.1
8.7
August 24
28.3
14.6
September 8
27.2
13.3
September 22
27.9
13.0 II
September 24
25.2
5.0
October 29
28.2
0.3
October 19
November 16
24.5
23.9
2.1
-0.4
January 30
,24.8
±4
1
The percentage change in the value of C as indicated by the observations when the late March value is taken as the standard, is given for all three lots in table 8.
When the data on these tables are put in the form of a chart (chart 4) and then an ideal symmetrical curve is drawn a number of interesting relations come into view.
In the first place the maximum of this curve in July is 13 per cent above the initial value of C. Second, during the month from the end of March to the end of April, C increases about 7 per cent; during the month from the end of April to the end of May, about 4 per cent, and from the end of May to the end of June, about 2 per cent. During July little change occurs and then the converse changes follow during the three months from August first to the end of October. During hibernation, November first to the end of March, the value of C is nearly constant.
These values are admittedly only approximations, but when so understood, they serve to show the general course of the seasonal changes.
680 HENRY H. DONALDSON
From the foregoing we are justified in concluding that there is a seasonal change in the relative weight of the central nervous system of the leopard frog, R. pipiens, and that this occurs regularly each year and in frogs taken from widely separated localities.
Moreover, if we know in any case the value of C for a colon}^ of frogs at a given date, it is possible in accordance with these results to determine approximately what the value will be for other representatives of the same colony, at any other season of the year.
Nevertheless in the first instance, the values of C for a given colony must always be determined by direct observation. We have seen that at similar dates the value of C for the Brandv
CHART 4
PER CENT CHANGES IN THE VALUE OF C.
FEB. MAR. APR. MAY. JUNE. JULY. AUG. SEPT. OCT. NOV. DEC. JAN.
Chart 4 Based on data in table 8. Also giving an ideal curve about which the several records are grouped. •
•Records from Chicago series. ° Records from Minneapolis series. X Records from Brandywine series.
wine frogs is 24.0, for the Chicago frogs 24.7 and for the Minnesota frogs 28.1. The differences between these series I refer to the general effect of the external conditions ( = food supply, abundance of water, etc.) but whatever the explanation is, such differences must always be anticipated.
Further, the individual variation in this character is large so that all determinations should be based on data from groups, and not on single eases.
On the other hand, although the frogs from a given locality or station may have the central nervous system developed in a proportion different from that found in frogs from another locality, yet frogs from the same locality tend to remain constant in this character.
CHANGES IN WEIGHT OP CENTRAL NERVOUS SYSTEM 681
This last statement is supported by my observations on the two European species, R. esculenta and R. temporaria, as given in table 9.
As is seen, the value of C in both the European forms is less than in the American R. pipiens (Donaldson, '08 and '10).
The point to be specially illustrated in table 9 is, however, that frogs from the same locality maintain the same values of C. Thus both series of R. esculenta, taken from the same station at Ziirich and examined in July, but with an interval of five years, show values of C nearly alike.
NUMBER OF
SPECIMENS
EXAMINED
VALUE OF C
SPECIES
Probable
Mean ; error of the
mean
R. esculenta
R esculenta
11
11
12 16
July 20, '04
July 6, '09
July 1, '04
August 20, '09
23.8
23.7
22.8 21.8
±0.46
±0.36
R. temporaria
R. temporaria
±0.47
±0.26
On the other hand, the two series of R. temporaria from the same station at Liverpool, give on July 1, 1904, 22.8 and on August 20, 1909, 21.9, a fall of about 4 per cent in the value of C, which, as explained above, is the sort of change to be expected, although the amount of it is larger than we should have predicted.
II. THE BIOLOGICAL INTERPRETATION OF THE SEASONAL
CHANGE IN THE RELATIVE WEIGHT OF THE
CENTRAL NERVOUS SYSTEM
In order to form a proper picture of the manner in which this change in the relative weight of the central nervous system as just described is brought about, it will be necessary to obtain a notion of the growth changes in the entire frog during the active period of each year. At present only two sets of observed facts are available: (1) The change in the percentage of water. (2) Changes in length : — but from these latter, changes in body weight can be fairly inferred.
1. The changes in the percentage of water
As to the percentage of water in the entire frog, I made the determination in the Chicago frogs by drying the animals for several weeks in the same oven which was used for the determination of the percentage of water in the brain and spinal cord.
From each series in the Chicago lot, with the exception of 1, 2, 3 and 38, one male and one female frog were taken; both were weighed fresh and then opened and the ova removed from the female when necessary. Care was taken of course that no other tissue was lost by the operation.
The results are given in table 10, first for single series, then for the averages of the several series grouped in the way in which the series appear in table 3. The entries for the series, 13, 14, 15 and 16, which series were not used for the determination of C, are also given, and it is interesting to note that there was nothing peculiar in the amount of water in the entire frog in these cases.
TABLE 10
On the amount of water in the entire frog. Chicago frogs, by single series and by
groups
Pebcentagb
OF Water
SERIES DATE
MALES
BODY
FEMALES
SERIES
GROUP
WEIGHT
SERIES
GROUP
WEIGHT
1
2
3
4
5
6
7
8
9
10
11
12
1901
March 29
April 4 April 10 April 17
April 24 May 1 May 8 May 15
May 22 May 29 June 5 June 12
78.7
80.1 79.7 80.1 82.1
81.1 81.4 80 3
78.7
78.7
1
80.5
] ■ 80.4
gm».
21.6
22,9
22.7
81.4
81.0
80.8 80.8 79.3
81.4 82.2 81.8 79.4
81.4
80.7
. 81.2
gms.
16.6 21.3
23.4
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM 683
TABLE 10— Continued
1901
June 19
June 26
July 4
July 11
July 18
July 24
July 31
August 7
August 14 August 21 August 28 September 4
September 12 September 18 September 25 October 3
October 10
October 17
October 24
October 31
November 14 November 27 December 13
1902 January 10 January 31 March 22
Percentage of Water
' BODV "^"^^^
BODY
8EKIES
WEIGHT
GROUP 1 SERIES GROUP
|WEIGHT
i
79.8 81.3 81.0
1
82.3
82.3
I
/
78.7
77.4
78.2
79.0
82.3
22.9
22.9
24.4
24.4
24.2
23.3
16.
.3 23.8
78.4
76.9
77.3
21.7
22.4
24.0
79.9 22.4
7 24.3
83.2 16.3
The entry for January 10 and 31, 1902, of M. 82.3 and F. 83.2 per cent seems very high — but the water in the nervous system of these groups was not found to be excessive. These high values are probably due to the fact, as suggested by the body weights, that these frogs were a year younger than those used for the rest of the series. I will return to this matter later. On following
684 HENRY H. DONALDSON
down the water determinations by groups, it is quite evident that the entire frog begins with a high percentage of water in April and May, which diminishes towards the midsummer (August) and then rises again during the autumn. Moreover as the records stand, they suggest that the percentage of water in the female, as compared with the male, is higher in the spring, lower from July to October and higher again during hibernation.
It may be added as bearing on this difference according to sex that I made a few observations on the percentage of water in the ripe ova of these frogs. These were taken early in the spring. The determinations are given in table 11.
TABLE 11
Percentage of water.
Females
IN ENTIRE FROG
WITH OVA
IN SAME FROG
WITHOUT OVA
1
IN OVA
Series 4
Series 5
Series 7
76.8
76.8
1 76.5
81.4
81.0
80.8
o6.4
63.2
52.3
The average of the values in table 11 is 57.3 per cent which is in general agreement with the old observations of Beaudimont and St. Ange ('47) giving in the eggs of Rana (esculenta?) the percentage of 55.7.
The data serve to show the relatively small amount of water in the ova and the effect of the presence of the ova in reducing the percentage of water in the entire frog. It is just possible that in late summer, at least, small quantities of young ova, considered at the time too insignificant to be removed, may have contributed to the lower percentage of water in the female at this season.
For the general course of this percentage during the season, as shown in table 10, it is difficult to give a complete explanation. Long ago, in his admirable study on the distribution of water, v. Bezold ( '57) showed that larger frogs (Rana temporaria) had a less percentage of water than smaller ones. His series ranged in body weight was from 3.0-61.0 gms. and the corresponding percentages of water were 79.77 and 74.31, with six intermediate determinations.
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM 685
In this series it is to be noted that the heaviest frogs are probably three years older than the lightest. At 22.7 gms., v. Bezold finds the percentage of water in the early summer to be 78.2, which is in good agreement with my July record of 78.7 (average of both sexes) for frogs the average body weight of which was 22.4 gms.
Since in table 10 the body weight values — save for April (females) and January (both sexes) — are nearly alike, the variations in the percentage of water cannot well be explained as due to size. On the other hand if the frogs were from eggs of the same year, except in the instances above indicated, the specimens examined later in the season must be older than those taken earlier. Advancing age would demand a fall in the percentage of water and this fall as we have seen, occurs. But the fall in turn gives way to a rise in the percentage of water after the beginning of September. To explain this latter result, I can merely suggest that it occurs as active feeding comes to an end and when the frog is less well supplied with food than in the early part of the summer, and thus it may represent a condition of underfeeding or starvation which has been shown by Moraczewski ('00) to cause the percentage of water in the entire frog to rise. It is evident therefore that age and food conditions at least have an influence on the percentage of water in the entire frog, but the results, like those given in table 10, cannot be fully explained until it is determined first whether there are additional important conditions, and second, how these conditions which we do recognize interact.
Before leaving this topic it is important to state that in both the brain and the spinal cord no systematic variation in the percentage of water can be observed during the active season. This statement is true for all three lots of frogs. Moreover, as shown in table 12, the averages for the percentages of water in the brain and cord (using the data in tables 3, 4 and 6) are nearly alike for all three lots of frogs.
2 In the paper cited above, Moraczewski's general conclusion 2, page 144, is contradictory to his tables. The above statement in the text is based on the tables and on the text on page 136 of the paper cited.
JOURNAL OP MORPHOLOGY, VOL. 22, NO. 3
686
HENRY H. DONALDSON TABLE 12
Average values of the percentages of water in the brain and spinal cord of the several lots of frogs, from data in tables 1, 3 and 6
Frogs from Chicago
Frogs from Minnesota
Frogs from the Brandywino
2. Changes in length
On the growth of the frog in length during the active season only two sets of data have been found. In the first instance Miss Dickerson ('07) gives pictures of Rana aurora at one, two and three years. On measuring these, I obtained the body lengths given in table 13.
Using the determinations made on R. pipiens — which Rana aurora resembles — and according to which the body length is 37.5 per cent of the total length — we obtain the calculated total lengths given in table 13.
From- a series of determinations of the relation of body weight to total length in R. pipiens, Dr. Hatai ('11) has developed the accompanying formula :
y = 158 Log(a; + 6.5) -63
in which y is the total length in millimeters and x the body weight in grams. This expresses the normal relation of total length to body weight in R. pipiens.
TABLE 13 Rana aurora
AGE
BODY LENGTH
CALCULATED
Total len?th
Body weights
yrs.
1.,
mm.
36 50 63
96
133
168
gms.
3.6 10.9 22 5
2
3.
' This formula was based on the measurements of several series of Chicago frogs,
R. pipiens. It fits the observations on the Minnesota frogs also. The observa
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM
687
When this formula is appHed to the foregoing data, we obtain for the given total lengths the body weights which are entered in table 13. The results show that for the years taken, the body weight approximately doubles during each active season. This completes the first instance.
The second instance is from Fischer-Sigwart ('97) who reports for R. temporaria the following body lengths at different ages: see 'body length' in table 14.
TABLE 14 R. temporaria
! BODY LENGTH
1
CALCULATED
Total length
Body weight
1
gms.
End of first year
! 20-25
68
0.8
End of second year
' 30-35
95
3.5
End of third year
(42-47)*
(128)*
S.O*
End of fourth year
1 55-60
1
163
22.0
♦Interpolated by H. H. D.
According to Boycott ('04) the body length in R. temporaria is 36.6 per cent of the total length.
If now we take for the determination of the total lengths the highest values for the body lengths as given in the foregoing table, we obtain the series of figures marked 'total length' in table 14. Using the data on body weight given for R. temporaria by Boycott ('04) in his table (p. 375) we obtain the approximations for the body weights which are given in the last column of table 14.
Here again the body weights are more than doubled from season to season during the last three years. The value of the foregoing calculations lies not in the exact numbers obtained, for these are in a measure open to correction, but in the indication which these numbers give of the rate of growth from year to year.
tions on the Brandywine frogs however show for a given body weight, total lengths about 4 per cent less than those determined by the formula. These last frogs are therefore heavier for a given total length or shorter for a given body weight than those on which the formula is based. The formula does not apply to frogs less than 3.5 gms. in body weight.
b88 HENRY H. DONALDSON
They show that the rate is such as to cause the body weight to double, or more than double, from season to season during the three successive annual intervals; a very peculiar interesting result when compared with the growth of mammals.
Having thus determined the growth in body weight from season to season, it is desirable to calculate the weights of tne central nervous system which correspond to the body weights found. We shall take but one instance — namely the last pair from the table 14 — as these represent records for which we have some control observations.
The weight of the central nervous system is determined by the formula based on body weight and body length, using in the first instance 20.2 for the value of C. This value of C was obtained in the following manner. On referring to table 9, page 681, it is seen that the value of C for R. temporaria on July 1, 1904, was 22.8. According to our present view of seasonal change, we should expect this to be the maximum annual value of C. If this be correct, then this value is 13 per cent greater than it would be in the spring or autumn; therefore at the two ends of the season we should expect the value of C to be 13 per cent less or 20.2. The weight of the central nervous system is therefore calculated accordingly, i.e., with C = 20.2 for the two ends of the season.
In addition to the two sets of values giving the body weight, total length and weight of central nervous system, first at the time of emergence of a given individual, and second at the time of its hibernation, there have been interpolated in table 15 the values for this same frog when half grown in body weight — that is weighing 15.0 gms., and in the first instance the weight of the central nervous system in this half grown frog is calculated using 20.2 as the value of C.
This table 15 gives us a notion of what would take place, if the frog increased by about two and a half times its initial weight in the course of the season, and at the same time underwent the normal correlated increase in body length and in the weight of its central nervous system — the relative weight of this latter remaining the same, i.e., the constant C remaining unaltered during the process and having the value of 20.2. From the Zurich series we have reason to think, (see table 9,) that the mid season value of
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM
689
C is not 20.2 but 22.8, as observed. When this latter value of C is taken, then the weight of the central nervous system at the mid season becomes 0.0941 gms., which is nearly the weight found at the end of the season. This value is entered in table 15 under C = 22.8.
The first comment on these results is that they are in good agreement with the direct observations on the weight of the central nervous system in R. temporaria (Donaldson, '08, table 9), and may therefore be used as a basis for further argument.
The foregoing computations have been repeated in the case of the data for R. aurora, (table 13), but as the results depend entirely on the fact that the frog more than doubles its body weight
TABLE 15
R. temporaria
TOTAL LENGTH
BODY WEIGHT
CALCULATED WEIGHT OF CENTRAL NERVOUS
SYSTEM
C=20.2
C=22.8
128
150 163
gms
8.0 15.0 22.0
gms
0.0614 0.0834 0.0968
gms
0.0941
from season to season, and as we do not have data to control them, it does not appear necessary to put down the numerical findings.
If now we attempt to picture how these growth changes which have been determined are related to one another in order to give the results found, the following appears.
As shown in table 15, the weight of the central nervous system at emergence is .0614 gms. and at hibernation .0968 gms., a gain of 57 per cent. For the mid weight value in this table, or a body weight of 15.0 gms., the weight of the central nervous system would be . 0834 if C remained constant. We know however that C rises in the first half of the season and in July is 13 per cent greater than in March. Taking C as 22.8 therefore, or 13 per cent more than its initial value, the weight of the central nervous system becomes .0941, or almost that found at the end of the
690 HENRY H. DONALDSON
season. We conclude from this that the growth of the central nervous system is precocious and takes place mainly in the first half of the active season.
The relations just described are plotted in chart 5. Here the shape of the curve for the increasing weight of the central nervous system is fixed for the last half of the season, but the form given to it for the first half is based on the assumption that growth must begin slowly and become rapid only later.
Until direct observations on the body growth can be made for the exact control of this curve, the form given here may stand as a probable representation of what occurs.
.10
- WEIGHT IN GMS.
.
CHART 5 ■
.09
r^"^
.08
/
.07
/
/ GROWTH OF
CENTRAL NERVOUS SYSTEM CALCULATED
■
.06
/
Chart 5 Curve for the growth of the central nervous system, R. temporaria. Based on data in table 15, using the value of C = 20.2 for March and October and the value of C = 22.8 for July.
It is to be remembered however that this curve as it stands is based on observations on R. temporaria, but from what we know about this species it seems most probable that it applies to R. pipiens also.
In connection with the phenomena just described, it may be well to review briefly the growth conditions for the frog at various seasons. When a normal frog disappears in hibernation, it is prepared for the experience. The digestive system has suffered involution and a considerable amount of fat has been stored in the fat bodies, liver and muscles. The frog emerges from hibernation with most of this stored material intact and lives on it largely during the breeding season and the earlier spring weeks.
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM 691
As this gradually becomes exhausted, feeding is resumed and with the advance of the season and the increase in food, the frog not only grows, but restores these reserves and prepares for the next period of hibernation.
This review of the feeding habits of the frog serves to emphasize the fact that the conditions for nutrition in the early part of the season are different from those to be found later, and in so far might be responsible for the peculiarities in the relative growth of the central nervous system which we have observed.
In addition there are several considerations which have a very immediate bearing on the foregoing results, and especially on their variability. It must always be remembered that we are working with an animal in which the regulation of both general and relative growth is poor: an animal very resi)onsive to the influence of external conditions — one that can'be chilled or warmed, dried or made moist, fed abundantly oi- left without food for long periods.
Thus, in a poor season, i.e., poor in insects, or in the water conditions, the frog may not exhibit its usual increase in size, may not store its full food reserve for the first half of the year to follow and so not only grow poorly in general, but also not be able to exhibit the usual relative growth of the central nervous system during the season which follows.
It is hardly necessary to elaborate these relations, enough having been said to indicate why frogs taken at the same date and in the same locality, may exhibit wide differences in the relative weight of the central nervous system. What we find in the case of any frog probably depends in large measure on the external conditions to which that individual has be(>n subjected, not only during the season in which it was caught, but also during the season which preceded. If one turn back therefore to table 1, it appears that in the Chicago series the minimal values of C vary irregularly from month to month; suggesting that some of the individuals grew very little — ^as this would be the readiest explanation of the absence of systematic changes — while the maximal values show more consistent changes, tending to follow the mean.
On the other hand, in the case of both the Minnesota and Brandywine frogs, both the minimal and maximal values tend to
692 HENRY H. DONALDSON
follow the mean values more regularly; Mimiesota. table 5, Brandywine, table 7.
HUMMARY
From the foregoing discussion, the following conclusions are drawn :
1. The relative weight of the central nervous system of the frog, Rana pipiens, changes during the active season, and such a change is probably characteristic for other species of frogs with like habits.
2. The relative weight of the central nervous system is low at the time of emergence, high in the midsummer (July) and low again at the time of hibernation. During hibernation it remains nearly constant. In the formula used to express the weight of the central nervous system, the absolute value of C is characteristic for the station from which the frogs come.
3. The range from minimum to maximum in the value of C is about 13 per cent, rising 7 per cent from the end of March to the end of April, 4 per cent more from the end of April to the end of May, and 2 per cent more from the end of May to the first of July, remaining stationary in July and then in reverse order falling month by month at a similar rate to the end of October.
4. This variation in the relative weight according to season is due to lack of coincidence between the growth of the central nervous system and the growth of the entire body.
5. In frogs from one to four years old, the body weight more than doubles during each active season. The precise form of the curve representing this body growth is not known.
6. The growth of the central nervous system is precocious in relation to that of the body, but in the absence of direct observations on the growth of the body, the form of the curve can only be indirectly determined as shown in chart 5.
7. During the active season, the percentage of water in the entire frog falls slightly from spring to summer and rises again from summer to autumn. These changes seem to be due to the combined effects of advancing age and varying food supply.
The conclusions just given apply primarily to the interpretation of the preceding observations, but secondarily they also
CHANGES IN WEIGHT OF CENTRAL NERVOUS SYSTEM 693
bear on the phenomena of growth as shown by vertebrates in general. The curve for the growth of the central nervous system of the frog as given in chart 5 has the general character of the corresponding curve for a mammal; but as is evident, this curve in the case of the frog must repeat itself from year to year, so that if we should plot the entire curve for the span of life, rather than for a single season, it would be represented by a sinuous ascending line in which the sinuosities would probal)ly diminish towards the upper end.
If we turn now to general body growth, which is closely correlated with that of the central nervous system, it appears that the poikilothermous vertebrates as a group must show a seasonal variation in growth in all latitudes where there is any marked seasonal change, and that the phenomena of hibernation, with the concomitant effects with which we have to deal, represent merely a special case of this seasonal variation.
If now we pass up the vertebrate scale we find in the temperate zones both hibernating mammals, as well as those in which the seasons seems to produce marked nutritional modifications, and finall,y we have Malling-Hansen's observations ('86) on Danish children from 9 to 15 years of age which show that the growth in stature is mainly in .the third of the year between the middle of April and the middle of August, while the third comprised between the middle of August and the middle of December is the one in which they gain nine-elevenths of their annual increase in body weight. This leaves the remaining third from mid-December to mid-April, i.e., late winter and early spring, as the one in which very little growth of any sort occurs. This seems to link the rhythmic growth in the frog with that in man. To be sure there are at present but very few data available, but such as we have suggest that within the annual cycle we should expect even in the higher vertebrates a distinct rhythm corresponding to the responses of the poikilothermous vertebrates, and still exhibited even by the group in which the regulation of temperature has been more or less completely attained.
Just one point more. The rate of growth in the frog, more than doubling its body weight for three successive years (as far as we have observations) shows that the rate in the frog does not
694 HENRY H. DONALDSON
fall off with anything like the rapidity that it does in man and some other mammals. This difference suggests a number of questions to answer which it will be necessary to take up the study of growth in hibernating mammals.
LITERATURE CITED
Beaudimont et St. Ange 1847 Sur les phenomenes chimiques de revolution
embryonnaire des oiseaux et des batraciens. Annal. de Chim. et de
Physique, 3rd series, vol. 21. V Bezold, A. 1857 Untersuchungen fiber die Vertheilung von VVasser, organ ischer Materie und unorganishen Verbindungen in Thierreiche. Ztschr.
f. wiss. ZooL, vol. 8, pp. 487-524. Boycott, A. E. 1904 On the number of nodes of Ranvier in different stages of
the growth of nerve fibers in the frog. J. of Physiol., vol. 30, pp. 370 380. DicKEKSON, Mary C. 1907 The frog book. Doubleday, Page & Co., N. Y. Donaldson, H. H. 1898 Observations on the weight and length of the central
nervous system and of the legs in bullfrogs of different sizes. Jour.
Comp. Neur. Psych., vol. 8, no. 4, pp. 314-335.
1902 On a formula for determining the weight of the central nervous
system of the frog from the weight and length of the entire body.
Decennial publications, University of Chicago, vol. 10.
1908 The nervous system of thb American leopard frog, Rana pipiens,
compared with that of the European frogs Rana esculenta and Rana tem poraria (fusca). Jour. Comp. Neur. Psych., vol. 18, no. 2, pp. 121 149.
1910 Further observations on the nervous system of the American leopard frog (Rana pipiens) compared with that of the European frogs
(Rana esculenta and Rana temporaria). Jour. Comp. Neur. Psych.,
vol. 20, no. 1, p. 1-18. Donaldson, H. H. and Schoemaker, D. M. 1900 Observations on the weight
and length of the central nervous system, and of the legs in frogs of
different sizes (Rana virescens bradi^^^ephala Cope). Jour. Comp.
Neur. Psych., vol. 10, no. 1, pp. 109-132. ' Fischer-Sigwart, H. 1897 Biologische Beobachtungen an unseren Amphibien.
Vierteljahrsch. d. Naturf. Gesell., Ziirich, vol. 42, Jahrg. 1897. Hatai, S. 1911 A formula for determining the total length of the leopard frog
(R. pipiens) for a given body weight. Anat. Rec, vol. 5, no. 6. Malling-Hansen, R. 1886 Perioden im Gewicht der Kinder und in der Sonnon warme. Kopenhagen. MoRACZEWSKi, W. VON 1900 Die Zuzammensetzung des Leibes von hungernden
und blutarmen Froschen. Arch. Anat. u. Phys., Suppl. Bd. zur phys.
Abthl. ToRELLE, Ellen 1903 The response of the frog to light. Am. J. of Physiol..
vol. 9, pp. 466-488.
THE PHYSIOLOGY OF CELL-DIVISION
IV. THE ACTION OF SALT SOLUTIONS FOLLOWED BY HYPERTONIC SEA-WATER ON UNFERTILIZED SEA-URCHIN EGGS AND THE ROLE OF MEMBRANES IN MITOSIS
RALPH S. LILLIE
From the Marine Biological Laboratory, Woods Hole, and the Physiological Laboratory, Department of Zoology, University of Pennsylvania
THREE FIGURES
INTRODUCTION
During the summer of 1909 at Woods Hole I found that membrane-formation and cleavage, leading in a small proportion of cases to the production of blastulae, could be induced in unfertilized eggs of Asterias and Arbacia by temporary exposure to isotonic solutions of various neutral salts. ^ Salts of sodium and potassium were chiefly employed, including chloride, bromide, nitrate, iodide and sulphocyanate ; last summer chlorate was also used. In the case of Asterias all of these were found to induce membrane-formation and cleavage in a large proportion of eggs. With Arbacia, however, only iodide and sulphocyanate showed a corresponding degree of effectiveness ; nitrate had wellmarked though less decided action, chlorate produced little and bromide still less effect, while chloride was almost entirely inactive; sodium acetate was found to act like chloride. The order of relative effectiveness of the salts, ranged according to the anions, is as follows: COO CH3 and CI <Br <C103 <N03^ Probably many of the bodies described in spermatocytes as yolk globules will prove to be mitochondria, as well as many bodies previously confused with idiozomes. However, Gross has distinguished in Pyrrhocoris yolk granules from mitochondria.
^'^ The large compact masses in the distal ends of spermatogonia, shown by Giglio-Tos in his fig. 1 and regarded by him as mitochondria, are clearly not such but mitosomes; he was misled by their taking the Benda stain, which shows how little service this stain is as a diagnostic.
SPERMATOGENESIS OF EUSCHISTUS ' 787
of spheres. But it must be said that none of these writers except Vejdovsky have paid particular attention to their origin. Pantel and de Sinety leave the question open, though they show that the 'pseudochromosomes' arise in close contact with the nucleus. Another group of investigators (Dumez, Janssens, Popoff, Wassilieff, Goldschmidt, Buchner, Jorgensen) hold them to be of nuclear origin, for the following reasons : In the pachytene stage of the spermatocytes the chromosomes are frequently, though not in all objects, definitely oriented, radiating in a 'bouquet stage' towards the distal pole of the nucleus, and the idiozome lies at that pole in the cell body; near that pole of the nucleus the mitochondria make their first appearance. This position of the mitochondria, close to a particular pole of the nucleus, is taken to mean that they are produced there by some nuclear activity. Most of these writers maintain that they are engendered by actual emigration of chromatin particles out of the nucleus at that pole, as indicated especiallj^ by Janssens, Wassilieff, Jorgensen and Buchner. This idea of chromatin emigration has been particularly instigated by Goldschmidt's view that the mitochondria belong in the class of chromidial formations. In Euschistus we found in the resting spermatogonia a particular chromatin plate upon the nuclear membrane, and this is invariably at the point where the idiozome lies; but in these cells there are no demonstrated mitochondria present around the idiozome. In the spermatocytes of this species there is another chromatin plate, here produced by the ends of two or three chromosomes, again always at the pole where the idiozome lies; in the close vicinity of this idiozome the first mitochondrial chains make their appearance. Pantel and de Sinety found in Notonecta that the mitochondria arise between the idiozome and the nucleus, therefore in relation to both. In Euschistus it is hard to make sure whether the mitochondria arise from the cytoplasm, from the idiozome or from the nucleus. But the fact is that they originate in the distal pole of the cell, close to the idiozome and the nucleus, therefore it is probable they are produced by either idiozome or nucleus or by a joint action of these. We saw previously that the idiozome probably
788 THOS. H. MONTGOMERY, JR.
stands under the influence of the chromatin plate. In Euschistus there is no evidence that the mitochondria are produced by emigrated chromatin particles, and indeed the idiozome intervenes between them and the chromatin plate. Also in this species they certainly have no relation to the idiochromosomes, which may lie upon any point of the nuclear membrane except the idiozome pole and they do not discharge any visible substance into the cytoplasm. ^i Were the mitochondria of strictly cytoplasmic origin it would be difficult to explain why they always arise at a particular pole of the nucleus near the chromatin plate and the idiozome. Therefore it seems a better working hypothesis to conclude that they are produced either by some chemical interaction of idiozome and cytoplasm, or of nucleus and cytoplasm, which would be, in either case, an ultimate nuclear origin. It is interesting to note that their period of early development in the spermatocytes corresponds with the period of conjugation of the chromosomes, and the latter process may be the initial step in their production. This is all in agreement with the concept of the nucleus as the particular fcrmative center of the cell. After the idiozome of Euschifetus has disintegrated, and the chromatin plate become disestablished, the mitochondria becomes scattered throughout the cell, and then they become larger and more prominent, evidently by autonomous growth. Thus it may well be that the nucleus directly, or acting through the idiozome, gives off ferments in small amounts to the cytoplasm and these ferments in their turn engender the mitochondria that later become selfperpetuating structures.
Another point of interest is how the mitochondria become distributed in the maturation mitoses. In neither Euschistus nor other forms is there evidence of autonomous division of them in mitosis ; they appear rather to become divided passively by the equatorial constriction of the cell; there is no other mechanism for their division, for they lie outside of the spindle and are
- i Both Buchner and Wassilieff hold the mitochondria to be produced by a pouring out of substance from the modified chromosomes, though Wassilieff's results
on Blatta have been contradicted by Morse.
SPERMATOGENESIS OF EUSCHISTUS " 789
little influenced by the centrioles. In Euschistus they lie quite irregularly in the spermatocytes, and generally each thread fails to divide in one of the maturation mitoses; further, it is a matter of chance at what point a mitochondrial thread becomes divided. They become irregularly divided in the two maturation divisions so that varying amounts of them become apportioned to the spermatids.22 That there is no accurate quartering of the mitochondria in a number of other species results from a study of the figures of various writers, such as the Schreiners ('05, Myxine), Meves ('00, Paludina, Pygaera), Gross ('06, Pyrrhocoris), Wassilieff ('07, Blatta). But in the bee and hornet (Meves, '07, '08), in Pamphagus (Giglio-Tos, '08), in Blaps (Benda, '03) and Stenobothrus (Gerard, '09) they appear more evenly arranged around the spindles, and in these objects probably become more regularly divided. But there is good evidence that in certain cases it is a matter of chance how they become divided, in which regard they differ markedly from the chromosomes.
The mitochondria in the spermatid of Euschistus coalesce to produce the true Nebenkern, which elongates and forms a pair of narrow bands lying on the sides of the axial thread and extending from the head probably to the end of the tail of the spermatozoon. A similar metamorphosis of the Nebenkern is known for flagellate spermatozoa in a number of invertebrates, while in the mammals the mitochondria engender a spiral filament around the middle piece. There is now, however, considerable evidence that in many forms of flagellate spermatozoa, such as those of mammals and molluscs, the whole tail enters the egg in fertilization, and is not left outside the egg (contrary to earlier observations); therefore it is probable all the mitochondrial substance of the sperm enters the egg; much more substance, accordingly, than merely the chromatin of the head.
Since then fertilization brings together the mitochondria of two parents, it now becomes of great importance to trace
22 I have previously intimated, ('10b) the possibility that the relative amount of the mitochondrial substance received might determine the sex-preponderance character of a sperm, a matter unfortunately very difficult to test.
790 THOS. H. MONTGOMERY, JR.
the history of the mitochondria during the fertiUzation of the egg, and especially the behavior of those furnished by the sperm.
PREFORMATION AND EPIGENESIS IN THE GERMINAL CYCLE, AND SEGREGATION OF THE GERM CELLS IN ONTOGENY
The discussion of embryologists upon preformation and epigenesis has treated mainly the phenomena of somatic differentiation, the factors regulating the growth of the embryo from the egg. If it be not too premature to speak of a consensus of opinion reached in this discussion, it would be to the effect that epigenesis and preformation are not mutually exclusive, but that both probably proceed at the same time.
Here it is my wish to call attention to the fact that in the history of the germ cells may be recognized both preformation and epigenesis, whereby the germinal cycles offer a certain parallel to the somatic.
In the spermatogenetic cycle a number of spermatogonia! generations occur, during which the number of the chromosomes remains constant and no marked cytoplasmic specializations take place. But in the spermatocytes remarkable differences arise suddenly: the chromosomes group themselves into gemini, the mitochondria increase rapidly in amount, the whole cell becomes larger, and frequently the centrioles take on unusual forms and positions (as upon the cell membrane). Somewhat similar changes occur in the oocytes, and more complex ones, owing to the production of yolk substance. The spermatids often undergo a marked metamorphosis. As one reviews the sum total of these processes the conviction arises that there are here in the completest form both preformation and epigenesis. The chromosomes are on the whole the most stable parts, apparently continuous from generation to generation, and though they may pass through marked changes in forming the sperm nucleus, they later emerge, in fertilization, under the same forms that they had previously. On the whole they seem to be the particular preformed bodies of the germ cells. But this is not the case with certain other cell constituents. For in Euschistus, which seems to exemplify in main features the sperma
SPERMATOGENESIS OF EUSCHISTUS 791
togenetic relations of most insects, a true idiozome arises in the spermatocytes, mitochondria develop around it, the idiozome disintegrates and a sphere appears, this sphere disappears entirely and in the spermatid another sphere is produced that originates another new body, the perforatorium.
It is clear that the first spermatocytes and oocytes are the most interesting cells of their respective cycles, for they exhibit the most significant processes — conjugation of chromosomes, reduction division, elaboration of mitochondria. In them the history of the cytoplasmic parts is markedly epigenetic. And another series of epigenetic changes is exhibited by the histogenesis of the spermatozoon.
There is then a parallel with the somatic cycle, which begins with an undifferentiated and terminates with a highly differentiated condition. The ripe ovum and spermatozoon are much more differentiated than the spermatogonium or oogonium; each of them enters, with the commencement of the growth period, upon its period of specialization.
The recognition of this resemblance may throw light upon the problem of the segregation of the germ cells. By what process is it that certain cells of the embryo are held back from somatic differentiation to become germ cells? In other words, why do not all the cells become specialized? The answer, it seems to me, is to be sought in the distribution of the specializations of the fertilized egg to the cells of the embryo. One set of specialized structures of the germ cells, the mitochondria, are now known to give rise to various important specializations of body cells. The mitochondria, that are elaborated, in greatest part at least, during the growth period of the germ cells, persist from the fertilized egg into cleavage stages, and ultimately transform into various specialized fibrillar structures. With this in mind the setting aside of germ cells from body cells could be explained mechanically as follows: any cleavage cell which failed to receive mitochondria, or failed to receive particular ones or a particular amount of them, would be incapacitated from engendering such somatic specializations, it would thereby become a germ cell. This might appear contrary to the idea
792 THOS. H. MONTGOMERY, JR.
that the body cells become different from the germ cells by a mechanical process of chromatin diminution, as in Ascaris. But this is quite conformable to our argument, for in Ascaris those cells which become body cells are the ones that include the cast-off chromosome ends in their cytoplasm, and it will probably be found that these ejected chromosome parts engender such cytoplasmic differentiations as characterize the body cells. Either this mechanism of the segregation of the germ cells may be admitted, or else one that would cause the mitochondria of prospective germ cells to remain unaltered or latent. Fortunately this is a matter that may be tested by observation. For if the first supposition be correct we should find that during cleavage an unequal distribution of the mitochondria occurs, just as we know occurs in the case of yolk granules. If the second be correct, then while the mitochondria are undergoing developmental changes in some of the cells, they should be found to remain relatively unaltered in others — the prospective germ cells.
LITERATURE CITED
Baumgartner, W. J. 1902 Spermatid transformations in Gryllus assimilis, etc. Kansas Univ. Sci. Bull. vol. 1.
Benda, C. 1891 Neue Mitteilungen uber die Entvvickelung der Genitaldriisen und iiber die Metamorphose der Samenzellen. Arch. Anat. Phys. 1898 Ueber die Spermatogenese der Vertebraten und hoheren Evertebraten. Verh. Physiol. Ges. Berlin. 1903 Die Mitochondria. Ergebn. Anat. Entw., Bd. 12.
Berghs, J. 1904 La formation des chromosomes h^terotypiques dans la sporo[^ genese vegetale. 1. La Cellule, tome 21.
Blackman, M. W. 1903 The spermatogenesis of the myriapods. II. Biol. Bull., vol. 5.
1905a The spermatogenesis of the myriapods. III. Bull. Mus. Comp. ZooL, Harvard, vol. 48.
1905b The spermatogenesis of the myriapods. IV. Proc. Amer. Acad. Arts and Sci., vol. 41.
1907 The spermatogenesis of the myriapods. V. Ibid. vol. 42. BoNNEViE, K. 1908 Untersuchungen iiber Keimzellen. I. Jena. Zeit. Bd. 41.
1908 Chromosomenstudien. 11. Arch. Zellforsch. Bd. 2.
SPERMATOGENESIS OF EUSCHISTUS 793
Boring, A. M. 1907 A study of the spermatogenesis of twenty-two species of
the Membracidae, Jassidae, Cercopidae and Fulgoridae. Journ.
Exp. Zool., vol. 4. BovERi, T. 1887 Zellen-Studien. 1. Jena. BouiN, P. 1905 Ergastoplasme, pseudochromosomes et mitochondria. Arch.
Zool. Exper. (4) tome 3. Brauer, a. 1893 Zur Kenntniss der Spermatogenese von Ascaris megalo cephala. Arch. mikr. Anat., Bd. 42. BucHNER, P. 1909 Das accessorische Chromosom in Spermatogenese und Ovo genese der Orthopteren, etc. Arch. Zellforsch., Bd. 3.
1910 Von den Beziehungen zwischen Centriol und Bukettstadium*
Ibid., Bd. 5. Chubb, G. C. 1906 The growth of the oocyte in Antedon, etc. Phil. Trans.
Roy. Soc. London, vol. 198. CoNKLiN, E. G. 1902 Karyokinesis and cytokinesis, etc. Journ. Acad. Nat. Sci.
Philadelphia, vol. 12. Cook, M. H. 1910 Spermatogenesis in Lepidoptera. Proc. Acad. Nat. Sci.
Philadelphia. Davis, H. S. 1908 Spermatogenesis in Acrididae and Locustidae. Bull. Mus.
Comp. Zool., Harvard, vol. 53. DiNGLEB, M. 1910 Ueber die Spermatogenese des Dicrocoelium lanceatum
Stil. und Hass. Arch. Zellforsch., Bd. 4.
DuESBERG, J. 1908 La spermatogenese chez le rat. Leipzig.
1910a Les chondriosomes des cellules embryonnaires du lapinetleur
r61e dans la g^nese des myofibrilles. Arch. Zellforsch., Bd. 4.
1910b Observations sur la structure du protoplasme des cellules v6ge tales. Anat. Anz., Bd. 36.
1910c Sur la continuite des elements mitochondriaux des cellules sex uelles et des chondriosomes des cellules embryonnaires. Ibid.
DuMEZ, R. 1902 Rapports du cytoplasme et du noyau dans I'oeuf de la Cytherea chione L. La Cellule, tome 19.
EiSEN, G. 1900 The spermatogenesis of Batrachoseps. Jour. Morph., vol. 17.
Farmer, J. B. and Moore, J. E. S. 1905 On the maiotic phase (reduction divisions) in animals and plants. Quart. Journ. Micr. Sci., vol, 48.
FicK, R. 1907 Vererbungsfragen, Reductions- und Chromosomenhypothesen, Bastardregeln, Ergebn. Anat. Entw., Bd. 16. 1908 Zur Konjugation der Chromosomen. Arch. Zellforsch. Bd. 1.
Foot, K. and Strobell, E. C. 1909 The nucleoli in the spermatocytes and germinal vesicles of Euschistus variolarius. Biol. Bull., vol. 16.
794 THOS. H. MONTGOMERY, JR.
GERARD, P. 1909 Recherches sur la spermatogenese chez Stenobothrus biguttulus (Linn.). Arch. Biol., tome 24.
GiGLio-Tos, E. 1908 I mitocondri nelle cellule seminali maschili di Pamphagus marmoratus (Burm.). Biologica, tome 2.
GoLDSCHMiDT, R. 1904 Der Chromidialapparat lebhaft funktionierender Gewebszellen. Etc. Zool. Jahrb., Bd. 21.
GoLDSCHMiDT, R. und PoPOFF, M. 1907 Die Karyokinese der Protozoen und der Chromidialapparat der Protozoen- und Metazoenzelle. Arch. Protistenk. Bd. 8.
Gregoire, V. 1904 La reduction numerique des chromosomes et les cineses de maturation. La Cellule, tome 21.
1905 Les resultats acquis sur les cineses de maturation dans les deux regnfis. Ibid, tome 22.
1910 Les cineses de maturation dans les deuxregnes. Ibid., tome 26.
Gregoire, V. et Deton. 1906 Contribution a I'etude de la spermatogenese dans rOphryotrocha puerilis. Ibid., tome 23.
Gross, J. 1904 Die Spermatogenese von Syromastes marginatus L. Zool. Jahrb., Bd. 20.
1906 Die Spermatogenese von Pyrrhocoris apterus L. Ibid., Bd. 23.
GuYER, M. F. 1900 Spermatogenesis of normal and of hybrid pigeons. Chicago.
Hacker, V. 1895 Die Vorstadien der Eireifung. Arch. mikr. Anat., Bd. 45.
1904 Bastardierung und Geschlechtszellenbildung. Festsch. f. Weis mann.
1910 Ergebnisse und Ausblicke in der Keimzellenforschung. Zeit.
indukt. Abstamm. und Vererbungsl., Bd. 3.
Hbidenhain, M. 1900 Die Zentralkapseln und Pseudochromosomen in den Samenzellen von Proteus. Anat. Anz., Bd. 18.
Henking, H. 1891 Untersuchungen liber die ersten Entwicklungsvorgangen in den Eiern der Insekten. Zeit. wiss. Zool., Bd. 51.
Hertwig, O. 1890 Vergleich der Ei- und Samenbildung bei Nematoden. Arch, mikr. Anat., Bd. 36.
Holmgren, N. 1902 Ueber den Bau der Hoden und die Spermatogenese von Silpha carinata. Anat. Anz., Bd. 22.
HovEN, H. 1910 Sur I'histog^nese du systeme nerveux peripheriqle chez le poulet et sur le role des chondriosomes dans la neurofibrillation. Arch. Biol., tome 25.
Janssens, F. a. 1905 Evolution des auxocytes mdles du Batrachoseps attenuatus. La Cellule, tome 22.
SPERMATOGENESIS OF EUSCHISTUS 795
Jordan, H. E. 190S The spermatogenesis of Aplopus mayeri. Publ. Carnegie
Inst., Washington. JoRGENSEN, M. 1909 Beitrage zur Kenntnis der Eibildung, etc., bei Schwam men. Arch. Zellforsch., Bd. 4. KoROTNEFF, A. 1909 Mitochondrien, Chondriomiten und Faserepithel der
Tricladen. Arch. mikr. Anat., Bd. 74. KouscHELT, E. 1895 Ueber Kernteilung, etc., bei Ophryotrocha puerilis.
Zeit. wiss. ZoU., Bd. 60. KoRSCHELT, E. und Heider, K. 1903 Lehrbuch der vergleichenden Entwick lungsgeschichte der wirbellosem Tiere. Jena. Lams, H. 1908 Les divisions des spermatocytes chez la fourmi (Camponotus
herculeanus L.). Arch. Zellforsch., Bd. 1. Lehhoss]6k, M. v. 1898 Untersuchungen liber Spermatogenese. Arch. mikr.
Anat., Bd. 51. McClung, C. E. 1900 The spermatocyte divisions of the Acrididae. Bull.
Univ. Kansas. McGiLL, C. 1904 The spermatogenesis of Anax Junius. Univ. Missouri Studies.
No. 2. Maziarski, S. 1910 Sur les changements morphologiques de la structure
nucleaire dans les cellules glandulaires. Arch. Zellforsch., Bd. 4. Medes, G. 1905 The spermatogenesis of Scutigera forceps. Biol. Bull, vol. 9. Meves, F. 1896 Ueber die Entwicklung der mjinnlichen Geschlechtszellen
von Salamandra maculosa. Arch. mikr. Anat., Bd. 48.
1897 Zelltheilung. Ergebn. Anat. Entw., Bd. 6.
1899 Ueber Struktur und Histogenese der Samenfjiden des Meerschweinchens. Arch. mikr. Anat., Bd. 54.
1900 Ueber den von la Valette St. George entdeckten Nebenkern (Mitochondrienkorper) der Samenzellen. Ibid., Bd. 56.
1902a Ueber oligopyrene und apyrenen Spermien und liber ihre Ent st hung, etc. Ibid., Bd. 61.
1902b Struktur und Histogenese der Spermien. Ergebn. Anat. Entw.,
Bd. 11.
1907a Die Spermatocytenteilungen bei der Honigbiene, etc. Arch.
mikr. Anat., Bd. 70.
1907b Ueber Mitochondrien bezw. Chondrokonten in den Zellen junger
Embrj^onen. Anat. Anz., Bd. 31.
1907c Die Chondriokonten in ihrem Verhaltnis zur Filarmasse Flem mings. Ibid.
1908a Die Spermatozytenteilungen bei der Hornisse, etc. Arch. mikr.
Anat., Bd. 71
1908b Die Chondriosomen als Trager erblicher Anlagen, etc. Ibid.,
Bd. 72.
796 THOS. H. MONTGOMERY, JR.
Montgomery, T. H., Jr. 1898 The spermatogenesis in Pentatoma up to the formation of the spermatid. Zool. Jahrb., Bd. 12.
1899 Cytological studies, with especial reference to the morphology of the nucleolus. Jour. Morph., vol. 15.
1900 The spermatogenesis of Peripatus (Peripatopsis) balfouri up to the formation of the spermatid. Zool. Jahrb., Bd. 14.
