Talk:Paper - Volumetric determinations of the parts of the brain in a human fetus 156 mm long (1915)
Volumetric determinations of the parts of the brain in a human fetus 156 mm. long (crown-rump)
F. C. Dockeray
Department of Psychology, University of Kansas
In the present communication there is reported a study of the volume of the main divisions of the brain as they are found in a fetus about four months old. The work was undertaken as a step in the history of the growth of the individual parts of the brain under the premise that a knowledge of their volume priority would indicate in a general way the functional prioritj^ of these parts. By means of the wax-plate reconstruction method it is possible to make an accurate enlarged model of the brain that can be separated into its chief component parts. Since such a model is made of wax of a uniform composition the relation by volume and by weight of the different parts can be determined both as to each other and as to the brain as a whole. This same method was used in determining the volume of the different parts of the opossum brain, by Professor Streeter and by Mr. H. A. Tash, who reported their results at the meeting of the American Association of Anatomists at Ithaca.^
The brain measured was taken from a male fetus measuring 156 mm. crown-rump, and 201 mm. total, length. The head measurements w^ere: Bitemporal, 48 mm.; occipito-frontal, 58 mm. These measurements were made on the fresh specimen. Its weight was 296 grams. The specimen was preserved in 10 per cent formalin, the skull having been opened to facilitate the penetration of the fixative. Subsequent^ the brain was removed, embedded in celloidin and prepared in serial sections 50 /jl thick,
1 streeter, G. L., 1911, Volumetric analysis of the brain of the opossum. Proc. Amer. Assoc. Anat.; Anat. Rec, vol. 5, p. 91.
every other section saved and stained with ahmi-cochineal. From this series a model was made enlarged five diameters after the well known Born method. Serial drawings were made with a projection apparatus on papers which were then incorporated in wax plates of such a thickness that the enlargement in all planes was the same ( X 5) . The drawings were then cut out from the plates and filed. This gave a model of the whole brain with the ventricles removed. The plates were then gone through a second time and the various parts cut awa}^ from each other so that their individual weights and volum.es could be separately determined. It was found that this could be done with considerable accuracy, and having the stained sections as a guide, it would have been possible to have carried the subdivisions further. But, having in mind both younger and older stages, it was decided that the adopted subdivision would prove most practical in the end. The results are given in table 1. In the first column of the table is given the weight in grams of the whole model and of its parts. In the second column is given the percentage of the total weight formed by each part, which would hold true for
TABLE
1
MODEL WGT.
PEH CENT OF
MODEL VOL.
ACTUAL
IN GMS.
TOTAL WGT.
INCC.
VOLUME
72.135
4.973
80.565
0.644
32.325
2.228
36.500
0.292
39.810
2.744
44.065
0.352
20.900
1.441
23.659
0.189
69.761
4.809
78.970
0.631
1287.673
88.776
1457.658
11.661
110.610
7.625
125.210
1.001
73.150
5.043
82.805
0.662
33 . 161
2.286
37.538
0.300
4.299
' 0.296
4.867
0.039
33.865
2.334
38.341
0.306
26.075
1.797
29.522
0.236
5.815
0.401
6.583
0.052
1.975
0.136
2.236
0.017
1143.198
78.815
1294.100
10.345
1450.468
100.000
1641 .929
i 13.135
Rhombencephalon
PvIeduUa and pons
Cerebellum
Mesencephalon
Diencephalon (inc. epiphysis)
Telencephalon
Basal ganglia
Caudate nucleus (inc. parolf. body and amygdaloid nuchuis)
Putamcn
Globus pallidus
Archipallium
Fornix and hippocampus
Paraterminal body
Olfactory bulbs
Neopallium
Total brain
the actual brain just as for the model. In the third column is given the volume in cubic centimeters of the whole model and of its different parts. Instead of determining the volume of each part separately it was found more practical to determine the specific gravity of the wax plates and then calculate the volumes from the weights given in the first column. In the last column is given the volume of the brain itself and of its parts. This was obtained by dividing the volume of the model by the amount of the enlargement, i.e., the cube of five diameters. It is to be remembered that this is the volume of the brain after it has been embedded and prepared in serial sections. The volume of the fresh brain could be obtained only by calculating the amount of shrinkage the specimen experienced in this process.