1901 A study of the chromosomes of the germ cells of Metazoa. Trans. Amer. Phil. Soc, vol. 20.
1903 The heterotypic maturation mitosis in Amphibia and its general significance. Biol. Bull., vol. 4.
1904 Some observations and considerations upon the maturation phenomena of the germ cells. Ibid., Bd. 6.
1905 The spermatogenesis of Syrbula and Lycosa, etc. Proc. Acad. Nat. Sci., Philadelphia.
1906 Chromosomes in the spermatogenesis of the Hemiptera Heteroptera. Trans. Amer. Phil. Soc. N. S., vol. 21.
1910a On the dimegalous sperm and chromosomal variation of
Euschistus, etc. Arch. Zellforsch., Bd. 5.
1910b Are particular chromosomes sex determinants? Biol. Bull., vol.
19.
Morrill, C. V. 1910 The chromosomes in the oogenesis, fertilization and cleavage of Coreid Hemiptera. Biol. Bull., vol. 19.
Morse, M. 1909 The nuclear components of the sex cells of four species of cockroaches. Arch. Zellforsch., Bd. 3.
Nichols, M. L. 1910 The spermatogenesis of Euchroma gigantea. Biol. Bull., vol. 19.
Otte, H. 1907 Samenreifungund Samenbildungbei Locustaviridissima. Zool. Jahrb., Bd. 24.
Pantel, J. et SiNETY, R. DE 1906 Les cellules de la ligne6 m41e chez le Notonecta glauca L. etc. La Cellule, tome 23.
Paulmier, F. C. 1899 The spermatogenesis of Anasa tristis. Jour. Morph., vol. 15, Supplement.
Platner, G. 1899 Beitrage zur Kenntniss der Zelle und ihrer Teilungserscheinungen. Arch. mikr. Anat., Bd. 33.
PoPOPF, M. 1907 Eibildung bei Paludina vivipara und Chromidien bei Paludina und Helix. Ibid,, Bd. 70.
Rath, O. vom 1895 Neue Beitrage zur Kenntnis der Chromatinreduction der Samen- und Eireife. Ibid., Bd. 46.
RtJCKERT, J. 1893 Zur Eireifung bei Copepopen. Anat. Hefte, Bd. 4.
1894 Die Chromatinreduction bei der Reifung der Sexualzellen. Ergebn. Anat. Entw., Bd. 3.
SPERMATOGENESIS OF EUSCHLSTUS 797
ScHAPFNER, J. H. 1897 The division of the macrospore nucleus in Lilium. Bot.
Gaz., vol. 23. ScHBElNER, A. and K. E. 1904 Die Ileifungsteilungen bei den Wirbeltieren Etc. Anat. Anz., Bd. 24.
1905 Ueber die Entwicklung der mannlichen Geschlectszellen von Myxine glutinosa. Arch. Biol., Bd. 21.
1906a Die Reifung der mannlichen Geschlechtszellen von Salamandra
maculosa (Laur.), Spinax niger (Bonap.), und Myxine glutinosa (L.).
Ibid., Bd. 22.
1906b Die Reifung der Geschlechtszellen von Ophryotrocha puerilis.
Anat. Anz., Bd. 29.
1907 Die Reifung der Geschlechtszellen von Enteroxenos oestergreni.
Vid. Selsk. Skr.
1908a Die Reifung der Geschlechtszellen von Zoogonus mirus. Ibid.
1908b Zur Spermienbildung der Myxinoiden. Arch. Zellforseh., Bd. 1 .
Stevens, N. M. 1905 Studies in spermatogenesis with especial reference to the 'accessory chromosome.' Publ. Carnegie Inst. Washington.
1906 Studies in spermatogenesis. 11. Ibid.
1910 An unequal pair of heterochromosomes in Forficula. Jour. Exp. Zool., vol. 8.
Stomps, T. J. 1910 Kerndeeling en Synapsis bij Spinacia oleracea L. Amsterdam.
Stricht, O. van der 1904 La structure de I'oeuf des mammif^res. Arch. Biol., Bd. 21.
1905 Structure de I'oeuf ovarique de la femme. Bull. Acad. Med. Belg.
1909 La structure de I'oeuf des mammiferes (chauve-souris, Vesperugo noctula). 3me Partie. TVIem. Acad. Roy. Belg., Bd. 2 h. 2.
Vbjdovsky, F. 1907 Neue Untersuchungen liber die Reifung und Befruchtung. Frag.
Van Beneden, E. 1883 Recherches sur la maturation de I'ceuf, etc. Arch. Biol., Bd. 4.
Waldeyer, W. 1906 Die Geschlechtszellen. in O. Hertwig's Handbuch d. vergl. u. exper. Entwickelungslehre d. Wirbeltiere.
Wallace, L. B. 1909 The spermatogenesis of Agalena naevia. Biol. Bull., vol. 17.
Wassilieff, a. 1907 Die Spermatogenese von Blatta germanica. Arch. mikr. Anat., Bd. 70.
AViLcox, E. V. 1895 Spermatogenesis of Caloptenus femur-rubrum and Cicada tibicen. Bull. Mus. Comp. Zool., Harvard, vol. 27.
JOURNAL OF MORPHOLOQT, VOL. 22, NO. 3
798 THOS. H. MONTGOMERY, JR.
Wilson, E. B. 1905 Studies on chromosomes. I. Journ. Exp. Zool., vol. 2. 1905b Studies on chromosomes. II. Ibid. 1906 Studies on chromosomes. III. Ibid., vol. 3.
Winiwarter, H. v. 1900 Recherches sur I'ovog^n^se et organogenese de I'ovaire des mammif^res (lapin et homme). Arch. Biol., Bd. 17.
Winiwarter, H. v. et Sainmont 1909 Nouvelles recherches sur I'ovog^ndse de I'ovaire des mammif^res (chat). Ibid., Bd. 24.
Yatsu, N. 1907 A note on the adaptive significance of the sperm-head in Cerebratulus. Biol. Bull., vol. 13.
ZwEiGER, H. 1906 Die Spermatogenese von Forficula auricularia L. Jena. Zeit.. Bd. 42.
PLATES
All the figures have been drawn to the same scale with the aid of the camera lucida at the level of the base of the microscope, with the Zeiss apochromatic immersion objective 1.5 mm. , and ocular 12; the original dimensions have been reduced one-fifth.
The following abbreviations have been employed:
c. centriole m. minute chromosome
ch. c. chromatoid corpuscles mit. mitosome (spindle remains)
p. larger idiochromosome pf. perforatorium
d. smaller chromosomes pi. plasmosome id. idiozome sp. sphere
PLATE 1
EXPLANATION OF FIGURES
1 and 3 from testis no. 103, all others from testis no. 282.
1 Penultimate spermatogonium, rest; (follicle 5).
2 The same, pole view of equatorial plate; (follicle 6).
3 Portion of a rosette of ultimate spermatogonia; (follicle 6).
4 Anaphase of ultimate spermatogonium; (follicle 4).
5 Later anaphase of the same; (follicle 4).
6-41 Successive stages of the early growth period of the first spermatocytes. All the cells from follicle 4 except the following: Figs. 23, 24 from follicle 5; figs. 10, 11, 15, 22 from follicle 6.
PLATE 2
EXPLANATION OF FIGURES
Successive stages of first spermatocytes, all from testis no. 282; all from follicle 4, except fig. 49 from follicle 5.
PLATE 3
EXPLANATION OF FIGURES
Successive stages of first spermatocytes, figs. 74, 75, 82, 83 from testis, no. 286; 86-93 from testis no. 265; the remaining figures from testis no. 282. Figs. 74-83, 86-89, 91-93 from follicle 6, the remaining from follicle 4.
69-86 Later prophases of the first maturation spindle.
87-93 First maturation spindle.
PLATE 4
EXPLANATION OF FIGURES
111 from testis no. 282, the others from testis no. 265. All the figures from follicle 6 except the following: 98 and 100 from follicle 4; 95-97, 110 from follicle5.
94 Metaphase of first maturation.
95-100 Anaphases of first maturation, fig. 98 being a polar view of a daughter chromosomal plate.
101 Polar view of metaphase of second maturation.
102-104 Lateral views of second maturation spindles.
105-106 Polar views of daughter chromosomal plates of the second maturation spindle.
107-111 Lateral views of later second maturation spindles.
112 An entire spermatid (on the right) nearly completely separated from its sister (shown in part outline on the left).
113-116 Earliest stages in the histogenesis of the sperm.
PLATE 5
EXPLANATION OF FIGURES
Histogenesis of the sperm. All the figures represent lateral views, except 145, which illustrates a transection of a sperm bead of the stage of fig. 144; the entire spermatozoon is shown in 117-126, in the others only the head and the proximal part of the tail are represented.
All the figures are of cells from follicle 6; 117-122 are from testis no. 265; 135-137 from testis no. 116; the remainder from testis no. 282.
T, H. Montg-omery, del.
THE LIFE HISTORY OF THE SCOLEX POLYMORPHUS OF THE WOODS HOLE REGION
WINTERTON C. CURTIS
From the Zoological Laboratory, University of Missouri
THIRTEEN FIGURES
CONTENTS
Introduction 819
Methods 820
The nature of the Scolex polymorphus 821
The normal occurrence of parasites in sand sharks taken at random 827
The infection of sand sharks with the Scolex polymorphus 829
Summary 849
Literature cited 848
INTRODUCTION
In attempting to determine the life cycle of Crossobothrium laciniatum, a cestode found in the sand shark, (Carcharias littoralis) of the Woods Hole region, attention was at once directed to the cestode larva known as the Scolex polymorphus, which seemed not unlikely to be the young of this species. The studies here described were begun in the hope that these two forms would prove to be a single species, since both are admirable for laboratory purposes and have been so used by students at the Marine Biological Laboratory for many years. It appears, however, that such a relationship does not exist; for the results of the experiments give strong, though perhaps not entirely conclusive, evidence that the Scolex polymorphus develops into the species Phoreiobothrium triloculatum (Linton) and show conclusivelj^ that it cannot be the young of Crossobothrium laciniatum.
819
JODRNAL OF MORPHOLOGY, VOL. 22, NO. 3
820 WINTERTON C. CURTIS
The material was collected and the experiments performed during the summers of 1903 and 1904 at the Marine Biological Laboratory and the United States Fisheries Bureau Laboratory at Woods Hole, Mass., and I am indebted to those in charge of these institutions for substantial aid in the prosecution of the work.
METHODS
The sharks used for infection and the squeteague from which the specimens of the Scolex polymorphus were obtained, were all taken during the summer months in the fish traps near Woods Hole. The sharks were marked by means of numbered copper tags, fastened to the dorsal fin and were kept in wooden fish cars about 5 X 6 X 14 feet, several of which were fastened together to form a float upon which much of the work could be conveniently carried on. At the outset it was necessary to devise some apparatus by which the animals could be held securely in a position convenient for any necessary operation and which could be manipulated with safety by a single person. The holder which was finally constructed consisted of a trough about four and one-half feet in length, formed by nailing two boards together and across their ends two shorter strips after the fashion of a farmer's 'pigtrough.' The top of this trough was covered by a hinged lid which, when fastened down, left the head and tail exposed but held the body of the fish securely along the greater part of its length. The cross piece of one end was hinged to the surface of the float where the holder was being used and to the free end was hinged a support, which, when swung out, held the contrivance at any desired height as the free end was lifted toward the upright position. By working rapidly, it was possible, with the fish thus held securely, to complete the necessary treatment in a few moments, and any more elaborate apparatus providing for the irrigation of the gills seemed unnecessary, since the sharks gave no indication that their vitality had been impaired by this brief exposure to the atmosphere. In operation this holder was safe and in every way satisfactor}. The shark was dipped up with
THE SCOLEX POLYMORPHUS 821
a hand net and placed in the trough ventral side uppermost. Upon fastening its lid, the holder was at once raised to a convenient position and the operator could then quickly introduce any desired object into the mouth cavity. In feeding, the jaws were pried open, a piece of food placed in the mouth and after being guided past the gill arches was thrust down the oesophagus by means of a billet of wood. The oil of male fern and the calomel, used in the attempts to rid the sharks of their previous infection, were given in gelatin capsules having a capacity of 2 cc. These were placed in a piece of galvanized iron pipe, which had been pushed down the oesophagus until it reached the stomach, and the capsule forced down the pipe with a small wooden rammer. This proceeding, or the giving of food, could be accomplished with comparative rapidity and the sharks were often back in the car within two or three minutes after being taken out. There was no sign of the regurgitation of food or drugs, though the bottoms of the cars were carefully inspected during the hours just after treatment, and both food and drugs were found in the digestive tracts of specimens which happened to be examined during the first few days. There can, therefore, be no doubt that both remained in the stomach when once introduced. Further details of methods used are given at the appropriate places throughout the paper.
THE NATURE OF THE SCOLEX POLYMORPHUS
Before describing the experiments in which the larval cestode known as the Scolex polymorphus was used to infect the sand shark, it may be well to offer a word of explanation regarding this form. The name S. polymorphus has been applied by Linton to a cestode larva found in a considerable number of fishes from the Wood Hole region. While most abundant in the squeteague (Cynoscion regalis) and the common flounder (Paralicthys dentatus) , he ha,s found it in varying numbers in some twenty-eight other teleosts, the list of which is given on page 413 of his 1901 report. This larva, which I have represented in figs. 1, 2, 3, 4, 12 and 13 of this paper, is frequently met with in a stage slightly
822 WINTERTON C. CURTIS
younger than the one here shown. In such a young condition it lives in the intestine of its teleost host, moving freely about and attaching itself by means of its four suckers. The slightly older stage which these figures represent and which was used in my experiments, is found in the cystic duct of some of these fishes, notably the squeteague and the common flounder. The specimens which I used for infection purposes all came from the first of these two hosts.
The name Scolex polymorphus originated with Rudolphi ('08) and has since been frequently applied to such larval forms taken from many different fishes, and by Van Beneden ('50) from crabs of the genera Carcinus and Eupagurus. Zschokke ('86), discussing such forms, showed that the mono-, bi- and triloculate types of bothria (figs. 2, 12 and 13 of the present paper), which can be discovered when any considerable number of individuals are examined, are only developmental stages of the same form, and his results indicated that the S. polymorphus with which he worked was the young of the genus Calliobothrium, and not of-Onchobothrium as had been previously suggested. These authors of course studied specimens from European waters.
Monticelli ('88) in an extensive paper upon the S. polymorphus of the region about Naples, showed that the larvae, which he found in a large number of fishes, though most common in the flounders, were the young of Calliobothrium filicolle. This conclusion was based upon the close anatomical resemblances between the more advanced larvae and young specimens of C. filicolle and upon experiments in which a species of Torpedo, after being freed of all parasites by starving (a method which his experience and that of the collectors at the Naples station had shown to be effective), was then fed for a time upon specimens of Arnoglossus known to contain the S. polymorphus. As a result of this, young specimens of the C. filicolle were obtained from the torpedoes so treated, and this taken with the anatomical resemblances seemed conclusive evidence. Monticelli has reviewed the literature exhaustively and he gives (p. 89) a list of thirty-six cases in which authors have appended various specific names to the term scolex and made as many different species
THE SCOLEX POLYMORPHUS 823
of this single form. He refers to these names as synonyms and apparently considers all the forms from whatever host to be the young of a single species.
Although my experiments indicate that the S. polymorphus of the Woods Hole region develops intoPhoreiobothriumtriloculatum and not into a species of Calliobothrium, a comparison of the larvae, as taken from the fishes about Woods Hole, with the description which Monticelli gives shows that the two types are closely similar, probably almost indistinguishable. The myzorhynchus, bothria with one, two or three loculi, two faint red pigment spots in the neck region of some of the specimens, the general shape of the body and the characteristic movements, are apparently identical. Only in the case of the hosts they inhabit is there any very apparent difference, which is of course necessitated by the differences in the piscine faunas of two such widely separated regions.
In his paper published in 1897, Linton suggested that the S.polymorphus perhaps represents the young of a number of different cestodes, and also that none of the fishes (teleosts) in which he has found it is the true host of either larva or adult. He considers such larvae when found in teleosts to be 'xencsites, ' or misplaced parasites, and thinks that the true intermediate hosts may be found among the species of crabs which frequent the feeding grounds of these fishes. In support of this view, he cites the fact that the existence of similar larvae in Crustacea has been recorded by Van Beneden ('59). Should this hypothesis prove correct, we should have a case where the transfer from such a host to a teleost fish, while not fatal to the parasite, still presents conditions under which it can develop but a little way beyond the stage alreadj^ attained. Basing his conclusion largely upon the presence of a median proboscis-like structure (the myzorhynchus) at the anterior end between the bothria, Linton ('97) expressed the opinion that our Scolex polymorphus is the young of the genus Echeneibothrium. This is not, however, as strong a clew as might seem, for in the adult Calliobothrium filicolle, to which Monticelli's larvae developed, this structure is quite degenerate, though well marked in the larva, while in the adult
824 WINTERTON C. CURTIS
of Calliobothrium leucartii and Calliobothrium verticillatum there is no trace of such a structure.
In a later paper on the parasites from the fishes of Beaufort, North CaroUna, Linton ('04) finds the Scolex polymorphus in many of the fish examined and speaks of the larvae as follows (p. 326) :
The larval cestodes, doubtless representing several genera, recorded in Parasites of the Woods Hole Region under the name of Scolex polymorphus, were found in thirty-four of the fifty-nine Beaufort fishes examined. As at Woods Hole these forms are found not only in the alimentary canals of their hosts but also in the cystic ducts of several. They are almost never absent from the cystic duct of Cynoscion regalis. In all cases, where these worms have been obtained from the cystic duct and from the intestine of the same fish, those coming from the cystic duct are larger, plumper, and more opaque than those from the intestine. Some of the older larvae suggest the genera Calliobothrium, Acanthobothrium and Phoreiobothrium.
Again, in speaking of the parasites of the sharp-nosed shark, Scoliodon terrae-novae, under Phoreiobothrium triloculatum, he says (p. 343) :
1 scolex, no segments yet developed; length 2 mm.; hooks small. This specimen looks very much like some of the more advanced specimens of Scolex polymorphus which have occasionally been found, save that the bothria have assumed the characteristics of P. triloculatum.
On page 359, under the parasites of the pipe fish, Siphostoma fuscum, he notes tliat the specimens of the Scolex polymorphus had bothria with two costae and rudiment at anterior end, suggesting loculi which occur at the anterior end of bothria in Echeneibothrium and Acanthobothrium; no red pigment."
On page 407 under the parasites of the toad fish,Opsanus tau, he speaks of a specimen which was "probably a young Calliobothrium," and of another which "had the characteristic bothria of Echeneibothrium and Rhinebothrium. Its prominent muscular proboscis, (myzorhynchus), if retained in the adult would place it in the former genus." Again, in another lot, ' ' The largest had bothria which resembled those of Calliobothrium and Acan
THE SCOLEX POLYMORPHUS 825
thobothrium, but without hooks." And finally, others which had "red pigment, two costae, one specimen noted with rudimentary hooks (Calliobothrium or Acan thobothrium)" and in another lot a specimen is recorded with the "rudimentary hooks and pigment spots."
Under the parasites of the sole, Symphurus plagusia, there are mentioned specimens which are "comparatively large, with two costae and red pigment like young Acanthobothrium, but without hooks."
Fig. 80, plate 12, shows a young specimen of Calliobothrium with rudimentary hooks, but otherwise much like the Scolex polymorphus.
In view of these later observations of Linton and Monticelli's results, one would expect the Scolex polymorphus from about Woods Hole to develop into one of the species of Calliobothrium found in this region, or perhaps some other species of the family Onchobothriidae, and this last is what I believe happens in the case of the larvae with which my experiments w^ere performed. Since two species of the genus Calliobothrium (C. verticilatum and C. eschrichtii) have been found in our region, by Linton ('99) who records this species from the dogfish (Mustelus canis), and since a considerable number of species belonging to most of the genera of the family Onchobothriidae have been described from Woods Hole by Linton, it would seem not at all unlikely that experiments made by feeding the Scolex polymorphus from the various teleosts to skates, dogfish and sharks might connect these larvae with other genera of the Onchobothriidae. Such experiments would be likely to give precise evidence for or against Linton's belief that the larvae represent the young of more than one form and they might give us data for further consideration of the whole question of xenositism, which Linton suggests is the condition of these larvae when found in teleosts.
In considering the possibility that the various forms of Crustacea are the true intermediate hosts in which the development was begun, I have made a careful tabulation of the food of these fishes as recorded mainly by Linton ('99), but also in more detail for a smaller number of fishes in the work of Peck ('95). This
826 WINTERTON C. CURTIS
tabulation is not given since it is merely a compilation and the important point ascertained can be briefly stated; namely that Crustacea of various sorts, shrimps, amphipods, isopods, crabs, etc., are an important food with almost all these fish. In cases like the squeteague and blue-fish, where they are not so important an item, it is noticeable that various crustacea-eating fish are a common food. This would account for the great numbers of the Scolex in the squeteague, which would then be like a sieve in which were retained many larvae which had come originally from Crustacea through the medium of another fish. The flounder, Paralicthys dentatus, is the other fish in which Linton has found the Scolex most abundant. Its food consists of smaller fish and a large proportion of various Crustacea, so that in this case the S. polymorphus might be obtained directly from Crustacea, or indirectly from another fish. I have also tabulated the food of the fishes from which Linton has not recorded the Scolex to see whether Crustacea form as large an element in their food supply, but no satisfactory facts can be gathered for the reason that the list of those containing the larva comprises the greater number of our smaller and more common fishes and because, as Linton expressly states, no systematic search has been made and hence the fact that the larvae have not been recorded from any fish may have little importance.
I quite agree with Linton's suggestion that this widely distributed larva, though it does not resolve itself into several easily recognizable types in the larval condition, may eventually be shown to represent the young of more than one cestode, and if I am correct in my conclusion that the S. polymorphus with which my experiments were made develops into Phoreiobothrium triloculatum, whereas the form with which Monticelli worked is the young of Calliobothrium filicolle, this may be the beginning of evidence which will give Linton's interpretation a secure foundation and thus the name 'polymorphus,' which seems originally to have been given because an individual larva of this sort can assume such diversity of shape, may come to have a new significance from the existence of many species under a guise which does not show differences by which each may be recognized.
THE SCOLEX POLYMORPHUS
827
The close resemblance of our S. polymorphus to the forms upon which Monticelli (op. cit.) worked, makes a discussion of the anatomy superfluous beyond what is shown by my figures and their explanations which have been made quite full. The differences are onlj'^ of a minor nature and hence this author's account is adequate for the anatomy of our forms.
THE NORMAL OCCURRENCE OF PARASITES IN SAND SHARKS TAKEN AT RANDOM
A knowledge of the normal content of parasites found in the sharks, as collected, was important both for its bearing upon the results of treatment which attempted to rid them of all parasites, and upon the results of any artificial infection with young cestodes. There are very few sand sharks examined which have not some infection with Crossobothrium laciniatum, which is the only cestode parasite known to infect the digestive tract of this host in considerable numbers. Some records by Linton ('99, p. 429) are here tabulated as quite representative of any dozen specimens taken at random.
Table from Linton's records
July 17, 1899..
July 21, 1899 . . .
August 9, 18E9
August 12, 1899
August 15, 1899
August 17, 1889
August 18, 1899
August 19, 1899
July 20, 1900 . . .
July 20, 1900 . .
August 12, 1900
August 13, 1900
From my own records, the results are similar, though the counts are usually higher because a careful search was being made for the small young specimens. The following table is from six sharks examined in 1904.
828
WINTERTON C. CURTIS
SHARKS
C. LACINIATUM
Adult
Young
July 28
July 30
1
34
20
30
30
35
50
50
10
6
50
August 4
August 5
25
30
During the years 1899-1910 this parasite has been used for study by the students in one of the courses given at the Marine Biological Laboratory at Woods Hole and we have always been able to obtain an abundant supply when several sharks were available. Sometimes the first shark opened has yielded all the material needed and it has never been necessary to examine more than four or five. Most of the actual counts recorded in my notes in 1903-04 were made upon sharks in which search was being made for specimens showing the early stages of proglottid formation and for this reason the sharks recorded are perhaps those which seemed, when first opened, to have an abundance of the parasites. Linton's records as given in the first table are therefore more fairly representative.
As a further example of their abundance, my notes record the examination on August 11, 1904, of ten sand sharks, taken in the traps on that date. Every one of the ten was infected and in only two cases was the number of the parasites noticeably small. Count was not made because it was evident that the amount of infection averaged substantially the sam.e as that shown by Linton's record.
From these data it is evident that one rarely finds a sand shark which has not some infection ; and from the fact that the worms may be found in all stages of development, from the specimens just beginning to form segments to the large adults which are shedding motile proglottids, one may conclude that the source of the infection has been in contact with the sharks within a quite recent period, if, indeed, it is not acting upon them throughout the summer.
THE SCOLEX POLYMORPHUS 829
THE INFPXrnON OF SAND SHARKS WITH THE SCOLEX POLYMORPHUS
The first attempt at infection of the sand sharks with the Seolex polymorphus was made with fish which had been held without food for a period of three weeks, a treatment which Monticelli ('88) found effective in ridding a species of Torpedo of its Calliobothrium. In all, eleven fish were so treated and each was then given all the larvae obtainable from twelve squeteague, a dose which was estimated at not less than five hundred larvae for each shark. Each fish was fed at the time of the infection and in the three weeks after infection, during which they remained alive, each was fed four times. For food, the flesh of the squeteague was used, an amount about equal to one-third, or one-half the bulk of a good sized fish being used at a feeding; my guide in this matter being the size of the pieces of food commonly found in the stomachs of recently captured sharks, which had fed under natural conditions. Judging from the rate of digestion, as observed on several occasions, this amount of food was unnecessarily large. Such an amount once a week would be ample for sharks in captivity. Moreover, the choice of food was not good; for one may be introducing almost any kind of a cestode larva by feeding the flesh of a teleost fish. In using for infection sharks which had not received any treatment, other than the three weeks' starving above mentioned, I was, of course, aware that one might expect each one of them to contain the normal infection of C. laciniatum. It seemed, however, that if the S. polymorphus from the squeteague did represent the larval form of Crossobothrium, it would develop readily in its normal host, and that the introduction of a very large amount of infection would perhaps give the fish thus treated so many young worms all the same size, as to show that they could only have come from the larvae introduced by the experimental infection.
Three weeks after the infection these eleven sharks were killed and their digestive tracts carefully examined. Each contained adult specimens of C. laciniatum in numbers sufficient to indicate that all the sharks had tj^eir normal complement of parasites and
830 WINTERTON C. CURTIS
therefore that the three weeks without food had produced no effect. In eight of the fish there were a considerable number of young specimens of C. laciniatum in all stages of proglottid formation, but as similar stages are commonly found in all sharks (Curtis, '03 and '06), their occurrence here was no evidence that they had come from the introduced Scolex polymorphus, and the diversity of their stages made any such interpretation out of the question. In these eight specimens there were found in addition to the young and adult C. laciniatum, an unusual number of individuals of the species Crossobothrium angustum, which Linton ('99) p. 426), records as a frequent parasite of the dusky shark, Carcharinus obscurus, and the blue shark, C. milberti. Although present in greater numbers than I have found in any other lot of sand sharks, these C. angustum were of all stages from young to adult and there seemed, therefore, no evidence which would connect them with the larvae which had been introduced by my infection. In the whole number of sharks I found upwards of fifty young of another cestode, all in about the same stage, and with well developed scolices and the segmentation into proglottids just beginning. Because of their conspicuous and characteristic bothria, these were at once recognizable as the young of the species Phoreibothrium triloculatum, a form which Linton has described from the dusky shark and which is represented in figs. 5, 6, 9, 10 and 11 of this paper. Unfortunately, my records give only the fact that each of these sharks contained some Phoreiobothi'ia and fail to give their distribution in the individual sharks.
The occurrence of this species in the sharks of this experiment would indicate, when taken alone, hardly more than that P. triloculatum is sometimes found in the sand shark, even though the only sand sharks in which I have found it are the ones previously infected with the Scolex polymorphus. The fact that the individuals were all of about the same early stage is more important, though not much stress can be laid upon it because of the failure of my records to state the distribution of these worms in the individual fish. I regard the results of this attempt at infection, which was the only one I was abl^to carry through in the
THE SCOLEX POLYMORPHU8
831
first summer of my work, as valuable only because they support the more satisfactory results obtained in the work of the following summer. A further discussion is, therefore, deferred until the results of the later work have been presented.
Having been unsuccessful in the attempt to reduce the numbei' of parasites by the process of starving, for as long a time as was available, if the sharks were to be used for any subsequent experiments, attention was directed at the beginning of my second summer's work to the discovery of an effective method by which the sharks could be entirely freed of their cestode parasites. For this, I used the oil of male fern, one of the most powerful vermifuges used in human and veterinary practice. Following this practice, the dose of the oil was followed after an interval of from twenty-four to forty-eight hours by one of calomel. The manner of introducing these drugs into the stomach of a shark is explained in the earlier section of this paper where the general methods of work are given. In the tables upon the pages which follow will be found the detailed results obtained with the several lots of sharks treated in this manner. The necessary general explanations will be given in the discussion of the first table, while in the discussion of the subsequent tables I shall give only the facts which the tables show.
TABLE 1
5 sharks, captured before July 2, kept wiihoul food until July 22
July 22, all sharks given 2 cc. each of oil of male fern July 23 the four surviving sharks given 1 cc. of calomel
DATE
NO. GIVEN 8HAKK
CONDITION OF
CE8TODES IN
REMARKS
IN THIS SERIES
SHARK AT DEATH
SPIRAL VALVE
July 23
No. 1
Dying
6 C. laciniatum found
dead
July 26
Nos. 2, 3, 4,
Killed
No signs of
Mucous mem
and 5
cestodes
normal
Table 1 shows the results obtained with five sharks which were starved from three to four weeks after capture and then given
832 WINTERTON C. CURTIS
2 cc. of oil of male fern, followed twenty-four hours later by about 1 cc. (measured dry) of calomel. The dates and other points of importance are shown in the several columns, under appropriate headings. For convenience of reference, each shark is given a number. The column which show^s the ' condition at death ' is inserted because specimens did not always survive the treatment and thus some of those examined had died from its effects. Such specimens are marked 'dead' while others which were killed because they seemed about to die are marked 'dying.' The specimens which are marked 'killed' are those which appeared to be in perfect condition when they were taken and killed for examination. The specimens found soon after death ('dead') and those which were killed when they seemed likely to die ('dying'), presented data of some value, for the reason that under normal conditions the death of the shark is not followed at once by the death and consequent disintegration of the parasites. One finds that living cestodes can be obtained from untreated sharks, which have been dead in the water, or lying exposed to the air upon the wharf for five or six hours and I have often seen, in my work with the students at the Marine Biological Laboratory, spiral valves left exposed to the air all day yielding an abundance of C. laciniatum which were still alive and seemingly about as active as if taken from a shark just killed, and these worms if put into sea water may live for as long a period as fortyeight hours. In no case of an untreated shark which, on being injured by rough handling in my cars or by the collectors when first captured, was killed before it died from the effects of this handling did I find the approach of death in the host killing the parasites. We may therefore conclude that when, in a shark which has died very recently, or in one which has been killed because death seems approaching, there are found dead Cestodes, the worms have been killed by the drugs and not bj'^ the actual, or approaching, death of the host. Such cases have for this reason a sufficient value to be considered in the series. The objections against them are, first, that they do not represent individuals taken at random, and second, that it is of no value to show that the worms are killed in the sharks which do not sur
THE SCOLEX POLYMORPHUS 833
vive the treatment while other worms are not killed in the sharks which do survive. It should be remembered, however, that such non-survivors do not represent the weakest individuals in any lot, for the strongest were probably the ones which fought hardest when put in the holder and such very active sharks received rougher handling and perhaps they sometimes died from this cause. With these reservations, the specimens found soon after death and those killed when about to die, may be cited as showing the immediate effect of the drug upon the parasites.
An important point, noted in some of the tables, is that about twenty-four hours after the dose of oil, that is, when the calomel was given, numbers of dead Crossobothria were squeezed from the anus as a shark was placed in the holder. Many entire worms of all sizes were thus obtained and these when placed in sea water slowly disintegrated, showing no signs of life, as they might have done if only stupefied by the oil.
Of the five sharks here recorded. No. 1 was not in good condition, though death did not seem near at hand, when it was killed the day after the oil was administered. It contained only dead C. laciniatum. Three days after the calomel was given the four remaining specimens were killed and examined, with the results which are shown in the table. They were all in good condition and the mucous membrane seemed quite normal. I may here state that my examinations of the spiral valves throughout this work have been most careful. In each case the outer wall was split longitudinally along one side and the cut continued down across the spiral folds to the opposite side. A similar cut was made across each fold along the middle of each half of the valve and the inner surface thus exposed as four parallel rows of triangular flaps, which were then examined one at a time under a lens. Where there was any such amount of chyme as to obscure the surface of the mucous membrane the valve was washed until clean and the washings examined in shallow dishes against a dark background.
Table 2 shows the results in five sharks, which were given a heavier dose of the oil of male fern, and the calomel after an interval of forty-eight hours. After the dose of oil, shark No. 1
834
WINTERTON C. CURTIS
TABLE 2
5 sharks, captured before Juhj 1, kept ivithout food 3 to 4 weeks
July 18, all sharks given 4 cc. each of oil of male fern July 20, the three surviving sharks given 1 cc. each of calomel
DATE
NO. GIVEN SHARK
IN THIS SERIES
CONDITION OF
SHARK AT DEATH
CESTODES IN
SPIRAL VALVE
REMARKS
July 19
No. 1
No. 2
Nos. 3, 4 and 5
Dying
Dead
Killed
1 strobilla
dead No signs of
cestodes
No signs of cestodes
Trflpps? nf nil
July 20
in stomach
Digestive tract
considerably
decomposed
Mucous membrane in good condition
July 21
TABLE 2a
/ shark, captured about June 15, and kept without food
July 1, given several cc. of oil of male fern diluted with ether.
DATE
NO. GIVEN SHARK
IN THIS SERIES
rONDITION OP
SHARK AT DEATH
CESTODES IN
SPIRAL VALVE
REMARKS
Julys
Killed
No signs oi
cestodes
seemed to be dying and was killed with the results indicated. No. 2 was found dead, but was a good deal decomposed and so the absence of worms may not mean much. The three surviving specimens, which were killed twenty-four hours after the calomel, were without any cestodes.
The results of these two experiments may be criticised on the ground that the sharks were not left alive long enough to show that more would not have died from the treatment, but my experience has shown that when the animals survived this treatment for two or three days the subsequent mortality was not likely to be greater than among any other sharks kept in confinement. Some of the sharks noted in other tables were kept alive a much longer time.
THE SCOLEX POLYMORPHUS
835
Before I began using the gelatin capsules a single shark which had been in captivity a few days was given some oil of male fern diluted with ether. How much actually got into the stomach I do not know, as some was spilled in pouring the mixture down the tube which was thrust into the oesophagus. No calomel was given. Five days later when the specimen was examined there were small traces of the oil still in the intestine and no worms were found.
20 sharks, captured July 4 to 7, kept without food
July 25, all sharks given 2 cc. each of oil of male fern
July 26
Julv 26
No. 1
No. 2 No. 3 No. 4
Dead
Dying Dying Dead
9 scolices and
bits of strobilae, all dead
2 dead
20 dead and in the faeces
4 dead and in faeces
Decomposition
not noticeable
in intestine
July 26
not recorded
July 26
not recorded
Decomposition
of intestine
not noticeable
July 26, the sixteen surviving sharks given 1 cc. each of calomel
Decomposition just begun
July 27
July 28 July 28
Mucous membrane normal
Mucous membrane normal
August 2, the twelve surviving sharks each fed on shark's flesh
August 7.
No. 9
Dead
No data
I Considerably I decomposed
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 3
836
WINTERTON C. CURTIS
TABLE 3— (Continued) August 15, the eleven surviving sharks each fed on shark's flesh
DATE
NO. GIVEN SHARK
IN THIS SERIES
CONDITION OF
SHARK AT DEATH
CESTODES IN
SPIRAL VALVE
REMARKS
August 28
August 2§
No. 10
No. 11
Killed
Killed
No worms
No worms
August 28
No. 12
Killed
1 Crossobothrium
laciniatum
August 28
No. 13
Killed
1 Crossobothrium
laciniatum
August 28
No. 14
Killed
2 Crossobothrium
young 2 Crossobothrium
scolices 1 Crossobothrium
laciniatum
laciniatum .
angustum
The six remaining sharks survived the treatment to this date and later, and were used for infection experiments. When examined they gave results as follows :
August 11
No. 15
•
Dead
No worms
Badly decomposed
August 22
No. 16
Dead
No data
Badly decomposed
August 29
No. 17
Killed
No worms
Normal
August 31
No. 18
Killed
1 Crossobothrium laciniatum
medium size
August 31
No. 19
Killed
8 C. laciniatum large, scolices
5 C. laciniatum, small
August 31
No. 20
Killed
7 C. laciniatum, medium size
10 C. laciniatum, small
Table 3 shows less satisfactory results than the foregoing, the net results being as follows: without parasites 11, found dead and so decomposed that no data were obtained 3; still infected 6.
If the survivors alone are considered the results are not nearly so good, for out of ten surviving specimens (Nos. 8, 10, 11, 12, 13, 14, 17, 18, 19, and 20) we have only four (Nos. 8, 10, 11, and 17) which are entirely free of cestodes, while there are six (Nos. 12, 13, 14, 18, 19, and 20) which are still infected. Of these six, Nos. 12, 13 and 18 have but a single parasite, so the chances are that some parasites have been eliminated, but Nos. 14, 19 and 20
THE SCOLEX POLYMORPHUS
837
show so many that one could not fairly claim any reduction as a result of the treatment. When, however, those specimens which were killed when they seemed in bad condition are taken into account, it is evident that many worms must have been eliminated, and the results indicate the elimination of many of the parasites from this lot of sharks, probably of all of them in those sharks which died from the treatment, but they also indicate that the dose as here given cannot be relied upon.
TABLE 4
12 sharks, captured before July 2, kepi without food
July 28, each shark given 2 cc. oil of male fern Julj' 29, each shark given ^ cc. calomel
August 30
No.
•
Dying
No worms
Traces of oil
and calomel
in gut
August 1
No.
2
Killed
No worms
Gut normal
August 3
No.
3
Killed
No worms
Gut normal
August 4
No.
4
Killed
No worms
Gut normal
August 16
No.
5
Dead
No data
Decomposed
August 11
No.
6
Killed
1 C. laciniatum, scolex
August 11
No.
7
Killed
3 C. angustum, small
3 C. laciniatum, small
August 18, the five surviving sharks each fed on shark's flesh
August 28
No. 8
Dead
No data Decomposed
August 29
No. 9
Killed
1 C. laciniatum, small
August 30
No. 10
Killed
2 C. laciniatum small found
at very anterior end of
spiral valve
August 30
No. 11
Killed
No worms |
August 30
No. 12
Killed
2 C. laciniatum, large and
with ripe proglottids
1 C. laciniatum, small
838
WINTERTON C. CURTIS
Table 4 shows five cases of entirely successful expurgation (Nos. 1, 2, 3,4 and 11). Specimens Nos. 9 and 10 had respectively one and two worms, while No. 12, since it has two large worms with ripe proglottids, does not justify the conclusion that the number of the parasites was even reduced, and No. 7 must be regarded in the same way.
TABLE 5 12 sharks, captured July 25 to August 6, kept without food
NO. GIVEN SHARK I CONDITION OF CESTODES IN
IN THIS SERIES | SHARK AT DEATH I SPIRAL VALVE
August 17, all sharks given 2 cc. each of oil of male fern
August 18
No. 1
Dying
No worms
Dead worms
from anus in
handling
August 18
No. 2
Dying
•• No worms
August 18
No. 3
Dead
No worms
Dead worms
from anus in
handling
August 18, the nine survivors given \ cc. each of calomel
August 30
No. 4
Dying
No worms
Gut normal
August 31
No. 5
Killed
No worms
Gut normal
August 31
No. 6
Killed
No worms
Gut normal
August 31
No. 7
Killed
No worms
Gut normal
August 31
No. 8
Killed
No worms
Gut normal
August 31
No. 9
Killed
No worms
Gut normal
August 31
No. 10
Active
No data
Escaped
August 31
No. 11
Killed
8C. laciniatum,
medium size
August 31
No. 12
Killed
4 C. laciniatum,
medium size
Table 5 has twelve sharks, less one which escaped during the handling. Of these eleven individuals, Nos. 5, 6, 7, 8 and 9 were without any infection, Nos. 1, 2, 3 and 4 did not survive, but showed that the drug had killed the parasites. Nos. 11 and 12 have respectively eight and four specimens of C. laciniatum and hence must be counted as against the effectiveness of the treatment.
THE SCOLEX POLYMORPHUS
839
It SO happened that no shark, in which the Crossobothria survived, was killed among the first in any lot ; as may be seen by reference to the dates on tables 1, 2, 3, 4 and 5. Thus by the 11th of August my records showed thirteen sharks which were killed when in perfectly good condition, six others which were killed, when it seemed that they were likely to die, and seven others, which died from the treatment and in not one of these had I found a single living Crossobothrium. This seemed to demonstrate the effectiveness of the single treatment with the oil of male fern, which was therefore continued, the sharks which survived it being kept for infection with the S. polymorphus. Later, when sharks began to appear in which some of the worms had survived, the season was so far advanced that there was neither
TABLE 6 9 sharks, captured August 6 to 9, kept without food
August 18, all sharks given 2 cc. each of oil of male fern
August 19, all sharks given ^ cc. of calomel
August 19, dead Crossobothria from anus of one specimen in handling
August 30 . August 30
^o worms
No data
Decomposed a little Escaped
August 30, gav^ the s^ven survivors 3 cc. each of oil of male fern
September 1
September 1 .
No. 3
No. 4
September 2
No. 5
September 2
No. 6
Ssptember 2
No. 7
September 2
No. 8
September 2
No. 9
Dead
Dead
Killed Killed Killed Killed Killed
iNO WOrulS
No worms
No worms No worms No worms No worms No worms
No noticeable decomposition
No noticeable decomposition
Gut normal Gut normal Gut normal Gut normal Gut normal
840
WINTERTON C. CURTIS
TABLE 7
Showing the results of infection experiments
DATE OF
shark's
CAPTURE
NO. GIVEN IN THIS SERIES
No.
1
No.
2
No.
3
No.
4
No.
5
No.
6
No.
7
No.
8
No.
9
DATE OF TREATMENT WITH O.M.F.
REMARKS
July 6. . July 6. July 6. July 1. July 15 July 6. July 6. Julyl.. July 1 .
July 25 August 1-3 August 1-3 August 1-3 August 1-3 August 1-3 August 1-3 August 1-3 August 1-3
August 4 August 4 August 10 August 10
August 10 August 10
August 9 August 8 August 8 August 11 August 10 August 9 August 9 August 11 August 11
August 10 August 10 August 10
August 10 August 10
Fed at time of infection
TABLE 7— Continued Showing the results as in tables 1 to i
NO. GIVEN
SHARK IN THIS
SERIES AS
ABOVE
CONDITION OF
SHARK AT
DEATH
CESTODES IN SPIRAL VALVH
August 12'.
August 16 . . . August 31 . . .
September 1.
September 1. September 1.
September 1. September 2. September 2.
No. 1
No. 2 No. 3
No. 4
No. 5 No. 6
No. 7 No. 8 No. 9
Dead
Dead Killed
Killed
Killed Killed
Killed Killed Killed
Many S. polymorphus found attached to the wall of the stomach and to the bit of the squeteague's stomach in which they were placed when introduced. The gut was slightly decomposed, but the larvae were still alive. No data. Badly decomposed Many small Phoreiobothrium triloculatum. See text p. 846 3 C. aiigubuuii., small 3 C. laciniatum, small
7 Phoreiobothrium triloculatum. 3 C. laciniatum, small.
8 C. lacininatum, large with the long neck region
5 C. laciniatum, small 10 C. laciniatum, small. 7 C. laciniatum, medium 3 C. angustum, small Many Phoreiobothrium triloculatum 1 C. laciniatum, scolex only. Many Phoreiobothrium triloculatum
THE SCOLEX POLYMORPHUS
841
the time nor the material for trying a modification of the treatment upon any considerable number of sharks. Only nine sharks were differently experimented upon and the results obtained are represented in table 6, the net results of which are; without infection, 5; found dead but in sufficiently good condition to give reliable evidence that the worms had been killed by the drug, 3; escaped, 1.
The results of all the expurgation experiments may be summarized in the table which follows:
TABLE 8 Showing net results from all experiments in treaimenl with oil of male fern
TABLE REFERRED TO
SURVIVING SHARKS
SHARKS WHICH DIED UNDER THE TREATMENT,
BUT IN WHICH THE WORMS WERE ABSENT OB
HAD BEEN KILLED BT THE OIL OF MALE FERN
No infection
Still infected
Killed when seeming
likely to die
Found dead
Total
1
2 3 4 5 6
4
3
4 5 5 5
6
4
2
1
1
3 1 4 3
1
4
5
5
17
10
11
8
26
12
13
5
56
In this table, a few specimens which were found dead and so decomposed as not to yield reliable data are omitted. The specimens which survived treatment are of course most important and stand in the proportion of 26 for and 12 against the success of the attempted expurgation. Of the non-surviving sharks we have 13 specimens taken while still alive and all showing that worms were killed, as a result of the treatment and some time prior to the death of the shark. When one compares the number of worms found in these treated sharks with the tables and records showing their prevalence in sharks which have had no such treatment, it will be quite evident that a large proportion of the parasites must have been eliminated even by the single dose.