The subdivisions that were used follow as far as possible the embryological subdivisions adopted by His. Their boundaries could in most cases be determined by the cell structure of the sections. In some cases it was necessary to depend on the surface configuration of the model. The landmarks utilized in carrying out this subdivision are herewith detailed :
Rhombeiicephalon. This was separated from the spinal cord as nearly as possible at a point post cephalic to the first cervical nerve. The cephalic boundary was determined by a plane just skirting the inferior colliculus and passing out ventrally just in front of the pons. Laterally this plane passes just in front of the brachium connecting the cerebellum and pons.
Cerebellum. This is plainly demarcated by its surface outUne, while the pons is determined more by its internal structure, the main characteristic being the densely massed nuclei. The cerebellum at this time consists of a well fissured vermis and the two lateral lobes which are fissured dorsally but are still smooth, ventrally. In removing it the floccular margin was included and also the brachium pontis on each side to the point at which it meets the pons. The removal of the cerebellum leaves the medulla and pons, whose weight and volume are given together.
Mesencephalon. The caudal Umit of the mesencephalon is the same as the plane marking the cephalic border of the rhombencephalon, which has already been given. Its cephalic limit is a wedge-shaped plane that projects in between the masses of the diencephalon. At the median Hne its boundary is marked dorsally by the posterior commissure and ventrally by a point post caudal to the mammilary bodies. From this median line the plane of division on each side extends backward so as to include the red nucleus with the midbrain and comes to the surface at a groove marking the antero-lateral margin of the superior colliculus. Owing to the advanced development of the colliculi and the retarded development of the peduncular portion, the mesencephalon is V-shaped as regards its ventral aspect, as well as its cephalic boundary.
Diencephalon. Its separation from, the mesencephalon we have ah'eady indicated. From the telencephalon it is separated bilaterally by the internal capsule, and a sharp line of demarcation on the surface is afforded by the stria terminalis. Ventral!}^ where this is not present the line of division is continued along the anterior margin of the optic tract. By this manner of subdivision there is comprised in this portion the optic tract and thalamus including the habenular nuclei and epiphysis and also the whole hypothalamus with the exception of the hypophysis, which had been removed.
Telencephalon. This includes all the remainder of the brain. It was subdivided into three main divisions as follows:
Ba.2al ganglia. At the end of the fourth month these structures are clearly defined and bear a relation that closely appro?qimates the adult. The putamen and globus pallidus are easily recognized in transverse sections. As for the lamina of capsule fibers that surround them, the incisions were made half-way, so that part of the fibers would go with the globus pallidus and part with the caudate nucleus. The caudate nucleus throughout its greater extent is likewise clearly defined. At its head and tail ends, however, it is complicated by fusing with the parolfactory body and amygdaloid nucleus respectiveh'. On this account these latter were included with it.
Archipallium. This includes, in the first place, the olfactory bulbs, which were removed at a transverse line at the point where they become free from the brain wail. This corresponds to both the bulb and stalk of the adult. The paraterminal body includes the gray substance where the olfactory bulb is attached and the region of the future septum pallucidum and the pillar of the fornix, which could not be easily separated from it. The remainder of the archipallium is made up of the body of the fornix and its fimbricated extension into the hippocampus. The hippocampus is easily recognized by its histological structure and by the way it bulges into the lateral ventricle. With it was included the dentate fascia and the uncinate bod3\ The corpus callosum was included with the neopallium.
Neopallium. This includes the remainder of the telencephalon and represents what we know in the adult as the convoluted cortex, together with the subjacent white matter and includes the corpus callosum, as we have just pointed out.
In conclusion I wish to acknowledge the courtesy of Professor Streeter, who kindly put the resources of the Anatomical Laboratory of the University of Michigan at my disposal for the purpose of this investigation, and gave me many helpful suggestions as the work progressed.
BOOKS RECEIVED
The receipt of publications that may be sent to any of the five biological journals published by The Wistar Institute will be acknowledged under this heading. Short reviews of books that are of special interest to a large number of biologists will be published in this journal from time to time.
A LABORATORY :\IAXUAL AND TEXT-BOOK OF EAIBRYOLOGY. By Charles W. Prentiss, A.M., Ph.D., Professor of Microscopic Anatomj' in the Northwestern University Medical School, Chicago. Octavo, 400 pages, 368 illustrations, many of them in colors. Philadelphia and London : W. B. Saunders Company, 1915.
Preface. This book represents an attempt to combine brief descriptions of the vertebrate embrj-os which are studied in the laboratory with an account of human embrj'ology adapted especially to the medical student. Prof. Charles Sedgwick ]\Iinot, in his laboratory textbook of embryology, has called attention to the value of dissections in studying mammalian embryos and asserts that "dissection should be more extensively practised than is at present usual in embryological work. . . ." The writer has for several years experimented with methods of dissecting pig embryos, and his results form a part of this book. The value of pig embryos for laboratory study was first emphasized bj' Professor Minot, and the development of my dissecting methods was made possible through the reconstructions of his former students. Dr. F. T. Lewis and Dr. F. W. Thyng.
The chapters on human organogenesis were partly based on Keibel and flail's Human Embryology. We wish to acknowledge the courtesy of the publishers of KoUmann's Handatlas, ^Marshall's Embryology, Lewis-Stohr's Histology and McMurrich's Development of the Human Body, by whom permission was granted us to use cuts and figures from these texts. We are also indebted to Prof. J. C. Heisler for permission to use cuts from his Embryology, and to Dr. J. B. De Lee for several figures taken from his Principles and Practice of Obstetrics. The original figures of chick, pig and human embryos are from preparations in the collection of the anatomical laboratory of the Northwestern University Medical School. My thanks are due to Dr. H. C. Tracy for the loan of valuable human nlaterial, and also to Mr. K. L. Vehe for several reconstructions and drawings.
C. W. Prentiss.
Northwestern University Medical School, Chicago, 111., January, 1915.
ox THE AA^IGHT OF THE ALBINO RAT AT BIRTH AND THE FACTORS THAT INFLUENCE IT
HELEN DEAN KING
The Wistar Institute of Anatomy and Biology
In the course of an extensive series of breeding experiments with the albino rat a large amount of data has been collected regarding the body weight of these animals at different stages of their growth. The records dealing with the weight at birth are given in the present paper: those of postnatal growth will be published later.
Two sets of observations on the body weight of very young albino rats have already been recorded. In a paper published by Donaldson in 1906 the average weight of 40 young male albino rats is given as 5.4 gi'ams, and that of 17 females is stated to be 5.2 grams. The more extensive records of Jackson ('13) give the average weight of 107 young male albino rats as 5.1 grams, and that of 109 females as 4.8 grams.
The body weights, as given above, are those of animals that were 'newborn' when weighed. The term 'newborn,' as Jackson states, covers the period in the life of the animal from the time of bu'th up to one day. As a rule, young rats begin suckling ver}^ soon after their birth, and not infrequently part of a Utter will have suckled before the rest have been born. The weight of 'newborn' animals, therefore, is probably not the same as the birth weight in many cases. To obtain the birth weight it is necessary that the animals be weighed before they have suckled, since the amount of food consumed during the first few hours of postnatal life very appreciably increases the body weight. One can tell very easily whether or not the young rats have suckled, as the skin of the young animals is quite
213
THE AXATOMICAL RECORD, VOL. 9, XO. 3 MARCH, 1915
214 HELEN DEAN KING
transparent and if milk is present in the stomach or in the intestines it can be seen very clearly through the body wall.
In the com^se of this investigation 113 litters of rats were obtained at or soon after birth and weighed before any of the individuals had taken food. The same course of procedure was followed in making the records for each of the litters. The young rats were first separated according to sex by the method devised by Jackson ('12). Animals of the same sex were then weighed together to a tenth of a gram and the average weight for each individual computed; if, however, there was a very marked difference in the size of the individuals of the same sex the rats were weighed separately and the weight of each recorded. In addition to recording the number of young, the sex distribution and average body weight of the members of each Utter, the exact age of the mother at the time the litter was born was noted, also her body weight after the birth of the litter and her general physical condition.
The complete series of records comprise data from five different strains of rats that are being bred in The Wistar Institute animal colony at the present time:
1. Stock albinos: Members of the general colony that supposedly represent the normal albino type as it exists at the present time.