842 WINTERTON C. CURTIS
In those sharks which were given the double dose of o. m. f. (table 6) not a single cestode was discovered, but as only eight specimens were thus treated the number of trials is insufficient to show that even this treatment may be regarded as always effective. It is perhaps too much to expect a treatment that will be entirely effective in every instance, nor is such a treatment necessary, for a straj" worm or two in one shark out of a dozen would still give us a result good enough for practical working purposes. Unsatisfactory though my results are, they do, I think, justify the belief that a little more experimenting along the line of repeating the dose one or more times would develop a method of treatment sufficiently effective for working purposes, i.e., will eliminate all the parasites, except in rare instances, without killing too many of the sharks. If such a method can be found, the sharks thus expurgated could be used in a variety of experiments. For example, by introducing young Crossobothria into a shark one might expect to find out more than we now know about the rate of growth and the maturing of proglottids in the cestoda, and another point worth examining would be the truth of the current idea that it is the inability of the cestode to survive in the wrong host which hmits the habitat of a given species of parasite to a single host, or to a few closely related hosts. This latter point might, indeed, be investigated by the infection of sand sharks with larval forms other than the ones known to occur in that host, but where the normal infection is so great the examination of the specimens would be much easier if the host was first freed of all infection. An insufficient supply of squeteague during the latter part of the second season made it impossible for me to make more than a few infections after the method of treatment, as above described, had been established. Only nine sharks were so infected and none of these was killed until about three weeks after the infection had been introduced. Hence the material is wanting to show my transitional stages between the S. polymorphus and the young specimens of Phoreiobothrium (figs. 9, 10 and 11) which I believe to have come from this larva. This lack of early stages resulted not because I chose, having only a limited number of
THE SCOLEX POLYMORPHUS 843
infections, to let them all run as long as possible, but because the sharks first infected were kept to run longest; while later infections were planned for the earlier stages, an arrangement which would give the most extensive time results in a limited time. When, however, the time came for infecting the sharks from which to secure the earlier stages, the squeteague could not be obtained in numbers sufficient to yield the larvae for infection purposes.
Of the nine specimens, two died in the cars and from one of these two no data were obtained, so there are only seven specimens from which entirely reliable conclusions can be drawn. The sharks used were all starved for three or four weeks after their capture and then given 2 cc. each of the oil, followed twentyfour hours later by about 0.5cc. of calomel. They were all specimens which had survived this treatment and the history of each individual previous to the time of the infection is given in the first half of table 7. Since there is no evidence connecting the genus Crossobothrium with the S. polymorphus, this table may also be used to furnish data on the success of the attempt at expurgation. The first part of the table gives the dates of capture and of the treatment with the drugs, between which events the sharks were kept in the cars without food. The dates at which they were fed are also shown for comparison with the dates of infection. Each shark is numbered, as in the previous tables, and these numbers are repeated in the second half of this table where the results of the examinations for parasites are tabulated in a manner similar to that followed in the tables for treatment with the oil.
In introducing the S. polymorphus into these sharks I took the portion of the cystic duct containing the larvae from twelve squeteague, and placed these in the little sac obtained by cutting off the end of a squeteague's stomach. This sac was turned wrong side out to avoid any possible injury from direct contact with the mucous membrane of the squeteague. The infection thus prepared was introduced into the shark's stom.ach by pushing it through a piece of iron pipe having a diameter sufficient to admit such an object without undue pressure. For food, I used
844 WINTERTON C. CURTIS
the flesh of another sand shark in preference to that of a teleost, since there is the minimum likelihood of finding any cestode larvae in the flesh of a large shark; whereas a teleost may, in addition to its normal parasites, contain almost any thing in the way of a 'xenositic' larva. All the sharks were fed shortly before or after their infection, as is shown in the first half of the table.
The latter part of table 7 shows the results when these same sharks were examined for parasites, and by reference to the number given each specimen, one may follow any one fish through the two parts of the table. The only shark examined soon after the infection is No. 1, which was found dead. In this shark, specimens of the S. polymorphus were found adhering to the surface of the shark's stomach and to the remains of the bit of squeteague's stomach in which they had been introduced. Although this specimen was not found until decomposition was quite in evidence, these larvae still showed some slight movements and had therefore survived in the stomach for a period of three days. Shark No. 2 died during my absence from Woods Hole and was so badly decomposed when found that no data were obtained. In the spiral valve of specimen No. 3 there were a very large number of young Phoreiobothrium triloculatum (figs. 9, 10 and 11). I collected and preserved some thirty-five of these worms, but this number represents only a small part of those present. Owing to their minuteness when only the delicate posterior end can be seen protruding from between the villi of the intestine and the tenacity with which their powerful hooks enable them to retain their hold, these larvae are often very difficult to detach from the walls of the spiral valve, though they may be large enough to be easily recognized. There must have been present in the shark many more than I collected; for taking into account the ones actually seen but not collected, I estimated at the time that there were a good many more than one hundred of these young worms in this single fish. A fact of perhaps more importance than their numbers is that in any one shark they were all in the same stage of development.
Shark No. 4 shows three specimens of C. laciniatum which are to be regarded as worms which survived the expurgation treat
THE SCOLEX POLYMORPHUS 845
merit. Seven small specimens of P. triloculatum were collected, but in this case it was evident that the valve contained a much smaller number of these than the previous one. My notes taken at the time state that no more were seen in addition to those actually collected, although there may have been some toward the anterior end where the villi are longest. Specimens Nos. 5, 6 and 7 all present evidence against the effectiveness of the vermifuge. In 6 and 7 the numbers of surviving cestodes is so great that no certain effect from the drug is indicated. There is perhaps some significance in the fact that here, as in some other cases where the worms are found surviving the effects of the oil, the proportion of young to adult specimens is somewhat increased. This may indicate that the drug is more effective with the large worms which have long bodies extending among the folds of the valve than with very young ones which may often be almost buried among the villi and so escape the full effects of the drug. In this table there are recorded from specimens 4, 5, 6, 7, 8 and 9 a total of 43 surviving Crossobothria. Of these only 9 are beyond the period of segmentation into proglottids and from what we know of the proportions in which the young and old are commonly found it would seem that the drug has destroyed more adult than young worms. This conclusion is of course based upon my belief that these young Crossobothria have not come from the introduced S, polymorphus.
In sharks Nos. 8 and 9 the same condition was found as in No. 3, namely, so many specimens of small P. triloculatum that the counting them was an impossible task. Later when I examined very carefully the preserved specimens I found that the Phoreiobothria from any one shark were of uniform size, but that when those from different sharks were compared there was some difference in the size attained. For instance those from specimen 9 are almost twice the size of the ones from specimen 3. This difference in size is noticeable only in the body portion of the larvae, the scoHces being of very uniform size.
The presence of the C. laciniatum and a few C. angustum, although there are more young specimens among them than one would expect to find in sharks taken at random, does not seem
846 " WINTERTON C. CURTIS
to indicate anything except the ineffectiveness of the attempt at expurgation. There can be no interpretation of the facts which would show that the S. polymorphus had given rise to these Crossobothria.
Despite the lack of intervening stages between the S. polymorphus and the specimens of P. triloculatum, which I found in these sharks, I think there is some pretty good evidence that the latter have developed from the introduced S. polymorphus. First, the examination by myself, and also by Linton, of a large number of sand sharks at various times has never shown that young or adult specimens of Phoreiobothrium triloculatum are to be found as regular parasites of this shark, or as frequent 'xenosites.' I have never found this form in any sharks except those which I infected with the S. polymorphus. It is very probable, however, that one might at any time find stray specimens of this worm, since the squeteague is a not uncommon food of this shark. The failure to find it in any of the sharks I have examined would indicate that it does not often survive, when introduced in nature along with the squeteague, and in my experiment only a small proportion of the larvae introduced have survived. Second, the specimens of P. triloculatum which I found in any one shark were of very uniform size, and unless we suppose that there is some limit to their growth, when in an abnormal environment (the wrong host), this would indicate that they all entered the shark at about the same time. In the third place this conclusion is also in line with the results of MonticeUi ('88) who has shown that the S. polymorphus of European waters develops into the genus Calliobothrium, which belongs to the same family as Phoreiobothrium, a fact which is further discussed in the portion of this paper which deals with the nature of the S. polymorphus.
Attention should here be directed to one point which has perhaps suggested itself from the examination of the figs. 7-13. This is that the specimens of the young P. triloculatum (figs. 9 and 10) are considerably smaller than some of the specimens of the S. polymorphus, from which I suppose them to have developed. This appeared to me at first a most serious objection, though upon further consideration it does not seem an insur
THE SCOLEX POLYMORPHUS S47
mountable one. The ragged and irregularly constricted ends of some of the specimens (figs. 7 and 8) suggest that part of the body is being lost, as in the case in the development of the strobila in those forms which have a typical bladder-worm stage. Again, while the body region is smaller than that of the larger specimens of the S. polymorphus (fig. 12) the scolex of the young P. triloculatum (figs. 9, 10 and 11) is considerably larger than that of the S. polymorphus. I should also add that fig. 12 represents one of the very largest of the larvae and has been killed under pressure to flatten the body and make it more suitable for a whole mount whereas the specimens of P. triloculatum were killed without pressure and the body has remained cylindrical. The S. polymorphus reaches this size (fig. 12) only in the cystic duct and the smaller specimens present a lesser disparity in size. I am inclined to think that in the case of such large specimens portions of the S. polymorphus body may be moulted off just as in the case of the bladder portion in the cysticercus. One constant feature of the larger specimens of the S. polymorphus has, perhaps, some significance in this connection. It is the occurrence of a denser region which terminates abruptly a little way behind the scolex (fig. 12). In a specimen stained and mounted whole after the carbonate of lime granules, which occupy so much of the parenchyma in the body region, have been dissolved out, one can distinguish this region as having the same denser appearance as the tissue in the body of the young Phoreiobothria (figs. 10 and 12) or the region of proglottid formation in young specimens of C. laciniatum. It is possible that the scolex and this region of the S. polymorphus are the most important in the formation of the adult worm and that part or the whole of the body region may be lost. It seems probable that larvae like the S. polymorphus have been derived from larvae of the cysticercus type and, if this be the case, it would not be surprising to have a part of the body region lost at this point in the development.
848 WINTERTON C. CURTIS
SUMMARY
Experimental infections of the sand shark (Carcharias littorahs) with the cestode larva known as the Scolex polymorphus indicate that the larvae used for these experiments developed into the species Phoreiobothrium triloculatum. It seems clear that the common tapeworm (Crossobothrium laciniatum) of this shark cannot come from the Scolex polymorphus, even though this larval type may represent the young of a number of cestodes, a possibility which is referred to in the third section of this paper.
Starving the sharks had no effect upon the cestodes, but by means of treatment with the oil of male fern followed by calomel the great majority of the parasites wese eliminated before the sharks were artificially infected. It seems probable that by a little more experimentation a method of treatment could be secured, which, for eliminating these parasites, would be sufficiently effective for all working purposes.
THE SCOLEX POLYMORPHUS 849
LITERATURE CITED
Bronn, Thikkreich 1894-1900 Cestodes.
Curtis, W. C. 1903 Crossobothrium laciniatum and developmental stimuli in the cestoda. Biol. Bull., vol. 5, no. 2, July, 1903. 1906 The formation of proglottids in Crossobothrium laciniatum. Biol. Bull., vol. 11, no. 4, Sept., 1906.
Linton, Edwin 1886 Notes on the entozoa of marine fishes of New England, with descriptions of several new species. Rept. Commissioner of Fish and Fisheries for 1886. Washington. Published in 1889. 1887 Notes on the entozoa of marine fishes of New England. Part II. Ibid, for 1887. Washington, Published in 1890.
1897a Notes on larval parasites of fishes. Proc. U. S. Natl. Museum, vol. 19, Washington.
1897b Notes on cestode parasites of fishes. Ibid., vol. 20, Washington. 1899 Fish parasites collected at Woods Hole in 1898. Bull. U. S. Fish Commission for 1899. Washington. Published in 1900. 1899 Parasites of fishes of the Woods Hole region. Ibid, for 1899. Washington. Published in 1901.
1904 Parasites of fishes of Beaufort, North Carolina. Ibid, for 1904 Washington. Published in 1905.
MoNTicELLi, F. S. 1888 Contributione alio studio della fauna ehninthologico del golfo di Napoli. 1. Richerche sullo Scolex polymorphus. Mitth. zool. Stat. Neapel, Bd. 8, p. 85-152.
Peck, James I. 1895 The sources of marine food. Bull. U. S. Fish Commission for 1895. Washington. Published in 1896.
RuDOLPHi, C. A. 1808 Entozoorum sive vermium intestinalium historia naturalis. Vols. 1 and 2. 1808 and 1809.
VAN Beneden, J. P. 1850 Les vers cestoides. Bull, de I'Academ. roy. de Belg. Tome 17, no. 1.
Wagener, C. R. 1854 Die Entwicklung der Cestoden nach eigenen Untersuchungen. Nov. Act. d. k. Leop-Carol. Akad. d. Naturf. Bd. 24, suppl., Breslau.
Zschokke, F. 1886 Le development du Scolex polymorphus. Arch. d. sc. phys. et. nat., 3 ser., Tome 16, Geneve.
PLATE 1
EXPLANATION OF FIGURES
1 Outline of a living specimen of the Scolex polymorphus. The stippled areas just behind the four bothria show the location of the faint red pigment spots seen in some specimens. Magnified about 45 diameters.
2 Specimens of the Scolex polymorphus drawn from the living specimen to show the characteristic mode of attaching with all four suckers to the bottom of a dish in which they are being examined and the manner in which they fasten to one another. Magnified about 45 diameters.
3 and 4 Outlines of the Scolex polymorphus drawn from living specimens to show other characteristic shapes. The area occupied by the large and small vascular trunks of either side and the terminal vesicle are added to fig. 3 as they appear in a stained specimen. The filiform appearance shown in fig. 4 is often seen as the larvae draw themselves over a surface by the characteristic movements of their bothria. Magnified about 45 diameters.
4a From a specimen stained and mounted whole, showing a portion of the body region on a large scale. The cuticle (cu), the larger and smaller excretory trunks (ivt) and the conspicuous longitudinal muscle fibres (w/) are shown. The stippling indicates the distribution of nuclei in the parenchyma as it appears in optical section. Along part of the upper margin are shown the minute projections which occur upon the cuticle in the posterior part of the scolex region and for a short distance along the body, finally giving place to the smooth cuticle as shown in the figure. Magnified about 200 diameters.
5 Scolex of a young Phoreiobothrium triloculatum, taken from a sand shark artificially infected with Scolex polymorphus (see shark no. 3 of table 7). This view shows the division between two neighboring bothria along the mid-line of the figure. The characteristic pairs of hooks are shown attached along the line separating the upper and middle loculi of each bothrium. In this figure the subdivision of the posterior margin of the bothrium into three loculi, from which the specific name is derived, is not seen because the contraction of the lower margin brings their surface at right angles to the general surface of the bothrium. The area on the mid-line of the figure toward the anterior end of the scolex, and marked by the closer stippling, appears in some specimens and may represent the *myzorhynchus' of the Scolex polymorphus. The minute spikes protruding from the neck region of the cuticle are shown in profile only. Magnified about 90 diameters.
6 The same as the last figure. One of the bothria is shown in front view and the three posterior loculi are expanded so as to show clearly. Magnified about 100 diameters.
PLATE 2
EXPLAXATIOX OF FIGURES
7 and 8 The posterior end of a young specimen of Phoreiobothriuni from a shark artificially infected with the Scolex polymorphus. These figures show the formation, either of psuedo-proglottids which are molted, or an irregular beginning of true proglottid formation like that of Crossobothrium laciniatum (Curtis,' 06). The figure shows so much irregularity that the process does not seem like true proglottid formation and the posterior end shows a peculiar outline which probably indicates that part of the specimen has been lost. Only the larger pair of the two water trunks is shown. Magnified about 70 diameters.
9 One of the largest specimens of Phoreiobothriuni triloculatum which was secured from the artificially infected sharks (see shark no. 8 of table 7). This is one of the few specimens which showed a segmentation of the posterior end sufficiently regular for comparison with the formation of the posterior proglottids in Crossobothrium laciniatum. The scolex of the specimen is shown in outline only. The area of the minute spikes is shown and back of this the two j^airs of water trunks which in such a view are superposed throughout the length of the body, are here shown in successive regions. Magnified about 70 diameters.
10 and 11 Two of the smallest of the specimens of Phoreiobothriuni triloculatum obtained from the artificially infected sharks (see shark no. 3 of table 7). The specimens represented in fig. 10 was ragged at the posterior end as though portions had been detached. Magnified about 70 diameters.
12 and 13 Show in more detail the features of the Scolex polymorphus. In fig. 12 the 'mj'zorhynchus' or proboscis-like structure is shown at the anterior end between the four bothria. The larger and smaller water vascular trunks are shown in successive regions and their convergence toward the posterior end where the vesesls become irregular and branched so that they can only be followed as an area of differentiation in the parenchyma. Each bothrium consists of a denser portion divided into three regions or loculi in the older specimens (fig. 12), and in those slightly smaller often showing only the line of division at the anterior end (fig. 13). In fig. 13 the opening of the terminal vesicle of the water tubes is seen in surface view, the posterior end of the specimen being slightly turned toward the observer. Magnified about 70 diameters.
852
IJFK HISTORY OF SCOI.EX POI.VMoRPITr w. c. criiTis
JOURNAL OF N:0HPH0L0GY, VOL. 22, NO. 3
853
DRAWN BY W. V. CURTIS
THE LIMITS OF HEREDITARY CONTROL IN
ARMADILLO QUADRUPLETS: A STUDY
OF BLASTOGENIC VARIATION
H. H. NEWMAN and J. THOMAS PATTERSON
From the Zoological Laboratory, University of Texas (No. lOS)
FIVE TEXT FIGURES AND EIGHT PLATES
CONTENTS
Introduction 855
Morphology of the integument 861
Meristic variation of the elements of the nine bands of armor 865
A Variation in the banded region as a whole 865
B Comparative variability of males and females 868
C Variation in the individual bands 870
D Fraternal correlation in the banded region as a whole; an index of the
limits of hereditary control 873
E Fraternal correlation in the individual bands 880
Atj^pical variation in the individual elements and in the bands 883
A Scute ' abnormalities' in the species ; their distribution and frequency . 883
B Hereditary control in connection with scute 'abnormalities' 888
C Band 'abnormalities' in the species; their distribution and frequency. 893
D Hereditary control in connection with band 'abnormalities' 897
Pairing, an intra-fraternal correlation; and its bearing on the problem of
hereditary control 905
The hereditary control of sex 910
General considerations 911
Summary 914
Bibliography 917
INTRODUCTION
To what extent or ivithin what limits are the definitive characters of the individual determined at the time of fertilization and in how far are the minutiae of organic structure to be considered as the product of individual variability beyond the limits of hereditary control? No more fundamental question could well be raised, and none more difficult of solution. The question has recurred
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 4 DECEMBER, 1911
855
856 H. H. NEWMAN AND J. THOMAS PATTEESON
in various guises, but the underlying inquiry has remained unchanged. In one of its older forms the question is phrased: "What is the relative potency of nature and nurture in development?" In some of its more modern phases it appears in terms of 'predetermination versus epigenesis/ 'blastogenic versus somatogenic variation/ and 'heredity versus environment.'
In spite of the antiquity of the problem very little direct evidence has appeared for its solution. So far as we have been able to ascertain, the only facts that seem to throw any clear light on the situation are those furnished by cases of human 'identical' twins. As long ago as 1875 Galton showed his appreciation of the value of such data in his paper entitled "The history of twins, as a criterion of the relative powers of nature and nurture." In a subsequent paper (Galton, '92) he made use of finger-prints as criteria for distinguishing between twins and as a suitable character for testing the degree of likeness and unlikeness of such pairs. That he had a deep insight into the underlying fundamental problems involved in the situation is shown by his remark: "It may be mentioned that I have an inquiry in view which has not yet been fairly begun, namely to determine the minutest biological unit that may be hereditarily transmissible. The minutiae in the finger-prints of twins seem suitable objects for this purpose."
In 1904, Wilder, interested in the same problem, presented a comprehensive review of the whole subject in his paper on "Duplicate twins and double monsters." His results are of great interest and will, we hope, assume a still greater significance in the light of the facts here presented. Wilder discovered a surprising degree of resemblance between the palm and between the sole patterns of duplicate twins and was able, he thought, to classify twins as 'fraternal' or 'duplicate' on the basis of this resemblance. In addition, the following phenomena were observed in the majority of cases, but were not universal:
" (1) A bilateral correspondence in the palms and soles of each individual of a set.
"(2) A reversal of the finger patterns in either of the right or the left indices.
LIMITS OF HEREDITARY CONTROL 857
(3) Differences occur more frequently on the left side." In his concluding paragraph, here quoted, are given in concise
form the chief results and conclusions derived from his studies
of duplicate twins:
The influence of the germ-plasm and its mechanism (i.e., the direct control exercised by heredity) is exerted upon the friction-skin surfaces only so far as concerns the general configuration, i.e., the main lines, the patterns and other similar features; the individual ridges and their details (minutiae) are apparently under the control of individual mechanical laws to which they are subjected during growth. //a?;e we ^/ien arrived at the limit of the control of the predetermining mechanism beyond which mechanical laws are alone operative; and is it then possible to hold that the modifications in this latter field are the results of individual experience, and that they are similar in the various members of a given species solely because of similar environment? To these and similar questions we can give no answer at present ; yet it seems likely that in the general subject of palm and sole markings, not only in man but in other mammals as well, we have a set of easily observed and very significant data which may yield important results to future investigators.
In addition to his data on palm and sole patterns Wilder furnishes us with rather elaborate physical measurements of four sets of duplicates. He realizes, however, that dimensional data are far less determinative than are other characters employed, since they are liable to fluctuations through numerous causes, both internal and external, and it could hardly be expected that the similarities here would be very striking." Yet some most striking physical resemblances have been brought out by different authors. Vernon ('03), for example, gives the data for two pairs of identical twins, one of which, aged twenty-three, showed an average per cent difference of 0.28 per cent; while the other, aged twelve, a somewhat greater difference of 0.71 per cent. Weismann presents the data on one pair of twin brothers, aged seventeen, who showed a per cent difference of 2.2, nearly ten times that of Vernon's first pair and about three times that of his second.
Wilder's measurements include a much wider range of characters than do the others cited and are therefore probably of somewhat greater value. Table 1 presents a brief summary of his data.
858
H. H. NEWMAN AND J. THOMAS PATTERSON
TABLE 1 Showing physical statistics of Wilder's twins
SET
NUMBER OP
CHARACTERS
PER CENT
DIFFERENCES
AGE IN TEARS
I
27
2.39
21.10
II
27
2.35
17.11
Ill
27
1.64
17.10
IV
5»
1.76
17.11
Mean
2.03
18.1
Although Wilder realizes that dimensional and other physical measurements should be taken during the younger life or at least before there is any marked difference in the experience," yet he apparently considers that in all of his four sets, whose ages range from seventeen to twenty-one years, these conditions are met with, as they are all those of young people." Seventeen or more years of post-natal life would seem, however, to be sufficient for the operation of nutritional differences, some of which might tend to cause originally identical characters to diverge, and others, originally divergent characters to converge. It would seem to be inadvisable then to attempt to draw the line between nature and nurture on the basis of physical characters so subject to modification by environment, for a variation in nutrition alone might readily produce all of the differences shown in the sets of physical measurements cited. Nutritional differences doubtless manifest themselves even before birth, and hence would tend to vitiate the results of physical measurements taken on new-born duplicates even if such were available. Consequently it would seem advisable to limit investigation to those characters which reach a definitive condition at an early period and which are subject to little or no modifications due to nutrition. The patterns of the friction ridges of the palm and sole are characters of this sort and should give highly reliable data as to the strength of hereditary control. But even this data, interesting and suggestive as it is, can be accepted only with a considerable amount of reservation; for it has, in common with all other data derived from
LIMITS OF HEREDITARY CONTROL 859
a comparison of human identical twins, the fundamental weakness that it is based on an assumption which is clearly beyond the possibility of proof. Because certain twins exhibit a most striking resemblance, it is assumed that they have been derived from a single fertilized egg. As Weismann ('02) expresses it, "Wir hahen nun alien Grund, die erste Art von Zwillingen von zwei verschiedenen Eizellen abzuleiten, die letztere Art von einer einzigen, welche erst nach der Befruchtung durch eine Samenzelle sich in zwei Eier getheilt hat."
Were it possible in a number of cases to determine by examination of the placental relationships of new-born twins whether they were duplicates or fraternals, and were it also possible to obtain data on these cases whose uterine history was known, we would have facts from which we could with confidence draw conclusions. Unfortunately, however, although monochorial twins have been observed at birth, no. interest has been manifested in their resemblances. In view of the lack of satisfactory criteria as to the origin of the twins investigated, the writers on these subjects have reasoned backwards from the facts of resemblance to the assumption of common origin, a procedure far from safe, but doubtless justified by circumstances. An arbitrary criterion is thus set up for the classification of twins ; and those that come up to specifications are classed as 'duplicates,' while those that fail to meet the arbitrary requirements are relegated to the rank of 'fraternals.' One cannot but be impressed, as he reads Wilder's monograph, with the author's feelings of uncertainty as he attempts to classify certain pairs. The following extracts indicate his attitude :
No. VII. This case has caused me considerable trouble, owijig to the preconceived notion that the marks ought to be found identical. The family emphasized the facial resemblance of these twins and when I first saw them they certainly looked much alike. One was, however, an inch taller than the other, and the facial resemblance, after a short acquaintance did not seem as great. . . . The case is plainly one of fraternal twins that resemble one another somewhat more than the average.
No. XIII. According to personal appearance these should be duplicates. I have never seen them, but the one who took the prints wrote: 'The Misses' are so similar in coloring, figure and features
860 H. H. NEWMAN AND J. THOMAS PATTERSON
that even their best friends confuse them. ' It must be confessed, however, that the differences in the formulae cannot be reconciled, and that the palms are, and remain, in respect to the main lines, very different.
- In the light of the results presented below we are inclined to
believe that Wilder was not justified in thus arbitrarily excluding from the category of ' duplicates' such cases as those referred to, for we have found not a few sets of armadillo foetuses which exhibit greater differences than some of Wilder's so-called 'fraternals.' Hence, although the results of his studies are valuable and highly suggestive, they are insecurely founded and therefore cannot be applied to the solution of the problem of the limits of hereditary control.
The material which forms the basis of the present investigation consists of a collection of advanced sets of foetuses (removed from the uterus with all of their placental connections intact) of a species of mammal, Tatu novemcinctum, in which we have demonstrated conclusively the existence of specific polyembryony ; and hence all sets of embryos, whether strikingly similar or not, are known to be the product of the division of a single fertilized egg. The basic assumption involved in the case of human duplicate twins is thus obviated, and at the same time it is possible to eliminate the factor of a diverse post-natal environmental experience by examining unborn foetuses, whose inter-relationships are shown by their placental connections. In the scutes of the banded region we have characters little if at all subject, even during gestation, to environmental control. We plan in the present paper to present an intensive study of the phenomena of blastogenic variation as exhibited by these integumentary elements, limiting our present investigations to the well defined banded region, believing that the conclusions arrived at from the study of one region wiU prove to be generally applicable, and that what is true for one character or set of characters will be found to apply in a general way to the whole organism.
In order that there may be no misunderstanding as to the kind of variates we are dealing with, it seems advisable to present in abbreviated form the results of a study of the morphology of the integument and of the variability of its elements as exhibited by
LIMITS OF HEREDITARY CONTROL 861
a large sample of the species. Without this data one would scarcely be in a position to appreciate the degree of resemblance or difference that exist among the foetuses of the different sets. Owing to the existence of an extensive curio industry, making a specialty of baskets shaped from the shells of armadillos, there has been afforded an exceptional opportunity for gathering a large mass of data on the variability of the species. By availing ourselves of the large stocks of basket-shells in the hands of various dealers we have been able to examine 1768 individuals for scute and band ' abnormalities' and to count the scutes in the banded region of over 500 shells, including those of all males and females sent to us alive.
MORPHOLOGY OF THE INTEGUMENT
The integument is one of the most characteristic features of the anatomy of the armadillo. For the most part it consists of a series of bony plates which are arranged so closely together as to form an almost continuous armor, especially on the dorsal and lateral parts of the body. When attacked the animal is able to retract itself well within this shell-like structure, much after the manner of a turtle, and although the belly and legs do not possess an armor, in the strict sense of the word, yet even here the skin is studded with horny scutes and the feet are armed with powerful claws. Altogether the integument of the armadillo forms a protective structure of high efficiency in an otherwise defenseless animal.
In our species, Tatu novemcinctum, five of the so-called armor shields described for armadillos are present. These are the cephalic, covering the front of the head; the scapular, overlying the shoulders; the thoraco-lumbar or banded region (sometimes called the movable zones), consisting of nine bands or incomplete rings; the pelvic, covering the hips; and finally the caudal shield, which consists of a series of rings surrounding the tail (fig. 17).
The elements composing the armor exhibit in each of the shields a somewhat different and more or less characteristic arrangement ; but since in this paper we are concerned with the study of varia
862 H. H. NEWMAN AND J. THOMAS PATTERSON
tion and heredity in but one of these shields, it is not necessary to enter upon a description of the elements of the others, except in so far as such an account would be of help in understanding the character of the elements in the particular region in question.
The nine bands are united to one another by strips of flexible skin which permit considerable motion in this part of the armor. The first band is attached in a similar manner to the scapular shield, while the integument joining the ninth band to the pelvic shield may be present only toward the ends of the band, in which case the middle of the band is firmly united to the shield. The soft parts of the body lying directly beneath the strips of flexible skin are not exposed, because of the fact that the bands overlap one another for a distance equal to about one-third their width, that is, the posterior margin of any band overlaps the anterior third of the succeeding band. On account of this overlapping the banded region as a whole presents a distinctly testudinate appearance.
Each band is composed of a number of elements, of which there are three kinds: (1) the thick, bony, dermal plates covering the under surface of the band and constituting its main body; (2) the thin, horny, epidermal scutes which cover the posterior exposed part of the band; and (3) the associated hair group.
Each bony plate is oblong in outline, with its long axis constituting the width of the band, and in the adult animal has an average width and length of 6 mm. and 30 mm., respectively. As we shall see later, the number of these plates varies in different bands as does also the number for the same band in different individuals, but in round numbers there are on the average about 62 plates to the band.
The epidermal scutes, unlike the underlying bony plates, are of two types, which we may call primary and secondary. The first of these is represented by a wedge-shaped area having a slightly convex upper surface, and with its base forming a part of the posterior margin of the band (fig. 20). The base of each scute has two notches which are situated so as to divide the margin into three areas of about equal width. In reality the notches are but the mouths of hair pits located in the underlying bony plate
LIMITS OF HEREDITARY CONTROL 863
and from which fine hairs extend. The plate also has two other marginal hair pits containing hairs, one situated well toward each corner of the base; but these emerge from beneath the scute, and consequently the margin of the latter shows no corresponding emargination, as in the case of the more centrally located pits.
The secondary scutes are also more or less wedge-shaped, but, unlike the primary ones, have their bases directed toward the anterior margin of the band (fig. 20) . In brief, their bases form the anterior limit of the exposed part of the band. The apex of the wedge is blunt and forms a small part of the posterior margin of the band. Through the median axis of the scute a faint groove extends from the middle of its base to a point near the apex. Upon the removal of the scute it is found that the groove is due to the depression of the suture between the two adjacent bony plates, directly above which the scute is situated. The secondary scutes are clearly composite structures; that is, they have been formed during the process of their evolution by the union of some three or four elemental units of regions such as are seen in the scapular and pelvic shields, which exhibit conditions more primitive than that of the bands.
In addition to the four hair-pits already alluded to, several others belonging to the associated hair group appear on the upper surface of the plate. The most prominent of these is a row of six or seven extending along each side of the primary scute marking. These pits possess very fine hairs which often extend above the surface of the band, especially in young animals. There are also faint indications of other hair-pits, both on the convex area of the plate as well as on its anterior unexposed third, but it is entirely beyond the scope of this paper to enter into a detailed description of them. It is sufficient here merely to note that each plate corresponds to a rather distinct hair area of other mammals, and consequently has a definite number of hairs included within its limits.
It will be evident from the foregoing account that for each primary scute there is a corresponding plate in which is imbedded a definite group of hairs. A count of the primary scutes will therefore also give a count of the plates and of the hair groups.
864 H. H. NEWMAN AND J. THOMAS PATTERSON
This whole complex will be, for purposes of brevity, designated the ' scute, ' because the scute is the index of all of the elements entering into the complex.
It has been suggested above that the bands have evolved from a more primitive condition of the integument — one in which the bony elements were not necessarily arranged into definite rows. This becomes obvious when one studies such adjacent parts as the scapular and pelvic shields. In each of these regions the general arrangement is much the same, but the transitional condition is more clearly brought out in the pelvic shield, and we shall therefore confine our account to this part.
In the central part of the pelvic shield the bony plates are hexagonal in outline, and are so closely crowded together that a solid bony structure is formed (fig. 16). At best the plates can only be said to be imperfectly arranged into rows. Toward the anterior margin of the shield, however, the serial arrangement into rows becomes more evident, and all of the plates show a distinct tendency to elongate in the antero-posterior direction. In the extreme anterior margin of the shield, or the part corresponding to a tenth band, they become distinctly oblong and greatly resemble those of the true bands. In many of them, however, one can still detect their hexagonal shape, although the anterior and posterior ends show but faint indications of their doublesided nature. Even in the last of the true bands, the ninth, the posterior end of many of the plates is still in the form of an obtuse angle.
On the upper surface of this same region of the armor the scutes show corresponding transitional conditions. The primary scutes are here slightly elevated above the more numerous secondaries and have their posterior ends capped with small white or unpigmented areas, giving to the entire pelvic shield a distinctly pebbled effect (fig. 21). As one passes forward on the shield there is noted a gradual change in the primaries from small polygonal areas to those with characteristic wedge-shaped outlines. This is particularly noticeable on the lateral aspects of the shield.
In the typical regions of the pelvic shield the secondary scutes are more numerous than in the bands. In place of the single
LIMITS OF HEREDITARY CONTROL 865
secondary we have here some four or five elemental units. In the neighborhood of the bands these gradually fuse together, and in the region immediately adjacent to the ninth band are usually typified by two pieces, one, a regular trapezoid in shape, forming the anterior part, and the other, triangular in outline, abutting against this posteriorly. In the true bands these two pieces fuse to form the characteristic secondary scute.
A great deal of interesting data might be given concerning the much more primitive condition of the integumentary elements as seen on the ventral side of the animal, but it must be sufficient merely to suggest one or two of their more salient features. On the belly the horny structures only are present, and these are associated with a group of five or six hairs, or even more. On the legs some few of the larger horny elements, especially those which have a tendency to be arranged into definite rows, are underlaid with true bony plates. Evidently the hair group is the most primitive element of the complex, and in connection with these elements have grown up the scutes and plates.
MERISTIC VARIATION OF THE ELEMENTS OF THE NINE BANDS OF
ARMOR
A. Variation in the banded region as a whole
It is our purpose to present in this section only so much of the results of our studies of the variability of the species as appears to be prerequisite for an understanding of the phenomenon of fraternal correlation. A concise tabulation of the distribution of the variates and a determination of the principal variation constants should serve the purpose in view as well as would a more detailed account.
That the characters dealt with show a high degree of variability is obvious if one examine the array exhibited in table 2. The number of scutes in the banded region vary all the way from 517 to 625, a range of nearly 20 per cent. In connection with our studies on fraternal correlation it will be of value to bear in mind this high species range of variability. In table 2 is presented the array of individuals investigated, comprising 508 adults and shells.
TABLE 2
Showing distribution of frequency of scutes in the banded region of 508 adults
NUMBER OF SCUTES
FREQUENCY
NUMBER OF
SCUTES
FREQUENCY
NUMBER OF
SCUTES
FREQUENCY
517
1
554
8
591
1
518
555
9
592
519
1
556
11
593
1
520
1
557
10
594
521
2
558
21
595
522
1
559
13
596
1
523
1
560
10
597
524
561
18
598
525
1
562
15
599
526
563
19
600
627
2
564
6
601
528
5
565
15
602
529
566
12
603
2
530
567
10
604
531
1
! 568
11
605
532
3
- 569
8
606
533
3
570
12
607
1
534
7
571
12
608
535
2
572
8
609
536
2
1 573
7
610
537
3
574
8
611
538
7
575
4
612
539
6
576
8
613
540
5
577
7
614
541
10
578
5
615
542
5
579
7
616
543
6
580
5
617
544
16
581
3
618
545
5
582
2
619
546
7
583
3
620
547
18
584
3
621
548
14
585
3
622
549
17
586
623
550
13
587
4
624
551
7
588
625
1
552
12
589
3
553
15
590
2
LIMITS OF HEREDITARY CONTROL 867
In grouping the variates shown in the table for purposes of seriation we have decided to fix the group size on as logical a basis as possible. The average range of variability of the sets of quadruplets is eight scutes. We shall therefore seriate the array in groups of eight for reasons which will become clear later. The polygon of variation obtained by this grouping represents a close
Fig. 1 Polygon of variation for the total number of scutes in the nine bands,
as determined from a seriation of 508 individuals. Class range = 8 scutes. The
solid line represents the observational and the broken line the theoretical normal
curve. In this and the succeeding figure the abscissae refer to number of scutes
and the ordinates to the numbei of individuals.
approximation to a normal array as may be seen from a comparison of the observed and the calculated curve (fig. 1).
The three important constants of the frequency polygon practically coincide; the median being 558; the mean, 558.2; and the mode, calculated from these two constants, 557.6. On the ex
868 H. H. NEWMAN AND J. THOMAS PATTERSON
treme right of the polygon will be noted several extreme variates, so far separated from the others as to be examples of discontinuous variation. The rejection of these extreme variates would render the three constants, mean, mode and median, practically identical, but it would seem inadvisable to take any liberties with the data. Rather we would prefer to accept a close approximation to the normal curve of frequency as an indication that we are dealing with a case of chance variation, little if at all disturbed by complicating factors.
From the above array have been calculated the standard deviation, 14.89 ±0.31 scutes; and the coefficient of variability, 2.685 ±0.32 percent.
In addition to these facts we are also led by analogy to infer that we have in the scutes of the ^nine bands of armor a variant which is inherited in the blended fashion. This inference is borne out by the evidence derived from an examination of the mothers of the various sets of quadruplets (table 6). Of the fathers we unfortunately know nothing.
B. Comparative variability of males and females
While the main mass of our statistical data came from an examination of baskets made from the dried and shaped shells, a considerable number of individuals were identified as to sex. The arrays dealt with in this section consist of the scute counts of animals shipped to us alive, and of the advanced foetuses in our collection. The larger number of females is explained by the fact that our first interest was centered on the facts of development and hence we ordered only pregnant females. The number of individuals is, however, sufficiently large to furnish a basis of comparison between the sexes. Table 3 indicates the frequency distribution of the variates of the two sexes. It will be seen at a glance that the array of males here presented is decidedly more variable than that of the females. Part of the disparity may be due to the occurrence of one set of foetuses, all of which have a scute count of over 600; but even if we arbitrarily exclude this aberrant set we do not materially reduce the variability of the
LIMITS OF HEREDITARY CONTROL 869
male array The mean of the female array will be seen closely to approximate that of the whole collection, while that of the males is several scutes higher. It is our impression that a larger collection of males would place the mean at about 559 or 560, which would indicate a slightly higher center of frequency for males than for females. Although the mode of the two sexes
Showing distribution of frequency of scutes in the banded region of 146 females
and 81 males
Females (146 Individuals)
Males (81 Individuals)
NUMBER OF
SCUTES
FREQUENCY
NUMBER OF
SCUTES
FREQUENCY
NUMBER OF
SCUTES
FREQUENCY
NUMBER OF
SCUTES
FREQUENCY
517
1
559
5
520
1
567
1
520
1
560
2
527
1
568
3
527
1
561
5
533
2
569
2
528
2
562
2
535
1
570
3
532
1
563
3
540
1
571
3
534
2
564
4
541
1
572
3
539
2
565
5
542
1
574
1
540
1
566
4
544
5
577
2
541
1
567
2
547
3
579
2
542
1
568
1
548
4
580
1
543
3
569
2
550
2
587
1
544
2
571
2
551
2
591
1
545
4
572
3
552
1
606
3
546
5
573
4
553
5
607
1
548
10
574
3
554
•1
621
1
549
2
575
1
555
4
550
4
577
1
558
1
551
4
578
3
559
3
552
4
581
2
560
1
553
9
582
1
561
2
554
3
583
1
562
2
555
6
585
1
563
4
556
5
589
1
564
1
557
5
590
1
565
2
558
6
596
1
566
1
Mean = 557.2
Mean = 561.3 scutes
Standard deviation = 13.35 scutes
Standard deviation = 17.71
scutes
Coefficient of variation = 2.39 per
Coefficient of variation = [
5.15 per
cent
cent
870
H. H. NEWMAN AND J. THOMAS PATTERSON
probably differs onlj^ to a slight extent, yet the standard deviation and coefficient of variation of the males are decidedly higher than those of the females, a condition that accords with our previous knowledge of the comparative variability of the sexes. The possible significance of this variational dimorphism will receive attention in the discussion of the hereditary control of sex.
C. Variation in the individual hands
In view of our studies of fraternal correlation among the foetuses it appears to be necessary to determine the scute variability, not only of the banded region, but also that of each of the individual bands. We will thus be able to determine whether there is a greater or less variability in individual bands than in the whole region under investigation; and in addition we can compare the variability of any one band with that of another. In table 4 is
Showing distribution of frequency of scutes in the individual bajids of 508 individuals
FREQUENCIES PER BAND
NUMBER OF SCUTES
1
3
13 26 33 65 102 86 83 44 30 12
5
4
I
I
j
I 3
1
5
7
16 58 71 109 104 56 49 15 11 2 2 2
4
18
27
53
70
116
80
70
47
15
4
2
1
1
I 7
3
5
15
40
49
85
102
82
66
34
21
4
2
1
6 10 18 53 96 89 84 68 37 25 14 2 2 3 1
2
6
9
21
63
65
99
86
69
47
28
6
2
4
LIMITS OF HEREDITARY CONTROL
871
given in convenient form the data, band for band. The arrays for the individual bands are considerably more satisfactory for purposes of seriation than that for the banded region as a whole; for the number of classes is comparatively small and consequently the number of counts sufficient to give a smooth cur.ve without any grouping of variates. As a sample of the type of curve derived from a seriation of the variates of the individual bands let us take that of band 1 (fig. 2),
00'
90
■ _,
V
80.
n
\
70.
r/
\
60
\
50
/••■■'
\
40
/
\
30
- /
\
20
\
10
yy'
^K..
55 5d 57 58 59 60 61 62 63 64 65 66 67 68 69 70 7!
Fig. 2 Polygon of variation for the total number of scutes in the first band, as determined from a seriation of 508 individuals. Class range = 8 scutes. The solid line represents the observational and the broken line the theoretical normal curve.
It will be noted that the mean, mode and median coincide and that the observed and the calculated curve are almost identical. The curves for the other bands are of the same type as that shown for band 1. These observations would seem to indicate clearly enough that in the distribution of the elements of the integument into bands we have a pure chance process controlled by fairly simple mechanical laws.
OP MORPHOLOGY, VOL. 22, NO. 4
872
H. H. NEWMAN AND J. THOMAS PATTERSON
In order that it may be clear that the sample of the species presented by the sets of foetuses studied is fairly representative, it would be well to compare the variation constants of the foetuses with those of the large sample of the species here dealt with. Table 5 furnishes the means of comparison. This table reveals to the student of biometry a number of interesting conditions not strictly pertinent to the present inquiry, but doubtless worth noting briefly :
1. With a few minor exceptions, all of which fall within the limits of probable error, there is a gradual decrease of the mean, mode and median from the first to the fourth band and then a gradual increase in these constants from the fifth to the ninth band.
2. The absolute variability, as indicated bj^ the coefficient of variation is greater in each of the bands than in whole banded region.
3. The proportional range of variability is likewise greater for the individual bands than for the banded region as a whole.
4. The standard deviation may be considered for convenience to amount to two scutes. This will be worth remembering when we come to study the band correlation among the foetuses of the various sets.
For the convenience of the reader it seems well to consider the subject of hereditary' control, as it is exhibited in connection with
Showing the vari
TABLE 5
T.tion constants of the individual hands of 508 adults
BANDS
MEAN
MEDIAN
MODE
BTANDABD
DEVIATION
COEFFICIENT OF
VAEIATIOK
RANGE IN
PER CENT
OF MEAN
1
62 ±0.066
62
62.0
2.223±0.047
3.o9±0.038
24.19
2
60. 55 ±0.065
61
61.9
2.217±0.046
3.66±0.038
23.3
3
60.44±0.063
60
61.76
2.12±0.045
3.51 ±0.037
1 26.46
4
60.23 ±0.062
60
60.92
2.08±0.044
3.45±0.035
j 24.9
5
61. 16 ±0.063
61
61.64
2.11±0.045
3. 46 ±0.035
24.54
6
61. 85 ±0.066
62
62.3
2. 23 ±0.047
3. 61 ±0.039
' 29.1
7
63. 09 ±0.066
63
62.82
2.22±0.O47
3.52±0.037
20.6
8
64.43 ±0.068
64
63.14
2.24±0.048
3.49±0.036
26.39
9
64.44±0.068
64
63.12
2.2 ±0.046
3. 57 ±0.038
23.27
LIMITS OF HEREDITARY CONTROL 873
the normal variability of the scutes in the banded region and in the individual bands, immediately following the variation data for the species, although it might appear to be more logical to complete the account of the conditions in the species before attacking that in the foetus sets.
D. Fraternal correlation in the handed region as a whole; an index of the liinits of hereditary control
Having presented the facts of variation for the species, we are now in a position to investigate the conditions exhibited by the various 'fraternities,' which consist normally of 'identical quadruplets,' i.e., of four individuals of the same sex enclosed in a common chorionic vesicle. We have in our possession at the present time about thirty sets of foetuses in a stage of development sufficiently advanced to permit of the determination of sex and of the accurate enumeration of scutes. Of these some few are of especial value on account of their more or less atypical conditions; a few others are characterized by an atypical number of foetuses. There are, however, ten sets of each sex in which the number of foetuses, the arrangement of bands, etc., is sufficiently normal to permit of statistical treatment. These twenty sets furnish the material for the present study of correlation and the limits of hereditary control. The enumeration of scutes may be relied upon to be accurate, since they were all counted independently by both of the writers and all points of discrepancy carefully examined and settled according to mutual judgment. A full tabulation of the scute counts of the twenty fraternities is presented in table 6.
As in our previous papers the four foetuses are numbered as follows :
Pair A / I. The lower individual of the right hand pair \ II. The upper individual of the right hand pair
Pair B /ill. The upper individual of the left hand pair \ IV. The lower individual of the left hand pair
Since each of these sets of foetuses is derived from a single fertilized egg it should be possible to determine exactly to what
874
H. H. NEWMAN AND J. THOMAS PATTERSON
TABLE 6 A — Female sets
FOETUS AND MOTHER*
liv
Mother.