2. Inbred albino rats: Animals, originally taken from the general stock colon}^, that have been closely inbred for many generations.
3. Extracted albinos : A strain of rats descended from albinos cast by F, hybrids of the albino and the wild Norway rat (Mus norvegicus) .
4. Piebald rats: A strain derived from Fj hybrids of the albino and Norway rat.
5. Extracted grays: A strain also derived from the Fi hybrids of the albino and the Norway rat.
Table 1 gives a general summary of the birth records arranged according to the strain of rats from which the litters were obtained. The data in this table show that, regardless of strain, the
WEIGHT OF ALBINO RAT AT BIRTH 215
weight of the albino rat at birth is considerably less than that of newborn animals as given by Donaldson and by Jackson. As a rule the male rat at birth is somewhat heavier than the female, as is the case in many other mammals including man.
TABLE 1
Showing the birth weight data for various strains of rats
.STR.UX OP R.\.TS
is
a a
- r.
7i
■I.
<
fa
TOTAL WEIGHT OF MALES IN GRAMS
TOTAL WEIGHT OF FEMALES IN GRAMS
AVERAGE WEIGHT , OF MALES IN GRAMS
^g5 < * 5
W o o
<
Stock albinos
Inbred albinos
Extracted albinos
12 73
8 19
1
113
95 644
43
147
9
47
311
21
80
4
48
333
22
67
5
215.6 216.1 1409.51410.8
' 88.2 88.2
386.1 324.3
22.9 27.7
4.59 4.53 4.20
4.82 5.72
4.50 4.23 4.00
Piebalds
Extracted gravs
4.84 5.54
Total
938
463
475
'
On comparing the data for the various strains of rats it is found that stock albinos, both males and females, weigh slightly more at birth than do the inbred rats. The differences between them are not sufficiently great to have much significance, especially as the number of stock litters that was weighed was relatively small. Extracted albinos weigh considerably less at birth than either the stock or inbred rats. This fact is not surprising considering that these animals grow much less rapidly than stock or inbred albinos and that many of them, particularly the females, fail to attain the average adult size of stock animals. Piebald rats, both males and females, have a birth weight that is greater than that of any of the albino strains.
The average weight of the piebald females, as given in table 1, is shghtly greater than that of the males. This is probably a chance variation, since in 12 of the 19 Utters that were weighed the average weight of the males exceeded that of the females. From the single litter of extracted gray rats weighed at birth
216 HELEN DEAN KING
one can obtain but little idea of the birth weight for the strain, yet the records are significant in that they show an average weight for both sexes that is close to the mean betw^een the birth weight of the wild Norway rat, which is about 6.4 grams for both sexes according to Miller ('11) and that of the albino rat.
In any litter of rats, as a rule, individuals of the same sex are practically of the same size and body weight at birth. Occasionally, however, very marked exceptions to this rule are found. In one of the litters of inbred rats the difference in the weights of the various individuals was so unusually great as to call for more than a passing notice.
The litter in question contained eleven individuals, four males and seven females. One of the males, which was the largest rat yet obtained at birth, weighed 7.5 grams; the other three males were nearly uniform in size, weighing 4.9 grams, 5.0 grams and 5.1 grams, respectively. The females in this litter also showed considerable variation in body weight. The largest of the seven females weighed 4.6 grams, the smallest weighed 2.9 grams. The latter is not the smallest birth weight for the rat that has been obtained, however, as in one case in which the mother of the litter was in the last stages of pneumonia when the litter was born, two of the three males in the litter weighed 2.6 grams each; the third male weighed 2.9 grams which was also the weight of the one female in the litter. Female rats do not seem to show as great a variation in body weight at birth as do the males. The largest female yet obtained weighed 5.9 grams at birth; the smallest weighed 2.7 grams.
It seems most probable that such marked variations in the size of the different members of a litter at birth as are shown above must ])e due to a difference in the age of the embryos at the time that parturition occurs, not to causes acting late in gestation. Evidence already presented (King '13) indicates that, under certain conditions, ovulation in the rat may extend over three or four days, possibly longer. Ova liberated at various intervals for several days would probably all be fertilized, as the period of heat in the rat exists for about one week (Miller '11). The body growth of the embryos is very rapid during the latter part
WEIGHT OF ALBINO RAT AT BIRTH 217
of gestation, and embryos that developed from the ova first set free might be expected to have a greater body weight than the embryos that developed from ova hberated late in the period of ovulation, since they would have a longer time in which to grow. Rats born at one period of parturition, however, must be very nearly the same age, as the smaller individuals show no evidence of immaturity other than in their smaller size. If there is considerable difference in the age of the embryos developing simultaneously in the uterus the more mature ones are born first, and the remaining ones are born from one to several days later when they have reached the proper stage of development (King '13).
The variations in body weight found among rats at birth are seemingly greater than those in 'newborn' animals. The largest male in Donaldson's series weighed 6.5 grams, the smallest weighed 4.3 grams: corresponding figures for the females give 6.2 grams as the heaviest weight and 4.2 grams as the lighest weight. In Jackson's series the weights of the males range from 3.4 grams to 6.6 grams and those of the females from 3.5 grams to 6.3 grams.
There is a possible explanation for the narrow range of variation in the weights of 'newborn' rats besides the obvious one that the series of animals weighed was too small to contain the extreme variates in body weight. Individuals that are very small at birth may increase in body weight and in size more rapidly during the first few hours of postnatal life than the members of the litter that have a heavier birth weight. This would tend to equalize the size of the individuals and so give all of them approximately the same chances of obtaining food. The early growth changes in the rat have not been studied sufficiently as yet to give evidence on this point.
In analyzing the data collected in connection with the birth weights with a view of ascertaining, if possible, some of the factors that help to determine the weight of the rat at birth, it has been considered advisable to make use only of the records for the 85 htters of stock and of inbred albino rats. The average weight of the young in the piebald and in the extracted gray litters is so much greater than that of the albinos that the effects
218 HELEN DEAN KING
of these records on the general averages, when taken in small groups, would be altogether disproportional to the number of individuals in^'olved. The birth weights for the extracted albinos, on the other hand, fall so far below those for the other albinos that they seem properly to belong in a class by themselves. It does not seem worth while to analyze the data for these strains otherwise than in the manner shown in table 1. The records are as complete as for the stock and inbred albinos, however, and are filed at The Wistar Institute.
THE EFFECTS OF THE AGE OF THE IMOTHER OX THE WEIGHT OF HER YOUNG AT BIRTH
Slonaker ('12) states that the age of the mother affects not only the number of young rats in a litter but also their weight at birth, young mothers being less prohfic than older ones. He makes no mention, however, of the extent of the data on which this conclusion is based.
Under the conditions existing in The Wistar Institute animal colony the female albino rat usually has her first litter when she is about three months old, and she is capable of bearing young until she is about fifteen months old. In order to study the effects of the age of the mother on the weight of her young at birth the reproductive period in the life of the albino female has been arbitrarily divided into the four following periods:
1. From 90 to 120 days: This is the age when young females are growing very rapidly and the time when the great majority of them cast their first litters.
2. From 120 to 180 days: During this time the female reaches the end of the rapidly growing period and becomes fully mature.
3. From 180 to 300 days: The female is at the height of her reproductive powers during this period and has attained full growth.
4. From 300 to 450 days: In this period there is a dying out of the reproductive power and little, if any, growth.
Table 2 shows the data for the 85 litters of stock and of inbred albino rats arranged in four groups according to the age of the
WEIGHT OF ALBINO RAT AT BIRTH
219
mother at the time that the Utter was born. The data, as arranged in this table, do not show the gradual increase in the body weights of the young from the first to the fourth group that one would expect to find if the age of the mother is the dominant factor in determining the bu'th weight of her young. If, however, we compare the average birth weight of the rats in Utters cast by females during the first reproductive period with that of the individuals belonging to litters born when the reproductive power of the mother is waning, it is found that the average body weights in the first group are considerably less than those in the last group. The difference between them, amounting to 0.3
TABLE 2
Showing the birth weight data for 85 litters of stock and inbred albino rats arranged according to the age of the mothers at the time that litters icere cast
gram in the case of the males and 0.2 gram in the case of the females is sufficiently great, I think, to warrant the conclusion that the weight of a litter at birth depends, to a certain extent, on the age of the mother. During the first reproductive period young females are growing very rapidly both in body size and in bodj^ length and presumably, therefore, the growth processes consume a considerable part of all available energy. Litters cast by females at this time contain, as a rule, few individuals and these are of relatively small size. In older females growth has practicaU}^ stopped and more energy can be used for the production of larger Utters containing individuals that have a heavier weight at birth.