Mother
<. A rudimentary emI bryo lies between m
I and IV
62 60 60
61 62 59
60 59 , 59
59 61 60
60 61 58
63 60 59
64 61 58
61 62 59
62
62 60
63 60
61 60 63 59
62 58
1 62 60
II 61 60
III 61 61
.IV 62 62
Mother 59 58
63 62 61
63 61 63
63 62 62
63 63 60
65 64 60
60
62 64 62 59 62 61 61 63 59 61
59
57 58 63
59 59 60
58 60 60
60 59 60
61 61 61 59 61 62 61 62
62 61
59 60 58 60
60 59
61
60 60 59 61 57
61 60 58 61 63 62 61 57 61 60 61 60 59 62 61 59 58 61 59 63 61 60 61
57 57 61 62 60 60
62 64
61 64
62 63 62 64 60 64 60.5 64
66 63 557
65 64 556
63 64 553
64 62 551
63 65 67 64
64 65
65 67 65 65
64 66
65 68
63 64
66 64
64 65
66 65
67 63
61
62
60
61
61
62
62
64
62
63
64
65
61
63
66 551 63 548 63 554 62 558 61 553 553 558 558
66
66
62
62
66
62
63
65
63
62
63
62
566
565
562
565
557
561
564
567
565
574
554
554.5
550
548
548
561
565
564
544
549
546
548
517
543
544 5
545
548
553
•In some of our first collections the mothers were not kept, cannot be given.
and hence the scute counts on these
LIMITS OF HEREDITARY CONTROL
875
TABLE 6 B — Male sets
FOETUS AND
MOTHliR
I j 63 58
|ix j 61 I 59
liii J 60 I 59
!iv I 59 I 58
JMother I 58 56
fli.. .
62
63
j 60
' 61
64
! 61 63 61 61
59 60 63 62
64
64 61 63
64 62 64 64 63
69 68 70 66 63
62 62 61 61 63
64
n.. ..:.::::::::::::::
59
)iiii
59
[jiv
[Mother
J I
62
64
In
61
„i
62
Iiv
61
•
■|i
58
,111
58
fi
62
|„
62
!ni...
61
IV. .
64
jMother
60
!xn::;;:::;::::::;;;::;
i: IMother
1
64
60
59
68
1
II
64
<li„
69
['iv....
67
iMother
60
\'i ..
61
138
n.. ..:::::::;:::::;::;
[iv.::::::;::::::::::::
Mother
62
I 5 I «
64 61 j 63 I 65
64 i 64 [ 62 [ 65
64 I 62 ! 63 64
61 63 I 62 I 65
63 i 65 65
61 j 63 I 64
63 65 65
64 64 65
58
56 j 56 58 I 61
57 1 57
62
61
62
61
61
62
57
60
61
56
60
56
58
58
63
62
64
63
61
58
61
62
, 59
60 j
59
59
59 ;
60
60
60
62
60
60
66
65
66
63
68
66
68
65.5
63 j
62
59
60
58
58
60 :
59 ;
60
61
61 [
59
67 65 571
67 65 571
66 65 570
66 I 65 570
57 60
60 j 61
61 I 59
58 ! 60 60 ! 59
63
61
65
64
63
63
j 62
65
65
62
63
63
62
64
64
57
62
64
60
62
63
62
61
66 !
62
65
68 i
63
65
67
63
63
65
62
65
68
59
63
65 ;
60
62
63
62
62
64
60
63
65
62
62
64
60
60
63 j
67
67
68
69
67
71
67
72
71
70
68
68
65
64
66 1
60
64
63
62
61
63
62
63
62
61
63
64
62
64
63
533 527 535 520
527
60 I 63 { 63 555
62 j 65 67 555 60 [ 65 I 67 ' 548 60 ' 63 j 63 .548
63 ; 67 65 566
66
565
66
565
65
565
63
557
63 544
63 542
66 547
66 548
67 572 66 579 66 563
65 570
66 555
551 552 560 559 550
71 606
70 606
71 621 70 I 607.5 67 I 573
63 553
63 j 550
64 554 64 551 66 562
876 H. H. NEWMAN AND J. THOMAS PATTERSON
extent the definitive number of scutes is determined at the time of fertihzation ; for presumably, in so far as the four individuals of a set are alike, this similarity was determined before they became separate individuals, and, in so far as they differ, their differences are due to epigenetic factors operating after the separation. Should they prove to be exactly alike in the number and arrangement of scutes we would be warranted in claiming that hereditary control was absolute; if we do not find them exactly alike we may claim that hereditary control is not absolute but only partial. As an index of the strength of hereditary control we have decided to employ Galton's coefficient of correlation, a constant indicating exactly what per cent of complete correlation or of absolute hereditary control exists.
A short method for determining the coefficient of correlation, and one which seems especially devised for cases like the one in hand, was described by Harris ('09). This method appears to have been elaborated by Professor Pearson for use in cases of symmetrical correlation tables in which both variates have the same mean and standard deviation. The correlation table is made symmetrical by using each individual of each set first as a 'subject' and then as a 'relative.' In the case of our twenty sets of foetuses such a table requires over one hundred vertical and horizontal columns and would not be suitable for reproduction here. As a matter of fact it is scarcely necessary to make a correlation table in order to derive the desired constant. One need only determine the positive differences between the subject and relative classes and their frequency. In the present case we find the positive differences and arrange them as follows:
Differences: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Frequency: 26 19 10 16 13 10 9 10 7 4 1 1 1 1 3 2
Multiplying the square of each difference by its frequency and adding the products, we obtain the sum of the squares of the differences [Sv"^]. Now since the negative differences are the same as the positive we may double the above sum and divide by the number of cases, 240, and thus obtain the standard deviation of the difference squared {(tv^), which is the only constant
LIMITS OF HEREDITARY CONTROL 877
not already determined that is needed for our calculation. The formula used is as follows:
^ 1 crV'
2 ^
/ 1 33.5 ,
Substituting we get: r = i 1 - — 9^62 ' "^ 0.9348.
We can use this coefficient of correlation as an index of the strength of hereditary control and can say concretely that, so far as the total number of scutes in the banded region is concerned, the individual is predetermined within certain narrow limits, or up to 93.48 per cent of complete determination. Beyond that point the variations in the number of scutes are due to differences in epigenetic factors, whose nature we do not pretend to understand.
No such close correlation as this has been determined for any of the ordinary blood relationships. The closest of all blood relationships, namely the fraternal, is represented by a correlation constant of only 0.4, a fact familiar to all who have read their Galton. Even homotypical parts of the same individual are correlated only a little more closely than are brethren, a fact which leads Pearson to conclude that the fraternal relation is only a special case of homotyposis. In brief there appears to be no other inter-individualistic relationship so close as that which we have found to exist between the individuals of our sets of quadruplets. If we desire to find correlations comparable in closeness with that determined in our material, we must seek them among the closest of intra-individualistic relationships, such as that existing between the antimetrically paired organs of the same individual. As an example of such constants we may cite the correlation coefficient between the right and the left sides of the carapace of Gelasimus, which is 0.947, or that between the lengths of right and left meropodite of the first walking leg of the same species, which is 0.918. These constants are obviously of the same grade as that determined for the scutes of our quadruplets. From these facts we are doubtless justified in concluding that the four individuals of each of our sets is morphologically the equivalent of a
878 H. H. NEWMAN AND J. THOMAS PATTERSON
single individual and that we are dealing with a special case of intra-individual variation.
Were the phenomenon of specific polyembryony in need of any further support, the discovery that the degree of correlation among the foetuses of a set is the equivalent of that found to exist between the right and left sides of the same individual, would in itself constitute a demonstration of its validity.
Showing the variability of seven -pairs of duplicate twins in terms of the coefficient of
variation
DFSIGXATION OF TWINS I COEFFICIENT OF VARIATION
Weismann' s 1 . 12
Vernon's i 0.22
0.34 1.43
Vei'non's ii. Wilder' s i.. Wilder' s n. Wilder' s iii
1.65 1.52
Wilder's iv i 1 . 17
Mean
In order to be able to institute a comparison between the strengths of hereditary control exhibited by armadillo quadruplets and human duplicate twins it becomes necessary to employsome method which does, not involve the use of the coefficient of correlation, for the twin data is too scattering and diversified for our purposes. The usual method employed by the various writers for indicating the degree of resemblance between twins is that mentioned in the introduction, namely the per cent difference method, but this can be used only in the comparison of pairs. A better method for our purposes involves the determination of the coefficient of variation for each pair or set and an averaging of these constants. Table 7 gives the data on the seven pairs of human identical twins referred to above. In table 8 are recorded the coefficients of variation deterndned for the twenty sets of foetuses. It will be noted that the degree of resemblance among the quadruplets is, on the average, nearly twice as close as that of the twins. This is perhaps no more than we should expect, even
LIMITS OF HEREDITARY CONTROL
'9
if we were certain of the origin of each pair of twins, for the twins have doubtless been much modified by post-natal environmental experience.
Incidentallj' it may be noted that the male sets exhibit a much higher absolute variability than the females sets, a fact which will be brought into discussion in a subsequent connection.
Perhaps a more equitable comparison between human twins and the armadillo sets w^ould be instituted if we were to consider the quadruplet set as consisting of two pairs of twins. According to this method of comparison the difference between the two species is greatl}^ increased, for the average of the forty pairs of armadillos is only 0.27 per cent, while that of the seven pairs of human twins is 1.62 per cent.
In concluding the present account of the strength of hereditary control as it appears to be exercised in the case of the total number of scutes in the banded region, it should be mentioned that further investigations now in progress, dealing with other regions of the armor and with certain dimensional variates, indicate
Showing the standard deviation of each of the sets of foetuses and iiidicating the
comparative variability of male and female sets with respect to the total
number of scutes in the nine bands.
Female sets
Male sets
.SET
standard
deviation
(in scutes)
coefficient •
of variation
(in per cent)
SET
standard
deviation
(in sctttes)
COEFFICIENT
OF VARIATION
(IN PER CENT)
2
2.38
0.43
1
2.48
0.43
23
3.69
0.66
4
0.5
0.08
95
2.5
0.45
97.'
5.86
1.10
98
1.5
2.16
0.26
0.38
112
116
3.5
3.46
63
99
0.61
119
2.59
0.46
117
2.38
0.43
121
2.86
0.50
120
5.70
0.99
122
1.92
•0.35
134
4.02
0.72
123
1.87
0.34
135
6.36
1.04
127
5.85
1.06
138
1.56
0.28
Total
27.32
4.89
35.82
6.31
Mean
2.732
0.489
3.582
0.631
880 H. H. NEWMAN AND J. THOMAS PATTERSON
that the degree of correlation found to exist in the banded region has its parallel in other regions ; in fact some of the results already obtained seem to indicate the existence of higher degrees of correlation than any yet determined.
E. Fraternal correlation in the individual hands
In one of our previous papers ('10) we had occasion to refer to the subject of pairing in the following words:
In this connection it should be mentioned that even where there is exact resemblance between the individuals of a pair in the total number of scutes in the nine bands of armor, there is no perfect correspondence with respect to individual rows. The resemblance in total numbers of scutes is, however, a matter of more importance than the exact manner of their arrangement into rows, which is a secondary process.
Further investigation has thrown light on this situation and we are now in a position to make a more satisfactory statement of the conditions referred to. By the application of statistical methods we have been able to compare the strength of hereditary control exercised over the total scute number and that over the arrangement of these scutes into bands. In order to do this it has been necessary to determine the correlation coefficient of each of the nine bands, using the same method which was applied to the whole banded region. The method pursued is illustrated in the case of band 1 (table 9) . The same method of procedure was carried out for each of the nine bands and the data and results are seen in concise tabular form in table 10.
Although the average coefficient of correlation for the individual bands is high as compared with that found for any relation other than an intra-individualistic one, it is decidedly low as compared with that shown to exist in the case of the total number of scutes in the whole region. This would seem to indicate that there is here a much wider scope for the operation of epigenetic factors. Evidently the alignment of scutes to produce the bands is, to a large extent, a mechanical process, involving a certain amount of shifting due to pressure, etc. In the earliest stages of scute formation it is probable that the primordia of these elements are arranged somewhat after the fashion seen in the abdominal
LIMITS OF HEREDITARY CONTROL
881
TABLE 9
Correlatioji and difference tables for the 20 fraternities, for band 1 Mean = 62.16
Standard deviation = 2.009 scutes Coefficient of variation = 3.23 per cent
Classes y
59
60 61 62
63 64 1 65
66
67
68
69
70
Total
Deviation rel.
class from mean
-2.16
-1.16-0.16
0.84
1.84
2.84
3.84
4.84
5.84
6.84
7.84
Classes
of X
Deviation from
mean
2
4
3
3
2
6
9
2
4
1
4
9
22
14
14
5
1
3
2
14
14
7
4
3
4
14
7 16 5
1
5
4
5
10
2
1
2
1 1 1
1
1 1
1
1 1
1
1 1
59
60
61
62
63
64
65
66
67
68
69
70
-3.16
-2.16
-1.16
-0.16
0.84
1.84
2M
3.84,
4.84
5.84
6.84
7.84
12
24
69
44
49
27
6
3
3
3
Total.
12
24
69 44
49
27
6
3| 3
3
240
V = positive diflferenoes of x ar
dy
1
2
3
4
5
6
7
8
9
10
11
2
4
3
3
6
9
2
4
1
22
14
14
6
1
.0
S(v)2
14
3
39
16
5
112
10
2
117
80
1
1
1
1
25
5
373
1
1
1
28
13
•
39
68
882
H. H. NEWMAN AND J. THOMAS PATTERSON
Showing coefficients of correlation and other constants in each of the nine bands of
armor
BAND
ME.\X
1
ffX2
<J-V2
COEFFICIENT OF
COHBELATION
1
2
3
4
62.16
(".1.03
60.05
GO 45
4.090
4.712
5.887
3.972
4.828 1
4.970
5.735 1
5 . 155 (
5.875
1
3.10
3.94
3.80
3.93
3.41
2.52
3.30
3.75
3.66
0.6201
0.5820
0.6769
5051
5
CO 97
6469
6
7
8
61.73
62.70
64 40
0.7458
0.7123
6363
9
fid 9.9
6880
Mean
6459
region, where the rows are far from straight and where it is not always possible to assign scutes to their respective rows. At such a stage in the development of the banded region we may assume that the ultimate position of certain scutes is determined only in a general waj^, and whether or not they come to be aligned with one or the other of two adjacent bands is determined by factors beyond the limits of hereditary control. Considering that the scutes are not only determined rather sharply for the whole region, but that, within this region, their primordia would also be regionally distributed under the influence of hereditaiy control, the coefficient of correlation is no larger than we should expect. But the correlation is so low as to preclude the possibility of any strong hereditarj^ control being exercised over the assignment of scutes to particular bands. This conclusion is in close harmony with that expressed by Wilder in connection with his studies of friction ridge patterns in human twins, and which is quoted in the introduction to the present paper. We might well ask with him : "Have we then arrived at the limit of hereditary control of the predetermining mechanism beyond ivhich mechanical laws are alone operative?" Perhaps we are in a position to give with somewhat greater assurance than was Wilder, an affirmative answer to the question.
LIMITS OF HEREDITARY CONTROL 883
ATYPICAL VARL\TION IN THE INDIVIDUAL ELEMENTS AND IN THE bands'
A. Scute 'abnormalities' in the species; their distribution and frequency
{1). Double scutes. In the individual scutes and in the bands certain so-called 'abnormalities' frequently occur, and while these may be but the expression of teratological phenomena, yet we believe that some of them at least are the result of more deeply seated factors and have a real phylogenetic significance, and consequently are worthy of consideration in a paper on variation and heredity. We shall speak of them as atypical variations.
The atypical variation of the scutes is expressed in at least three different ways, one of the most frequent of which we shall designate as the 'double scute.' In this type two contiguous primary scutes are fused along their adjacent sides, and the double structure thus formed has five or six hairs at the free or posterior end (figs. 6a, la, 8a). Evidently there is a suppression of two or three hairs consequent upon this union. All stages of fusion have been observed in our material, and in the figures just referred to is seen a series of three, taken from one specimen showing different degrees of the process. The nature of the origin of these double scutes is made clear when the region affected is examined from the under side, where the double bony element is always clearly expressed (figs. 66, 76, 86). Sometimes the bony elements are of equal size and occupy the full width of the band, but more frequently one of the plates is greatly reduced, as if there had been a tendency to crowd it out (fig. 86).
The double scutes occur in about ten per cent of the animals — or to be more exact, in a total of 516 individuals examined for this particular point 51 showed double scutes. Ordinarily there is but one of these to an animal, but four exceptions to this have been found, as follows: one with four double scutes, one with three, and two with two each.
For the purpose of locating exactly the various scutes, we have chosen to begin their enumeration always on the left margin of the band. Thus scute ' 10' of band 5 would be the tenth scute counting from the left margin of this band. Since it is obvious that the double scute is the product of what was originally two distinct elements, it has been given the value of two in our collected data, though we designate it by the number corresponding to its first or left-hand element. This point can be made clear by a reference to the numbering in figs. 6 to 10.
The double scutes are generally distributed over the bands as can be seen in table 11. There is, however, a tendency for them
TABLE 11 Showing distribution of ^double scutes'
DISTRIBUTION IN BANDS
LATERAL DISTRIBUTION
Left Middle
Right
1
2, 4, 7, 11,23, 24, 31, 42
10, 10, 14, 15, 27, 29, 57
12, 25, 27, 29, 41, 53, 57, 59
3, 19, 19, 23, 54, 54, 55, 60 2, 2, 2, 32, 44, 53, 57
2, 5, 6, 25, 27, 28, 56, 57, 58
3, 12, 24, 50 2, 15, 36 36,65
1
4 3
1
2
4
1
3 3 3 2
2
2 3
1 1 3 1 1 1
1
3
4
4
4
5
3
6
3
7
1
8
9
1
Totals
56 cases
22
16
18
Note, in the left-hand half of the table the numbers in the horizontal rows lying opposite the bands designate the first elements upon which the double scutes fall. Thus in the first band eight animals had these scutes, falling on elements 2, 4, 7, etc., respectively.
to be localized in the anterior two-thirds of the banded region, or in bands one to six. Perhaps of more importance is their lateral distribution, which can be showm by dividing each band into three equal parts — two of which represent the left and right thirds and one the middle third — and determining the number of double scutes falling within each of these general divisions. The data thus collected is shown in the right half of the table, from which it is clear that there is a tendency for these scutes to be localized toward the margins of the armor. They fall most frequently
LIMITS OF HEREDITARY CONTROL 885
on scute '2' (six cases) on the left, and scute '57' (four cases) on the right. However, it should be pointed out that there are but four cases of exact coincidence (one involving three specimens) by which we mean double scutes falling at the same point in the same band in two or more individuals.
(2). Incomplete scutes. The second type of atypical scute variation is just the opposite to that of the preceding, in that it is probably the product of a spHtting of what was originally a single scute, and results in the formation of an incomplete element. This type appears so rarely that not very much importance can be attached to it. In the 516 specimens examined only three cases of incomplete scutes were found. These occur as follows : Specimen one, a female, between scutes 52 and 54 of band 6 (the small plate is counted as a whole one) ; specimen two, a shell, between scutes 14 and 16 of band 8; specimen three, also a shell, between scutes 56 and 58 of band 5. The first and second of these are sketched in figs. 9 and 10, respectively. The most interesting point brought out is that in each case one of the scutes lying adjacent to the incomplete one has but three hairs instead of the typical four, while the small scute has in each case but a single hair associated with it. The presence of three hairs in an adjacent scute would not alone be cogent proof that the primordium of such a scute had spHt to give rise to the incomplete element, as will be evident when the next section of the paper is reached. However, all of the other relations in these specimens point toward a fission process.
{3). Three-hair type of scute. The third type of atypical variation is perhaps the most fundamental of all, because it involves a distinct change in the morphological unit of the scute. In this type the bony plate underlying the scute has but three hairs at its posterior end instead of the customary four. It was first detected in a shell which had an unusually high number of scutes in the bands; in fact the one representing the most extreme case of high numbers that we have so far found. When the shell is first observed the attention is at once attracted by the narrowness of the scutes, and upon examining them the interesting fact is revealed that the majority show the three-hair type (fig. 22).
We have studied this shell in considerable detail and have compiled the data secured in table 12. Here it will be noted that 391 of the 625 scutes have but three hairs each, and furthermore that there is a strong tendency for the 3's to be distributed on the lateral parts of the shield. However, the most strikingfact revealed in the table is indicated in the fourth column, where the total number of hairs for each band is given. Bands 2 and 8, which show the extremes in scute variation, have exactly the same number of hairs, while the two extremes in hair variation, bands 7 and 9, are but six hairs apart in their totals.
TABLE 12 Shell 2, with 625 scutes showing distribution of S-hair and 4-hair scutes
BANDS
DISTRIBUTION OF 3'S AND 4'S IN BANDS
LATERAL DISTRIBUTION OF 3'8
3-Halr Type
4-Hair Type
Number of
Hairs
Left Middle
Right
1
^:::;:::;::;
4
5
6
7
8
49
34
37
38
38 49 41 58
22
33
31
30
30
22
27
235
234
235
234
234
235
231
234
19
14
18
21
19
23
18
91
16
12
5
3
6
4
7
ifi
14
8
14
14
13
22
16
9.1
9
47 24
237 23 8 16
Totals
391
234
2109
176
77 1 138
1
The question naturally arises : Can the high number of scutes exhibited by this specimen be accounted for by the fact that so many of them are of the three-hair type? Undoubtedly it can, but the affirmative answer necessitates the assumption that the integumentary primordium out of which each band arises contains a definite number of hair follicles, which are later distributed into groups of 3's and 4's according to the propensities of the formative scutes. It is evident upon this assumption that a tendency for the scutes to form about groups of four hairs would result in the production of a shell having a low number of scutes, while
a tendency to form about groups of three would produce just the
opposite result. A reference to table 13, in which the total counts
on five specimens are recorded, indicates that, although the animals may vary considerably in the total number of scutes, yet
TABLE 13 Showing -pro-portion of 3-hair and 4-hair scutes in a small sample of the species
SPECIMEN
NUMBER
OF
SCUTES
3-HAIR
TYPE
4-HAIR
TYPE
NUMBER
OP
HAIRS
LEFT
MIDDLE
RIGHT
Shell 2
9221
9223
9217
9204
625
577
545
534
529
391
159
77 19 26
234
418
468
515
503
2109
2149
2103
2110
2091
'11
33
13
77
29
9
2
138
60
35
10
13
there is very little difference in their total hair counts. In the two extreme cases, shell 2 and female 204, there is a difference of 96 scutes, but in total hair counts they vary by only 18 points, which is a very inconsiderable difference w^hen one considers that there are between three and four times as many hairs as scutes. Undoubtedly data on a larger number of animals would reveal a much greater variation; but the fact remains that, however variable the species may be with respect to the scutes (and their associated plates), it is comparatively stable with respect to the hairs of a given region. A statistical study of the hairs might be made, but their enumeration is difficult; for, on the one hand, not a few of the adults possess armors from which many of the hairs have been lost by abrasions through contact with rocky dens, and on the other hand, even advanced foetuses do not have hairs sufficiently developed to allow a census of them being taken. For this and other reasons it has seemed best to confine our studies to the variability and heredity of the scutes, and not to attempt a compilation of statistics on hair counts, which at best could be only imperfect. It is perhaps sufficient then merely to have indicated the line along which the evolution of the species is apparently directed. The fact that in all of the regions showing a primitive condition of the scutes, as in the case of the belly,
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 4
888 H. H. NEWMAN AND J. THOMAS PATTERSON
the hair groups are larger than in the more highly differentiated region like that of the bands, is prima facie evidence that here the direction of evolution is from the four-hair to the three-hair type, with a consequent increase in the total number of scutes. These high numbered specimens with their many scutes of the three-hair type are of further interest because they are probably to be regarded as mutations, in the sense that they represent discontinuous variations. This in part can be made clear by an inspection of table 1. At the lower end of the series therein represented the specimens occur with a frequency sufficient to be explained on the basis of an ordinary fluctuating variability, and this is also true of the other end of the series up to the specimen with 596 scutes; but beyond this point specimens are of rare occurrence and consequently are separated by wide gaps. However, neither the rarity of their occurrence nor their wide separation is sufficient to place such specimens in the .category of mutations, because there is always the possibility that a larger number of the species would furnish variates enough to fill up the gaps. If we would therefore explain these as saltations we must look to another source of information, namel}^, to their behavior in inheritance. This topic will receive attention in a succeeding section.
B. Hereditary control in connection with scute ^ abiiormalities^
{1). Double scutes. Attention has been called to the comparative rarity of double scutes in the species and to the fact that in only a very few cases do double elements fall in the same place in the same band. It is rarely the case even that the same general region of a band of more than two specimens shows the peculiarity in question. When, therefore, we find in as many as three out of four members of a set of foetuses a double scute in almost precisely the same locality, we are forced to the conviction that even these elements, which might be defined in the words of Galton as Hhe minutest biological units, ^ are predetermined in the undivided germ cell.
Two sets of foetuses show the conditions most clearly and they are described in detail below:
LIMITS OF HEREDITARY CONTROL 889
Set 121 (Mother without any double scutes)
Foetus I. No double scutes
Foetus II. In band 2, scute 29, double
Foetus III. In band 2, scute 28, double
Foetus IV. In band 3, scute 28, double
It is to be noted that in three of the four individuals the double element occurs practically in the same spot. In view of the fact that, even in perfectly normal sets, the number of scutes in the same row varies widely among the individuals, it would hardly be expected that a predetermined doubling could strike the same numbered scute in any two members of a set. The regional hereditary control in this case is therefore somewhat more accurate than we would expect to find it, in view of the laxity of hereditary control in the matter of the arrangement of scutes into bands. The fact that in one of the foetuses the double scute occurs on the third band, while in two other foetuses the doubling affects the same or contiguous scute of the second band, only goes to bear out the idea that the process of scute alignment is to a large extent a mechanical process, involving the shifting backward or forward of the individual primordia under the influence of pressures. It is only to be expected then, in the light of these considerations, that we should find a case now and then where a readily recognizable element like a double scute should indicate by its position that it may have been shifted from one row to another.
On account of the small extent of its 'abnormalities' this set has been used in the study of correlation (table 6, A) as well as for an example of a peculiar type of band arrangement, (p. 874).
Set 12S {Mother without atypical scutes of any kind)
Foetus I. In the last row of the scapular shield, scute 30 is double; scute 2 of band 2, double; scute 10 of band 3, double
Foetus II. In the last scapular row scute 28, double; an inconfplete scute like that shown in fig. 9, occurs in band 6 between scutes 2 and 3
Foetus III. In the last scapular band scute 14, double; scute 49 of band 6, double
Foetus IV. Scute 31 of band 4, double
In this set the following points are noteworthy:
1. In none of the specimens examined in the statistical study of the frequency and distribution of double scutes was a case of doubling found in the last scapular row; hence the likelihood of the conditions described being due to coincidence is extremely remote.
2. The location of the double scute is evidently not so rigidly defined as in the former case, since the double scute in foetus III is rather far separated in position from that of either of the other foetuses. In the two individuals of the natural pair A, of the left hand side (right in the figure), however, the position of the double scute is almost precisely the same. One can detect the difference in position only by counting the scutes. Careful examination of the two right hand individuals in fig. 24 will make this clear. In the subsequent discussion of the phenomenon of pairing this circumstance will receive further attention.
3. It is very unusual, as was indicated in the general discussion of scute ' abnormalities, ' for an individual to have more than one double scute. When, therefore, three of the four quadruplets have two or more double scutes and the other has one we are inclined to suspect that they are all predetermined.
4. In this and the last set the ' three-to-one' proportion is shown in several ways: (a) In set 121 three show a double scute and' one lacks it; (b) in set 123 three have the scapular double scute and one lacks it; (c) in the same set one of the four has three double scutes, one has only one double scute, and one of the four has an incomplete scute in place of a double scute. Many other cases of a similar kind are noted in connection with both normal and atypical sets.
The following points although possibly of no real significance should be noted: (a) In foetus iv, the adjacent of foetus i, occurs a double scute in the same position in its band as that which occurs in the scapular region of foetus i, but four bands posterior to the latter, (b) In foetus ii occurs an incomplete scute occupying the same position in its band as does one of the double scutes of foetus I, but situated four bands posterior to the latter, (c) In foetus III a double scute of the sixth band is in position almost
exactly the 'mirror image' of the double scute in the third band
of foetus I.
These facts may indicate various degrees of imperfect hereditary control in connection with the localization of these minute 'abnormalities,' but if this be the case we are dealing with a state of affairs too complex to admit of solution with the present material.
There are several other sets which exhibit scute 'abnormalities.' These may for convenience be brought together into a compact table (14). Here the number of the double or incomplete elements in the various columns refers to the position of each in the band; 'D' indicates double scute and 'S' incomplete or split scute:
Although in the members of these sets the double or incomplete elements would appear to occur almost at random, so far as their position within the nine bands is concerned, there are some cases in which, within a given pair at least, the hereditary control is fairly obvious. Table 14 further indicates that a close morphological relationship exists between the two types of ' abnormality' and that they may readily be imagined to have a common hered
TABLE 14
Shoiving the distribution of double and split scutes in the 20 sets of foetuses
99.
116.
119. 127.
135. 138.
I. . II. .
IV.
/ni.
I IV.
58D
9D
IID
15D
25D
41D
46D 25D
60D
COS
37D
54D
3S
9D
13D
892 H. H, NEWMAN AND J. THOMAS PATTEESON
itary basis. It would seem to be profitless to attempt any detailed analysis of the conditions in the sets tabulated, but there are some points worth study that have not been mentioned.
Every case in which two or more individuals show a scute ' abnormality^' is almost undoubtedly a case involving some phase of hereditary control, for on the basis of chance there would be very few sets more than one individual of which would have a double scute.
In concluding this account of the hereditary control of double scutes attention might be called to the fact that in this connection we have excellent material for testing Galton's '^ minutest biological unit capable of hereditary transmission." Whether or not these minute pecularities of the scutes are directly inherited from the immediate parents we cannot at present determine, but we are certain that the conditions are blastogenic in the sense that they are predetermined in the fertilized egg. If this be the equivalent of hereditary transmission we may rightly claim to have found just the sort of unit that Galton was looking for. It seems unlikely that any smaller unit could be predetermined.
(2). Three-hair type of scute. In only one set of foetuses (set 135) have we found that peculiar condition described as a possible discontinuous variation or mutation. Although we are unable to count the hair primordia in the foetuses, we are confident that the great majority of them are of the three-hair type. This assumption is justified by the fact that the scutes have the same narrow appearance that is seen in adults in which the three-hair type prevails, and by the additional circumstance that we have never found an individual with over 600 scutes in which the threehair type of scute did not largely predominate. As an additional piece of evidence in favor of the idea that the condition in question is a mutation, it seems highly probable that it is inherited in the exclusive fashion and is dominant. If we suppose that the foetuses show a blend between the mother, an individual with 573 scutes, and the unknown father, it would necessitate the positing of a male parent with a scute number far in excess of any we have found to occur in the species. That the lower scute numbers, which are made up from shells showing comparatively few three
LIMITS OF HEREDITARY CONTROL 893
hair scutes, are inherited in the blended fashion is shown by a comparison of the counts of the mothers and those of the foetuses. In nearly every case the latter show marked divergences from the former, which would seem to indicate a total lack of exclusiveness about the inheritance.
Attention might also be called to the fact that in this set three individuals have almost identically the same number of scutes while the fourth has decidedly more.
C. Band 'abnormalities' in the species; their distribution and
frequency
The atypical variations in the bands consist primarily of extra or supernumerary bands or parts of bands, and show a rather marked degree of regularity in that they occur repeatedly in about the same regions of the armor. They are found most frequently in the first or second band, or in both, and may occasionally be seen in the region of the eighth or ninth band, but rarely appear in bands three to seven. Their frequency has been determined from a study of 1768 specimens, in which 60 are abnormal, or about 3.4 per cent of the total number of individuals examined.
On account of the apparent nature of their origin, as well as for convenience in description, we have chosen to consider these 'abnormalities' under three headings, as follows: (1) Fusions, in which parts of two bands have united to form a single structure; (2) Splittings, in which a series of elements of a band have divided transversely to produce a double row; (3) Additions, in which a band or part of band, either from the scapular or pelvic shield, has been added to the thoraco-lumbar shield, thereby increasing the banded region, and in case of the contribution or an entire band, producing a ten-banded animal.
{!). Fusions. This kind of variation is found in 31 of the 60 'abnormal' specimens, or in more than fifty per cent of the cases. It usually expresses itself in one of two ways. In one type two adjacent bands have their scutes at one side of the armor united to form a single structure. This type may be spoken of as 'unilateral,' in contrast to the second or 'bilateral' type, in which the
894 H. H. NEWMAN AND J. THOMAS PATTERSON
fusion of the two bands occurs on both sides of the armor. A very good example of this latter type is shown in fig. 11 in which seven scutes are involved on the left side and four on the right.
The unilateral type occurs 25 times in the 31 cases of fusion, and appears 15 times on the left side and 10 on the right. It is distributed as follows among the several bands: 18 times between bands 1 and 2; 3 times between bands 8 and 9; and once each between bands 2 and 3, 4 and 5, 5 and 6, 6 and 7. In almost every case of unilateral fusion the scutes at the extreme lateral margins of the bands are involved, and in only a few specimens is it situated well within the margin (e.g., in fig. 18).
The six remaining cases of fusion are all of the bilateral type, and in each instance the two fusions are practically symmetrical, both as to position and extent. This is true even when they are situated toward the median line of the armor. All the bilateral fusions observed are between bands one and two.
(2). Splittings. The second kind of atypical band variation is the splitting, which as already stated consists of a transverse division of the elements to give rise to two bands. It is not always an easy matter to determine, especially in complex cases where two bands seem to be greatly confused, whether we are dealing with a fusion or a splitting; but a rather safe, though not universal, rule to follow is to count the elements in each row of the double regions, and if these be equal in numbers it is safe to decide that one is dealing with a splitting (fig. 12).
Using this as our principal method of determination we have classified sixteen specimens as belonging to this group. Five of these are unilateral (three on the left and two on the right) and eleven bilateral. One of the most striking features of the latter type is the strong tendency to be almost bilaterally symmetrical. For example in fig. 12 is shown one in which the splitting begins in the seventh scute on each side, and the specimen fails to show symmetry in its bilaterality only in having the splitting slightly less extensive on the right than on the left side. Another point worthy of note is the fact that splittings are confined almost entirely to the first band ; thirteen of the fifteen cases occurring here.
LIMITS OF HEREDITARY CONTROL 895
The fourteenth and fifteenth cases appear in the sixth and ninth bands, respectively.
Several of the cases of spUtting are rather complicated, but not to such an extent as to render their classification uncertain. One of these is shown in fig. 13, and it will be noted that while the ' abnormality' is clearly of the bilateral type, yet it is slightly complicated by having an additional splitting just to the right of the left-hand split.
(5). Additions. The final class of band variations is additions, by which is meant the adding of a band or part of band from either of the shields lying adjacent to the banded region. Six cases of this type have been observed, and five of these are from the scapular shield. An excellent example of this type is seen in fig. 19. The right half of the posterior row of scutes has dropped down from the shield and forms a perfect band on this side of the armor. One can follow the scutes of the added half-band across to the left side where they clearly form the posterior row of the scapular shield, so no question can exist regarding the origin of such 'abnormalities.'
It is easy to imagine how a completion of the dropping would result in adding an entire new band to the banded region, and the four or five cases of ten-banded animals that we have observed may have acquired their extra band in this manner. However, it is possible that additions may take place from the pelvic shield; in fact, our sixth case of additions has probably come to exist in this way, because the pelvic shield of this particular specimen is foreshortened. It should be noted also that a ten-banded animal could be produced by a complete splitting of one of the nine regular bands.
In four of the seven cases of band variations, not classified in the above groups, not sufficient data were taken to permit an exact determination of their characters ; they are merely recorded as being 'abnormal.' Two of the remaining three are almost exactly the same, except that one is in the first band and the other in the ninth. The latter is sketched in fig. 14. The ' abnormality' is located approximately in the middle of the band, and consists of a region of four double scutes. At first one would be inclined
896 H. H. NEWMAN AND J. THOMAS PATTEESON
to suggest that it had been produced by a spHttmg of four scutes, but upon a closer inspection it will be seen that the upper row is closely associated with the left half of the band, and the lower row with the right half. In the light of the ' theory of concrescence' one is tempted to suggest that it has been brought about through a failure of the embryonic primordium to affect a perfect meeting in this region, and consequently the inner ends of the two half-bands have slipped past each other for a short distance.
In this connection attention should be called to a certain rather rare atypical condition which sometimes appears either in the ninth band or in the first row of the pelvic shield. The peculiarity consists of a 'jog' at the middle point of the band or row, and in all probability has come about in a manner like that suggested just above. A specimen showing this condition in the first row of the pelvic shield is seen in fig. 23.
The final 'abnormality' to be considered is shown in fig. 15. It really is a double peculiarity, in that both the eighth and ninth bands are involved. In the eighth band there are three small primary scutes lying just anterior to numbers 53-56 of the main scutes of the band. These small scutes do not affect the bony plates and must therefore be looked upon as very rudimentary in character. In the ninth band a somewhat similar ' abnormality' is seen, lying at scutes 52-56 of this band. It consists of four small scutes, which however affect the bony plates, as can be seen in the sketch of the under surface of this region of the band (fig. 15 h). It cannot be said to be a splitting because there are five plates in the lower or main row and only four above.
In concluding this section of the paper two facts brought out in the foregoing account should receive especial attention, because of their direct bearing on what is to follow. (1) In all those cases of atypical variations that we have designated as bilateral there is a very strong tendency toward symmetry in the affected regions both in position and extent. This is particularly clear in specimens classed as splittings, as the citation of a few cases will make evident: (a) In shell no. 8 band 1 has 57 scutes, and the two splittings occur in scutes 6-14 on the left and 44-51 on the right — were the right-hand split in scutes 44-52 instead of
LIMITS OF HEREDITARY CONTROL 897
44-51, it would be a perfect example of bilateral symmetry; (b) in shell no. 21, in which band 1 has 59 scutes, the splits are in scutes 8-13 and 46-52 — here again a change of one point on the right side would make a bilaterally symmetrical 'abnormality'; and (c) in shell no. 13 there are 63 scutes in band 1, and the splits are in scutes 4-16 and 48-59 — a close approximation to symmetry. These conditions remind one of Wilder's results in duplicate twins, already referred to in the introduction. (2) The second point to be emphasized is this, that although the atypical variations in the bands show a marked degree of regularity in occurring so frequently in the first and second bands, yet they display a great diversity in the fact that scarcelj^ any two of them are exactly alike. This diversity for the species is striking when considered in the light of the fact that in 1768 specimens examined but two sets of coincidences have been found, and these occur in the simplest type of 'abnormality.' Since many of the shells upon which these data were taken are from the same locality it is not improbable that some of these cases are either brothers or sisters, and not coincidences. The fact that only two sets of these supposed cases of coincidences, and these of the simplest type (very slight fusions), have been found in 1768 individuals will serve to emphasize the importance of fraternal correlation as brought out in a subsequent section of the paper.
D. Hereditary control in connection with hand 'abnormalities'
Considering the comparative rarity of band 'abnormalities' in the species we have been fortunate to secure a collection of sets of foetuses, among which appear examples of practically all of the types of malformation described above. As in the case of meristic variation in the normal scutes and in the matter of double scutes the precision of hereditary control is much more marked in some cases than in others; in some the conditions are fairly simple and in others highly complex. The sets are described in detail and discussed separately, in so far as the special conditions of each case are involved. The general significance of manj' of the observations can be discussed to advantage only after all
the data has been presented and the underlying problems presented for discussion.
Set 64- (Mother normal) . Foetus i appears at first sight to be perfectly normal, in so far as band arrangements are concerned; but examination reveals the presence of ten full bands. In view of the fact that all of the other members of the set show more or less extensive regions of splitting in the first band we are forced to conclude that the extra band in this individual has been produced by a process of band splitting carried to a completion.
Fig. 3 Diagrams of the 'abnormal' bands in the four foetuses from female no.
64. In this as in the two succeeding figures, the Roman numerals i-iv refer to the
individual embryos from which the sketches were made, while the arable indicate
the number of scutes in each of the regions to which they are adjacent.
This is really only a bilateral expression of the condition seen in the right hand half of the first band of foetuses in and iv. That there are 64 scutes in the first half-band and only 62 in the second might seem to militate against the idea of splitting, but there occur in our collection several undoubted cases of splitting where the number of scutes in the two series produced by the split are unequal. The split first band of foetus i is shown in fig. 3, i.
LIMITS OF HEREDITARY CONTROL 899
Foetus II shows a condition entirely different from that described for its partner, namely a bilaterally symmetrical, regional splitting of the first band, beginning four scutes from the margin on each side and involving 15 double rows of scutes in each case. The condition is diagrammatically shown in fig. 3, ii.
Foetus III exhibits a decidedly asymmetrical regional splitting of the first band, being a mixture of the two pure types shown in foetuses i and il The left half of the band is identical with that of foetus II, while the right half is the duplicate of that in foetus I (fig. 3, III).
Foetus IV is identical with foetus iii, in so far as the splitting is concerned, but shows in additidn to the latter a fusion of the anterior row of the completely split half with the opposite half of the last scapular row of scutes. This condition has been observed in the shells of several adults and has been interpreted as a case of the incipient addition of a band to the banded region by means of a 'drop-down' from the scapular shield. The present condition could hardly be interpreted in that way in view of our knowledge of the conditions seen in the other members of the set. It would appear that we have here a case of an epigenetic process, involving a secondary fusion of two unrelated half bands, the right half of the first band and the left half of the last scapular row (fig. 3, iv).
This is in some ways the most extraordinary set in the collection and suggests a number of theoretical considerations for general discussion. The main points that should be noted while the facts are fresh in mind are as follows:
1. The splitting process involving the first band occurs in all members of the set, which would indicate that this much was predetermined.
2. The regional splitting involves in all four cases (twice in foetus ii) exactly the same number of scutes, 15 in each case; and these are located precisely the same distance from the margin every time. The precision of hereditary control is, in this regard, nothing short of marvelous, since it is perfect.
3. There are evidently two distinct expressions of the splitting tendency, the complete and the incomplete. The distribution of the two types is quite impartial in so far as their frequency is concerned, for, out of the possible eight lateral halves of the four foetuses, four are occupied by each type. It would appear from this that each was equally strongly predetermined ; but the exercise of hereditary control in the distribution of the two types among the four foetuses is somewhat haphazard and reminds one strongly of a pure chance combination of two elements selected in pairs, hke, for example, the Mendehan ratio of F 2's, D-2Dr-r.
4. Apart from the 'secondary' fusion of one of the split half bands with the scapular shield, the members of the natural pair B (hi and iv) are strikingly identical. Evidently hereditary control within the pair is more accurate than in the whole set. The explanation of this condition must be discussed in the subsequent section on pairing.
Set 96 {Mother normal). This case is somewhat simpler than the last in that it involves only one type of 'abnormality,' namely a simple fusion of the first two bands. We have decided to call the condition a fusion, because, including the two which are united, there are only nine bands. The strictly marginal character of the band unions would also serve to indicate a fusion, for we have found no cases of incipient marginal splitting in our examination of adult shells.
Foetus I shows a unilateral fusion of comparatively sHght extent, involving only 5 scutes and confined to the right side. There are 57 free scutes in band 1 and 58 in band 2 (fig. 4, i).
Foetus II shows a bilateral fusion of small extent, involving 7 scutes on the right and 4 on the left. There are 51 free scutes in band 1 and 49 in band 2 (fig. 4, ii).
Foetus HI shows an extensive unilateral fusion, involving 21 scutes and confined to the right side. There are 40 free scutes in band 1 and 39 in band 2 (fig. 4, iii).
Foetus IV shows the same condition as does foetus iii except that there are 19 fused scutes instead of 21. There are 41 free scutes in each of the first two bands (fig. 4, iv).
The following points may be noted:
1. The fusion in the first two bands occurs in all individuals but differs in its extent and in the unilaterality or bilaterality of its
expression. Evidently a marginal fusion of bands 1 and 2 is predetermined, but here the predetermining influence ceases to operate and epigenetic factors of some sort determine whether the character shall have a unilateral or a bilateral expression and to what extent the fusion is to be carried in each case. It will be noted that hereditary control is much less precise in this case than in set 64.
2. The 'three-to-one' proportion once more presents itself in that three members of the set have the fusion on the right side only, while one has it on both sides.
FijT. 4 Diagrams of the 'abnormal' bands of the four foetuses from female
no. 96.
3. There is a distinct pairing on the basis of the extent of the fusion, pair A (i and ii) showing the character in a much less extensive form than pair B (iii and iv). Within the pairs the resemblance is very close.
iSet 101 {Mother normal). The conditions seen here are in many respects equivalent to those described for set 64. All four foetuses show a rather advanced stage of splitting in the first band. At first we thought that foetuses iii and iv were quite
902
H. H. NEWMAN AND J. THOMAS PATTERSON
normal, but further examination reveals our mistake, for each of these two individuals possesses ten well defined bands. The condition seen in one of the foetuses is so obviously a splitting that we are compelled to interpret the extra band in each of these foetuses as having been produced by a complete splitting of the first band.
Foetus I shows an extensive bilaterally symmetrical and somewhat complex regional splitting of the first band. On both sides, beginning 6 scutes from the margin, there occurs a splitting involving 9 scutes, and in the central part of the band there is another
Fig. 5 Diagrams of the 'abnormal' bands of foetuses i and ii from female no. 101.
split region of 17 scutes. Separating the median from the two lateral splittings are paired unsplit regions, each involving 8 scutes. In the ninth band there is another band peculiarity, consisting of a 'jog' in the middle of the band similar to that'shown in fig. 23. There are 34 scutes to the left of the break and 31 to the right. The region involving the fusion is illustrated diagrammatically in fig. 5, i.
Foetus II shows a more advanced stage of the splitting in that the process has gone on to completion on the right side and has extended to within 6 scutes of the margin on the left side (fig. 5, ii). As in foetus i, there occurs here also a ' jog' in the ninth band, which is approximately the 'mirror image' of that in the partner individual, in that there are 34 scutes on the right of the break and 30 on the left.
LIMITS OF HEREDITAEY CONTROL 903
Foetuses iii and iv both show a completed sphtting of the first band into two, thus producing ten bands; and they also lack the irregularity in the ninth band.
Attention is called to the following points:
1. All four members of the set show an extensive splitting of the first band. The splitting w^as predetermined in so far as the location within a certain band is concerned, but the extent or manner of its expression appear to have been beyond the limits of hereditary control.