220 HELEN DEAN KING
THE INFLUENCE OF THE BODY WEIGHT OF THE MOTHER ON THE BIRTH WEIGHT OF HER YOUNG
Under normal conditions body weight and age are closely correlated in the rat, as Donaldson's investigations have shown. Body weight, however, being easily affected by changes in environment or in nutrition is, to a certain extent, independert of the age factor and indicates very clearly the physical condition of an animal. Rats that are heavy for their length and age are usually in excellent health; those that are light in weight are generally ill, as certain diseases — for instance, 'pneumonia' — may be shown by a rapid drop in the weight of the animal before any other symptoms of illness are manifested. The body weight of a female, as indicating her general physical condition irrespective of age, may possibly, therefore, be a factor that would tend to influence the birth weight of her young.
As age is the factor that so largely determines body weight in the rat, it has seemed advisable to group the records for the body weights of the females according to age. To do this it is necessary to know the body weights that are normal for various ages. Table 3, compiled from an extensive series of unpublished data collected in the course of my breeding experiments, gives the normal weight of stock and of inbred albino females that correspond with the age groups used as the basis of analysis in the previous section. Inbred albino females are shghtly heavier
TABLE 3
Showing the body weight of stock and of inbred albino rats normal for different age
groups
AGE OP FEMALES IN BATS
NOEMAL WEIGHT OF STOCK ALBINO FEMALES
NORMAL WEIGHT OP INBRED ALBINO FEMALES
90 to 120 days 120 to 180 days 180 to 300 days 300 to 450 days
148 to 173 grams 173 to 195 grams 195 to 219 grams 219 + grams
156 to 175 grams 175 to 199 grams 199 to 221 grams 221 + grams
for a given age than are stock albino females, as is shown in table 3. Since the great majority of the birth weights recorded are of litters belonging to the inbred strain, the body weights
WEIGHT OF ALBINO RAT AT BIRTH
221
of the inbred females have been used as the basis for the grouping of the data in the present instance.
Table 4 gives the various birth weight records arranged according to the body weight of the mothers at the time that parturition occurred. If the body weight of all the females had been normal for the age at which their litters were cast, table 4 would be practically a duplicate of table 2, where the data are arranged according to the age of the mothers. A comparison of the two tables shows, however, a very different distribution
TABLE 4
Showing the birth weight data for stock and inbred albino rats arranged according to the body weight of the mothers at the time that the litters were cast
^
^
fc m
fe z
H Z
^ !5
Q
O S
O "
K "
K S
^3
g
K a
Eh
o
S n
BODY WEIGHT OF FEMALE
O
a "
a H » >3
^:i
K
w to n m 2 «
m
i-i
<
^ o o
S
b.
H
H
■4
-<
To 175 grams
27
209
98
111
423.8
440.7
4.46
3.86
175 to 200 grams
23
25
222 233
108 111
114 122
491.3 508.7
506.8 524.7
4.55 4.58
4 47
200 to 220 grams
4.30
220 + grams
10
75
41
201.7
158.0
4.91
4.64
of litters in the corresponding groups. The first group in each table happens to contain the same number of litters, but in table 4 this group comprises 9 litters from females that were over 120 days of age when parturition occurred. These females had fallen below the normal weight for their age and were, presumably, not in especially good physical condition. The second group in table 4 contains 6 litters from females younger than 120 days, and 4 litters from females that had passed 180 days of age when the litters were cast. In this group, therefore, 10 of the 23 litters belonged to females that did not have a normal body weight. Eighteen of the 25 litters belonging to the third group of table 4 were cast by females younger than -180 days of age and two of them by females older than 300 days. In the last group only thi-ee of the 10 litters came from females that had a body weight normal for the age of which the litters were cast.