2. There was evidently a fairly rigid hereditary control in the matter of the various limits of the incomplete sphttings, as one may judge by the facts that the two sides of foetus i are exactly identical and that six scutes constitute the marginal unsplit region in every case where the splitting is incomplete. This would indicate a precision of hereditary control almost as remarkable as that noted in set 64.
3. There is a pairing of foetuses on the basis of the extent of the splitting, pair A (i and ii) showing an incomplete splitting, and pair B (iii and iv) a complete one. Pairing is again evident in the matter of the irregularity in the ninth band, pair A (i and ii) showing it, pair B (iii and iv) lacking it.
4. The 'three-to-one' proportion appears again in that three individuals are bilaterally symmetrical and one is unilateral in the expression of the splitting process.
Set 118 {Mother normal). This is the only set in our collection where only one member of a set shows a band 'abnormality.'
Foetuses i, ii and iv are normal.
Foetus III shows a fusion of small extent between the first two bands, involving only three scutes at the left hand margin.
The extent of the 'abnormality' is so slight that it seems scarcely worth while to discuss the possible bearings of the condition. It would appear best to consider the case as one of incipient fusion, which may have arisen epigenetically in one of the four individuals. In the next generation it might have been inherited in a more pronounced form. There is, however, the alternative explanation which is developed in the discussion of the theoretical causes of pairing.
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 4
904 H. H. NEWMAN AND J. THOMAS PATTERSON
Set 121 (Mother ahnor7nal). This is one of two observed cases of the direct inheritance of an atypical condition from the mother. In this case the mother had a right marginal fusion of the last two bands, involving 4 scutes.
Foetus I is normal.
Foetus II shows a 'jog' in the ninth band similar to that shown in fig. 23 and in set 101. There are 33 scutes on each side of the break and hence the latter is median.
Foetus III is normal.
Foetus IV shows a 'jog' in the ninth band just like that in foetus II except that there are 33 scutes on the right and 35 on the left.
The 'abnormality' in the mother is entirely different in character from that shown by the two foetuses, but in that it is unilateral and involves the ninth band in both mother and foetuses, there would seem to be ground for believing that there is some genetic connection between the peculiarities seen in the two generations. It is quite possible that the parental and filial conditions may have no hereditary connection, but there must nevertheless be some predetermining mechanism controlling the occurrence of the same peculiarity in two of the members of a set.
It should be noted that the two diagonally placed foetuses seem to be paired with respect to the band irregularity, and that the same individuals were noted as paired on the basis of scute counts. This condition strengthens the hypothesis that occasionally there may occur such a shifting of the blastomeres as to bring about a false pairing in the arrangement of the quadruplets without interfering with their inherited potentialities. For a more detailed discussion of this situation reference is made to one of our former papers on this subject. ^
Set 123. Mother shows a deep crease in the scapular shield, between the ninth and tenth scapular rows (counting from the posterior margin of the shield). This crease is apparently due to the suppression of part of a row of scutes.
All four foetuses show the same peculiar crease in precisely the same region. The photograph (fig. 24) will serve to emphasize
1 Newman and Patterson, 1910. This Journal, vol. 21, no. 3, p. 401.
LIMITS OF HEREDITARY CONTROL 905
the striking identity of the four quadruplets with respect to the pecuharity on question. This is the only undoubted case of the direct transmission of a definite scute peculiarity from the mother to the offspring which we have encountered. In this case it seems clear that hereditary control and hereditary transmission are equally precise, and it appears probable that most of the cases we have described are inherited either from the unknown fathers or from some recent ancestor.
PAIRING; AN INTRA-FRATERNAL CORRELATION; AND ITS BEARING ON THE PROBLEMS OF HEREDITARY CONTROL
The most difficult question that confronts us in attempting to gain an insight into the operation of the predetermining mechanism concerns not so much the resemblances between the quadruplets as their differences. Why are they not all exactly alike and why are there closer resemblances between some individuals of a set than between others?
It has been shown that, with respect to their uterine connections, the four foetuses are definitely paired, in that two of them are attached to each of the lateral placental discs of the chorionic vesicle. Embryological investigations have revealed indications of a much earlier and more intimate pairing. With very few exceptions the resemblances between the foetuses are strictly in accord with their placental pairing, the closest resemblances occurring between the two individuals attached to the same placental disc. This is not a mere matter of arbitrary judgment, but is based on a comparison of the inter-pair and intra-pair correlation. By the use of the difference method previously employed the two constants have been determined as follows:
Inter-pair coefficient of correlation = 0.9257 Intra-pair coefficient of correlation = 0.9517
It will be noted that the correlation between the individuals of opposite pairs (inter-pair correlation) is lower than that determined for the whole set (.9348), while the correlation between the individuals of the same pairs is decidedly higher; in fact this is probably the highest correlation constant determined for any
906 H. H. NEWMAN AND J. THOMAS PATTERSON
organic relation. How surprisingly alike are the members of the pair may be realized by an examination of table 6. A glance down the column of totals will reveal the frequency with which exact identity exists between the paired individuals and how rarely such is the case between individuals of opposite pairs.
It might be claimed that any method of pairing would give a closer correlation than that derived from a treatment of the whole set. That this claim is without justification is readily shown by the experiment of pairing the foetuses in different ways and determining their correlation constants. There are two other possible ways of pairing the quadruplets. We may put together the adjacent members of opposite pairs, associating foetus i with iv and II with III, or we may pair them diagonally, associating i with iii and II with iv. The result of the experiment is shown below :
Correlation coefficient, true pairs (i-ii) and (iii-iv) =0.9517 Correlation coefficient, false pairs (i-iv) and (ii-iii) = 0.9270 Correlation coefficient, false pairs (i-iii) and (ii-iv) = 0.93G4
It will be noted that the correlation constants for both of the false pairs differ only to a slight extent from that of the whole set; which shows conclusively that there is no intra-fraternal correlation on the basis of either of these artificial groupings.
The pairing relation is brought out even more strongly in the study of atypical scute and band conditions. In each set dealt with reference has been made to the instances of pairing as they occurred, and in several eases it was shown that pairing was exhibited in several different ways in the same set. No more convincing evidence of the reality of pairing could be asked for than is afforded by the conditions seen in sets 64, 96 and 101.
The fact of pairing would seem then to need no further demonstration; but its underlying causes and its relation to the mechanics of hereditary control are problems that require much consideration.
It must be frankly admitted that so far we have failed to prove conclusively that each foetus in a set is the product of a single blastomere of the four-cell stage. The youngest embryonic vesicles appear to be still single individuals so far as any visible
LIMITS OF HEREDITARY CONTROL 907
demarkation of embryonic primordia is concerned; but the absence of a visible line of separation between the quadrants of the vesicle does not preclude the possibility that each quadrant may be formed exclusively, or nearly so, of cells derived from one of the first four blastomeres. It seems very likely, in fact, that the cleavage products of each blastomere would continue to occupy the relative position held originally by the parent cell, in spite of the various complex developmental processes that ensue, just as the cells derived from the first two blastomeres, in many organisms, retain their bilateral positions and go to form the right and left sides of the individual. To this extent then we are probably justified in tracing back each of the quadruplets to one of the first four blastomeres. If this be granted it results logically that each pair must be the product of one of the first two blastomeres.
On this basis alone are we able to offer any reasonable explanation of the observed phenomenon of pairing or to attempt an answer to the question : Why should there be a closer resemblance between paired than non-paired individuals? Our original explanation of the condition naively assumed that the first cleavage would divide the fertilized egg into two somewhat unequal parts, and that the second cleavage would divide these half eggs into quarters more nearly equal than were the halves. In brief we assumed that the first cleavage gave products more variable than the second. Is there any basis for such an assumption? Is there, as development proceeds, a progressive decrease in the variability, real or potential, of daughter cells? It has been discovered from a study of the intra-individual variability of certain plants, notably Ceratophyllum, (Pearl, '10), in which whorls of leaves are successively produced by the apical bud or growing point, that the leaves of the first whorl are the most variable, least closely correlated, and that later ones vary less in an orderly progression. Similarly, if we consider that in the first cleavage the armadillo ovum divides itself into two potential individuals and that each of these in turn divides into two more individuals, we would expect to find a decreasing variability among the like parts produced at each successive division. In this case, however, the production of quadruplets destroys the twins and we can dis
908 H. H. NEWMAN AND J. THOMAS PATTERSON
cover the variability of the products of the first division only by determining the inter-pair correlation in our sets of quadruplets. This constant should be lower than that determined between the individuals of the several pairs, the intra-pair correlation. This is exactly what we have done, and the results, as given, are in accord with the hj^pothesis. An interesting test of the validity of this explanation of pairing might be made in connection with the foetuses of Tatu hybridum, one of the South American armadillos, which has most commonly eight or more polyembryonic foetuses. Here the production of individuals has gone at least one step further than in our species, and it should be possible to find out whether the products of the last division are more closely correlated than are the paired individuals in our species or than the products of the previous division. It is to be hoped that some investigator to whom the material is available may see fit to satisfy our curiosity on this point. If our hypothesis should prove to be well founded we shall have furnished an interesting illustration of the law of decreasing variability with the production of like parts, which is probably a corollary of the more general law that variability decreases progressively with advancing development, a law made clear by Vernon in his book on Variation in animals and plants."
What are the logical consequences of accepting such an explanation of pairing, and what light may the ideas expressed throw on the problem of the mechanics of hereditary control? Two courses are apparently open. We may consider the fertilized armadillo egg as heterogeneous in structure, so that its four quadrants, which occupy the positions of the four blastomeres, bear the different materials which determine the differences between the four foetuses. On this basis it would be difficult to explain the paired relation, and still more difficult to accept the necessary^ consequences of the assumption that, where certain at^'pical conditions occur in all four foetuses, the physical basis of the 'abnormality' must have been repeated four times over in as many quadrants of the egg. It would involve too severe a strain on one's credulit}^ to ask him to believe that in sets 121 and 123, for
LIMITS OF HEREDITARY CONTROL 909
instance, the double scute primordium occurred separately in three of the quadrants of the uncleaved oosperm.
The alternative hypothesis involves the assumption that the differences are due to the lack of complete accuracy in the bilateral distribution of the hereditary materials (probably chromosomes). During the process of cleavage by means of the mechanism of hereditary transmission, whose visible operations are probably intimately associated with the mitotic figure, the various materials which condition the development of the definitive structures are distributed more or less unequally to the two daughter cells. The next cleavage involves another unequal distribution of materials, but the inequality is lessened ; and presumably, as the cleavage process continues, either the distributing mechanism becomes more exact through practice or else the material distributed becomes progressively^ more homogeneous, and hence less apt to produce variability.
The Mendelian-like ratios which we have noted so often may be no more than fancied parallelisms, but we have been unable to dismiss the phenomenon without some speculation as to its possible significance. The true Mendelian ratio of 'three-toone' is the result of the segregation of grand-parental characters, a fact which has suggested the thought that we may have here a condition involving in some way, which we are at present unable to understand, the interaction of dominant and recessive ancestral characters. The only alternative explanation of the presence of a character in three individuals of a set and its absence in the fourth involves the assumption of latency, whose aid w^e hesitate to invoke, because we fear that its generous assistance has already been presumed upon over much. After all there is still much to explain in connection with the problem in hand and we can scarcely expect to be able to solve some of its mysteries without the aid of breeding experiments. In spite of the many difficulties that confront us in connection with attempts to keep these animals alive in confinement and to have them breed under experimental control, we feel that breeding is imperative and that success will come in time.
910 H. H. NEWMAN AND J. THOMAS PATTERSON
THE HEREDITARY CONTROL OF SEX
The only character that seems to be rigidly controlled — which does not vary at all in the members of a given set of foetuses — is the character of sex. Whether the set shall be male or female is apparently settled at the time of fertilization and has its physical basis presumably in a dimorphism of the spermatozoa. Sex having been determined, there is no room for individual variation with regard to this character, unless there be such a thing as a more or less pronounced maleness or femaleness. If degrees of sex do exist they no doubt express themselves in terms of fertility.
Along with the primary character of sex are predetermined all of the secondary sexual characters, but these are not so rigidly controlled. Probably no more highly variable characters exist than those that are associated with sex. In the armadillo the two sexes are remarkably alike with regard to somatic characters. So similar are the two sexes that we have never been able to tell them apart without examining the genitalia. A statistical study of the scutes has, however, revealed a sexual dimorphism in the scute numbers. The mean number of scutes in males is a little higher than that in females, but the difference is one of only four or five scutes in the whole banded region. A more pronounced dimorphism exists in the comparative variability of the two sexes. Not only in the species, but within the confines of the various fraternities, the males are decidedly more variable than the females. Ts there any structural dimorphism of the hereditary materials that could be held to be in any way correlated with this variational dimorphism?
A preliminary study of the spermatogenesis of this species has revealed that a dimorphism of the spermatozoa probably exists. If the results of a more extensive study support this tentative conclusion, it will be possible to bring the question of the production of male and female sets into line with the general conclusions already reached on sex-determination in the other forms in which a dimorphism of the sperms brings about a balanced chromosome complex in the female and an unbalanced one in the male. Can there be any underlying connection between the bal
LIMITS OF HEREDITARY CONTROL 911
anced chromosomal relations of the female and variational stability on the one hand, and between the unbalanced condition of the chromosomes in the male and its variational instability on the other? If such a connection exist we have, in addition to sex, another character whose physical basis may in some way be associated with the dimorphism in the chromatin content of the fertilized egg.
A further suggestion might be made in this connection. It seems to be an established fact that the male sex is the more highly specialized, for males depart more widely from the juvenile type than do females. Might it not be possible that the unstable equilibrium of the male chromatin complex lies at the basis of the higher specialization of the male; for a condition of instability would involve an increased potentiality for progressive change? Is there any more or less justification for these suggestions than there is for the generally accepted idea that sex is in some way causally associated with the presence or absence of the accessory chromosome, or its equivalent?
GENERAL CONSIDERATIONS
In his chapter on blastogenic variation Vernon quotes from Weismann the statement that the individual is determined at the time of fertilization, or, in other words, the individuality of the organism results from the fact that the germ-plasm is composed of the paternal and maternal ids which are brought together in the egg cell." As evidence of the vahdity of this statement the author brings up the facts about human identical twins and on the basis of these facts claims that heredity is potentially decided at the time of fertilization."
In the present paper we have shown that the individuality of the organism is not precisely determined at the time of fertilization, but that the characters are hereditarily controlled only within certain limits. These limits we have been able to define with respect to a number of different characters.
Our results are based on the assumption that the degree of divergence shown between the four foetuses is the index of the
912 H. H. NEWMAN AND J. THOMAS PATTERSON
variational potential of the fertilized egg. We assume that if the egg which produces a given set of quadruplets could be made to produce just one individual, this individual would have a potential range of variability equal to that exhibited by the set of quadruplets. Whether or not this assumption is justified depends to some extent on whether our ideas about the operations of the mechanism of hereditary control are sound. If one attribute the differences between the foetuses to the fact of the unequal distribution of the hereditary materials, he would seem to be forced to the conviction that, were a single foetus to develop, there would be no chance for the operation of such a factor, and hence the heredity would be settled at the time of fertilization. This conclusion does not appear to be so necessary when we consider that, even where the egg produces only one individual, it must still divide into two, four, etc., blastomeres, and that each division affords an opportunity for an unequal distribution of hereditary materials. In the first cleavage, instead of producing two distinct individuals, the egg divides its material into two bilateral halves, which are destined to produce the right and left sides of the definitive body. Now it has been shown in another connection that the correlation between the antimetrically paired organs of the same individual is of the same grade as that shown to exist between the quadruplets. This fact simply confirms the idea that the variability of the sets of foetuses gives a reliable measure of the variational possibilities of the fertilized egg. Specific polyembryony doubtless furnishes a special case of intra-individual variability, in which the original individual breaks up into several strictly homologous and independent parts.
We have realized from the beginning of our work on this subject that the results obtained might appear to some biologists to be explicable on the basis of similar or different environmental influences operating during gestation. For these reasons we have been careful to select for study structures little if at all likely to be influenced by environmental factors. It will doubtless be claimed by some that the foetuses are almost identical because they have developed under almost identical conditions, and that the slight differences are the result of equally slight differences in
LIMITS OF HEREDITARY CONTROL 913
position, nutrition, etc. That the environmental factor cannot be seriously considered is shown when particular cases are examined. What kind of an environmental stimulus, for example, could operate to produce such a definitely localized, minute peculiarity as the double scutes in the individuals of sets 121 and 123? One could hardly imagine that a slight difference in the amount or character of the maternal nutriment would produce such a condition, nor could any mechanical factor, such as position, pressure or contact with the amnion, be held accountable for so definitely localized a character occurring in several individuals.
Again, in the matter of atypical band arrangements, it would appear equally absurd to attribute resemblances or differences to environmental factors. Consider, for example, the conditions in set 64. Here the region of incomplete splitting is so definitely localized and involves so fixed a number of scutes that one would have to posit some kind of environmental influence which would be able to cover just so many scutes and place itself just so far from the margin in every case. It would also be necessary to explain why two of the four individuals should show the effects of the influence bilaterally and why the other two should show it unilaterally.
These and many other cases that might be examined point to the untenability of the position taken by the proponents of the efficacy of environmental factors, and strongly fortify the position here taken, that the characters dealt with are purely of blastogenic origin. We conclude then that we have without question made an advance in the direction of determining the limits of hereditary control. We have shown with what degree of exactness the numbers of certain integral variates, the scutes, may be predetermined and how much room is allowed for the play of epigenetic factors. We have demonstrated that the alignment of scutes into bands is very largely controlled by mechanical factors. We have indicated the degree of exactness with which certain 'abnormalities' may be hereditarily controlled, and how small a biological unit is capable of predetermination.
One cannot but be impressed, how^ever, with the diversity of conditions seen in the different sets. In some of the normal sets
914 H. H. NEWMAN AND J. THOMAS PATTERSON
it appears that hereditary control is almost perfect, as, for example, in set 4, where there is a difference of only one scute among the four foetuses; in other sets, 97 for example, the variation among the foetuses is so great that, were thej^ not found to be enclosed in a common chorion, one would hardly believe them to be related. The same conditions prevail in connection with the atypical scutes or bands. In some sets the 'abnormality' is repeated in the different individuals with the utmost fidelity of detail, while in others neither the extent of the region affected nor its location is at all rigidly defined. In a word, one can speak definitely as to the limits of hereditary control for only one set at a time. What appears to be true for one case does not apply exactly to another. Even here, then, where one would expect the phenomena of variation and heredity to exhibit almost diagrammatic simplicity, we find a high degree of complexity and lack of uniformity; and one is again compelled to acknowledge that nature is baffling in her manifoldness of expression and in her freedom from the trammels of exact laws.
SU^LMARY
1. The data derived from human identical twins cannot serve as a criterion for the determination of the limits of hereditary control, for two reasons: (a) The origin of the two individuals from a single fertilized egg is assumed from the fact of resemblance.
(b) The comparison between the twins is made only after years of post-natal life.
2. Armadillo quadruplets furnish a reliable substitute for human identical twins, because: (a) The phenomenon of specific polyembryony has been demonstrated for the species, (b) The unborn foetuses, with all placental connections intact, are used.
(c) The scutes of the nine bands of armor, which are the objects of the present investigation, are elements that reach their definitive number and arrangement long before birth, and hence are excellent for the study of heredity.
3. A study of the morphology of the integument reveals the fact that the integumentaiy unit is a complex element made up of a bony plate, a horny scute and a well defined hair group.
LIMITS OP^ HEREDITARY CONTROL 915
These are so closely associated and so definitely inter-related that a study of one element furnishes an index of the variability of all. For convenience, the external element, the scute, is chosen for statistical study. When the term 'scute' is used the whole complex is to be understood.
4. A statistical study of the species variation in the nine bands reveals the facts that we are dealing with a highly variable character which fluctuates according to the laws of chance. The mean, mode and median practically coincide, and the observed and the theoretical variation curves are very similar.
5. Males are decidedly more variable than females with respect to the characters studied.
6. The variability of the bands taken separately is proportionately greater than that of the banded region as a whole. In each band, however, the variation in the numbers of scutes appears to take place according to the laws of chance,
7. The twenty sets of normal foetuses furnish an ideal array for the study of fraternal correlation. Taking each set as a fraternity, and deahng with the total number of scutes in the banded region, a correlation coefficient of .9348 is obtained. This is taken as an index of the strength of hereditary control with respect to the character in question. The only correlation constants at all comparable with this are those derived from a study of the antimetrically paired organs of the same individual. This fact confirms the idea of the polyembryonic origin of the quadruplets, and show^s that, morphologically, we are dealing with four parts of one individual.
8. The correlation coefficients determined for the individual bands are comparatively so low that the conclusion is reached that the process of scute alignment is largely mechanically determined and hence beyond the limits of hereditary control.
9. A study of the atypical variation in the banded region shows that there are several types of scute 'abnormality,' double scutes, split scutes and the three-hair type of scutes; and also several types of band 'abnormality,' fusions, splittings and additions. All of these conditions are comparatively rare and highly diversified in detailed expression.
916 H. H. NEWMAN AND J. THOMAS PATTERSON
10. Practically all of the types of atypical variation are found in the present collection of foetuses. They are shown to be predetermined, in some cases, with remarkable precision, and in other cases, only in so far as their general character and location are concerned,
11. A further statistical study of pairing serves to demonstrate the truth of this relation. A mechanistic interpretation of pairing is offered, involving the idea that the differences in the four foetuses of a set may be due to the inexactness of the distributing mechanism in cleavage. The paired condition is thought to be an illustration of the general law that variability decreases with the production of like parts.
12. Sex is the only character absolutely predetermined, but the suggestion is made that the greater variability of males, both in the species and within the several sets, may be associated with the lack of balance in the chromosome complex.
13. In reply to the possible objection that the variabihty of the four foetuses of a set is not necessarily an index of the possible range of variability of the ovum, it is argued that the variation between the right and left sides of the body of a single individual is of the same grade as that shown to exist among the quadruplets, and hence, from the variational standpoint, the single foetus is on the same footing as the quadruplets.
14. The conditions described are shown to be inexplicable on the basis of varying environmental factors acting during gestation.
15. The lack of uniformity in the different sets of foetuses, with respect to the limits of hereditary control, leads to the realization of the complexity of the problem and to an acknowledgment that we are still far from a complete understanding of the factors involved.
LIMITS OF HEREDITARY CONTROL 917
BIBLIOGRAPHY
Davenport, C. B. 1904 Statistical methods. New York.
Galton, F. 1875 The history of the twins, as a criterion of the relative powers of nature and nurture. Journ. Anthrop. Inst.
1892 Finger-prints. Macmillan, London.
Harris, J. Arthur 1910 A short method of calculating the coefficient of correlation in the case of integral variates. Biometrika, vol. 7, pp. 214-218.
Newman, H. H. and Patterson, J. Thomas 1909 A case of normal identical quadruplets in the nine-banded armadillo, and its bearing on the problems of identical twins and of sex determination. Biol. Bull., vol. 17, no. 3, August.
1910 The development of the nifie-banded armadillo from the primitive streak stage to birth; with especial reference to the question of specific polyembryony. Jour. Morph., vol. 21, no. 3.
Pearl, Raymond 1910 Intra-individual variation and heredity. Proc. Seventh Intern. Zool. Congress.
Pearson, K. 1909 Determination of the coefficient of correlation. Science, N.S., vol. 30, pp. 23-25.
Vernon, H. M. 1903 Variation in animals and plants. London.
Weismann, a. 1893 The germ-plasm. London.
Wilder, H. H. 1904 Duplicate twins and double monsters. Amer. Jour. Anat., vol. 3, no. 4.
1908 The morphology of cosmobia; speculations concerning the significance of certain types of monsters. Amer. Jour. Anat., vol. 8, no. 4.
Figs. 6-8 These figures show the upper and lower surfaces of three double plates
in various degrees of fusion. X f .
Figs. 9 and 10 Two specimens showing incomplete plates. X f .
Fig. I5a Right-hand ends of the 8th and 9th bands of a shell showing two slight
'abnormalities,' for a description of which see text. X |.
Fig. 15b Under surface of the affected regions of the preceding. X f .
Fig. 16 Under surface of a portion of the anterior half of the pelvic shield. Note the hexagonal shape of a majority of the bony plates. These are all perforated by one or two small holes through which blood vessels and nerves pass. X f.
922 H. H. NEWMAN AND J. THOMAS PATTERSON
LIMITS OF HEREDITARY CONTROL
923
18
Fig. 18 A shell showing a fusion bet\Yeen bands 1 and 2. X i.
Fig. 19 A shell showing an addition to the banded region from the scapular
shield. X A.
Fig. 20 Photograph of a portion of the left side of a shell. This shows the condition of the primary and secondary scutes. X f .
Fig. 21 Photograph of the upper surface of the pelvic shield (also part of the banded region), to show the pebbled effect. X h
Fig. 22 Left side of the shell with the high count of scutes (625), many of which
are of the 3-hair type. A number of these can be made out in the photograph,
especially in the first and second bands. X \.
Fig. 23 The middle portion of the pelvic shield of a' specimen showing a 'jog' in the region of the ninth band. X f .
Fig. 24 Dorsal view of the anterior ends of the litter of embryos from female no.
123. The embryos all show a well marked crease in the middle of the scapular
shield. The mother also had this same peculiarity. Note the double scute in the
middle of the last row of the scapular shield of each of the two embryos lying on
the right. Those two embryos are members of one pair. X |.
EXPERIMENTS ON THE CONTROL OF ASYMMETRY
IN THE DEVELOPMENT OF THE SERPULID,
HYDROIDES DIANTHUSi
CHARLES ZELEXY
SEVEN FIGURES
INTRODUCTION
The following experiments- were made as part of a study of the factors controlling asymmetry in the Serpulid, Hydroides dianthus. The operations were performed on young individuals in which asymmetry was just starting to develop and in which the adult mechanism for reversal of the opercula was not yet present. The results are interesting not onlj- in connection with the problem of the cause of the original asymmetry and of the reversal of the opercula in adults but also as bearing on the extent of agreement between regeneratory and ontogenetic stages. A more defxuite formulation of the problems follows.
1. The first functional operculum of young animals is developed before the first rudimentary operculum. In the absence of any special mechanism in the form of a rudimentary operculum does reversal of position follow removal of the functional organ at this stage? What bearing does the behavior following such operations have on the question of the origin of the original asymmetry and the cause of reversal of the opercula in adults?
2. The first functional operculum of the young is different in type from that of the adult. Does the regenerated operculum
1 Contributions from the Zoologictil Laboratory of the LIniversity of Illinois, No. 8.
2 The experiments were performed at the Biological Laboratory of the Bureau of Fisheries at Woods Hole, Mass., during July and August, 1909. I am indebted to Dr. Francis B. Sumner, the director, and to Professor Raymond C. Osburn, acting director during a part of my stay, for many courtesies. A preliminary report of the experiments was given before the Central Branch of the American Society of Zoologists at the Iowa City meeting, April 8, 1910.
927
928 CHARLES ZELENY
resemble the one removed or does it develop directly as an operculum of the adult type?
3. Further, since the first operculum is a modification of a branchia and therefore originally has a respiratory function, is its removal followed by regeneration, first as a branchia which only later develops the opercular modification, or is the opercular modification regenerated directly?
For a descriotion of the opercula in adults and for experiments on reversal, reference is made to two former papers (Zeleny '02, '05). The development of the opercula in the young is treated in the second of these (pp. 38 to 54). A brief outline of the necessary points must suffice here.
ASYMMETRY IN ADULTS
In the adult Hydroides dianthus there is a large functional operculum or tube plug on one side of the body, either right or left, and a small rudimentary operculum on the other side (fig. 1). The functional operculum (fig. 1, F) has a stout stalk ending in a hard chitinous enlargement consisting of a serrated cup, from the centre of which rises a circlet of curved spine-like projections. The genus Hydroides is characterized by the presence of these two separate circles of serrations or projections in its functional operculum. The rudimentary operculum (fig. 1, E) is a small bud-like structure corresponding in position exactly with the functional operculum of the opposite side of the body. Near the base of each opercular stalk there is a well defined line, the breaking line or breaking joint.
When the cup of the functional operculum is removed, the rudimentary operculum of the opposite side develops into a functional operculum snnilar to the one which had been injured. The stalk of the inj ared operculum meanwhile drops off at its breaking joint and a rudimentary operculum develops in its place. There is thus, as a final result of the operation, a reversal in portion of the two opercula. This reversal may be repeated several times.
CONTROL or ASYMMETRY IN HTDROIDES
929
o
m '
Fig. 1 Hydroides dianthus. A, dorsal view of right-handed specimen, showing relations of parts. Ends of branchiae and functional operculum not given (X6); B,C, diagrams of anterior surface of head of left-handed and right-handed specimens (X6); D, branchia viewed from inner surface (X25); E, rudimentary operculum (X 30). F, functional operculum (X 30).
The process is very similar to the reversal of the two chelae in Alpheus as described by Przibram ('01 to '07), Wilson ('03), Zeleny ('05) and Stockard ('10). In Alpheus there is a larger socalled 'snapping' chela on one side of the body and a smaller 'cutting' chela on the other side. These chelae differ from each other both in size and in structure. When the snapping chela is
930 CHARLES ZELENY
removed, the cutting chela changes mto a snapping chela and in place of the former snapping chela a cutting one is developed.
In Hydroid'es the presence of the functional operculum in some way inhibits the growth of the rudimentary operculum iato a functional one. Likewise in Alpheus the presence of the snapping chela inhibits the change of the cutting chela into a snapping chela. It is evident that the rudimentary operculum in the one case and the cutting chela in the other need only the proper stimulus or removal of an inhibition to develop at once into organs like their mates. These smaller organs are therfore in one sense merely stages in the development of the others.
Various suggestions as to the explanation of this inhibition have been made. In Hydroides the injury to the large operculum produces activity on both sides of the body, bringing about the immediate grow^th of the rudimentary into a functional operculum and a start in the same direction to form a rudimentary operculum on the side of the injury. The functional operculum is reached only on the side wdth the earlier start, the presence of the one functional operculum restraining the possible development of another functional one. This view is supported by the result obtained w^hen the head of Hydroides is removed. In this case, as also in a part of the cases in which the two opercula are removed without injury to the body proper, a functional operculum is developed on each side of the head though the size of each is usually reduced. In such a case the two opercula get an equal start and neither one is able fully to restrict the development of the other.
The cases of similar chelae in the adult Alpheus and Homarus may be explained in a similar way by coincident development. Under the ordinary conditions of removal of both chelae the advantage of greater blood-supply and probably other features lies on the side of the stouter snapping or crushing chela. Therefore reversal does not occur in such cases. When, however, there is a secondary advantage of just the proper strength occurring to the side of the former smaller chela two equal chelae may develop. The probability of this explanation is strengthened by the fact that in those cases in which two equal snapping or crushing chelae
CONTROL OF ASYMMETRY IN HYDROIDES 931
were obtained by Wilson ('03) and Emmel ('08) there was an additional factor, djifference in time of removal of the two chelae, injury of the nerves of the chelae or the removal of the walking leg or legs on the side of the former more slender chela (see Stockard '10, discussed below).
Wilson has suggested that the reversal may be under nervous control and he made a study of it in Alpheus by cutting the nerves going to the chelae. His results, as he himself recognized, are, however, not conclusive, since other influences, such as disturbed blood-supply, are not eliminated. Wilson also suggests that the snapping chela may develop always on the side of the body which has the greater amount of material to start with and which therefore has the greater body of nutrition directed to it. Stockard has tested this hypothesis as applied to the whole group of appendages on the two sides, by removing the walking appendages on the side of the cutting chela in the case in which the snapping chela is removed. The operation leaves the greater mass of appendage 'material on the side of the chela removed. Nevertheless, practically all of the Alphei showed reversal as in cases without removal of walking appendages. Variations of these experiments showed the same negative result. The removal or non-removal of other appendages on either side has no proved relation to the phenomena of reversal. It should however be said that Stockard's experiments do not test 'the essential point of the hypothesis of difference in amount of material or effective mechanism for growth in the form of blood-siipply, etc., between the stumps of the two chelae themselves. The walking legs may not be directly concerned in the difference of activities of the two sides at the first chela level and the hypothesis of difference of materials at the first chela level be unaffected thereby. Furthermore, if there be a correlation between these different levels, is it proper to assume, as Stockard has done, that the smaller amount of nutritive activity will be on the side regenerating other appendages. As a matter of fact, the opposite assumption is indicated by some experiments (Zeleny '09), and Stockard's experiments themselves, as far as they show any difference in results between the two cases, point in the same direction (Stockard '10).
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 4
932 CHARLES ZELENY
ASYMMETRY IN YOUNG INDIVIDUALS
The adult asymmetry of Hydroides is preceded by a symmetrical stage. With the development of a functional operculum the first asymmetrical phase is assumed, followed in turn by a normal reversal in position and a change in structure to the adult type. One or more additional reversals may then take place. It was shown that these reversals occur in nature and furthermore it is probable that they may come periodically without being preceded by any definite injury to the functional operculum other than the wear of ordinary use.
The free-swimming larva of Hydroides, after attachment to a solid object, secretes a tube around its body and develops branchiae on its head. There is a definite stage in which. the branchiae are all alike and symmetrically arranged with respect to the median line (fig. 2). Additions to these branchiae take place at the lower edge of each lateral group. At a stage slightly older than the one shown in fig. 2 the young serpulid closely resembles in its branchial characters an adult of the genus Protula (Fritz Miiller '64). There is no trace of an opercular modification in any of the branchiae.
The branchia next to the dorsal one on the left side then begins to form a terminal cup with a single row of serrations (figs. 3 and 4). The branchial filaments are still present and the respiratory function is evidently still retained along with the new tubeplug function. The opercular branchia at this stage may be compared with the functional operculum of adults of the genus Apomatus, in which also the opercular enlargement is on a stalk bearing branchial filaments though the enlargement itself is different in structure. Apomatus has in addition a rudimentary operculum of the same type as the functional, i.e., borne on the end of a branchia which retains its respiratory filaments (fig. 6, A, B, C, D,). It is important to note that the operculum in Hydroides first appears on the left side of the body while in adults it is sometimes on the right and sometimes on the left.
During the following period the branchial filaments disappear from the opercular stalk and at the same time the corresponding
CONTROL OF ASYMMETRY IN HYDROIDES
933
Fig. 2 Young Hydroides dianthus with six branchiae (individual C, no. 2780) .
The most dorsal pair has not yet developed its pinnules. The levels of the cuts
made in the operation are indicated by C.
Fig. 3 Hydroides dianthus, age 34 days; stage with four pairs of branchiae. The next to the dorsal one on each side is shown in full. The left one shows the beginning of the knob of the functional operculum. The right one drops off later and the first rudimentary operculum is developed from its stump.
934
CHARLES ZELENY
branchia of the opposite side drops off, the break appearing at the base of its stalk. In its place a bud-like rudimentary operculum is developed (fig. 5). There is then on the left side a functional operculum with a naked stalk and a terminal cup' bearing
Fig. 4 Hydroides dianthus, age 26 to 28 days but further advanced than the individual shown in fig. 3. Stage with four pairs of branchiae. Figure shows the
three most dorsal pairs. The ventral pair is directed away from the observer
and is not shown. The opercular knob of the left side is notched and its stalk
still retains the respiratory pinnules. The branchia next to the dorsal one of the
right side has eight pinnules but differs from the other branchiae, except the opercular one, in the absence of new pinnule buds. A, level of cut in individual A,
(no. 2778); B, level of cut in individual B, (no. 2779); Ba, regenerating functional
operculum one day after operation in individual B.
a single row of serrations, and on the right side a rudimentary operculum. The characteristics are those found in the adults of the genus Serpula except for the fact that in Serpula the functional operculum may be either right or left (fig. 6, E).
CONTROL OF ASYMMETRY IN HYDROIDES
935
After a period of activity in this stage lasting several davs the functional operculum drops off. The loss is accompanied by the development of the rudimentary operculum into a functional operculum not like the one that had been lost but more complicated in structure, with two distinct and quahtatively different
Fig. 5 Hydroides dianthus. Dorsal view (X 38). On the left side is functional
operculum of Serpula type with naked stalk and one row of serrations. On right
side is rudimentary bud developed from the base of the cast-off second branchia
of that side.
rows of serrations. This operculum is similar to that of the adult Hydroides. In place of the operculum that had dropped off a rudimentary operculum is developed. The condition is now like that of the adult except that some adults have the functional operculum on the left and others on the right. It is therefore neces
936
CHARLES ZELENY
sary to assume further reversal or reversals of position during later development and perhaps throughout adult life.
The original asymmetry is apparently always of the same nature, i.e., the functional operculum always develops on the left
Fig. 6 A, B, C, D, Apomatus ampullifera. Adult. A, dorsal view showing
functional and rudimentary opercula and branchiae. Pinnules not represented
(X 5) : jB, tip of non-operculate branchia (X 17). C, tip of rudimentary operculum
(X 17): D, tip of functional operculum (X 17): £?, distal portion of functional operculum of Serpula vermicularis (X 19).
side and from a particular branchia. Adult individuals differ in character because some have had an odd and some an even number of reversals.
CONTROL OF ASYMMETRY IN HYDROIDES 937
EXPERIMENTS ON ASYMMETRY IN YOUNG INDIVIDUALS
The first functional operculum appears before the first rudimentary (fig, 4). During the stage in which the opercular stalk retains its branchial filaments there is as yet no modification of the bran<^hia of the opposite side. It ends in a point just like that of other branchiae. Buds of new pinnules are however absent (figs. 3 and 4). There is at this time no perfected mechanism for reversal as in the adult. The removal of the functional opercuIdoi, or at an earlier stage the removal of the end of the branchia which later develops the opercular enlargement, was accomplished in several individuals (figs. 2 and 4). I.i only a few of them, however, was the operation clean cut and -the further history of the experiment followed.
The character of the material did not allow experimenting with narcotization. It was therefore necessary to keep watch of the young animals under a binocular microscope, holding a needle knife blade above the opening of the tube. With exceptional good fortune it was possible to remove the terminal cup of the operculum in a few instances without seriously injuring the rest of the animal, though in most cases the neighboring branchia of the same side was also injured.
There was no reversal of opercula, though interesting developments followed. In place of the removed functional operculum a new one like the one removed was developed. There was no loss of the old stalk, the regeneration taking place directly at the cut surface (fig. 4, Ba). It might have been expected that, in case regeneration occurred, the new operculum would at once grow out as one of the Hydroides type as found after the first natural reversal. This, however, did not occur. The regenerated structure therefore repeats the stage of the removed structure and does not pass on to the next ontogenetic stage (fig. 7, A).
The effect upon the branchia of the opposite side is also interesting. The terminal part of the branchia develops a small knob, approaching in character a rudimentary operculum of this species but formed from a group of cells which in normal growth never develop opercular enlargements, but are, in fact, lost when
this branchia drops off to make place for the rudimentary operculum developing in its place (fig. 7, B, C). A branchia with a small knob-like rudimentary operculum at its end is thus formed. This condition is never found in normal development in this species but is a normal feature of the adults of the genus Apomatus which have also functional opercula with stalks bearing branchial filaments (fig. Q, A, B, C, D).
Fig. 7 A, regenerated functional operculum of Serpula type. Individual A
(no. 2778), on August 3rd. B, branchia next to the dorsal one of right side as modified, following operation on the corresponding branchia of the left side. Individual A, (no. 2778) July 31. C, enlarged tip of B. D, rudimentary operculum on
left side of individual A as developed after the functionaloperculumof the Serpula
type had dropped off, August 14. E, functional operculum of Hydroides type as
developed on the right side of individual A after the functional operculum of the
Serpula type on the left side had dropped off, August 14.
It might have been expected that the removal of the functional cup would accelerate the breaking off of the corresponding branchia of the other side of the body and be followed by the development of a normal rudimentary operculum. As a matter of fact, the acceleration of opercular development is found, but in a way entirelj- different from the normal. A group of cells at
CONTROL OF ASYMMETRY IN HYDROIDES 939
the tip of a braiichia is stimulated to develop. This condition, however, lasts only a short time. The branchia with its terminal enlargement breaks off as does the corresponding one in normal development and a rudimentary operculum grows out in its place. WTiile this is taking place the stalk of the regenerated functional operculum loses its filaments and the resultant condition is like that of the Serpula stage of normal ontogeny (fig. 5).
Outline of typical individual experiments
Individual A. {No. 2778)
July 1—1909. Egg fertiUzed.
July 27 . Condition of opercula (fig. 4).
Left side. Second branchia has opercular enlargement. Respiratory pinnules still present.
Right side. No opercular modifications. Each branchia ends in a long tapering filament.
Operation. Distal half of second left (opercular) branchia and distal third of first left branchia removed (fig. 4, level A). July 31.
Left side. Removed parts are regenerated. The new operculum resembles the removed one.
Right side. The next to the dorsal (second) branchia has developed a small ovoid enlargement at its end (fig. 7, B, C). The whole branchia now resembles a rudimentary operculum of the kind found in adults of the genus Apomatus (fig. 6, ^, C). August 3.
Left side. The new opercular stalk has lost its respiratory pinnules. The terminal cup retains a single circle of serrations (fig. 7, A). The structure resembles an adult operculum of the genus Serpula (fig. 5).
Right side. The next to the dorsal (second) branchia with its small terminal enlargement has dropped off and a rudimentary operculum has developed in its place (fig. 5). August 7.
Left side. Functional operculum of Serpula type (fig. 5).
Right side. The former rudimentary operculum is going forward in its development to a functional operculum with two rows
940 CHARLES ZELENY
of serrations {i.e., of the Hydroides type). The new operculum is about half developed, August 14 Left side. The functional operculum of the Serpula type has dropped off and a rudimentary operculum has developed in its place (fig. 7, D).
Right side. .The functional operculum of the Hydroides type is fully developed (fig. 7, B).
Individual B. (No. 2779)
July 1. Egg fertilized.
July 28. Condition of opercula as in no. 2778 on July 27.
Operation. The opercular enlargement at the end of the branchial operculum was removed (fig. 4, level B). July 29.
Left side. Removed opercular cup is regenerating (fig. 4, Ba). August 3.
Left side. A functional operculum of the Serpula type is fully developed. The respiratory pinnules have already disappeared (fig. 5).
Right side. The next to the dorsal (second) branchia has dropped off and a rudimentary one has developed in its place (fig. 5). August 7.
Left side. Functional operculum of Serpula type.
Right side. The rudimentar}^ operculum has developed into a half-grown operculum of the Hj^droides type. August 14 Left side. Rudimentary operculum (7, D).
Right side. Functional operculum of Hydroides type (", E).
Individual C. {No. 2780)
July I. Eggs fertilized. July 28.
Condition of opercula and branchiae. Three pairs of branchiae are present. The most dorsal pair has not yet developed its pinnules. There is no sign of opercular enlargement (fig. 2).
CONTROL OF ASYMMETRY IN HYDROIDES 941
Juhj 29.
Operation. The distal half of the left second or future opercular branchia and the tip of the left third branchia were removed as shown in fig. 2. August 3.
Left side. From the branchia next to the dorsal one a functional operculum of the Serpula type has developed. The stalk has already lost its pinnules (fig. 5).
Right side. Rudimentary operculum (fig. 5). August 7.
Left side. Rudimentary operculum.
Right side. Functional operculum of Hydroides type, onehalf developed. August 14- '
Left side. Rudimentary operculum (fig. 7, D).
Right side. Functional operculum of Hydroides type. Not fully developed as yet (fig. 7, E).
The experiments thus answer the three questions proposed at the beginning of the paper.
1. Is there a reversal of opercula as a final result of removal of the first functional operculum before any rudimentary opercular structure has been formed? The experiment shows that there is no reversal of opercula. A new functional operculum develops in place of the removed one, and on the opposite side of the body the final result is a rudimentary operculum like the normal one. Before this final condition there is, however, the interposition, as a result of the operation, of a new kind of opercular modification, namely, a rudimentary opercular enlargement at the end of a branchia. The enlargement is formed from cells which do not normally produce an operculum. The formation of the structure is directly stimulated by the operation. It has, however, no permanent result, the operculum thus formed in no way preventing the dropping off of the branchia.
2. Is the removal of the first functional operculum folio v\'ed by the regeneration of a structure of the same type as the one removed or does the later or adult type develop at once? It is shown that the new structure is like the one removed. It does not skip to
942 CHARLES ZELENY
the next succeeding stage. At this step, as probably at other steps in the ontogenetic process, removal is followed by a repetition of the part removed. The stages need not follow in a definite prescribed order. A stage may be repeated when the necessary stimulus is given. In normal ontogeny the first functional operculum has a period of existence as such which ends by its breaking off near the base. In its place a rudimentary operculum is developed, an operculum which is a preliminary stage in a new functional, the second of the Hydroides type, which develops only after a long period of latency. If, however, the first operculum be lost early in its life, there is no skipping of the later phases leading up to a normal loss, but the first stages are repeated and an operculum like the one removed results.
3. Is the removal of the first functional operculum followed by regeneration, first as a branchia which only later develops the opercular enlargement or is the opercular modification developed directly? The opercular modification is developed directly. An enlargement at the cut surface of the stalk is evident very soon after the operation. There is no trace of the development, first, of a pointed end like that of the original branchia, but opercular growth proceeds directly at the regenerating surface. There seems thus to be no repetition of the ontogenetic branchial stage at this regeneration (fig. 4, Ba).
The observations and experiments on young Hydroides contribute the following data on the factors controlling asymmetry of the animals.
1. The animal is originally symmetrical and the appearance of the asymmetrical structure, always on the left side, can not be due to the preponderance of nutrition on one side as a result of a larger amount of material on that side. There is further no indication that the original asymmetry is due to the character of the tubes or the nature of their curvature.
2. The removal of a functional operculum is not necessarily followed by the development of a rudimentary operculum on the same side. The larger mass of material, following the removal, was on the opposite side of the body. Nevertheless, the func
CONTROL OF ASYMMETRY IN HYDROIDES 943
tional operculum developed from the cut surface. A. small opercular knob at the end of the branchia of the opposite side was, however, developed as a result of the operation.
3. The. result of the operation was an animal with the same symmetry as a normal individual.
SUMMARY
1. The removal of the first operculum of Hydroides dianthus in its early stages before the development of a rudimentary operculum is followed:
A. By the regeneration of a new functional operculum of the same type as the one removed. There is no reversal of opercula such as occurs in the adult.
B. (1). The branchia occupying the positiou of the future rudimentary operculum and which in normal development shows no opercular modification develops an opercular enlargement at its end as a result of the operation.