222 HELEN DEAN KING
According to the records as given in table 4, the a\'erage weight of young rats at birth increases directly as the body weights of the mothers increase. ^\Tien the body weights of the females are below 175 grams the average weight of the young males in the litters is 4.46 grams. This average rises to 4.91 grams for the males in the litters cast by females weighing 220 grams or more. Records for the female young do not show quite such uniformity as in the case of the males, since the weights of the individuals belonging to the third group are less than those of the individuals in the second group. The weight of the females in the fourth group, however, is greater by 0.78 grams than that of the females belonging to the first group. This difference is considerably greater than that shown by the males in the two groups.
That these results depend to a considerable extent on the age of the mothers of the litters there can be little doubt, since body weight is closely associated with age and in the noi'mal animals the range of variation is not very great.
If, however, we disregard the age factor and take the body weights of the mothers as indicative of the physical fitness of the animals, it is evident that the heavier females, being in excellent condition, tend to produce young that are larger at birth than the young cast by females that are relatively light in weight and probably, therefore, not in very good condition.
From this analysis of the data it follows that it is not the body weight of the female in itself, but the factors on which the weight largely depend, i.e., age and physical condition, that have a pronounced influence on the birth weight of the young,
THE INFLUENCE OF THE LITTER SIZE ON THE WEIGHT OF THE
YOUNG AT BIRTH
The size of a litter of albino rats depends, seemingly, on a number of different factors, one of which is undoubtedly the age of the mother. Very young females and those that have passed their prime have smaller litters, as a rule, than females at the height of their reproductive powers. The physical condition of the mother is also a factor that apparently affects the litter size,
WEIGHT OF ALBINO RAT AT BIRTH 223
as females in poor condition rarely have a large litter, and if the number of young exceeds the average for the species several of them are usually stillborn. Litter size therefore is another factor so inseparably linked with the age and physical condition of the mother that its influence on the birth weight of the young must be considered in connection with the other factors involved. Unpublished data for over 1000 litters of stock albino rats show that the average litter contains seven j^oung. The size of the litter varies greatly in different cases, the range being from one to thirteen.
TABLE 5
Showing the birth weight data, for stock and inbred albino rats, arranged according to the size of the litter. G, litters cast by females in good physical condition; P, litters cast by females in poor condition
SIZE OF LITTER
&<
m 1
OS
<
S '
X
a
z
■$■
to !
J
< < 1
S^
H
1
B* -• ir -I ac
K « , a S aw
^ < ID \ ~„*' -ST.
o or less
[4(0) 18 ' 9 9 45.1 39.6 o.0l\ I 4.40l
\5(P) 24 10 14 37.3 50.8 3.73/3.62;
6 to 8 30 217 111 106 512,6 464.7 4.61 4.38
9 or more 46 480 228 252 1030.01075.7 4.51 4.26
In order to analyze the birth records on the basis of litter size, the litters have been arbitrarily divided into three groups: small htters, containing five or less young; medium sized Utters, containing six to eight individuals; large litters with nine or more young.
The arrangement of the data according to the above classification is shown in table 5. The data, as shown in this table, seem directly opposed to the generally accepted view that animals belonging to small litters weigh more at birth than those belonging to large litters, since the average weight of both males and females is considerably less for the small litters than for the very large ones. The record cards show, however, that five of the nine small htters were cast by females that were in poor physical condition. In these five litters the average weight
224 HELEN DEAN KING
of the males was 3.73 grams and that of the females was 3.62 grams; in the fom' htters from females apparently in good condition the average weight of the males was 5.01 grams and that of the females was 4.40 grams. Only seven of the 85 htters weighed were cast by females in poor condition, and five of them, as shown above, belong in the small litter group. The other two were litters of medium size, and if their records were omitted from table 5 the average bkth weight of both males and females in this group would be raised shghtly.
It seems justifiable, therefore, to disregard the data for the fi\-e small htters cast by females in ill he'alth and to take the records for the four litters cast by females in good condition as representing the bhth weights for the small litter group. If this be done, the data in table 5 show that the average weight of the 3^oung in small litters greath^ exceeds that for the individuals in large htters: individuals in medium sized litters weigh close to the mean between the weights of the animals in the small and in the large litters. This rule holds true for animals of both sexes, and seems to indicate that the birth weight of young rats depends to a considerable extent on the size of the litter, irrespective of the other factors that may be involved.