(2). The opercular enlargement is developed from cells which in normal development do not form an operculum.
(3). The enlargement is of the rudimentary type such as is found in adults of the genus Apomatus.
2. The regenerated functional operculum is like the one removed. Regeneration does not go direttly to the next succeeding or Hydroides type of the operculum. On the other hand neither is there a repetition of the preceding stage, the opercular enlargement appearing directly without the interposition of a branchial tip, later to be modified into an opercular structure.
3. The results point to the conclusion that reversal of opercula in the adult is dependent upon the presence of a specialized structure, the rudimentary operculum, capable of developing rapidly into a functional operculum. In the absence of such a special structure the regenerating tissue on the old functional operculum side retains its supremacy, getting an earlier start and inhibiting the development of a similar structure on the opposite side. The injury to the functional operculum does, however, initiate an
944 CHARLES ZELENY
opercular modification in the branchia which stands in the position of the future rudimentary operculum. This process, however, is not sufficiently rapid to gain the upper hand and cause the development of a functional operculum.
LITERATURE CITED
Brues, C. T. 1904 The internal factors of regeneration in Alpheus. Biological Bulletin, vol. 6, p. 319.
Emmel, V. E. 1908 The experimental control of asymmetry at different stages in the development of the lobster. Jour. Exp. Zool., vol. 5, no. 4.
Morgan, T. H. 1904 Notes on regeneration. Biological Bulletin, vol. 6.
MtJLLER, Fritz. 1864 Fur Darwin, pp. 76-77.
Przibram, H. 1901 Experimentelle Studien liber Regeneration. Archiv fiir Entw.-Mech., Bd. 11, 1901.
1902 Experimentelle Studien iiber Regeneration. II. Archiv fiir Entw.-Mech., Bd. 13, 1902.
1905 Die 'Heterochelie' bei decapoden Crustaceen. Archiv fiir Entw.Mech., Bd. 19.
1907 Die 'Scherenumkehr' bei decapoden Crustaceen. Archiv fiir Entw.-Mech., Bd. 25.
Stockard, C. R. 1910 The question of reversal of asymmetry in the regenerating chelae of Crustacea. Biological Bulletin, vol. 19, no. 4, Sept.
Wilson, E. B. 1903 Notes on the reversal of asymmetry in the regeneration of the chelae in Alpheus heterochelis. Biological Bulletin, vol. 4, pp. 197210.
Zeleny, C. 1902 A case of compensatory regulation in the regeneration of Hydroides dianthus. Archiv fiir Entw.-Mech., Bd. 13.
1905 Compensatory regulation. Jour. Exp. Zool., vol. 2, no. 1.
ANATOMICAL ILLUSTRATION BEFORE VESALIUS'
WILLIAM A. LOCY
From the Department of Zoology, Northwestern University
TWENTY-THKEE FIGUBES
The study of anatomical illustrations before Vesalius is not chiefly of antiquarian interest. It brings under consideration a momentous period of intellectual development when the scientific spirit was awakening and struggling for better expression. The examination of the human documents containing the early attempts at pictorial representation of the results of observation, have a peculiar interest for those who are still engaged in observing and recording results by the graphic method. Moreover, the consideration of these crude sketches reveals to us the conditions under which scientific men worked, the mental habit of the period, the educational practice in science and the degree to which accurate observation in anatomy prevailed. Nothing else shows more definitely the state of anatomical knowledge of the time, that which is covered and rendered ambiguous in the text stands exposed in the sketches — these graphic indices show the degree of fidelity to nature of the observer and his mental bias in the matter of interpretation.
1 The notable interest of Professor Whitman in the historical phases of his science makes it appropriate that one of his students should offer in his memory a study of anatomical illustration before Vesalius — a study in the awakening of the scientific spirit.
Doctor Whitman was a pioneer in the United States in inaugurating university instruction in the history of comparative anatomy and of generation (see the Clark University Register for 1890). It is with pleasurable reminiscences that the writer acknowledges the influence of Dr. Whitman in the development of his mental interests. The friendly as well as the preceptorial relations with this leader of biological thought were a source of stimulus, especially as regards the philosophical outlook on nature, and the growth of a disposition to view current biological thought and attainment in the light of its historical development.
945
946 WILLIAM A. LOCY
The pursuit of science from the historical standpoint has appealed only to a limited number, and there is needed at present a sympathetic recognition by scientific men, in general, that this affords a worthy field of research. This conception is being promoted by the relatively new movement in European universities, that has resulted in the appointment of professors of the history of medicine and natural science, to the establishment of periodicals devoted to researches in the same field and to the foundation in Leipzig of an Institut fiir Geschichte der Medizin. All this, and the growing disposition to provide a historical background for courses in biological study, is a sign that there is to be a widening of the field of biological research. It is to be hoped that the time is near when this line of study will be a recognized division of biological research, running parallel with other forms of biological investigation, and pursued as a research subject by examination of the original sources.
The attempts at pictorial representations of anatomy began before the invention of printing, as is shown in the pen, crayon and chalk drawings of anatomical subjects found in the medical manuscripts stored in the libraries at Berlin, Paris, Oxford, Munich and other places. A rich series of these manuscript anatomical sketches has been brought to light by Karl Sudhoff and his collaborators, and reproduced by photographic methods in the Studien zur Geschichte der Medizin and in the Archiv fiir Geschichte der Medizin. These resurrected manuscript sketches have thrown a flood of light on the sources of early anatomical illustrations. A genetic connection has been established between some of them and the earliest printed anatomical figures.
The question arises in connection with the early printed illustrations: Are these sketches crude representations of actual dissections, or are they based upon earlier traditional diagrams? They are in reality mixed as to the source. Many of the earliest printed anatomical figures that were thought to be original are traceable to manuscript sketches that were based upon reading of the anatomical descriptions of the Arabian and of thfe classical authors. Other printed illustrations based partly on observation show departures from the traditional schemes. There is.
ANATOMICAL ILLUSTRATION 947
however, even in the improved sketches, a mixture of observation and tradition, with a stronger inchnation to preserve the traditional than to let go of it and depend on observation.
The date at which sketches of anatomical subjects were first used is uncertain. There is a tradition that Aristotle employed anatomical plates in his teaching, but no remnants of them are known. There are known, however, manuscript illustrations of anatomy dating back to the twelfth century, and furthermore, some of the manuscript sketches of the early part of the fourteenth century have a recognized genetic connection with the earliest printed illustrations of anatomy. For example, Sudhoff has recently published copies of the diagrams used by De Mondeville, about 1304, to illustrate his lectures at Montpellier, and the connection between these pictures and those published by Peyligk in 1499, and by Hundt in 1501 is undoubted. There are other known correlations between manuscript sketches and early printed figures that will be mentioned later in connection with a consideration of the printed sketches.
The earliest printed illustrations of anatomy occur in the Fasciculus Medicine of Ketham of 1491, and, from that time to the publication of the Fabrica of Vesalius in 1543, there are about one hundred different anatomical cuts. Some of these pictures are duplicated in different treatises so that the enumeration of figures in the different printed books would exceed this number. This statement does not include the seven hundred to eight hundred anatomical sketches of Leonardo da Vinci, none of which were published until much later. In addition to the printed books of the period there were anatomical plates printed and sold separately. To this latter group belong the figure of the skeleton by Richard Helain, printed in Niirnberg in 1493, and its modification, by Griininger of 1497, the anatomical plates of John Schott of 1517, the plates of Vesalius of 1538, etc.
The pictures of this period are little known to anatomists, accordingly it is the printed illustrations of anatomy from 1491 to 1543 that are to be brought under consideration in this paper. The writer has had for personal examination the printed books containing the pictures referred to with one or two exceptions
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 4
948 WILLIAM A. LOCY
that will be noted below. The subject of manuscript illustrations is not attempted, since very few ol these have been available, and the references to manuscript sketches are drawn chiefly from the publications of Sudhoff.
In reviewing the old anatomical treatises, points of bibliographical interest emerge, and comparison of the texts brings out some features of interest to scholars. No attempt has been made however to embrace bibliographical notes and textual comparison. The boundaries of the paper are limited to a consideration of the character and quality of the earliest illustrations of anatomy with the further aim to determine to what extent these are based on observation, and to add some comments on the conditions or the time as they affected the development of observation in science. The pictures are not comprehensive enough in their range to show all the anatomy of the period, but they are significant in showing the spirit of the time, the dependence on descriptions and the lack of a positive anatomy based on observation.
Sources. The printed books published before 1543, that contain anatomical figures are medical treatises, anatomical texts and surgeries. I have had for examination, chiefly in the Surgeon General's Library at Washington and the John Crearer Library at Chicago, the primary sources named below. The books are designated by date and abbreviated title only, since the full titles are often long and cumbersome.
I have examined thirteen copies of the anatomical treatise of Mundinus, Anatome omnium humani corporis interiorum meinbrorum. Of these, seven were published separately and six M^ere incorporated with other writings in the Fasciculus Medicine of Ketham. The collection embraced: two copies of the Melerstat edition of Mundinus, published in Leipzig about 1495, 39 leaves, 4°, with one illustration; and one copy each of the editions of G. Lincium, Venice, 1494, 22 leaves, no illustrations; F. Picium, Freiburg, 1507, 23 leaves, no illustrations; J. Adelphus, Strass^ burg, 1513, 40 leaves, figure of the zodiacal signs as related to regions of the body and a rough sketch of the heart; the large annotated edition of Berengarius, Bologna, 1521, 528 leaves (1056 pages), 21 woodcuts and the extensively illustrated edition
ANATOMICAL ILLUSTRATION 949
of J. Dryander, Marburg, 1541, 70 leaves, 45 woodcuts, one repeated, and two on one plate. These books are all of quarto size. In addition I have had six copies of the Incipit Anatomia Mundini in the editions of Ketham mentioned below.
Six editions of Ketham's Fasciculus Medicine (or Medicinse) came under observation, all of folio size; Venice, 1495, Latin edition; Venice, 1500, Italian; Venice, 1500, Latin; Milan, 1509, Itahan; Venice, 1522, Latin; Venice, 1522, Italian. All these contain woodcuts to be mentioned below.
The plate of the skeleton by Richard Helain, Paris and Ntirnberg, 1493, 53 cm. high, from the library of Dr. Mortimer Frank of Chicago.
J. Peyligk, Philosophie Naturalis Compendium, containing the Compendiosa. capitis physici declaratio, which is the illustrated part, Leipzig, 1499, folio, frontispiece and thirteen separate anatomical illustrations in the text.
Magnus Hundt, Antropologium de hominis dignitate natura et proprietatibus, Leipzig, 1501, 4°, 120 leaves, 19 figures, one being repeated. Two copies of this rare book came under observation, one in the Surgeon General's library and the other in the library of Dr. Mortimer Frank of Chicago.
Phryesen (Fries, Frisen, etc.), Spiegel der Artzney, three copies, the Strassburg edition of 1519, folio, Dutch, 4 figures; Strassburg, 1529, German, 141 leaves, one picture in addition to the illuminated title page, and the same, Strassburg, 1532,
Berengarius (Carpus), his Commentaries on Mundinus (mentioned above) Bologna, 1521, 4°, 528 leaves and 21 woodcuts. Three editions of his Isogogse Breves : Bologna, 1523, 4°, 80 leaves, 23 figures; a small pocket edition, 1530, 132 leaves, with 24 very crude, small woodcuts copied from the edition of 1523; Venice, 1535, 4°, 63 leaves, 19 cuts.
Petrus d'Abano, Conciliator differentiarum philosophorum, 1526, containing the first printed pictures of the abdominal muscles copied from the edition of 1496.
Leonardo da Vinci, I Manoscritti di Leonardo da Vinci della Reale Biblioteca di Windsor, etc., Paris, 1898, 1901, etc.; ten of the twenty-four volumes contain anatomical sketches and
950 WILLIAM A. LOCY
notes, 223 plates and upwards of 750 figures, with an introduction by Duval. These anatomical illustrations, executed about 1510, are in all particulars the most notable contribution to anatomy before Vesalius.
J. Dryander, Anatomia Mundini, and other old writers, Marburg, 1541, 70 leaves and 44 illustrations.
W. H. Ryff, Anotomi (very long title), 1541, woodcuts.
For collateral reading the treatises of Choulant, Chievitz, Hopf, Hyrtl, Roth, Pagel, Sudhoff, Toply, Weindler and Wieger have been of especial service. In Wieger are found photographic reproductions of visceral dissections from Reisch's Margarita philosophica, 1503 and 1504, forming a link in the development of anatomical sketches. I am greath'^ indebted to the contributions of Sudhoff for general enlightenment, for knowledge of the manuscript sources and for an illustrated account of Brunschwig's Anatomy in his Chirurgie, 1497.
Mundinus. (Mondino, etc. ; the Romanized form of his name is used here because his book was chiefly printed in Latin.) The anatomy of Mundinus (Anatome omnium humani corporis interiorum membrorum, De omnibus humani corporis interioribus membris anatomia, Incipit Anathomia Mundini, etc.), although not the first treatise on anatonw to be illustrated, is the natural starting point for a consideration of pre-Vesalian anatomy. Appearing, in manuscript, in 1316, it was the first professional treatise on anatomy after more than eleven centuries of Galen. On account of the extensive use in medical schools it forms the genetic link between the ancient anatomy and that of the renaissance period. It was the forerunner of the anatomical treatises that appeared before the epoch-making book of Vesalius.
Mundinus, on account of the influence of his teaching and of his treatise, looms large in the background of historical anatomy. He helped to overcome the opposition to dissection and he is usually credited with having brought the practice into general recognition. Although he was a pioneer in the restoration of anatomy, his wsiy had been prepared by others. De Mondeville as early as 1304 had been illustrating his lectures on anatomy at Montpellier; the Senate of Venice had decreed in 1308 that a
ANATOMICAL ILLUSTRATION 951
body should be dissected annually; William of Salicit, Richardus and others had dissected before Mundinus.
The purpose of his book was to simplify the teaching of anatomy and it was designed primarily for his students. (As he says: "proposui meis scholaribus in Medicina quoddam opus componere. ' ') It was so highly esteemed that it had a general use for upwards of two centuries and often was used as an introduction to Galen or in connection with his anatomical writings. It came to be prescribed by legislation as the required textbook of anatomy in Italy. Before the invention of printing it was copied and extensively circulated among medical students. Mundinus was a great favorite with the students who came under his instruction. He seems to have been a man of engaging personality gifted with powers of clear exposition. His book is well arranged and terse in description. Although he states that prior to its composition he had dissected three human bodies, it is too much to say that it was an original treatise based on personal observation. He merely brings into systematic form the teachings of Galen with some modifications of his own. Roth and others have pointed out that in his compilation he did not make use of a pure text of Galen in the Greek, but, on the contrary, employed impure Latin and Arabic translations. He does not succeed in overcoming the influence of tradition and of dialectic compilation. With Galen he enumerates five lobes in the liver and perpetuates other errors that observation on the human body should have corrected. His book is also burdened with the terminology of the foreign texts; the stomach, for illustration, is designated the myrach, the peritoneum as the cyphach (siphac), the omentum as zirbus, the mesentery as eurachus, etc., etc. The key to the influence of the book of Mundinus is not its originality but its wide circulation; it is conspicuously lacking in evidences of independent observation.
The book was first printed in small folio form in Padua, in 1478, and, between that date and 1580, when the last edition was published, not less than twenty-five editions are known. These are usually annotated and commonly in quarto form. The thirteen editions of Mundinus examined, excepting those in Ketam
Fig. 1 From the Melerstat edition of INIundinus, Leipzig, 149;^
ANATOMICAL ILLUSTRATION 953
vary from an edition of 22 quarto leaves to the extensive edition of Berengarius, of 1521, containing 40 commentaries and 586 leaves. The 21 illustrations in the last mentioned will be considered under Berengarius.
The only edition to be mentioned at present is that of Dr. Melerstat, printed in Leipzig about 1493-95. The book was published without date or indication of place. The copy in the Surgeon General's library at Washington has a note saying that it was published, probably in Leipzig about 1493. It has 39 leaves, including the title page, with a letterpress of 3f x 6 inches. This was the first edition to be printed with a woodcut which is shown reduced in fig. 1. The original is 31 x 8f inches. It represents a teacher of anatomy seated and reading from a textbook, while, in front, his demonstrator is engaged with a visceral dissection. The sketch of the viscera is highly diagrammatic. On the table is seen the large curved knife like that exhibited in the pictures of surgical instruments of the period. This picture shows the method of teaching anatomy at that time. Often the reading was done without any subject before the hearers, at other times dogs and other animals were used for demonstration, and, on rare occasions, a human body was dissected in public anatomies. This picture is a type of many others found both in manuscripts and in early printed books. Sometimes in these pictures students are shown grouped around the dissecting table, but the teacher is always seated and reading from a text. The academic dress is a feature of them all and affords an index to the costume worn by teachers and students at different schools and at different periods of time. For several similar pictures see Choulant, Chievitz, etc.
Ketham. (Johannis de Ketaz, etc.) The Fasciculus Medicine (also Medicinse) of J. de Ketham is believed to be the first printed medical treatise to be illustrated. The first edition printed in Venice in 1491 contained six woodcuts and the subsequent editions contain usually nine or ten. Prepared under various editors, there are several editions but they are similar as to text and illustrations. The book is of folio size and is a collection of medical writings embracing sections on the means of recognizing
954 WILLIAM A. LOCY
diseases by the various colors of the urine; the practice of venesection and blood letting; comments on surgery; the figure of female anatomy, showing a foetus in the uterus; advice regarding diseases and, in 1493 and thereafter, the anatomy of Mundinus.
Only two of the illustrations can be classed as anatomical, that showing the location of the viscera in the female (fig. 2), and preparation for opening the body cavity, first introduced in 1493, in connection with the Incipit Anathomia Mundini. The other illustrations show: the circle of 21 urine glasses, with circles indicating the four temperaments; the signs of the zodiac as related to parts of the body, as in the figures in the old almanacs ; the points on the body for blood letting; a sick man on a couch; the wounded man, showing cuts, impact of clubs, etc. The drawings were made by Petrus de Montagnana.
Fig. 2, reduced from the edition of 1491, gives a fair conception of the quality of the pictures. This figure is borrowed from Wieger, since I have not had the edition of 1491 for examination, but have examined the corresponding figure in various editions beginning with that of 1495. The sketch shows in outline the position of the viscera; the uterus is represented as opened and containing a foetus. In 1493 the drawing of the female figure was modified by observations, and after that date the illustration bears the inscription 'Tratta dal Natura. '
A very interesting connection between the printed copies of this book and its manuscript sources has been brought to light by Sudhoff. He found about 1907, in the Bibliotheque Nationale at Paris, a neatly written Latin manuscript of quarto size, and 54 leaves, which belongs to about the year 1400. In this manuscript is a complete series of the Ketham pictures of 1491, and much of the Ketham text. After folio 45 in this manuscript is an anonymous treatise that agrees substantially with the Fasciculus Medicine of 1491. The text and figures of the Paris manuscript are not assembled as in the Ketham of 1491, but the text is in places identical, and the printed figures are evidently copies of the manuscript sketches. The way in which this collection of writings came to bear the name of Ketham is a matter of conjecture. Sudhoff thinks likely that there was a Johannis
Fig. 2 From Ketam's Fasciculus Medicine, Venice, 1491 (after WiegerJ
Fig. 3 The skeleton of R. Helain, Niirnberg, 1493 (after Wiegei-;
ANATOMICAL ILLUSTRATION 957
de Ketham, who, about a century before 1491, assembled the drawings and text, and, that when printed for the first time they bore his name, but of this we have no certain knowledge. This whole collection is. probably derived from earlier manuscript sources in French, German and Italian.
Prior to the publication of Ketham, there was printed in 1485, in De proprietatibus rerum of Barthalomaeus Anglicus, a woodcut of some anatomical interest. Standing in front of a walled garden is the figure of a man with the abdominal cavity opened and a very diagrammatic representation of the viscera. Within the garden the figure of Eve is appearing before the Lord from the side of the sleeping Adam.
R. Helain. Anatomical figures on separate plates were published as early as 1493, the first one to appear being a representation of the skeleton. It is probable that plates of this kind were exposed in barber shops and bath establishemnts, and that they were also purchased by medical practitioners and by the curious of the general public. A cut of the earliesr known picture of this kind is shown in fig. 3, which is copied from Wieger, although I have since seen a copy of the original in the library of Dr. Mortimer Frank of Chicago. It is attributed to a Paris physician, Richard Helain, and w^as printed in Niirnberg in 1493. Whether or not it was also printed in Paris is not known. The original plate was 53 cm. high. It seems to have been drawn from a partly dried specimen and the drawing is in many particulars fantastic. Among the curious features are, the dark abdominal portion, the expanded pelvis, the divided lower jaw and numerous teeth (17 on the lower jaw), the bones of the feet and the 'oslaude' of the skull. This 'os laude' or 'os capitale relaude' is an apochryphal bone, and its designation will puzzle those acquainted with classical Latin not a little. We might expect it to be os laudis but in the corrupted Latin of the period the termination e is commonly used for ae and we conclude that it is ^os laudae.'
This anatomical plate is referred to by Hyrtl, Wieger, and others as the work of Ricardus Hela. There is probably a mistake in the name, since Sudhoff , by a careful search of the records of the Paris physicians of this time, was not able to find the name of
958 WILLIAM A. LOCY
Hela, but instead found that of Richard Helain. The plate should probably be attributed to him. This picture formed the basis for a modification by the pubhsher Griininger in 1496-97 (fig. 4) which was printed in Brunschwig's Chirurgie, in 1497, and in various other texts. The picture (fig. 4) is however taken from Phryesen's Spiegel der Artzney, 1519.
The earlier manuscripts show a considerable number of sketches of skeletons, some of which resemble the drawing of Helain. (See Sudhoff, Studien zur Geschichte der Medizin, Hft. 4.)
Petrus d'Abano. In the Conciliator differentiarum philosophorum of Petrus d'Abano there appeared in 1496 the first printed illustration of the abdominal muscles. This is shown, considerably reduced, in fig. 5, w^hich is taken from a reproduction of the woodcut by Sudhoff in its original dimensions (5| x 6| inches). I have had for examination the 1526 edition of the Conciliator in which the same figure occurs, slightly modified and reduced in size. In that edition, in the 199th differentia, on page 231, is a description of the eight abdominal muscles. The picture is evidently made with the help of a dissection. There were earlier editions of the Conciliator in 1472, 1476, 1483 and 1491, but there is no picture in any of these; the first illustrations occur in the third Venetian edition of 1496. There is also a manuscript edition of the Conciliator near the beginning of the 14th century that speaks of eight abdominal muscles from Greek and Arabian sources.
Sudhoff, in pointing out that the figure of 1496 was based on a dissection, locates that dissection in Bologna. He found a copy of Mundinus wdth a marginal pen drawing of these muscles by a student, dated Bologna, 1494.
Brunschwig. In the interval between 1496 and the appearance of Peyligk's illustrated treatise came the publication of Brunschwig's Chirurgie in 1497. A few pages of this is devoted to anatomy, and in it w^e find a picture of the Griininger skeleton (fig. 4), which was a modification of the Helain skeleton of 1493, and also a picture of the wounded man showing visceral anatomy. This picture is one of the series showing the de\ elopment of illus
Fig. 4 The Gruninger modification (1497) of the skeleton of Helain. Printed
in Brunschwig's Chirurgie, 1497, and in other later texts. This cut from Phryesen's
Spiegel der ArtzBey, 1519
Fig. 5 First i)rinted .sketch of the abdominal mu.scle.s, from the Conciliator
DilTerentiarum of 1496 (after Sudhoff)
ANATOMICAL ILLUSTRATION 961
trations of internal anatomy. It is reproduced by Sudhoff in the Archiv fiir Geschichte der Medizin, Bd. 1.
Peyligk. In 1499 was published the Philosophie Naturalis of Johannes Peyligk which contains ten figures of separate organs of the body besides one large figure showing internal anatomy of head, thorax and abdomen. Peyligk's book is the compilation of a jurist of Leipzig. It is a fine folio of 96 leaves, 8^ x llf inches, with the letterpress 4f x 8 inches. The last twelve pages are embraced under the title Compendiosa capitis physici declaratio principalium humani corporis, etc., and it is this part alone that contains the anatomical illustrations. The frontispiece, which is printed on the reverse of the last page of the Philosophie Naturalis, is shown reduced in fig. 6, the original being 3| x 7f inches. In this figure we see the three cavities (venters) of the body indicated: the upper (supremis), containing the animal members; the middle (medius), containing spiritual members and the lower (inferioris), containing the natural members. The head shows only the ventricles of the brain as conceived of at that time. The thoracic cavity has a diagram of the lungs, the heart, the trachea and the oesophagus. Below the diaphragm, which is indicated as an oblique line passing across the trunk, there is represented the stomach, the spleen, the intestines and the liver with two blood-vessels. The liver is represented with five lobes according to Galenic tradition, and the gall-bladder is shown as a pear-shaped vesicle on the liver. In addition to this large diagram of the organs in situ, the text is embellished with sketches of the separate organs. Fig. 7 shows a picture of the page containing the figure of the stomach, oesophagus and intestines. Fig. 8 shows the separate illustration of the heart; the manuscript notes in this copy are also to be seen. All these figures, manifestly, are diagrams and not sketches from nature. Since they are the earliest printed illustrations of separate organs, it is an interesting matter to locate their source. Are they purely fanciful sketches based on descriptions of earlier writers?
The source of Peyligk's figures remained for a long time undetermined, and the assumption was generally made that they were schematic mental pictures, derived from reading the anatomical
descriptions of Arabian and classical writers, and transferred to paper. The sketches are certainly schematic and show the influence of tradition but they were not produced by Peyligk. The speculation of, Stockton-Hough that they came from an illustrated Mundinus of 1498 is unfounded. There is no known illustrated Mundinus of 1498 and the suggestion is probably due to a confusion of the Mundinus text in Ketham's Fasciculus of 1495. Several of the sketches are now traceable to the diagrams of Henri de Mondeville, and used by him about 1304 in illustrating his anatomical lectures at Montpellier. Pagel made known in 1889 that de Mondeville had employed sketches and, in 1890, Nicaise reproduced the miniature sketches of entire figures showing internal anatomy. He saj^s that de Mondeville also made use of sketches of separate organs of which all trace had been lost. These separate sketches have now been unearthed and were published in 1908 by Weindler, and, in a separate article, by Sudhoff. Those reproduced by Sudhoff embrace eighteen manuscript figures, nos. 1 to 7 found in the Royal Librarj^ at Berlin and nos. 8 to 18 in the Royal Library at Erfurt. The resemblance of some of the figures of Peyligk to these manuscript sketches of de Mondeville, leaves room for no reasonable doubt that the latter were the sources from which the Peyligk figures were drawn. It is uncertain how the pictures of de Mondeville originated. Sudhoff suggests that possibly de Mondeville began illustrating his earliest lectures at Montpellier by making diagrams of traditional anatomical sketches. Of this we have no certain knowledge, but we have, at' any rate, the sketches of separate organs of de Mondeville to add to his miniature pictures of entire anatomical figures that were previously known.
Hundt. The next printed anatomical illustrations to come under notice are those of Magnus Hundt, a Leipzig anatomist. His Antropologium de hominis dignitate, etc., published in 1501, is a rare quarto of 120 leaves, with a letter-press 3j x 5f inches. It contains nineteen illustrations, one of which is printed twice. The sketch of the viscera in situ is shown in fig. 9. There is another large figure in the book (the one that is repeated) showing the ventricles and the general location of physiological function
Fig. 6 From Peyligk's Compendiosa, 1499
Fig. 7 Part of page from Peyligk, showing sketch of stomach and intestines, 1499
of the brain. The original of fig. 9 is 3| x 5j inches. It is in some particulars more crude than the corresponding figure of Peyligk. In the thoracic cavity one sees the undivided lung, the heart, the trachea and the oesophagus. In the abdominal cavity is the large many-lobed liver with the gall bladder on its surface, the pouch-like stomach, the spleen, the loops of the intestine, and, pushed to one side, the kidneys, the bladder and the testes. The blood vessel connected with the liver is the S^ena chilis' and the blood vessels to the kidneys are the 'venae emul
Fig. 8 Sketch of the heart from Peyligk, 1499
gentes.' In the figure that I have photographed the iris of the eye is black, while in that reproduced by Sudhoff from the copy in the Leipzig library the iris of the right eye is indicated by a white circle.
The book is provided with text-figures of separate organs copied from Peyligk's treatise. The figures, however, are not printed from the same blocks; they are nearly identical but careful inspection will show slight differences in the lines. Fig. 10 shows the sketch of the liver with a part of the text. It is the same as the corresponding figure in Peyligk but is not quite so carefully engraved.
Gregor Reisch. In the 1504 edition of Reisch's Margarita Philosophica there is an ilhistration (see fig. 11) showing new details in internal anatomy of the thorax and abdomen. Although the anatomy is very crude it is an improvement over the corresponding figures of Pejdigk and Hundt. The kidneys and bladder are represented in a more nearly normal position. The lungs are
Fig. 9 Visceral anatomy from Hundt 's Antropolosium, 1501
divided into lobes; the liver, stomach, spleen and intestines are still very untrue to nature. The nomenclature of the period is shown in the names attached to the organs, the lung, 'pulmo,' the heart, 'cor,' the liver, 'epax,' the kidney, 'ren,' the bladder, 'vesica,' etc. In 1503 a similar picture had appeared in the edition of the Margarita Philosophica from the printing house of
Fig. 10 Part of page from Hundt's Antropologium, 1501
John Schott. In this earher picture the ureters are not shown, and the intestine is represented as connected with the bladder. I am indebted to Wieger's treatise for the sketch of 1504, that appeared in the Margarita Philosophica published by Grieninger. This figure was reproduced in other texts and the original of this cut (fig. 11) is in Brunschwig's Destillirbuch.
Leonardo da Vinci. With Da Vinci we come to the one man who, before Vesalius, showed independence in observation and
Fig. 11 Visceral anatomy from Reisch's :\Iargarita Philosophica, 1504. This cut from Brunschwig's Destillirbuch, 1512 (after Wiegerj
notable fidelity to nature in his sketches. Although the larger number of his anatomical drawings were made about 1510 they were not fully published until 1898 and 1901. They bear internal evidence of having been made from actual dissections. It is well known that he became associated with Delia Torre who projected a treatise on anatomy for which Leonardo was to supply the drawings. Nevertheless, Da Vinci had studied anatomy independently before his association with Delia Torre, and he contin
Fig. 12 Anatomical sketches from I Manoscritti di Leonardo da Vinci, 1510 (after the Paris facsimile edition)
ued to make dissections and anatomical sketches after the death
of the latter in 1506. We may assume that his method was improved by intimate collaboration with a professional anatomist,,
but we must recognize that this extraordinary man was a master
unto himself. The association with Delia Torre was not merely
that of an artist working under an anatomist who exposed the
parts and required sketches made under his direction. It was
Fig. 13 Anatomical sketch by Leonardo da Vinci, 1510
rather the cooperation of two anatomists, one of whom was gifted with great powers of artistic delineation. Antonio de Beatis had it from Leonardo's own lips, about 1510, that he had dissected not less than thirty human bodies, both male and female. Leonardo projected a comprehensive work on anatomy of which he speaks in his History of Painting, and also in his manuscript notes. The notes and drawings bear testimony that this treatise was not designed merely for artists but was to be, as well, a work for medical students and for the professional anatomist.
The working drawings and notes for this projected work are preserved as a part of the manuscript collection in the Royal Library at Windsor Castle. They were published in Paris in 1898 as Foglio A of Leonardo's IManoscritti. This sumptuous volume
Fig. 14 Anatomical sketch l)y Leonardo da Vinci, L510
contains 245 anatomical sketches, reproduced as fac-similes both as regards the sketches and the paper upon which they are drawn. The notes are translated into French. His other anatomical sketches, also in the library at Windsor Castle, were published
Fig. 15 Anatoinicul sketch by Leonardo da \'inci, 1510
in eight additional volumes in 1901. This does not include the
volume on the anatomy of tlie horse. The range of the drawings
is astonishing; the entire collection embraces more than 750
separate sketches, some of them being several times repeated.
The notes accompanying the sketches, always written from right
to left, are, usually, descriptions of the figures, but, sometimes,
are general reflections regarding the plan of his projected book.
That he read anatomy is evident since he specifically corrects
some misstatements of Mundinus. Leonardo ]:>laced great reli
Fig. 16 Sketches of sections of the brain by Da Vinci, 1510.
ance on good figures, declaring them to be essential to the understanding of anatomy. Some of his delineations of muscles haA^e been so frequently reproduced that they are well known, but it is not so generally known that he made deep dissections of all kinds including the viscera and the brain.
The reproduction of a few of Leonardo's sketches will serve to show their quality, and will at once reveal the fact that the3' are totally different from any other sketches of the period. These drawings are not made from anatomical descriptions, but from
Fig. 17 Plate of tlie skeleton published by J. Sehott, 1517 (after Wiegei-;
original observations. The drawings speak for themselves, accordingly brief comments will suffice.
Fig. 12 shows a page with four separate figures and manuscript notes. The heart and chief blood vessels of the thorax and abdomen, and the sketches of the lungs are obviously made from actual dissections of the human body.
Fig. IS Anatomical sketches from Phryesen's Spiegel der Artzney, 1517.
This cut from a Dutch edition of 1519
Fig. 19 Title page of the Commentaries on Mundinus by Berengarius, 1521
ANATOMICAL ILLUSTRATION
977
DE ANATOMtA
wm'?^^^^m:immi£Mm)
Fig. 20 Figure of the skeleton from the Isogogse Breves of Berengarius, 1523
978
AVILLIAM A. LOCY
gura uidcnf
ofTaptis po/
fterioris hois
uidef et due
calue i in qtu
dextraurco/
ronalis pmif
fura'qeiprc
fupiori:ui&
Cigitt3lis:qc
imcdioiufcc
laude comif/
fura:qelpce
iftnoxi'.A late
rib'* et uidef
due comiffU
reamefupra
noiate in ana
tomia cranei
q funt fupra
pmifluras fqf
mofasexiitef
J>pe'aufes;r5
ifte font fete
ifenfibiles : a
finjfttisealia
calua*. in q ui
dc£ madibu/
Iej& pats Co/
miffute coto
nalist&'due
comiffute in
fta fagtttale
ad unu latus
exiftentes: &
uf unu OS de
duob' ofTib' _ ^™„_— _««>,-«..-_.
paris:qdr eft a regioc cciUi feu ab oflc di^u vomu hzki tidcn'ij^Uiu cv
pitisuccfasamein.
Fig. 21 Posterior view of the of Berengarius, 1521
skeleton from the Commentaries on IVIundinus
ANATOMICAL ILLUSTRATION 979
Fig 13 shows the chief blood vessels of the neck and the adjacent region.
Fig. 14 shows a deep dissection of the blood vessels of the thigh.
In fig. 15 we have a rather comprehensive dissection of the thorax and abdomen with the alimentary canal removed. The original of this figure is 13^ x 22 inches.
A limited number of drawings can give no adequate conception of Leonardo's work in anatomy. His sketch of the stomach and intestine is a good drawing of the relative size and the normal arrangement of these viscera. In the delineation of muscles it is not merely the superficial layers that engage his attention, he shows details of the arrangement of the tendons on the toes and fingers, a number of cross-sections of the leg at different levels, the muscular architecture of the heart, etc. Among his many pictures of the bones, he correctly draws vertebrae from various aspects, and the bones of the fore-arm in pronation as well as in other positions. He made sketches of the dissection of nerves. His figures on generation show uteri opened, with contained foetuses, and the placental connection.
Before leaving his work, however, we should have one of his sketches of the brain as shown in fig. 16. Here one sees, on the left, a median sagittal section, and, on the right, a horizontal section. These sketches show fairly the extent to which the brain had been dissected up to the year 1510.
Other contemporary or nearly contemporary artists, as Michael Angelo, Raphael, and Albrecht Dtirer, made anatomical sketches, but not so comprehensive as those of Da Vinci, and the details regarding which it is not necessary to consider.
Johannes Schott. In 1517 there appeared from the publishing house of John Schott at least two anatomical plates, one representing a skeleton and the other a sketch of the internal anatomy of the body. The picture of the skeleton is shown in fig. 17. It is still very crude in its execution, but in some particulars is an improvement on the earlier printed figures. The skull is better drawn than in the plates of Helain and Griininger (figs. 3 and 4), but it still shows the spurious 'os laude, sive capitale.'
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 4
980 WILLIAM A. LOCY
There are other marked deficiencies as in the arm and wrist, where the carpal bones are enumerated as eight, but are not drawn, etc., etc. The sketch of the internal dissection was published in several texts as in the Phryesen of 1518 (see Chievitz, p. 90), 1529, etc. Choulant reproduces a similar but not identical figure, also from Phryesen, the plate of which bears the date 1517.
Phryesen. (Fries, Friesen, etc.) In 1518 there appeared in Strassburg the Spiegel der Artzney of Laurentius Phryesen, containing two plates. The one is a copy of the Griininger skeleton (see fig. 4), and the other a visceral anatomy, surrounded by six figures of the anatomy of the brain and one of the tongue. This cut appears in the different editions of Phryesen with some modifications. Fig. 18 shows a copy of this plate from a Dutch edition of Phryesen dated Strassburg, 1519. The original woodcut is 5^ X 7f inches. The edition of 1529 contains another picture of the visceral dissection, — the same as shown in Chievitz, fig. 31, — that lacks the marginal sketches of the brain, and is also somewhat different in other details. The figures of the brain in the Spiegel der Artzney, except for those of Leonardo, are a new departure in anatomical illustrations.
Jacohus Berengarius Carpensis (Carpus). Berengarius has often been heralded as the greatest anatomist between Mundinus and Vesalius, and, if we except Da Vinci, the assignment of this rank to him is perhaps justified. Whatever may be said of his alleged dissection of more than one hundred bodies, the illustrations of Berengarius are not original, nor are they based on good observation. They bear resemblance to sketches in the manuscripts of the fourteenth century and to printed pictures in earlier publications, as the Conciliator differentiarum (1496), Margarita philosophica (1504), etc. As hfas already been said, we find that all sketches of the period, with the sole exception of those of Da Vinci, show interrelationships with manuscript illustrations as well as with earlier printed figures. As Roth has pointed out, the anatomical writings of Berengarius are compilations without credit being given to the original sources, and there is inharmony between his text and the illustrations, — a circumstance that is, at times, adverted to by himself. It is altogether likely that the
ANATOMICAL ILLUSTRATION 981
cuts were inserted by his publislier from such pictures as were available.
The first anatomical publication of Berengarius was an extensive series of commentaries on Mundinus. In this the text of Mundinus is printed in larger type, and the forty commentaries in smaller, but so extensive are the annotations that the book is brought up to a thick quarto volume of 1056 pages. This book, published at Bologna in 1521, is rare and a cut of the title page is shown in fig. 19. The size of the original is 4f x 7 inches; the border is red and the enclosed printing black. His commentaries contain at times corrections to Mundinus, and show the results of some observations mixed with dialectic compilations from the earlier writers. In the 21 illustrations the dependence on tradition is very marked.
He soon branched out for himself and wrote an introduction to anatomy, designated Isogogse breves, etc., which was first published in 1522 and followed by a modified edition in 1523 which is the only one well known. In addition to a copy of his commentaries of 1521, I have had for examination three editions of the Isogogse breves; that of 1523, Bologna, 4°, 80 leaves with 23 woodcuts; an edition of 1535, Venice, 4°, 63 leaves, 19 plates, and a small pocket edition (2f x 4j inches letter-press), dated 1530, and containing 24 figures. The illustrations are wretched copies of those of the edition of 1523, the increase in their number, by one, is owing to the separation of tw^o figures that appear on one plate in the larger edition. This appears to have been a relatively cheap edition for students.
More than one-half the illustrations of the commentaries are reproduced in the Isogogse of 1523 and new ones are added. Most of the plates in the edition of 1523 are provided with an ornamental border, added to a double line boundary, while the plates of the commentaries of 1521, are limited by a single line border. Roth reproduces a full-size figure of the skeleton from the Isogogse of 1523, but his plate lacks the ornamental border.
Fig. 20 is a representation of the skeleton from the Isogogse of 1523 in which the ornamental border has been retained, but the marginal description, present in the original, has been cropped
982 WILLIAM A. LOCY
off. This curious figure has 13 ribs, widely expanded pelvis and a spurious fissure in the frontal bone.
The posterior view of the skeleton, shown in fig. 21, is taken from the commentaries of 1521. It has a single-line border and the marginal note has been retained. The basin-like pelvis appears more fantastic than in the preceding figure. The skull shows two spurious furrows on the parietal bones, the presence of which seem to have confused Berengarius. The two best illustrations in Berengarius are those of the bones of the hand and of the foot. The close resemblance of these pictures to drawings of Leonardo da Vinci gives ground for the suspicion that they were in some way based upon his sketches. Although this is a mere conjecture, these two figures are on a different plane of accuracy from any other illustrations in the Isogogse breves.
Dryander (also known as Johann Eichmann). This professor of anatomy at Marburg published, in 1537, an Anatomiae h. e. corporis humani dissectionis pars prior, etc., illustrated by 20 plates that were based on dissections. I have not seen a copy of this work, but have examined his edition of Mundinus and other earlier writers, published in 1541, which contains most of these earlier figures, some new ones, and 18 figures copied from Berengarius. The copy at my disposal contains 45 figures, one plate of which is repeated. Some of Dryander's illustrations are a considerable improvement on those of Berengarius. Fig. 22 shows his sketch of the alimentary tube, the original woodcut being 4j x 6 inches. The drawing of the caecum and the vermiform appendix shows that it is based on observation, but the figure is not so good as that of Da Vinci of the corresponding parts.
Walther Hermann Ryff- In 1541 appeared Ryff's Anatomi with a very long and cumbersome title. This book, of which I have examined a copy in Dutch and one in German, was published after the first plates of Vesalius (1538) and before the appearance of the famous Fabrica (1543). It, with the Dryander mentioned above, lies on the border line of pre-Vesalian illustrations of anatomy. One of Ryff's illustrations of the arterial circulation, reproduced in fig. 23, gives a fair idea of the appearance of his sketches.
ANATOMICAL ILLUSTRATION 983
Other anatomical illustrations of the pre-Vesalian period embrace the very rare copper plates of Canano, showing bones and muscles of the arm. Choulant says that, prior to 1543, one book
Fig. 22 Anatomical sketch from Dryander's edition of Mundinus, 1541
of the work was published, containing 27 illustrations, but the work was never completed. Between 1536 and 1543 there were several plates of anatomical figures placed on the market. These so-called Tliegende Blatter' include the six Tabulae anatomicae
984
WILLIAM A. LOCY
Mmtmi/€mufacm m kfc5m()un|
Fig. 23 Anatomical sketcli from Ryff' s Anatomi, 1541
of Vesalius that appeared in 1538 and were a forerunner of his great work of 1543.
There were also during the period other anatomists of more than usual insight, as Achillini, whose anatomical treatises were
ANATOMICAL ILLUSTRATION 985
not illustrated, and, therefore, do not properly come under consideration here.
Summary. The chief printed illustrations of anatomy before Vesalius may now be chronologically arranged, omitting different editions of the same work with slight modifications of the figures :
1491. The arrangement of the viscera in the human female in Ketham's Fasciculus Medicine.
1493. The skeleton of Richard Helain.
1493 (?). A demonstration of visceral anatomy in the Melerstat edition of Mundinus.
1496. The abdominal muscles, in the Conciliator differen tiarum.
1497. Plate of the Griininger skeleton.
1497. The wounded man with internal anatomy in Brunschwig's Chirurgie.
1499. Anatomy of the three body cavities, together with figures of the separate organs, in Peyligk'sCompendicsa.
1501. Similar illustrations in Hundt's Antropologium.
1503-'04. Organs of the thorax and abdomen in Reisch's Margarita Philosophica.
1510 (?). Leonardo da Vinci, more than 750 sketches of human anatomy; not, however, pubhshed.
1513. Mundinus, zodiacal signs and rough sketch of the heart.
1517. Plates of visceral anatomy and of the skeleton published
by Johann Schott.
1518. Skeleton and visceral anatomy in Phryesen's Spiegel
der Artzney. 1521, '22-'23. Berengarius, commentaries on Mundinus and Isogogse breves.
1536. Fliegende Blatter.
1538. Six Tabulae Anatomicse of Vesahus.
1537. Dryander, Anatomiae, corporis humani, etc.
1541. Dryander, edition of Mundinus, with 45 illustrations. 1541. Ryff, Anatomi.
986 WILLIAM A. LOCY
The survey of these printed sketches of anatomy, covering a century-and-a-half before Vesalius, brings into notice the relatively slow progress. While we remember that this is the period of the awakening of the scientific spirit, still, the drama of intellectual progress does not unfold as rapidly as we might expect. ^Vhy, after the revival of dissection under Mundinus, and why, especially, after the introduction of printing, was there not more rapid progress? Some seek to find an answer in the difficulty of getting material for dissection and others in the opposition of the church, but the thing that held anatomical science in check, was not so much the lack of opportunity to dissect as the mental habit of the time. The disposition to dissect was not especially strong. That internal hunger for the analysis of nature at firsthand was not of dominating insistence. The effects of tradition and of education had to be overcome, and the gradual assimilation of new methods and new ideas was necessarily slow. Those who would have done better under gifted and inspired leaders were perplexed and too closely bound by the mental habit of the time to map out and follow an independent course. Thus, the retarding influence was generic rather than specific. Independent spirits of great originality were rare then, as now, and it seems natural that the habit of imitation should have so long perpetuated anatomical sketches of poor quality. Da Vinci was the only man whose product exhibits great originality and independence. His anatomical work was on the plane of that of Vesalius but his sketches were not printed until long after.
The practice of dissection by medical men was not so actively opposed by the church as is generally supposed. A superficial reading of the bull of Pope Boniface, de Sepultis, issued in 1300, has led to the statement that it was directed against the practice of dissecting for scientific purposes, but it was, in reality, a proscription of the practice of dismembering the bodies of dead Crusaders, in order that their bones might be more readily transported home for burial in consecrated ground.
The practice of plagiarism was widespread during this period. Publishers and authors engaged in it in a wholesale way; both sketches and text were commonly copied without credit being
ANATOMICAL ILLUSTRATION 987
given. The ethics of the rights of intellectual property were unrecognized. The earliest printed sketches were derived, from manuscript sources and, these, in turn, were based upon the traditional anatomy, chiefly of Galen and his commentators. Now and then a touch of original observation was added to the traditional figures but they were not perfected. Dependence on authority was still the deep-seated method of the intellectual life, and the rise of independent observation was slow. But, the better intellects were opposing it, and with all these limitations the light of the renaissance was breaking. Dependence on authority was giving way, and, finally, thanks to the work of his predecessors, Vesalius was able to establish a new method based on observation and reason. With the publication of his Fabrica in 1543, there was ushered in the era of good illustrations of anatomy. The prevailing mental habit of the time was now at least partly overcome and the era of independent observation was started.
WILLIAM A. LOCY
BIBLIOGRAPHY
The ^uU titles of the printed books from 1491 to 1543, containing anatomical cuts, and listed in the body of this paper as 'Sources,' are often long and cumbersome. They will be found in the Catalogue of the Surgeon General's Library, in the lists of printed books in The British Museum, in Hain's Repetorium Bibliographicum, in Haller's Bibliotheca Anatomica, etc. The other books used chiefly as references are:
Ball 1911 Andreas Vesalius the Reformer of Anatomy.
Baas 1889 Outlines of the history of medicine. New York.
Chievitz 1904 Anatomiens historic. Copenhagen.
Choulant 1852 Geschichte der anatomischen Abbildungen. Leipzig.
HoPF 1904 Die Anfange der Anatomie bei den alten Kulturvolkern, in Abhandl. zur Ges. der Medizin. Breslau.
Pagel 1898 Geschichte der Medizin. Berlin.
PuscHMANN 1902-05 Handbuch der Geschichte der Medizin. Jena.
Roth 1892 Andreas Vesalius Bruxellensis. Berlin.
SuDHOFF 1907'-08 Various articles in Studien zur Geschichte der Medizin, Leipzig; and in Archiv fur Geschichte der Medizin. Leipzig.
TofLY 1903 Geschichte der Medizin, in Puschmann's Handbuch. Jena.
Weindler 1908 Geschichte der gynakologisch-anatomichen Abbildung. Dresden.
WiEGER 1885 Geschichte der Medizin und ihrer Lehranstalten in Strassburg vom Jahre 1497 bis zum Jahre 1872. Strassburg.
MINIMAL SIZE REDUCTION IN PLANARIANS THROUGH SUCCESSIVE REGENERATIONS
S. J. HOLMES
It is a familiar fact that very small pieces of fresh water plan arians will regenerate and give rise to minute individuals closely resembling the original form. I have endeavored to ascertain how far reduction of size in Planaria maculata may be carried without causing a failure to give rise to a normal individual. With very small pieces of an adult planarian there is a large proportion of cut surface which produces an injurious effect and there are also various mechanical impediments to regeneration; these facts, combined with the specialized condition of much of the tissue, conspire to restrict the regenerative capacity of the parts. In order to eliminate somewhat these factors the device was resorted to of subjecting the animals to a number of successive divisions. A planarian was cut into fifteen or twenty pieces; when these had regenerated into small planarians they were again cut into several pieces. These regenerated into still smaller individuals which in turn were divided, the process being continued until forms were reached which were so small that complete regenerations were no longer obtained. In this way, wheri the minimal size limit was approached, regeneration became very slow and many of the pieces lived for months without restoration of the missing parts. As a general rule it may be said that the smaller the piece the more slowly the restorative processes take place. In this way the proportion of cut surface was reduced, the tissue kept in a more plastic condition, and the whole process of regeneration made easier.
Eugene Schultz has studied the reduction in size of planarians from starvation. He found that planarians could be reduced in this way to one-tenth or one-twelfth of their original size. A study of the size of cells of various kinds showed that there was
990 S. J. HOLMES
little reduction in size either of the cells or their nuclei as a result of starvation; the diminution in the size of the body was produced mainly by their reduction in number. Various organs suffered unequally in this process. Organs of copulation, sex ducts and vitellaria were among the first to disappear, the eyes degenerated, and there was a marked reduction in the number of parenchyma cells. The muscle cells suffered little decrease and the number of muscle bands remained unaltered; there was little reduction in the nervous system. The male sex cells were among the least altered. Cells of the intestinal epithelium and the outer ectoderm, while reduced in number, were not reduced disproportionally to the body as a whole.
The study of small regenerated planarians was undertaken in order to ascertain how far the various organ systems would suffer on account of reduction in size and how far the results might be parallel with the effects of starvation. Through successive regenerations it is possible to carry the reduction very much farther than can be done by the withdrawal of food. While starvation may reduce the animal to one-tenth or one-twelfth its original size, by the method of successive regenerations it may be reduced to roVo or TsVo its original size. Many of these very minute animals had practically the same form as the adult. Several specimens were sectioned and careful measurements were made of several kinds of cells and compared with measurements of corresponding cells of individuals of ordinary size. Ectoderm cells, parenchyma cells, cells of the intestinal epithelium were of the same size as in the larger worms. The muscle cells, while less in length, were nearly as thick as in the larger worms. The nuclei were also not reduced in size and therefore bore the usual relation to the size of the cells. The gonads, sex ducts, copulatory apparatus and vitellaria could not be found. The muscular system is well developed, the outer layers being present and only a little thinner than in normal individuals. The number of dorsoventral strands in a cross-section is not more than about one-fourth that of larger specimens and they contain fewer fibers. The alimentary canal has but a very few short branches. The cells are not shrunken as occurs in starved individuals and
MINIMAL SIZE IN PLANARIANS 991
cross sections of the diverticula appear much as in the larger worms except for the reduction in the number of cells. The size of the brain and the diameter of the nerve cords bear about the same ratio to the rest of the body as in large individuals. In a few cases there was but one eye instead of two and this was not a median one as sometimes occurs in small individuals but was in the position of one of the lateral eyes. The eye is of normal size, in relation to other parts and it has essentially the usual structure, but there is a great reduction of the number of the retinal cells.
The pharynx, which bears about the same relative proportion to the body as in larger planarians, is composed of the same epithelial and muscular layers. The relative proportion of the parenchyma and digestive organs is little altered in the smaller individuals. The relative thickness of the outer epithelium is however much greater, since it is composed of but one layer of cells which have the same size in the large and the small planarians. Pigment cells occur sparsely scattered over the dorsal surface and appear of enormous size in relation to the rest of the body. On the whole, the small individuals are strikingly like the larger ones in general form and the relative proportions of the systems of organs.
Observations were made on the movements and reactions of these minute forms. Methods of locomotion, exploring movements of the head, reactions to light and contact, responses to mechanical stimuli, righting movements, and various other activities, even down to the most delicate details, were carried out in practically the same way as in individuals of normal size. These facts indicate how effectively the functional unity of the organism is maintained notwithstanding the enormous reduction in the number of its cells.
One factor which probably determines the mininal size which may be attained is the fact that the size of the cells cannot be reduced and there must be a certain number of kinds of cells to preserve the physiological unity of the organism. There must be nerve cells, muscle cells, parenchyma cells, epithelium, etc., if the planarian is to be a planarian. The work of the organism, like
992 S. J. HOLMES
that of a factory, can be performed on a large or a small scale, but as there are many functions to be discharged, and as one cell cannot do the work of another, a point naturally has to be reached somewhere when a further reduction of the number of cells brings operations to a standstill. Matters might work out, however, in a different way by effecting a general simplification of structure, such as occurs in the reduction of Hydra. This simphfication of structure, which has been compared to a reversal of embryonic development, does not proceed in the planarian very far. The loss of sex ducts and associated organs is of doubtful significance, since these parts often atrophy at certain periods in adult individuals. In attempting to carry reduction below a certain size the cut ends of the pieces heal over and there is little further change; the organism does not transform itself into a simple embryonic stage, and we cannot with safety speak of the reversal of developmental processes (if it be really such) beyond perhaps a few retrogressive steps.
THE GEOTROPISM OF PARAMOECIUM
E. H. HARPER
From the Zoological Laboratory, Northwestern University
FIVE FIGURES
Are there any free-swimming organisms which are oriented to gravity by means of the shape or contents of the body in the same way as the axis of orientation of certain eggs is determined by the difference of specific gravity of the opposite poles? The body of Paramoecium caudatum suggests the possibility of its being oriented to gravity by the difference in buoyancy of the two ends. The posterior end is broader and the anterior end indented by a deep groove. Is the body of Paramoecium ' sternheavy, ' and if so, does it account for the ordinary negative geotropism of the animal, the anterior end having a tendency normally to point upward? Various explanations have been propounded for the geotropism of Paramoecium, such as difference in water pressure at different depths, difference in pressure on the lower and upper surfaces and finally Lyon,i ascribes the geotropic response to the internal st'muli of heavy particles, making the body of Paramoecium the analogue of the statocyst of higher forms.
To test the writer's hypothesis, it was thought it might be possible to increase the difference in specific gravity of the two ends by making the animals ingest heavy particles. These would at first lie toward the posterior end near the mouth, making the body more 'stern-heavy,' and so possibly increasing the negative geotropic tendency.
iLyon, E. P., 1905. On the theory of geotropism in Paramoecium: Amer. Joiirn. Physiol, vol. 14, pp. 421-432.
993
994 E. H. HARPER
EXPERIMENTS WITH SUBSTANCES OF HIGHER SPECIFIC GRAVITY
The experiments weie carried out as follows: The water used was ordinary tap water boiled to drive away gases. A control experiment was in all cases conducted side by side with the treated specimens. Ordinary test tubes were employed for the experiments. 'Iron by alcohol' was used, being ground up fine in an agate mortar. In order that the number of Paramoecia might be practically equal in the two test tubes, the water containing the animals was measured into equal portions. Large numbers were used in order to facihtate observation and comparison of aggregations. In the control test tube was placed a quantity of finely ground iron and after this had settled to the bottom, which occurs very quickly, the measured quantity of Paramoecia were introduced into the tube. The others were put in the agate mortar with finely divided iron and stirred for a definite time so as to keep the particles in suspension. In some cases the control animals were similarly stirred before placing them in the test tube in order that mechanical agitation should affect the results equally if at all, since this is known to change the reaction, sometimes, to positive.
The amount of iron ingested is readily observed with the microscope, and a short treatment may be sufficient to cause the ingestion of a considerable quantity. They were allowed in different experiments to ingest the iron for intervals varying from fifteen seconds to five minutes before transferring to the test tube The control animals and the treated ones were placed in the tubes at practically the same time, so that there would be no difference in time interval to allow for in comparing the movements and aggregations of the animals in the two tubes. As a method of recording the observations, which were made with the naked eye and hand lens, the approximate distribution was indicated on a diagram. This could be made sufficiently accurate to indicate any decided difference that might be noted in the regions of greatest aggregation as well as the general distribution in the two tubes.
THE GEOTROPISM OF PARAMOECIUM 995
Many factors influence the movements of Paramoecia so that some cultures aggregate quickly at the top in a ring and others remain distributed, and in some cases move downward. The various unknown factors have been disregarded and comparisons made only to determine whether the treated specimens exhibited a stronger negative geotropic tendency than the ones not treated. The placing of iron filings in the control tube was to eliminate any possible chemotactic factor. Finelj^ divided bismuth and nickel were also tried, but iron was found entirely satisfactory for the purpose of the experiment.
One condition of the experiments is that the iron filings after a time become distributed through the endoplasm more evenly, so that any difference in behavior is to be looked for before this change occurs. A second condition is that an excessive amount of iron may overload the animals apparently and cause them to aggregate at the bottom. The Paramoecia rid themselves of the iron after the course of a few hours without apparently harmful effects from a small amount.
One more precaution needs to be stated. The whole quantity of iron in the mortar must be removed with the Paramoecia, or a possible loss might occur of some individuals whose tendency was to remain on the bottom, or else the iron must be repeatedly rinsed to remove all of them.
The experiments have been repeated so often that certain results appear as typical, and these will be reported just as recorded.
Experiment 1 Control tube referred to as No. 1 ; treated individuals as No. 2. Ingestion of iron for five minutes.
Five minutes after the beginning of the experiment the distribution in the two tubes did not noticeably differ.
In ten' minutes there was about an equal collection gathered at the bottom of each tube. To observe the size of these aggregations better the tubes were slightly shaken.
Twenty minutes later there was a large collection at the top of No. 2, very few at the top in No. 1 . Looking down from above gave the appearance shown in fig. la; fig. lb gives a side view of the same.
JOURNAr OF MORPHOLOGY, VOL. 22, NO. 4 DECEMBER, 1911
996 E. H. HARPER
In this experiment, on account of the long treatment, a considerable number had apparently taken on an overload of iron, which would account for the aggregation at the bottom of No . 2.
In one hour and twenty minutes nearly all had collected at the bottom in both tubes, but a collection still remained at the top in No. 2, none in No. 1 (fig. 2)
Experiment 2. Animals from the same culture, same day. Ingestion of iron for thirty to forty seconds.
Fifteen minutes later, there was an aggregation in the bottom of No. 1, and a few scattered all the way up the tube In No. 2 there was a slight aggregation at the bottom and a considerable collection above the middle of the tube.
In twenty minutes the collection appeared as shown in fig. 3. The smaller number at the bottom of No. 2 was referable to the fact that less iron had been ingested than n the former experiment Notwithstanding the normal tendency to go downward, there was a decided tendency among the treated animals to go upward.
Experiment 3. This is a sample of those experiments with cultures in which the normal tendency was to go upwards and form a more or less dense ring at the top of the tube. Ingestion of iron for one minute. In ten minutes a much denser ring was formed at the top of No. 2, and this inequality remained (fig. 4). The Paramoecia in No. 2 moved more slowly than in No. 1. When the test tubes were corked and inverted, no air being allowed to enter, there was the same collection at the top, but not ring formation.
The interpretations of Experiment 3 might differ. The denser ring at the top of No. 2 might be explained as an entrapping of the animals in this region, resulting from their slower movements, on the principle by which Jennings explains aggregations due to chemical influences. At any rate the fact remains that the treated animals swarmed to the top more quickly and formed a denser ring there than in the control. These experiments are less crucial than those with animals having a normal tendency to go downward.
THE GEOTROPISM OF PARAMOECIUM 997
EFFECTS OF INGESTION OF SUBSTANCES HAVING A LOWER SPECIFIC GRAVITY
Various substances were tried, but finally paraffii^was selected as best adapted for the purpose. Finely di^^ided paraffin was obtained by melting and cooling in hot water several times till the paraffin was thoroughly washed. Then a small amount was melted in a flask of hot water and shaken and then suddenly cooled under the tap. The fine, suspended particles would soon rise to the top. For the control experiment paraffin in particles too large to be taken in was used.
Experiment 4. Both Nos. 1 and 2 were shaken at the same time for several minutes and the tubes then allowed to stand. In No. 2 the animals at the end of fifteen minutes were aggregated densely at the bottom at rest. They remained so, under observation, for several hours. After twelve to twenty-four hours all would be found scattered through the tube. In the control no noticeable effect was produced by the paraffin. The dense aggregation at the bottom of No. 2 was obtained with cultures of the different types (fig. 5).
The inference that the posterior end was buoyed up by the paraffin particles, so as to orient the animals downward, is the conclusion of the writer, for the present at least. *
DISCUSSION OF RESULTS
The above experiments seem to place the gravity orientations of Paramoecium on the same plane as the orientation of the axis of certain eggs by specific gravity.
Lyon compares Paramoecium to a statocyst. It is not easy to see how an animal revolving continually on its axis could react to the localization of an internal stimulus. While such a stimulus might conceivably act effectively in an antero-posterior direction, the explanation offered in this paper .seems simpler, at least to the writer.
Lyon centrifugated Paramoecia strongly into a tube ending in a capillary bore and found that the animals moved with the anterior end outward, and that certain heavier particles were driven into the anterior end.
998 E. H. HARPER
The inference that the anterior end is heavier is contrary to what the shape of the body would indicate, unless the heavier particles are located anteriorly. The writer wishes to suggest as an explanation of Lyon's experiment that in strong centrifugation the same effect is produced at the outset as by mechanical agitation, i.e., the reaction changes to positive, Jensen^ showed that with weak centrifugation the animals moved centripetally.
It is conceivable that the pull of gravity on the heavier posterior end may produce a tipping effect which is able to orient passively but is too weak to stimulate. When, however, the centrifugal effect exceeds the pull of gravity and produces too sudden an orienting tendency, this may act as a stimulus to the animal to resist as in the ordinarj^ rheotropic reaction against the current. When strongly centrifuged the animal takes a position so that it moves in the water just as the water moves past it in the rheotropic response. In other words, if it allowed itself to be oriented by the centrifugal force with the posterior end in advance, its relation to the water would be the reverse of what it is in the rheotropic response. The writer repeated Lyon's centrifugation experiments and the explanation here given, namely, that the animal is able to react at the outset of centrifugation, seemed to him the ntost satisfactory explanation of the fact that all are found to move with the anterior end outward.
So also in shaking, on account of the difference in buoyancy of the two ends, the heavier end will move more rapidly, and this may become effective as a stimulus, causing the movement downward.
If the explanation here given hold, we have in the normal, quiet, geotropic reactions of Paramoecium an example of a purely mechanical tropism. The term tropism would be here apphed to a passive orientation not involving the irritability. If it be desirable to use the term 'tropism' for such a kind of orientation,
^Jensen, P. Ueber den Geotropismus niederer Organismen; Arch. f. d. gesPhysiol., Bd. 53, pp. 428-480.
THE GEOTROPISM OF PARAMOECIUM 999
it is evident that it forms a separate class from those that involve a change of irritability. When, however, a stronger force is substituted for gravity so as to produce a sudden orientation, the irritability is affected, and the animal reacts to the change.
THE FORMATION OF THE SPERMATOPHORE IN
ARENICOLA AND A THEORY OF THE ALTERNATION OF GENERATIONS IN ANIMALS
ELLIOT ROWLAND DOWNING '
FOUR PLATES AND SEVEN TEXT FIGURES
CONTENTS
Location of the gonads 1002
Methods 1002
Limits of the gonads 1004
General statement of spermatophore formation 1006
Breeding habits 1007
Degeneration and phagocytosis of the gonad 1008
Extra blood vessels for aeration 1009
Origin of the spermatogonia from peritoneal cells 1011
Structure of the gonad 1012
The spermatophores, their formation 1013
Their discharge into the body fluid 1014
Their analogy to blastulae and gastrulae 1015
Their behavior when ripe 1016
The giant spermatogonia 1017
Spermatogonia! macromeres and micromeres 1018
The spermatophore originates from one primary spermatogonium by a process
of cleavage and invagination 1021
The spermatophore an individual — the gametozoon 1022
The alternation of generations 1023
Botanic use of the term 1023
Usual zoological significance 1023
Alternation and chromatin reduction 1023
Independent phenomena 1024
Chamberlain's theory of the alternation of generations in animals 1028
Objection to it 1029
Limits of the gameto- and sporo-generations 1029
Illustrated by graphic life histories 1030
Sexuality and reduction 1033
Adjacent phenomena 1033
Not necessarily causally related 1033
The primitive animal type and the development of the alternation of generations as found in Arenicola 1034
1001
1002 ELLIOT ROWLAND DOWNING
The gametozoon a 2x form 1037
The common plant and animal prototype 1037
The primitive position of the reduction phenomenon 1038
Its shift toward typical plant and animal positions 1038
Reduction and tetrad formation 1039
Beard's hypothesis of alternation of generations in animals 1039
Bibliography 1041
LOCATION OF THE GONADS
The gonads of the ArenicoHdae are located on blood vessels which run diagonally across the surface of the nephridia.
The typical somites of these worms are composed of five annuli, one of which bears the setae. Three annuli are anterior and one posterior to this setigerous one. There are six pairs of nephridia in Arenicola cristata, situated in the fifth to the tenth somites inclusive, a pair for each somite. Each nephridium consists of a funnel, body and bladder. The funnel has a sagittate dorsal lip set with ciliated plates and a lobed convex ventral lip; between the lips is the nephrostome. The body is club-shaped and connects, near its larger anterior end, with the funnel, at its posterior end with the bladder. The bladder is roughly spherical and opens to the exterior by a narrow tube through the nephridiopore.
Each funnel is attached by the outside of the dorsal hp to the ventral surface of an oblique muscle, at some little distance from the attachment of the muscle to the body wall; so that the apex of the arrow-shaped funnel points downward and forward. The body of the nephridium passes up, back, and outward from its juncture with the funnel, to the bladder, which lies against the body wall close to the line of insertion of the oblique muscles to the sides.
METHODS
The ordinary method of pinning the animal out for dissection so pulls the nephridia that the shape, particularly of the delicate funnel, is distorted. The method followed has, therefore, been (1) to stupefy with 70 per cent alcohol, adding it rapidly, drop by drop, to just enough sea water, in a long dish, to cover the animal. Stupefaction, with complete relaxation of the powerful muscles of the body wall ensues, in A. cristata, in from ten to thirty
THE SPERMATOPHORE IN ARENICOLA 1003
minutes; in A. claparedii, in from three to eight minutes. (2) With a hypodermic syringe, sufficient preserving fluid is injected to distend the body ; the whole worm is then immersed in the preservative. After hardening, the nephridia, when dissected out, have the shape and relations described above.
Details of the methods of preservation will be reserved for a later paper in which the chromatin changes and other histological matters will receive attention. It will be sufficient now to state that testes preserved in strong Flernming, Bouin and vom Rath fixing fluids have given best results. The forming spermatophores have been studied in fresh body fluid, stained with methylen blue or neutral red; or in smear preparations killed by exposure to osmic acid fumes and fixed in Merkel or in vom Rath; or from sections prepared by squirting body fluid, freshly drawn, immediately into hot corrosive-acetic or Flemming, then hardening in alcohol in the usual way and sectioning in paraffin. Warm iron haematoxylin and Bordeaux red or saureviolett and fuchsin have been among the most successful stains.
As the author has elsewhere published a description of the relations of the blood vessels to the nephridia in the Arenicolidae, the following brief description will suffice here. Each nephridium of A. cristata is supplied with blood by a branch of the ventral blood vessel. This afferent vessel, on approaching the nephridium, branches to the setal sac and gill (if present) , to the integumentary vessels, notably the dorsal-longitudinal, and to the nephridium. The branch to the nephridium enters the anterior angle of the sagittate funnel and after traversing it, passes on to the upper surface of the body of the nephridium, which it crosses diagonally from the anterior inner to the posterior outer side. Peripherad to the nephridium, it joins the nephridial longitudinal vessel. From the point of emergence from the funnel on to the body of the nephridium to its juncture with the nephridial longitudinal vessel, the blood vessel is designated the gonadial vessel, since upon it the gonad is found. Gonadial tissue is also found, to a slight extent, upon the nephridial longitudinal just anterior to its attachment to the gonadial vessel (plate figs. 1-6). These figures show the location and relative size of the gonads in the several species.
1004 ELLIOT ROWLAND DOWNING
They are taken from typical nephridia and are drawn to the same scale. In each case the individual selected was an average sized worm and was taken at a time of the year, too, when the particular species was approaching the maximum of its breeding activity, so the gonad should have its maximum size. The extent of the gonad evidently varies greatly. It is confined to a relatively small area on the gonadial vessel in A. grubii and claparedii. It is much more extensive in A. cristata and about equally so in A. marina. The gonad achieves its greatest size in A. ecaudata. The relatively immense testes of this species are due to the fact that the sperm are retained within them until nearly mature while in the other species the spermatogonia are early discharged into the body fluid, there to undergo the major part of their development. The large bladders of the nephridia of A. grubii seem similarly due to the fact that they are used as storage rooms for the sexual products after their formation, while in A. cristata, marina and claparedii these are held in the general body cavity.
LIMITS OF THE GONADS
The gonad usually surrounds the blood vessel. It appears as a light yellow incrustation, varying in thickness with the season. As it decreases in size it occupies a more and more restricted area on the posterior portion of the gonadial vessel. Not infrequently the other blood vessels adjacent to the nephridia, the dorsal longitudinal, the nephridial longitudinal and the nephridial branch of the afferent, are covered with a similarly appearing incrustation, but on microscopic examination, the material is found to be chlorogonous, never gonadial. Furthermore, the gonad is confined to the blood vessels of the second to the fifth nephridia, inclusive, in A. cristata. Over a hundred males have been carefully examined : in 6.2 per cent, one or both of the first nephridia had the gonadial vessel slightly to plainly coated with the yellowish incrustation; 11 .3 per cent of them had the gonadial vessel of the sixth nephridium coated. Such nephridia have in all cases been removed and sectioned and with one possible exception, the sections show the incrustation to be chlorogonous. In a single instance the gonadial vessel of a sixth nephridium had upon it a few cells looking
THE SPERMATOPHORE IN ARENICOLA 1005
like degenerating spermatogonia. As these worms were selected at intervals throughout the year so they would be representative, we may safely conclude that the vessels of nephridia one and six never bear active gonads, while there is slight evidence that the gonadial vessel of the sixth nephridium bears a degenerating gonad.
The limits of the gonads have not been as carefully studied in the other species of the Arenicolidae is in A. cristata. Yet I have exmained, macroscopically, several dozen specimens of each of the other species, claparedii, ecaudata, grubii and marina and have examined microscopically the blood vessels when any doubt could exist, with results confirmatory of the statements of previous authors, notably Gamble and Ashworth, as follows: A. ecaudata has thirteen pairs of nephridia in setigerous segments 5-17, A. marina six, in segments 4-9, A. grubii and A. claparedii each five pairs in segments 5-9.
Presumably the Arenicolidae have evolved from a more generalized polychaete in which nephridia and gonads were segmentally repeated organs. Both organs have gradually been confined to a smaller and smaller region. This gradual reduction seems to be well illustrated within the group as indicated by the number and position of the nephridia given above. Moreover, according to Gamble and Ashworth, the first pair of nephridia of A. marina are frequently absent and the last pair occasionally. Lillie remarks of the nephridia of A. cristata that The two earliest formed pronephridia, those of somites iv and v, degenerate at a comparatively early period in the development. The remaining six pairs (in somites vi-xi, inclusive) are directly transformed into the definitive adult nephridia." Fauvel ('99) states that the number of nephridia in A., ecaudata is only occasionally thirteen; that twelve is the usual number, the last pair of nephridia being absent from his specimens. In the two hundred and more specimens of A. cristata examined I have found only three cases of variation in the number of nephridia. In two of these the first pair of nephridia was wanting; in the third only the funnel was present in the sixth right. I have yet to find variation in the other species.
1006 ELLIOT ROWLAND DOWNING
The gonads are more restricted than the nephridia. As noted above, the first gonadial vessel in A. cristata never bears gonads, the sixth rarely and then degenerating cells only. In the other species the first nephridium never has a gonad. Here then is a case in which the genital cells show an evolutionary character, the tendency to restriction, more emphatically than the somatic cells.
GENERAL STATEMENT
A section through the testis of any of the Arenicolidae shows, ordinarily, a mass of cells of two or three sizes (fig. 7-9) : These are the spermatogonia of successive generations. The larger ones lie adjacent to the blood vessel. At the periphery of the gonad spherical bunches of spermatogonia or occasionally single ones are seen to be loosening from the general mass preparatory to discharge into the body fluid (except in A. claparedii). In this fluid the further divisions of the cells result in the formation of hollow spheres of spermatogonia (fig. 10), the last generation of which grow to spermatocytes. These, by the customary two divisions, become the spermatids (fig. 11) the cells still adherent in the spherical masses, which are meanwhile however altering their shape and becoming saucer-shaped, in A. cristata (fig. 12) slightly biconvex in the other species, except in A. claparedii, in which the successive divisions occur in the testis and the sperm are discharged into the body fluid. These sperm masses are the spermatophores, (fig. 13). The body fluid of A. cristata is loaded with these spermatophores (or with the eggs) except for a short time just after the discharge of the sex products, and in the other species, except claparedii, for weeks before the eggs are deposited. There are present, of course, other elements, body cells, chloragogue cells, etc., but the dominant objects are the eggs and the spermatophores. Finally the tails of the sperm are stiff and are aggregated into conical masses (fig. 12).
Toward the close of the breeding season, I have found, both in A. cristata and A. claparedii, exceptionally large spermatogonia discharged singly from the gonad, in addition to the customary masses described above. These develop in a somewhat different and highly instructive manner, as will be described later.
THE SPERMATOPHORE IN ARENICOLA 1007
BREEDING HABITS
The breeding season at Woods Hole lasts, for A. cristata, from about the first of May until the end of August, attaining its maximum during June. The cylindrical jelly strings containing the eggs are found in the shallow water over the littoral mudflats at low tide, one end attached at the burrow of the worm. The string lies on the bottom almost afloat. From field and laboratory observations I conclude that the male lies adjacent to the female during the discharge of the eggs and simultaneously discharges the sperm through the nephridiopores. The following facts support the statement: (1) I have repeatedly captured both male and female at an egg string when the latter was just beginning to appear. That one frequently fails to find both animals is, I presume, due to the fact that they burrow with extreme rapidity. If the tail of an animal be exposed with one stroke of the digger it often disappears before the next stroke can be taken and only very hurried work makes capture possible. When two worms are at the same burrow the chances of getting both are not great. It is to be remembered also that it is necessary to capture the worms in order to determine the sex as there are no external differences. To determine the sex without killing the animals I have examined a small drop of the body fluid withdrawn from the body cavity by means of a hypodermic syringe. The presence of either eggs or sperm can usually be determined by the naked eye. (2) The discharge of eggs and sperm has been seen to occur through the nephridiopores in worms kept in pans in the laboratory. (3) I have been reasonably sure that male and female were cooperating in the formation of the egg string in animals kept in aquaria in the laboratory. At best, however, the details of the process are obscure, since the animals, even when close to the glass, are pretty well covered with sand.
The conspicuous egg strings make A. cristata the easiest species to locate when depositing the eggs. The other species probably lay the eggs in the sand and debris among which they live. The times of their sexual maturity are fairly well established. A cristata is found to mature at about the same time at Naples as at
1008 ELLIOT ROWLAND DOWNING
Woods Hole, i.e., June to August (LoBianco). A. marina is captured with mature eggs and sperm at Woods Hole in the early spring, April and May. It is found mature on the English coast in the spring (the laminarian variety) and summer (the littoral variety) according to Gamble and Ash worth. My specimens of sexually mature A. ecaudata were taken at Plymouth in April as also were specimens of A. grubii. The latter and A. claparedii I found mature at Naples in May. Both of them have been found breeding much earher there (LoBianco), A. claparedii beginning as early as November and continuing throughout the winter and spring. By using a haemocytometer I have estimated the number of spermatophores per cc. of body fluid in a mature male at about forty million. Each spermatophore will average in the neighborhood of a thousand sperm. A good sized male A. cristata will easily contain twenty-five or more cc. of body fluid, that is, a quadrilUon sperm, ready to be discharged when fertilization is to be accomplished.
DEGENERATION AND PHAGOCYTOSIS
At the close of the breeding season the body cavity of the male contains very little sperm. In only 10 per cent of the September specimens of A. .cristata taken at Woods Hole was sperm present in any quantity In 60 per cent so little was presen t that it was impossible to determine the sex except by sectioning the gonads. During August and September the gonad is at its minimum size. This I have determined in two ways: First, by macroscopic examination. The gonad is apparent on the blood vessel even to the naked eye. Examination of the animals throughout the year, with record of cases in which the gonad is plainly evident on nephridia two to five, shows that in August and September the gonads show least frequently. They become plainer during the fall, achieve the maximum size from December to March and then gradually become smaller again. Second, the same results have been reached by making serial sections of nephridia for each month and making camera lucida drawings of the largest cross section of the gonad as compared with the blood vessel on which it lies. The third and fourth nephridia are used preferably for the com
THE SPERMATOPHORE IN ARENICOLA 1009
parisons as they show the maximum gonad development. In September the blood vessel shows only a thin line of gonadial material in a very limited area. This grows rapidly from month to month until the gonad is of large size and is giving off spermatophores into the body cavity. Sections from the December and January worms show the maximum relative cross section of the gonad. It gradually decreases as its substance is given off into the body fluid as forming spermatophores during spring and early summer. During June the body cavity has its maximum of sperm. By the last of July and in August fibrous degeneration is evidently going on in the gonad (fig. 8-9), and the disintegrating remnants of gonadial tissue are being ingested by abundant phagocytes (fig. 14).
Though specimens of the other species have not been collected in such quantity as A. cristata throughout the year, yet enough of each has been seen to make quite certain that the description given will equally well apply to them, making allowance for the changed breeding period, always excepting A. ecaudata.
By October the degenerative changes have ceased in A. cristata and multiplication of the gonad cells has begun again. A month before the gonad attains its maximum size the peripheral cells of the gonad are beginning to break away in small masses and float in the body fluid. The later development of these spermatophores goes on in the body fluid. In October the margin of the gonad appears entire in section; in February and later the margin appears very ragged as the masses of gonad material are constantly discharging (compare figs. 7 and 8 and text figs. 1 and 2).
EXTRA BLOOD VESSELS
It will be seen then that the maximum size of the gonad does not coincide with the height of the breeding season. This is marked rather by the greatest abundance of the mature spermatophores in the body fluid. In September the body fluid contains few or no sperm. By October they are appearing and steadily increase so that by November or December the fluid is crowded with the masses of developing sperm. An interesting development of the circulatory system goes on simultaneously with this
1010
ELLIOT ROWLAND DOWNING
accumulation of growing spermatophores. During December the diagonal muscles in the region of the first nephridium especially and to some extent in the second are apparently becoming ' hairy.' This appearance is due to the numerous long, fine bloodvessels attached at one end to a large vessel, the other floating freely in the body fluid. Similar vessels have been noted before
Text fig. 1. Outline of the testis of A. cristata in October. The gonadial material lies around the blood vessel. The dotted area corresponds to that shown in detail in fig. 7, pi. 2.
Text fig. 2. Outline of the testis of A. cristata in February. The dotted area corresponds to that shown in fig. 8, pi. 2.
in A. claparedii and A. grubii by Gamble and Ashworth (p. 518) with no suggestion as to their probable function. They seem evidently a device for the better aeration of the body fluidand the ehmination of its wastes while it is heavily loaded with the developing spermatophores. They disappear in A. cristata in early summer when the sperm are fully formed.
THE SPERMATOPHORE IN ARENICOLA 1011
ORIGIN OF THE SPERMATOGONIA
Lillie says of A. cristata:
.The early germ cells in connection with each nephridium become distinguishable soon after the appearance of the blood vessel of the latter and arise as a proliferation of the peritoneal cells of its walls. They appear first on the anterior and first formed portion of the vessel, i.e., in the region immediately adjoining the posterior angle of the funnel. The germ cells usually appear on their respective nephridia in the order of formation of these organs, i.e., in order from before back.
This peritoneum out of which the germ cells differentiate is derived from large teloblastic nuclei located at the posterior portion of the embryo in the growing zone, which nuclei Lillie thinks are the homologues of the definite teloblast cells found in such forms as Clepsine and Lumbricus. They in turn are apparently direct descendants of 4d, one of the fourth quartette of blastomeres derived from the macromeres at the sixth cleavage, as described by Child.
I have endeavored, in Arenicola, to discover some constant characters in the germ cells, which, appearing also in certain of the peritoneal cells, would enable me to trace the germ cells back through successive generations to the derivatives of 4d. But in this I have had no success and we must conclude that, as far as optical characters are concerned, the germ cells are indistinguishable by present methods from the other peritoneal cells. In other w{)rds, the differentiation of the germ cells is probably called forth by stimulation of adjacent cells due to progressive inherent changes.
Such a differentiation of the germ cells from the peritoneum is common in the annelida. Die Bildungstatten der Geschlechtsproducte gehoren bei den Ringwiirmern genetisch den epithelialen Wandungen des Coloms an und erscheinen in Folge diesen als directe Abkommlinge der Mesodermstreifen oder des secundaren Mesoderms." (E. Meyer, Studien iiber den Korperbau der Annehden, iii.)
JOURNAL OF MORPHOLOGY, VOL. 22, NO. 4
1012 ELLIOT ROWLAND DOWNING
Gamble and Ashworth state ('00, p. 31) in regard to A. marina that In large Arenicola, at certain seasons, the vascular process has no gonad and it is possible, as Ciienot ('91) suggests, that a formation of the amoeboid corpuscles of the coelom takes place at this point when the animal is not breeding." If this be true, evidently the gonad cells must form anew, as a proliferation of the peritoneal cells in adult life as well as in the embryological development and the distinction of germ cells and soma would be hypothetical. If so, too, the annual appearance of the germ cells must be due to a cyclical change in the organism contemporaneous with or due to the seasonal change without. Of course this must be true of their rapid increase anyway.
I have examined many specimens of A. cristata taken in September in which the body cavity showed no sperm, sectioning nephridia on whose blood vessel naked eye examination showed no gonad, but have always found under the microscope some gonadial tissue. So that I feel reasonably certain that in this species at least, after once the primitive spermatogonia appear in the embryological development, they do not disappear.
STRUCTURE OF THE GONAD
These first few germ cells, then, formed from the peritoneal cells, multiply with rapidity until the entire gonadial vessel is covered with a thick incrustation of them. Division of the peripheral cells now becomes more rapid, so rapid in fact that the daughter cells do not grow to the size of the parent cells before ■ division again ensues. There thus result zones of cells diminishing in size to the periphery of the gonad where they are being discharged into the body fluid. The largest cells, those adjacent to the blood vessel, are about 12/x in diameter in A. cristata (11. 67^ the average of over one hundred cells). The next smaller size are 9.36/i in diameter, then 7.55/x and the outermost 6.02/i. In A. claparedii and A. grubii the cells adjacent to the blood vessel are smaller, only about lO/x in diameter.
Such an ideal arrangement of the cells is never found throughout the gonad. At some points the largest sized cells will at times
THE SPERMATOPHORE IN ARENICOLA 1013
delay division until well out toward the periphery. The intermediate sizes may be scattered, with no apparent order, throughout the organ. Not infrequently, cells of the third or even second generation reach the edge and are broken from the mass to float in the body fluid separately or in larger or smaller groups. But frequently the typical arrangement described will hold for large regions of the gonad. The peripheral cells tend to cohere into roughly spherical masses of from ten to fifty or so cells all in the same stage of division and these break away into the body cavity. But smaller masses of cells may be detached, even single cells. In the body fluid division continues rapidly as will be described shortly.
THE SPERMATOPHORE
Formation
The largest spermatogonia, immediate descendants of the peritoneal cells, may be called the primary spermatogonia. During the year, except just before the height of the breeding season, these primary spermatogonia multiply, and after division, the daughter cells grow to the size of the parent cells except toward the periphery of the gonad. In the late fall and winter, in A. cristata collected at Woods Hole, the gonad is largely made of the primary spermatogonia; it presents quite a solid appearance. In the spring, however, division of these cells ensues so rapidly throughout the gonad, as it always does at the periphery, that the secondary spermatogonia do not have time to become as large as the primary ones before they divide in turn. Now the division of the cells derivative from a single primary spermatogonium, after it starts on its course of rapid subdivision, seems always to be roughly synchronous. The gonad taken during the fall and winter presents a somewhat mottled appearance along the margin when sectioned and stained, due to the proiTiinence of these masses of cells in division among the relatively inactive primary spermatogonia. In the spring this mottled appearance pervades the whole gonad, for here and there a primary spermatogonium will start on its course of rapid subdivision, giving rise to two, then four, eight, etc., cells, which at first lie close together
1014 ELLIOT ROWLAND DOWNING
as a solid mass, but later have a cavity at the center, the segmen tation cavity. These forming spermatophores, for such they are^ move to the periphery in order to break away into the coelomic fluid. So in spring and early summer the gonads have, when sectioned, a very ragged appearance. Deep bays and systems of lacunae run into the mass so that the spermatophores may be discharging not only at the periphery but are moving out from many spots in its interior (figs. 7 and 8) .
There is frequently protoplasmic material at the center of this early spermatophore. It is derived from the disintegration of cells which, while the spermatophore is forming in the body of the gonad, come to assume a central position. Degeneration goes on as the cells move toward the margin of the gonad so that by the time the margin is reached little is left of the central cells. Occasionally strands of protoplasm run from some of the cells to a central point (fig. 15), suggesting a figure such as Calkins has shown for Lumbricus, as if the cell walls formed peripherally before they do centrally. This is an exceptional condition, however, and I am confident that no such interpretation is to be put upon it in Arenicola. Figure 16 is much more usual. In the more mature spermatophores one can, by carefully tapping the cover glass, cause the spermatogonia to break away from the central portion, leaving it as a colorless sphere containing occasionally a few granules. This does not ordinarily stain, has shght consistency and seems like a drop of slightly viscid fluid. Not infrequently even this disappears and the spermatogonia break away without leaving any residual mass. The central cells are then absorbed very early and the central remnant is possibly excretory in its nature, the products of anabolism that have diffused into the cavity.
Discharge of the for7ning spermatophore
The cells are customarily discharged, as has been stated, from the surface of the testis into the body cavity in roughly spherical masses. The constituents measure about 6m in diameter in A. cristata. Three divisions at least follow and probably more,
THE SPERMATOPHORE IN ARENICOLA 1015
since the cells are growing at the same time they are dividing in the nutritive body fluid. Finally, the last generation of spermatogonia is formed, measuring about 2.9^1. They transform into the spermatocytes of the first order, with a diameter of 3.2fx. These divide to form the spermatids with a diameter of about 2/x. In A. claparedii and A. grubii the spermatids are somewhat larger, measuring some 2.4ju in diameter; in the other species the cells vary very little in size from the measurements given for A. cristata.
Analogy to blastulae
There is thus formed a hollow sphere of cells, reminding one of the blastula of the segmenting egg (fig. 23). This likeness is even more emphatic when one follows the history of a single spermatogonium when freed from the surface of the testis. It will be recalled that as well as the spherical masses of several dozen cells, smaller masses, even individual cells, are discharged into the coelomic fluid, so that one finds in this fluid all stages in the formation of the spermatophore from the one-celled condition up to the completed structure. Naturally the resulting spermatophores vary very decidedly in size according as they have origin in a single discharged spermatogonium or in a coherent mass of such cells. Measurement of the mature sperm masses gives a variation, their diameters ranging from one to eight.
The single spermatogonium, floating freely in the coelomic fluid, divides into two, then four, eight, sixteen ceUs, etc., simulating in general appearance the cleavage of an egg. Further discussion of this matter will be taken up below when the division of certain giant spermatogonia is considered (figs. 17-25).
During the process of cell multiplication the forming spermatophore is constantly increasing in diameter. Since the cells form only a single layer, the smaller they become the thinner the wall of the sphere is and hence the larger it may be with a given amount of protoplasmic material. This protoplasmic material must also increase in amount during the spermatogonial divisions by appropriation of nutritive material from the coelomic fluid, for the spheres increase in diameter more than the mere thinning of their
1016 ELLIOT ROWLAND DOWNING
walls would account for. This is not true for the later stages, however, for with the formation of the spermatids a decrease in the size of the individual cells accompanies the transformation of the spermatids to the spermatozoa.
Contemporaneously with the change of the spermatids to spermatozoa, or even beginning when the cells are yet spermatocytes of the second order, a change in the shape of the mass of cells occurs. In A. cristata the spherical mass invaginates in a manner that forcibly suggests the invagination of some egg blastulae to form the gastrulae (fig. 24). The gastrula-like mass remains cup-shaped, the mouth wide open, the hps never approximating to suggest a closure of the blastopore. Usually before invagination is complete, the cup begins to flatten out, becoming saucershaped, which is the form of the mature spermatophore in this species (fig. 25).
This phenomenon is apparently merely analogous to the process of gastrulation in the egg. Presumably the physical relations between the cell mass and the surrounding medium happen to be such that invagination ensues with regularity in this one species. Body fluid freshly drawn by means of a hypodermic syringe shows these gastrula-like forms, as do also preparations made by fixing the coelomic fluid in a variety of fluids. In other species no such thing happens, but the spherical mass of cells merely flattens out to form a biconvex spermatophore. There is however in the spermatogenesis of this group a phenomenon which is really homologous to the segmentation and gastrulation of the egg and which will be considered in the discussion of the giant spermatogonia.
The ripe spermatophore
As the spermatids transform into spermatozoa the cells elongate, their long axes at right angles to the surface of the saucershaped or biconvex mass (fig. 25). The nuclei stain with increasing intensity. The tails of the sperm appear as stiff rods, finely attenuate, held rigidly at right angles to the head and gathered
THE SPERMATOPHORE IN ARENICOLA 1017
together into one or several bundles (fig. 12). The head of the sperm is about 2(x in length, 1.22^ in transverse diameter, while the tail is ten to twelve times as long as the head.
When the coelomic fluid is drawn by means of a hypodermic syringe and placed in sea water, if the animal is not a perfectly 'ripe' male, the spermatophores remain intact, the rigid tails perhaps moving slightly but stiffly. If, however, the sperm masses be quite mature, those in this condition will show movements, the immature ones remaining quiescent. The tails move at first stiffly through varying arcs, the point of attachment to the head as the center. The bimdles of tails disentangle and all the tails come to lie at right angles to the surface of the spermatophore. Now vigorous movements of the tails ensue, stiffly at first; then the tails become supple and undulatory movements begin. The spermatophore now disintegrates and the sperm swim away. This process is much more rapid if eggs be also added to the sea water and more rapid still if some of the slime from the surface of the body of the female be put into the water.
THE GIANT SPERMATOGONIA
The fact has before been mentioned that toward the height of the breeding season the margin of the gonad bears exceptionally large spermatogonia which are discharged singly into the body cavity. The primary spermatogonia in A. cristata are usually about 11 to 12ju in diameter. But these giant spermatogonia achieve a diameter of some 17/i before they are freed from the surface of the gonad to undergo their farther development in the coelomic fluid. During July and August they appear in the gonads of A. cristata at Woods Hole. Similar cells were found in the gonads of A. claparedn collected at Naples the last of May and in A. grubii taken at Plymouth in August. At this time the margin of the gonad is very ragged, and fibrous degeneration with phagocytosis is going on in the parts adjacent to the blood vessels (fig. 8).
1018 ELLIOT ROWLAND DOWNING
SPERM ATOGONIAL MACROMERES
The giant spermatogonium contains a very large nucleus and a prominent nucleolus (fig. 17). When shed into the body fluid the cell undergoes an interesting development. It divides unequally, producing one large and one small cell, (fig. 18). The small cell next divides, giving a three-cell stage. Unequal division of the large cell then occurs, producing a four-cell stage seen in polar view in fig. 19. It is so evident that these early stages in the division of the giant spermatogonia are at least roughly similar to the cleavage stages of the egg in Arenicola that we may modify the nomenclature of the latter to describe clearly the former. Just the order of cleavage of the four cells I have been unable to determine except that the three small cells divide before the large one, differing in this respect from the division of the egg blastomeres. When all have divided we have an eight-cell stage consisting of four large cells, the spermatogonial macromeres, and four small cells, the spermatogonial micromeres. The position of the cells (figs. 20, 21), shows that the third division has evidently been a dexiotropic one as it is in the egg cleavage. Up to the sixteen-cell stage it is reasonably certain (and I think for at least one additional cleavage) that the cell lineage of these giant spermatogonia is homologous to that of the egg. Spermatogonial blastulae and gastrulae form much as in egg development. The figure of the sixteen-cell stage, (fig. 22) as indeed all these figures of the cleavage of the giant spermatogonia, are camera lucida drawings done with exceptional care, under a one-sixth inch objective and a one-half inch ocular at the level of the table. It is needless to multiply sketches as they would simply be duplicates of the admirable figures already given by Child for the cleavage stages of the egg.
Shortly after the sixteen-cell stage the macromeres disappear from the surface and migrate into the segmentation cavity. If one shghtly crush the spermatophores in the body fluid under a cover glass, the great majority will show the blastophore exuding from the center of the spherical masses, yet a few will show four to six rather large cells which escape from the cavity. This is
THE SPERMATOPHORE IN ARENICOLA 1019
especially true at the height of the breeding season. Spermatophores containing such cells are produced, I take it, from the spermatogonia that are liberated singly from the gonad, notably the giant spermatogonia. The large cells are the macromeres and possibly some of the first quartette — in other words, the homologues of the mesentomeres. The spermatophores consisting of a hundred or so cells never show such differences, all gi^dng under pressure, the same so-called blastophore, apparently a drop of fluid, perhaps enclosed in a very deUcate sheath, the fluid staining fairly deeply with methylen blue but scarcely at all with neutral red or other stains tried. It seems reasonably certain therefore that the invaginated mesentomeres disintegrate promptly to form the nutrition for the developing spermatophore.
It is manifesth^ difficult to determine with exactness the order of cleavage and the relations of the cells. These enlarged spermatogonia occur only at the height of the breeding season and then make up. a very small per cent of the developing spermatophores in the body fluid since these have been accumulated by the ordinary method for months. One must determine what occurs by the chance finding of successive stages, a laborious process, since the percentage of the desired material is so small. Presumably the cleavage of the giant spermatogonia might be watched if one could keep the body fluid under normal conditions. But it coagulates in the course of a few minutes after removal, the bodyfluid cells cohere in masses and all other cells, too, cease their activity. ^Mien stages are found in the development of these giant spermatogonia it is not easy to determine the exact relations of the cells, for the cell mass is small and transparent, even if stained; furthermore it is difficult to manipulate the cell mass without breaking it as it is only about one four-hundredth the size of the developing egg. Still I am confident of the above statements, as I have worked with the living material, stained whole mounts and sections. The results stated have been repeatedly confirmed during several summers at Woods Hole, working on A. cristata and have also been confirmed with the living and fixed material of A. claparedn and A. grubii.
1020 ELLIOT ROWLAND DOWNING
After my attention had been caught by the pecuhar egg-hke cleavage of these giant spermatogonia which float freely in the body fluid, I hunted carefully for the early developmental stages of the normal sized spermatogonia that are occasionally set free singly in the coelom. So far as I can find, their development is the same as that just described for the giant form.
One other interpretation might be suggested for these giant cells, namely, that they are tiny eggs which cleave in the body fluid to a certain point and then disintegrate. It is well known, of course, that such hermaphroditism of the gonad occurs when degenerative changes are going on in it. But such an explanation seems negatived in this case by the following considerations —
1. If they are developing eggs they would be undergoing cleavage long before thej^ reach normal size, in fact when only about one four-hundredth of the size of the egg when it is normally ready to be fertilized and begin cleavage.
2. The normal sized spermatogonia undergo a similar cleavage when they are liberated singly in the body fluid.
3. At the height of the breeding season, when these so-called giant spermatogonia are present, there are also found quite frequently giant spermatozoa whose volume bears about the same relation to the volume of the normal sized sperm as the volume of the giant spermatogonia bears to the volume of the usual primary spermatogonia.
It is to be noted that these exceptionally large spermatogonia appear toward the close of the breeding season. It is at this time that the body fluid is supercharged with spermatophores, evidently taxing the respiratory and excretory organs to the limit of their capacity. For it is at this time that the blood vessels develop in numbers like a thick growth of hair on the first and second nephridia and the adjacent muscles in A. cristata and in similar positions in the other species. At this time, too, the gonad is invaded by phagocytes, while the portions adjacent to the blood vessel suffer fibrous degeneration. It seems quite likely then, since the respiratory organs are taxed to the utmost, that an oxygen starvation sets in in the gonadial tissue, inducing fibrous
THE SPERMATOPHORE IN ARENICOLA 1021
degeneration and phagocytosis. The growth of the giant spermatogonia may be due to the same general causes, the accumulation of wastes so changing the osmotic relations between cell content and the surrounding m.edium that an increase in size results, either from accumulation of materials customarily excreted or through increased absorption from the nutritive fluids that bathe the cell.
THE SPERMATOPHORE DEVELOPED FROM ONE SPERMATOGONIUM
In the light of these facts regarding the development of the giant spermatogonia and those of normal size that float freely in the body fluid, it is now worth while to review the formation of the ordinary spermatophores — those that are developing from the cells in the body of the gonad. A primary spermatogonium divides into two, four, eight cells, etc. Some of these cells move to the center of the mass and disintegrate to form the blastophore whose substance is absorbed as nutrition by the surrounding cells and is replaced by more or less excretory matter. Meanwhile the mass of cells is migrating toward the margin of the gonad. Arrived there, the nearly hollow cluster is given off into the body fluid where division of the component cells continues until the last generation of spermatogonia is formed. By a slight growth the cells are changed into the spermatocytes of the first order. The two customary divisions of the cells ensue and the spermatids are formed and then change to sperm. Meanwhile the spherical mass has changed its shape to the saucer-shaped or lenticular mass of the adult spermatophore.
It may seem an unwarrantable assumption that the ordinary spermatophores are the result of the segmentation of' a single primary spermatocyte. The idea was suggested, rendered probable perhaps, by the development of the giant spermatogonia. It seems proven by the fact that all the cells in a given group manifest the same stage of division. It certainly presents a striking appearance, (figs. 8 and 9), to have a group of cells — a spherical bunch — all in the early prophase, for instance, when the adjacent cells manifest no sign of division. The possibihty has
1022 ELLIOT ROWLAND DOWNING
been considered that, perchance, some influence emanates from the central cell of a fortuitously accumulated mass as this cell divides, which, passed on to the adjacent cells, causes them to divide also. One is at a loss however to see why the influence should not be passed on to the still more peripheral cells so that all the cells of the gonad would divide more or less in unison.
It is proper to speak of the division of these cells as synchronous only in a general way. It must not be taken to mean that each phase of division occurs in all simultaneously, merely that certain prophase and telophase conditions, that anyway last for a considerable time, are frequently found in all or nearly all cells of the group at once, so that the cells of the group in some stage of division, not necessarily exactly the same, will contrast with the surrounding tissue.
One finds these synchronous cells more or less disconnected at times, as if the stress of the neighboring growing cells had broken the integrity of the mass. Still the general harmony of division is maintained, as would be the case if the blastomeres of a four- or eight-cell stage in a developing egg were separated by mechanical means. When this does occur the separated spermatogonial blastomeres apparently proceed to form spermatophores of a fourth or an eighth the normal size, thus readily accounting for the previously noted variation in the size of the spermatophores.
THE SPERMATOPHORE THE GAMETOZOON
It seems evident, then, that the development of the primary spermatogonia in the gonad, the same spermatogonia, when shed singly into the body cavity, and the giant spermatogonia are all in accord and are sufficiently suggestive of the development of the individual derived from the egg to make the hypothesis quite plausible at least that we have in Arenicola an alternation of generations. The primary spermatogonia are asexual spores, each of which, cleaving in a manner quite analogous to the cleavage of an egg, produces an individual, the spermatophore, all the cells of which are transformed into gametes.
THE SPERMATOPHORE IN ARENICOLA 1023
THE ALTERNATION OF GENERATIONS
There follows the union of the sperm and the egg, the second individual in the alternation, what we ordinarily know as the adult worm. This individual it is that produces the spermatogonia or oogonia, or in other words, the asexual spores.
Botanic use of the term
The conception of the alternation of generations has developed in its clear-cut simplicity among the botanists. It is that in the life history of a form there are two generations, one of which produces sexually, the other by asexual spores. It is tj^pified in the bryophytes and most pteridophytes. In the higher plants the sexual or gametophyte generation is gradually reduced so that it is only in relatively recent times that its existence as such has been recognized in the phaenogams.
Usual zoological significance
The term alternation of generations has been used by zoologists in a totally different sense. Two generations occur in many animals, the so-called sexual and asexual. The latter originates from a fertilized egg; the former arises by budding or a similar process from the asexual generation. There is thus an alternation of a generation that reproduces sexually with one that is never sexual, but the latter does not reproduce by asexual spores as is the case in plants. It is unfortunate that the same term is used for both processes. The asexual generation in the animal alternation is much more comparable to the sporophyte which is produced in propagation by cuttings or by runners. I am using the term alternation of generations strictly as it is understood by botanists.
ALTERNATION OF GENERATIONS AND REDUCTION
Now in all except the lowly plants, in all the archegoniates and even in many algae this alternation of generations is accompanied by the phenomenon of chromatin reduction, and reduction seems
1024 ELLIOT ROWLAND DOWNING
always to occur with definite relation to the alternation. The gametophyte, the plant that gives rise to the sexual elements, bears the reduced or haploid number of chromosomes. The sporophyte, the generation that produces the asexual spores, has the diploid or somatic number. Reduction occurs at the time the asexual spores are produced. So generally is this true, that for a time in many botanical papers, the presence of the diploid number of chromosomes was looked upon as a criterion that the cell possessing this number belonged to the sporophyte; or is a gametophyte cell if it has the haploid number, and the conclusion reached in a study of the archegoniates is forced, on a priori grounds, to cover the thallophytes as well. Thus Yamanouchi speaking of Williams' work on Dictyota says The fertilized egg nucleus gives rise to an asexual plant with double the number of chromosomes and consequently a sporophyte generation." (Bot. Gaz., vol.42: p. 431). And again, quoting from Davis, "Morphologically we can distinguish sporophyte plasm from gametophyte plasm by the double number of the chromosomes." (Am. Nat., vol. 39: p. 456).
In subjecting this life history [of Coleochaete] one of the green algae to what is regarded as a critical test of the two generations it has been discovered that this special spore-producing body is not a sporophyte. The test has to do with the number of chromosomes in the nucleus, a number which is definite for each plant species. The chromosomes are doubled in number l3y the fusion of the sperm and egg to form the oospore; and this means that in some other point in the life cycle the number must be reduced again. Accordingly the sporophyte, which arises from the oospore, is characterized by the double or 2x number of chromosomes in its nuclei ; and the gametophyte, which gives rise to the gametes, is characterized by the reduced or x number of chromosomes. Text book of Botany, Coulter, Barnes, Cowles; vol. 1, p. 32.
Alternation and reduction independent
Recently, however, cytological studies on botanical material have thrown serious doubt on this conception. Reduction and the alternation of generations are, even in plants, independent phenomena. I shall briefly cite three lines of evidence in proof of this proposition. The few papers to which I refer will give references to abundant literature.
THE SPERMATOPHORE IN ARENICOLA 1025
1. It is proven by cytological studies on aposporous and apogamous material. By 'apospory' (Vines, Journal of Bot., 78, p. 355) is meant the direct production of a gametophyte from the tissue of a sporophyte without the intervention of a spore. 'Apogamy' (DeBarry, Bot. Zeit., '78, p. 449) means the growth of a sporophyte as a vegetative outgrowth from the gametophyte. This definition is tentative, as later writers, Strasburger, Farmer and Digby, etc., are not yet agreed on the limitations of apogamy and parthenogenesis.
Farmer and Digby succeeded, in four forms with which they experimented, in inducing apogamy, causing the omission of sporo genesis. They derived the prothallia directly from abortive sporangia or from pinnae. Such gametophytes have approximately the diploid instead of the usual haploid number of chromosomes. They conclude therefore that there is no necessary relation between the periodic reduction in the number of chromosomes and the alternation of generations."
Again, Yamanouchi, in his study of apogamy in Nephrodium, obtained a sporophyte with the haploid instead of the customary diploid number of chromosomes.
2. The independence of the alternation of generations and reduction is further demonstrated by the fact that reduction may occur before, after, or during the sexual act, that is, in either the sporophyte or the gametophyte generation. I realize that in such a statement of the argument I am begging the question. I am merely stating the facts as they appear from my standpoint.
Whether the sporophyte and gametophyte of the archegoniates are phylogenetically continuous with the spore-bearing and gamete-bearing generations of the algae or whether the sporophyte of the archegoniates is a new structure, developed out of the fertilized egg and unrelated to the spore-bearing generation of the ancestral alga is a moot point. Botanists are far from agreed as to the course of the evolution of the higher plants from their algal ancestors. The Chlorophyceae has been designated the probable ancestral group and both Chara and Coleochaete are pointed out by different investigators as probable connecting links. With equal conviction other botanists, notably Schenk recently,
1026 ELLIOT ROWLAND DOWNING
discard these forms as pathways of ascent and adopt the Phaeophyceae as the most likely progenitors of the higher plants. Still others believe some common ancestors of these groups, a form now extinct, to have been the starting point of the archegoniates. Disagreeing over the probable course of evolution, they are equally at variance on the moot point mentioned above.
Insuperable difficulties seem, so far, to stand in the way of tracing the evolution of the sporophyte of higher plants from the so-called rudimentary sporophyte which develops from the fertilized egg of such forms as Oedogonium, Ulothrix and Coleochaete. In these forms the egg, after fertilization, breaks up into a number of separate cells, each of which gives rise to a new plant, thus functioning in a way suggestively like the asexual spores of the archegoniates. In Coleochaete, the only one of the lot that approaches the Hepaticae in structure sufficiently to be considered a probable ancestor, this structure can not be considered a sporophyte unless it be one with the x number of chromosomes instead of the 2x, in which case it is difficult to see how it gives rise to the sporophyte of the higher forms, which is usually characterized by the 2x number.
In quite as many algae, Sphaerella, Volvox, Vaucheria, Chara, Fucus, Dictyota, etc., the fertilized egg, possibly after a rest period develops directly into the spore-bearing generation. It may be invidious for a zoologist to suggest that the sporophyte of the higher plants has arisen from this class of algae, by the inclusion of the spore-bearing generation of an alga within the gametebearing generation, somewhat as in Volvox one individual is included within the other. And I will not even venture the suggestion but will merely call attention to the fact that botanists are still not in sufficient agreement as to the course of the origin of the sporophyte in the archegoniates to prejudice, by their plant evidence, a zoologist against a theory along this line for animals. It is such a point of view that I take, namely, that in animals, and possibly also in plants, the spore-bearing and gamete-bearing generations of the protozoa (and algae) are phylogenetically continuous with the sporozoon (or sporophyte) and the gametozoon (or gametophyte) of the higher forms.
THE SPERMATOPHORE IN ARENICOLA 1027
With such a position in niind, I have a right to take evidence on the independence of reduction phenomena and the alternation of generations from the algae and protozoa. Moreover I have the precedent set by eminent botanists who use Coleochaete, Fucus, Polysiphonia, etc., as examples of the alternation of generations in the algae and base their theories of the rise of the phenomenon on such algal evidence.
Karsten has shown that the mitoses in the zygote of Spirogyra are reduction mitoses. The same is true of the Desmidaceae. These mitoses occur, of course, after the fusion of the egg and spel-m. In Fucus, reduction occurs in the division of the antheridial and oogonial initials; in the case of the egg, the reduction is three mitoses prior to fertilization. (Yamanouchi.) In the Dictyotaceae, the cells of the tetrasporangium are the seat of the reduction division (Williams) . Yamanouchi has shown that in Polysiphonia it is in the division of the tetraspore mother-cells that reduction occurs, w^hile Wolfe claims that in Nemalion reduction occurs at the time of carpospore formation. Davis is so impressed with the fact that among the algae reduction occurs at so many different phases of the life history that he concludes that the phenomenon has a multiple origin. He says:
All of these cells in being the seat of reduction mitoses are analogous to the spore mother cells of archegoniates, but that would not warrant their being considered homologous with the latter structures. There is, on the contrarv, good reason to believe that, in plants, reduction phenomena became established as features in the life histories of a number of groups quite independently of one another, as the evidence indicates was also true of the processes of sexual evolution and the differentiation of the sporophyte generations. (Am. Nat. 43: 109).
Similarly among the protozoa we find the reduction phenomena occurring at various times. In Adalia (Siedlecki) and Monas (Prowazek) reduction occurs by the formation of chromidia in the gametes before they unite. Neresheimer and Metcalf find that in Opalina reduction occurs before the gametes form. Schaudinn has shown that in Actinophrys the cytoplasm of the conjugating individuals unites and then a reduction of the nuclei occurs.
JOURNAL OF MORPHOLOGT, VOL. 22, NO. 4
1028 ELLIOT ROWLAND DOWNING
Dangeard thinks that reduction occurs during germination in Chlamydomonas. Prowazek's work on Polytoma indicates that the same is true in this form. Illustrations might be multiplied. Hertwig, years ago, after careful study of many protozoa, concluded that either before, after or during the sexual union there is a reduction division.
3. Among the protozoa and algae already cited, it is evident that reduction occurs when there is no alternation of generations.
CH.iMBERLAIN'S THEORY
I have at some length presented what seems to me good evidence of the proposition that the alternation of generations and reduction are independent phenomena. I have been anxious to make the matter emphatic, otherwise, in any comparison of the alternation of generations in the higher plants and animals, unwarranted conclusions are reached. Thus Chamberlain, in proposing a theory of the alternation of generations in animals says.
The egg with the three polar bodies constitutes a generation comparable with the female gametophyte in plants; similarly, the primary spermatocyte with the four spermatozoa constitute a generation comparable with the male gametophyte in plants. All other cells of the animal constitute a generation comparable with the sporophyte generation in plants, the fertihzed egg being the first cell of this series.
In support of this theorv I shall present two lines of evidence: (1) the gracliial reduction of the gametophyte in plants, with the constantly diminishing interval between the reduction of the chromosomes and the process of fertilization; and (2) the phenomena of chromatin reduction in both animals and plants.
Briefly, his argument is this: that since, in the higher plants the gametophyte is gradually reduced, producing a condition apparently identical with that in animals, the egg and its three polar bodies and the spermatids of animals are to be regarded as tetraspores and the gametophyte generation is undeveloped except as represented by these cells: that these cells are further proven tetraspores because, in their formation, reduction occurs in a manner very like the reduction in the formation of the tetraspores of the higher plants.
THE SPERMATOPHORE IN ARENICOLA 1029
Objections to Chamberlain^ s theory
The facts adduced in the formation of the spermatophore of Arenicola and the evidence cited above to show the independence of reduction and the alternation of generations lead me to doubt both the validity of his argument and the accuracy of his conclusion. We are concerned at present only with his argument.
The fertilized egg of a lily is the first cell of the sporophyte, whether it ever divides at all. Consequently, we regard the zygospore of Ulothrix or Spirogyra and the fertilized egg of Vaucheria or Oedogonium as sporophytic structures, even if the first division of the zygote should be meiotic, as seems probable. From such a simple beginning, we believe that the more complex sporophytes with more conspicuous alternation have been developed. (Am. Nat. 44: 603).
LIMITS OF GAMETOPHYTE AND SPOROPHYTE
The gametophyte and sporophyte generations must be marked off, then, it seems to me, by some hmits independent of the phenomena of reduction. A much safer means of definition is to return to simple conceptions and designate that generation the gametophyte which had its origin in the asexual spore and which terminates with the formation of the egg. The sporophyte generation begins with the fertiUzed egg and terminates with the formation of the asexual spore. Presumably all would agree that this means of distinction is the best; the only reason for relying on any other criterion is the difficulty of applying this one.
Yet to insist that the conclusion reached in the study of the archegoniates that the 2x number of chromosomes marks the sporophyte, a law not without exceptions, even in the archegoniates themselves, shall apply in the algae, protozoa and higher animals too, seems to me unwarranted, for it drives one to the conclusion that as great a reduction has occurred in the gametophyte among protozoa and algae, Fucus, for instance, as has occurred in the whole evolution of the archegoniates from lowly Hepaticae to the most speciahzed angiosperms; and because there is another possibihty, namely, that the reduction phenomenon, in thecourseof evolution, shifts its position with reference to the boundaries of the gametophyte and sporophyte generations in the Ufe history of plants and animals- — a possibility that, to me, seems more plausible. In algae and protozoa alike we have the usual asexual reproduction followed by the sexual method, when some change in the physical or chemical condition, either of environment or organism, is an active cause. This is apparent from Calkins' studies on the conjugation of Paramoecium, Kleb's work on the formation of the gametes in Hydrodictyon and similar papers that followed these pioneer investigations. In these forms, as in most algae and protozoa, there is an intercalation of sexual among asexual generations, rather than an alternation. But let those conditions which produce sexuality recur with rhythmic regularity and an alternation is developed such as we have in Dictyota. We may presume that, phylogenetically, the higher forms exhibiting the alternation of generations have arisen from those lower ones that possess an intercalation of the sexual among the asexual generations. Further discussion of the point may be deferred and the alternation in Arenicola presented more fully.
Graphic life histories
In Arenicola the individual male is a sporozoon giving rise to certain cells, the spermatogonia, which are really the microspores. These cleave in a manner homologous with the cleavage of the egg and give rise to the gametozoon or the spermatophore whose development is a curtailed recapitulation of the primitive gametozooic generation. At the end of this generation reduction occur^ during gametogenesis. The union of the gametes initiates the sporozoic generation again. This life history may be graphically represented by the conventional diagram I, (text fig. 3).
In the higher dioecious plants we should have a very similar graphic life history, see diagram II, except for the fact that reduction occurs at a different point. Comparing merely these two life histories, it seems difficult to homologize the generations, simply because so important a phenomenon occurs at such widely separate points. Not only does reduction occur at different phases of the life history of the higher plants and animals, but
Fig. 3. Diagram I. Schematic representation of the spermatogenesis and alternation of generations in Arenicola.
Fig. 4. Diagram II. Scheme representing the life history of a phaenogam.
Fig. 5. Diagram III. Scheme of the life history of the moss or other bryophyte.
Fig. 6. Diagram IV. Scheme of the life history of such an alga as Spirogyra or such a protozoon as Chamydomonas.
1031
1032 ELLIOT ROWLAND DOA\'NING
this is true, too, in the algae, as abeady pointed out. Let us review the facts down into that group with diagrams.
The mosses, (diagram III) , show a sporophyte and gametophy te generation of about equal importance. The egg and sperm are usually borne on the same gametophyte.
The sporophyte becomes less and less prominent in the archegoniates as we approach the algae. Among the green algae the intercalation of a sexual generation seems the rule rather than the alternation of generations. The life history of Spirogyra, of the Diatomaceae, and of such protozoa as Chlamydomonas and Poly
Fig. 7. Diagram V. Scheme of the life history of Adelia, Monas, etc.
toma would be graphically shown in diagram IV. In such forme as Adelea, Monas and Actinophrys among the protozoa, in no algae so far as I know, reduction occurs during the fusion of the gametes ; the life history is given in diagram V (text fig. 7) .
Now which of these life histories most nearly approximates primitive conditions? The question involves a discussion of the relation in phylogeny of the alternation of generations, sexuality and reduction. The evidence already given shows that the latter two phenomena antedate the alternation of generations, for both are found among simple animals and plants that have not achieved the alternation.
SEXUALITY AND REDUCTION
Many biologists are inclined to regard sexuality and reduction as causally related. Thus Davis says " Chromosome reduction as a physiological process seems to be a corollary of sexual nuclear functions." (Bot. Gaz., vol. 43: p. 109.) If either were the effect of the other, we should expect reduction to bear a fixed relation to the sexual act. As it is we have seen that it may occur before, during or after such act. It would be better to say, then, that reduction is an adjunct of sexuality and both are probably corollaries of some fundamental cause yet to be determined.
It has been suggested by Biitschli and later writers that the union of the gametes is a process of rejuvenescence. In other words, the sexual act is induced by exhaustion of the organism. We know that chromidia are due to degeneration produced by exhaustion and otherwise (Dobell, '07). Further, a very simple type of reduction has been found in some protozoa (Monas, Adelea, etc.) to consist of the extrusion of chromidia. The very incomplete state of our knowledge regarding reduction, chromidia and sex, particularly among the lower plants and animals, makes the proposal of even a working hypothesis premature. I call attention to the exhaustion of the organism through repeated asexual divisions as a possible cause of both sexual union and reduction, merely to illustrate the possibility that both are corollaries of some underlying single cause.
About the only justifiable conclusion regarding the phylogenetic relation between the sexual act and reduction is that they arose as adjacent phenomena. This seems probable, since in primitive forms, both plants and animals, they occur close together in point of time. It would seem, on a priori grounds, too, that if a sexual union occurred in forms that had been producing asexually, a reduction in the amount of chromatin would promptly occur if it had not occurred previous to the act, so as to restore customary conditions as speedily as possible.
1034 ELLIOT ROWLAND DOWNING
THE PRIMITIVE ANIMAL TYPE
Eotanists seem agreed that the significance of the persistent gametophyte in the higher plants is that it represents a return to a primitive type. The conclusion seems a proper one. We would similarly expect the gametozoic generation, then, to represent a primitive animal type. The spherical spermatophore, arising from a spermatogonium as a spore might cleave, must find its counterpart in the phylogeny of the higher animals as the thallus-like gametophyte of the higher plants indicates a thailine ancestry for them.
Possibly the Volvocales approximate most closely to such an hypothetical ancestral form. The group is one that has been appropriated by many zoologists as showing marked animal affinities, and is admittedly a widely aberrant type by all botanists. Volvox reproduces asexually; certain cells of the central cavity, the so-called parthenogonidia, reproduce by fission, and the offspring divide and subdivide, the products cohering in spherical masses to form colonies like the parent. These are freed when the parent colony disintegrates. Volvox also reproduces sexually. Certain cells differentiate as eggs and come to lie in the interior of the colony. Other cells, similarly discharged to the interior, develop within them numerous sperm. The sperm break out of their containing cells and fertihze the eggs. The oospore remains inactive for some time as a resting spore but finally develops a colony like the parent. My preparations are not yet sufficiently clear to permit of working out all the chromatin changes, but I am quite positive that reduction occurs in Volvox, as it does in animals, during gametogenesis.
In the Dicyemidae, looked upon by many, zoologists as the connecting link between the protozoa and the metazoa, the process of reproduction is suggestively similar, except that the asexually produced young bore out from the parent to an independent existence instead of awaiting the death of the parent. The center of the dicyemid is occupied by a cell, in which, besides the nucleus, there are one to several embryo cells from which the asexual individuals arise. These cells are evidently the homologues of
THE SPERMATOPHORE IN ARENICOLA 1035
those partheno^onidia found in the Vol vox colony. Max Hartmann, in describing the reproduction of the Dicyemidae, speaks of this central cell as the agametangium because in it develop the asexual, or as he calls them, the 'agamic' individuals. After several generations of these agamic individuals, there arise the sexual or the gametic individuals. They arise from the embryo cells (Keimzellen of Hartmann) by a process of cleavage very similar to the development of the agamic individuals, except that in the female the reproductive cells are split ofT early. The female as well as the male sexual individual grows within the asexual parent. The eggs are developed from the embryo cells of the female. They are freed from the gametic individual into the socalled agametangium of the agamic individual by the death of the gametic form. The egg matures by forming polar bodies and is fertilized by sperm which have discharged from the male gametic individual by its death and disintegration.
To harmonize Hartmann's description with the language I have used in giving the alternation of generations in the Arenicolidae I have only to change his terms slightly; (I have adopted the terminology proposed by Beard, cited later). Call his agamic individuals the sporozoon and the Keimzellen spores; his gametic individual, the gametozoon; then the dicyemids make concrete our conception of how the alternation of generations of the higher forms arose from such simple ones as Volvox, even though the line of ascent may not actually have passed through the Dicyemidae.
Imagine that in a Volvox-like form the sexual colony or gametozoon develops eggs and sperm before it is discharged from the colony which reproduces asexually and the condition of the dicyemids is practically achieved. Now imagine that a regular alternation of generations is established instead of the intercalation of an occasional sexual generation in the midst of the dominant asexual reproduction of the Dicyemidae and a condition is established that needs little if any modification to give the alternation of generations as we find it in the ^Arenicolidae, as follows :
The adult Arenicola, a sporozoon, corresponds to the sporophyte colony of Volvox or the agamic individual of the dicyemid. At some time in its life history cells are developed within its central cavity — the primitive germ cells or spore mother cells. These develop spores which we know as spermatogonia or oogonia. Such spores, in Volvox, develop new colonies which may be sexual and which are freed when the parent disintegrates. In the Dicyemidae, if a gametic individual develop from such spores (or Keimzellen) it is retained in the parent where it disintegrates to free its egg or sperm. In a word the gametozoon degenerates prematurely within the sporozoon. Now in Arenicola, I take it, the development of the spermatogonia into the spermatophores is the development of the gametozoa. Gametozoon degeneration begins before its development is complete and the sperm are produced by a short cut. Instead of developing an individual, within which some cells form the clusters of sperm, its cells form the sperm cluster immediately. In this genus, as I have already shown is the case in Hydra, the sperm seems to form within the spermatid, reminiscent perhaps of the primitive condition found in Volvox of forming the sperm within the cell.
For the sake of my theory, I should like to agree with Tannreuther in the spermatogenesis of Hydra. He claims that the spermatozoa develop in groups, each group enclosed within a single cell or cyst. But the clearness of my own preparations seems to negative completely such an interpretation. I can only reiterate what I have already published on the spermatogenesis of this form. It is too bad, for Hydra would make even a better transition form than it does, between such a spermatogenesis as we have in Volvox and the typical animal spermatogenesis in which the spermatid transforms into the sperm in its entirety rather than developing the sperm within it.
Hartmann, in the paper on the Dicyemids, describes briefly and figures a development of the fertilized egg that is very similar to the development of the asexual spores. The same parallelism is noted in many botanic papers, the oospore developing much as an asexual spore does in its early stages. To find, then, in Arenicola that the spermatogonia or asexual spores develop to produce the spermatophore or gametozoon in a manner homologous to the development of the fertilized egg is to strengthen the position taken that the spermatophore is the gametozoon.
THE SPERMATOPHORE IN ARENICOLA 1037
THE GAMETOZOON, A 2X FORM
Botanists, reasoning on the basis of the fact that the disappearing gametophyte in the higher forms possesses the haploid number of chromosomes, have been led to assume, not only that the gametophyte represents a reversion to the primitive type, but also that this primitive plant possessed the x number of chromosomes. Reasoning in an analogous manner, we should be forced to assume, that, since the gametozoon possesses for most of its life history, the 2x number of chromosomes, so the primitive animal type did not have the reduced number. This distinction between plants and animals has long been recognized. ' ' So far as groups of plants above the thallophy tes are concerned, the period of chromosome reduction has been found to be always associated with sporogenesis and never with gametogenesis as in the case of animals." (Yamanouchi; Polysiphonia, p. 43.) The difference may help to trace the gradual separation of the plant and animal types in the course of evolution. Upon this distinction, for instance, Dobell rests his belief in the plant affinities of the Phytomonadina (The structure and life-history of Copromonas subtilis, p. 112). The discovery already mentioned that Volvox has reduction occurring during gametogenesis would justify, in a measure, the classification of the form as an animal rather than as a plant.
THE COMMON PROTOTYPE
Presumably plants and animals have come from a common ancestor. Now in all higher animals reduction occurs near the close of the gametozooic generation. It occurs, in all higher plants, near the close of the sporophyte generation. In the common ancestor it must have occurred at a point between these two extremes, possibly during or in close connection with the conjugation of the gametes. Such a possibility is rendered probable from the fact that, in thallophytes and protozoa, the reduction occurs at variable times in the life histories, usually as an adjunct of the union of the gametes, as if that variation were dominant which later becomes fixed in the two prevaihng plant and animal types.
1038 ELLIOT ROWLAND DOWNING
ORIGINAL POSITION OF REDUCTION
If reduction originally occurred in the primitive common ancestor somewhere closely adjacent to the union of the gametes, the reduced number of chromosomes would exist for only a short period and might not occur at all. The major part of the life history of the forms would possess the somatic or diploid number. This is now the case for most animals and for many of the algae as already shown. In plants, then, the place of reduction in the life history has been shifted; the phenomenon has been postponed. In animals it occurs much nearer its original position.
Reduction shifted
Since in practically all animals and in many algae, the phenomenon of reduction occurs before or during conjugation of the gametes, the preponderance of evidence appears in favor of such a position in the primitive common ancestors of plants and animals. It is all the animals and many algae against the higher plants in favor of such an hypothesis. Text fig. 7, then, might nearly represent primitive conditions. Evidently following the gamete-bearing generation with itsdefinitenumber of chromosomes would come a spore-bearing generation with the same number of chromosomes. This is true now in the Conjugales, Coleochaete, etc., and I believe in Volvocales and the animals. True, in Conjugales, Volvocales, etc., there are several spore-bearing generations following each other in succession. But if, for any reason, the alternation of generations be established by the omission of all but one of these spore-bearing generations, there would be left the gametophyte and sporophyte generations as I conceive them to exist in Arenicola, only that the reduction has shifted from a position like that of fig. 7 of the text to a place before the conjugation of the gametes. For plants the shift in position has been, in the opposite direction — a shift that is seen progressing in text fig. 6 and completed in fig. 5.
In Chamberlain's theory of the alternation of generations in animals, he maintains that the shift in the animal group has
THE SPERMATOPHORE IN ARENICOLA 1039
been in the same direction as in the higher plants. As pointed out by Coulter and Miss Pace, the end result attained in plants by the gradual reduction of the gametophyte generation to the point of complete extermination, in such forms as Pandanus, is entirely similar to the condition found in animals. Because the end results are similar is not prima facia evidence that the means of achievement in the course of evolution have been the same ; it is very evident that in the higher plants such a condition as is found in Cypripedium, etc., is the result of a reduction of a gametophyte generation with the x number of chromosomes, for all steps in the process are evidently traceable. But nowhere in the animal kingdom, not even among the protozoa, is there any evidence of a corresponding gametozoic generation with a reduced number of chromosomes.
REDUCTION AND TETRAD-FORMATION
It is true that there is a striking similarity between the formation of the tetraspores in most plants and the development of the spermatids from the spermatocyte. It is rendered doubly suggestive by the fact that reduction occurs during both processes. Yet, even in the plants themselves, we are not warranted in concluding that all cells in which reduction occurs are homologous. Reduction occurs without tetraspore formation in a sufficient number of cases, as in Lemanea, Chantransia, etc., to show that there is no fundamental phylogenetic association involved. The customary appearance of a fourfold division at time of reduction may be based on some fundamental property of carbon compounds, possibly on the tetravalent condition of carbon itself; in which case it would not be strange to find tetrad formation common in plants and animals without assuming that, when occurring, a morphological homology is indicated.
BEARD'S HYPOTHESIS
In the preceding pages I have noted one hypothesis of the alternation of generations in animals — that proposed by Chamberlain, and I have given my reasons for discarding it. Another
1040 ELLIOT ROWLAND DOWNING
hypothesis of the alternation of generations in animals has been proposed and vigorously advocated by Beard. He recognizes an antithetic alternation of generations in animals. He, too, identifies the primary germ cells as the equivalents of the sporemother-cells of plants." He regards the larva or phorozoon as the asexual generation, the homologue of the sporophyte. He derives the gametozoon, the adult animal, from it by apospory. He is forced to conclude, then, "that the final reduction of chromosomes has been deferred to a later portion of the life cycle in metazoa as compared with plants."
The same objections apply to Beard's hypothesis as to Chamberlain's, namely, that there is no evidence in fact of the successive steps in the postponement of reduction in animals similar to that so constantly apparent in plants. So far as we know, the process always occurs closely adjacent to the union of the gametes, continuing, in the higher animals, in much the same position in the life cycle that it occupies in the lower animals and plants. Whereas, in plants, it is evidently shifted from tliis primitive position and the successive steps of the shift are traced with some degree of certainty in living forms.
Furthermore, it seems to me, an impossible task to articulate his theory with what we know of the reduction phenomena and the development of the protozoa and mesozoa. He says:
The sexual generation of plants is at best a miserable failure from the morphological point of view. . . . The higher one ascends the smaller it becomes until in the higher plants it has almost reached the vanishing point, without, however, being able to disappear entirely.
In the animal it is the larva, the phorozoon, or asexual generation which makes the bravest show in the lower metazoa; ... In the higher forms it becomes reduced.
But how will a relation be established between the larva of the metazoa and the asexual generation of the protozoa? The one should pass over into the other. It seems unwise to adopt a theory which demands a hiatus at this point, when it is easy to blaze a possible, uninterrupted trail along which evolution may have proceeded by way of such forms as the Volvocales and the Dicyemidae, if we adopt the hypothesis I have proposed.
THE SPERMATOPHORE IN ARENICOLA 1041
BIBLIOGRAPHY
Allen, Chas. E. 1905 Die Keimung der Zygote bei Coleochaete. Bcr. dcutsch. bot. Gesells., Bd. 33: p. 286.
DeBarry, a. 1878 Ueber apogame Fame und die Erscheinung der Apogamie im Allgemeinen. Bot. Zeit., Bd. 36: pp. 449-487.
Beard, J. 1902 Heredity and the epicycle of the germ-cells. Biol. Centralbl., vol. 22: pp. 321-328, 353-360, 398-408.
Beard, J. and Murray, J. A. 1895 On the phenomena of reproduction in animals and plants. Ann. of Bot., vol. 9: pp. 441-468.
BtJTSCHLi, O. 1876 Studien uber der ersten Entwicklungsvorgange der Eizelle,
der Zelltheilung und die Conjugation der Infusorien. Abh. d. Senckenb.
naturf. Gesell. Fr. a. M., Bd. 10: pp. 213-452. Calkins, Gary N. 1895 The spermatogenesis of Lumbricus. Jour. Morph.,
vol. 11, no. 2: pp. 271-302. Chamberlain, C. J. 1905 Alternation of generations in animals from a botanic
standpoint. Bot. Gaz., vol 39: pp. 137-144.
1910 Nuclear phenomena of sexual reproduction in Gymnosperms.
Am. Nat., vol. 44: pp. 595-603. Child, Chas. M. 1900 The early development of Arenicola and Sternapsis.
Arch. f. Entw. der Org., Bd. 9: pp. 587-717. Coulter, John M. 1908 Megaspores and embryo sacs. Bot. Gaz., vol. 45:
pp. 361-366. Coulter, Barnes, Cowles 1910 Text book of botany. Chicago. Dangeard, p. a. 1898 Sur les Chlamydomonadin^es. C. R. Ac. Sci., Paris,
Tome 127: p. 736. Davis, B. M. 1905 On the plant cell, vi and vii. Am. Nat., vol. 39: dd. 449 500 and 555-600.
1909 Origin of the Archegoniates. Am. Nat., vol. 43: pp. 107-111. DoBELL, C. C. 1907 Physiological degeneration in Opalina. Q. J. Micr. Sci.,
vol. 51: pp. 633-646.
1908 The structure and life-history of Copromonas subtilis. Q. J. Micr. Sci., vol. 52: pp. 75-120.
1909 Chromidia and the binuclearity hypothesis. Q. J. Micr. Sci., vol. 53: pp. 279-326.
Downing, E. R. 1905 The spermatogenesis of Hydra. Zool. Jahrb., Abt. f. Anat. u. Ont., Bd. 21: pp. 379-426.
1909 The connections of the gonadial blood vessels and the form of the nephridia in the Arenicolidae. Biol. Bui., vol. 16: pp. 246-258.
1042 ELLIOT ROWLAND DOWNING
Farmer, J. B. and Digby, L. 1907 Studies in apospory and apogamy in ferns. Ann. of Bot., vol. 21: pp. 161-199.
Fauvel, p. 1899 Arenicola ecaudata. Mem. Soc. Sci. Nat. Cherbourg., Tome 31: pp. 101-186.
Gamble, F. W. and Ashworth, J. H. 1900 The anatomy and classification of the Arenicolidae. Q. J. Micr. Sci., vol. 43: pp. 419-569.
Hartmann, Max 1906 Untersuchungen tiber den Generationswechsel der Dicyemiden. Brussels. Also in Mem. of Royal Belgian Acad., N.S., vol. 1.
Hertwig, R. 1896 Ueber Kerntheilung, Richtungskorperbildung und Befruchtung von Actinosphaerium eich., Abh. d. k. bay. Akad. d. Wiss., Miinchen, ii Kl. 19: pp. 1-104.
HoYT, W. D. 1909 Alternation of generations and sexuality in Dictyota dichotoma. Bot. Gaz., vol. 49: pp. 55-57.
Karsten, G. 1909 Die Entwicklung der Zygoten von Spirogyra jugalis Ktzg., Flora, vol. 99: p. 1.
LiLLiE, Ralph S. 1905 The structure and development of the nephridia of Arenicola cristata Stimpson. Mitth. a. d. zool. Stat. z. Neapel., Bd.l7: pp. 341-405.
LoBiANCO, S. 1899 Mitth. a. d. zool. Stat. z. Neapel., Bd. 13: p. 484.
Meyer, Ed. 1901 Studien iiber der Korperbau der Anneliden, iii. Mitth. a. d. zool. Stat. z. Neapel., Bd. 14: pp. 247-585.
Metcalf, Maynard M. 1908 Opalina. Its anatomy and reproduction, with a description of infection experiments and a chronological review of the literature. Arch. f. Protist. Bd. 13: pp. 195-374.
Neresheimer, E. 1907 Die Fortpflanzung der Opalinen. Arch. Protistenk., Suppl. 1: pp. 1^2.
Pace, Lulu 1907 Fertilization in Cypripedium. Bot. Gaz., vol. 44: pp. 353374.
Prowazek, S. von 1901 Kerntheilung und Vermehrung der Polytoma. Ost. bot. Zeitschr., Bd. 51: p. 51, etc.
1901 Flagellatenstudien. Arch. Protistenk. 2: pp. 195-212.
Schenk, H. 1908 Ueber die Phylogenie der Archegoniaten und der Characeen. Engler's Bot. Jahrb. Bd. 42: pp. 1-37.
ScHAUDiNN, F. 1896 Ueber die Copulation von Actinophrys sol. Sitzber. Akad. Wiss., Berlin. Bd. 1 : pp. 83-89.
Strasburger, Ed. 1904 Die Apogamie der Eualchemillen und allgemeine Gesichtspunkte die sich aus ihr ergeben. Jahrb. wiss. Bot., Bd. 41: pp. 88-164.
Siedlecki, M. 1899 Etude cytologique et cycle evolutif de Adelea ovata Schneider. Ann. Inst. Pasteur, Tome 13: pp. 169-192.
THE SPERMATOPHORE IN ARENICOLA 1043
Tannreuthiok, G. W. 1909 Observations on the germ-cells of Hydra. Biol. Bnll., vol. lH: pp. 205-209.
Williams, J. Lloyd 1904 Studies in the Dictyotaceae. Ann. of Bot., vol. 18: pp. 140-160 and 183-204.
WoLFK, J. J. 1904 Cytological studies in Nemalion. Ann. of Bot., vol. 18: pp. ()()7-630.
Yamanouchi, Shigko 1906 The life history of Polysiphonia. Bot. Gaz., vol. 42: pp. 401-449. 19()S Apogainy in Xciphrodiu.u. Bot. ( laz., vol. 4.5: pp. 289-318.
JOURNAL OP MORPHOLOGY. VOL. 22, NO 4
PLATE 1
EXPLAXATIOX OF FIGURES
1 Latero-dt.rsal view of the second left nephridium of A. cristata. X 15.
2 Latero-ventral view of the second left nephridium of A. crist ta. X 15.
3 Latero-dorsal view of the third left nephridium of A. grubii. X 15.
4 Latero-ventral view of the fourth left nephridium of A. marina. X 15.
5 Latero-dorsal view of the third left nephridium of A. claparedii. X 15.
6 Latero-ventral view of the third left nephridium of A. eeaudata. X 15. The bladder end shows a latero-dorsal view, but the rest of the nephridium is iwisted by the weight of the gonad.
PLATE 2
EXPLANATION OF FIGURES
7 Section through the testis of A. cristata in October. It corresponds to the dotted portion of text fig. 1. The cells are some 7.5 ju. in diameter.
8 Section through a testis of A. cristata in February. It corresponds to the dotted portion of text fig. 2. Fibrous degeneration and phagocytosis are apparent.
9 Section through the testis of A. cristata; numerous spermatophores are forming.
EXPLAXATION OF FIGURES
10 A spermatophore from the body fluid of A. cristata. These cells are the sixth generation of spermatogonia from the primary spermatogonia. Each divides to form the last spermatogonial generation. The cells have an average diameter of 3.5m. All are in the equatorial plate stage. X 340.
11 A spermatophore from the body fluid of A. cristata. The cells are the spermatids just after the division of the spermatocytes of the second division. They are in the early anaphase. X 340.
12 A mature spermatophore of A. cristata. Smear preparation fixed in vom Rath's fluid and stained in saureviolett and iron haematoxylin. X 625.
13 Body fluid of A. cristata under low power, showing spermatophores in various stages (a, b, c), and coelomic cells (d), leucocytes and chloragogue cells. X 100.
14 A phagocyte from testis of A. cristata in April. X 930. Note the ingested cell.
15 An early stage in the formation of a spermatophore, from the body fluid of A. cristata. Diameter of the spermatophore 32 )j.; average diameter of the cells 7.5.
1049
PLATE 4
EXPLANATION OF FIGURES
16 Section of a more typical spermatophore. X 1250.
17 A giant spermatogonium. X 1800.
18 A giant spermatogonium dividing, the two-cell stage.
19 A giant spermatogonium dividing, the four-cell stage, polar view.
20 A giant spermatogonium dividing, the eight-cell stage, side view.
21 A giant spermatogonium dividing, the eight-cell stage, polar view.
22 A giant spermatogonium dividing, the sixteen-cell stage, polar view.
23 A later stage in segmentation of a spermatogonium, a spermatophore from the body fluid of A. cristata.
24 The 'invagination' of the forming spermatophore in A. cristata.
25 Section through the saucer-shaped spermatophore of A. cristata. The cells are spermatids, IJ X 4/z.