Additional evidence that the size of the litter influences the birth weight of the young is furnished by the records for the largest Utter of albino rats as yet obtained. This litter, which belonged to the inbred strain, contained 17 individuals, 10 males and 7 females. The young rats in this litter were several hours old when first examined and they had all suckled, so that it was not possible to obtain their birth weights. The average weight of the 17 individuals at the time they were found was 2.7 grams. At birth, therefore, these rats probably weighed not more than 2.5 grams each. The smallest bu'th weight for the rat that has as yet been taken is 2.6 grams. In this litter, therefore, the very small size of the individuals was undoubtedly due- to the exceptional size of the litter, as the mother of the litter was large for her age and seemingh' in the best of condition when the litter was born.
WEIGHT OF ALBINO RAT AT BIRTH 225
In a study of the weight of guinea-pig Minot ('91) found that the size of the pigs at birth depends to a considerable degree on the number of young in a litter; the larger the htter the smaller the pigs at birth. According to ^Nlinot, the litter size influences the birth weight by changing the length of the gestation period.
When the htter is large the gestation period is shortened and therefore the young pigs do not have as long a time in which to grow as is the case when the litter is small and consequenth^ the gestation period longer. The birth weight of guinea-pigs, therefore, does not depend, according to ^Nlinot, on "the ratio of food supply and demand" but on the length of gestation.
In the rat, as in the guinea-pig, the length of the gestation period varies considerably in different cases. Normally it is from 21 to 23 days,, but in lactating females it may be extended to 34 days (King '13). As far as I am aware, no attempt has been made as yet to ascertain the relation between the number of young in a litter and the duration of gestation in the rat. When a lactating female becomes pregnant the number of young in the second htter is certainly a factor of considerable importance in determining the length of the gestation period, for gestation is prolonged from one to thirteen days if the number of young carried is very large. How the birth weight of the yomig is affected by this prolongation of the gestation period remains, to be determined. Growth is so rapid during the latter part of gestation that the extension of the period for even one day might be expected to materially increase the size of the embryos unless other factors tend to check growth at a certain stage of development.
THE RELATIOX OF THE LITTER SERIES TO THE WEIGHT OF THE
YOUNG AT BIRTH
To arrange the records for the birth weights according to the position of the Utters in the litter series would seem to be merely another way of testing the effect of the age of the mother on the weight of her young at birth, since it is impossible to eliminate the factor of age in carrjdng out such a plan. Female rats show very great individual differences, however, in regard to the
226
HELEN DEAN KING
time when their htters are produced. Seme rats begin breeding before they are three months old; others do not have their first litter until they are four or six months old. Certain females will cast a litter every month for several succeeding months; others never have a litter oftener than every two months and not infrequenth' tlii^ee or four months will intervene between htters even when the females are apparentlj^ in excellent condition.
The plan of the breeding experiments at present under way requires that every breeding female shall have four htters. Some females have the required number of litters by the time they are six months old ; others not until they are a year or more old. While, therefore, the range of variation in the age of the mothers is comparatively shght for the first and for the second pregnancies, it may be extended over several months before the third and the fourth litters are produced.
The litter series must always be, in a sense, an age series, yet the individual variations in the frequency of the litter production are sufficiently great, I think, to make it worth while to study the birth weight records from the point of view of the number of the pregnancy.
Table 6 shows the records for birth weights arranged according to the litter series. An analysis of the litters from several hundred females has shown that the first of a rat's four litters
TABLE 6
Showing the birth weight data for stock and inbred albino rats arranged according to the position of the litters in the litter series
,
a ?:
b. m
6- Z
H Z
f z
^
B
^
o a
O «
X "
s
e
^ ■<
o
c
i^
a
n
LITTER SERIES
a a
S ° «
2
Hi
w
<
w
VERAOE OF MA ORAM3
VERAGE OF FEM GRAMS
■<
z
s .
f
H
<
<
1
i
' 22
8.5
187
91
96
400.0
390.2
4.39
4.06,
2
21
S.9
188
82
106
370.8 452.7
4.52
4.27
3
25
0.4
236
125
111
576.6, 492.5
4.61
4.43
4
17
7 5
128
60
68
278.61 296.5
4.64
4.36
WEIGHT OF ALBINO RAT AT BIRTH 227
is usually the smallest, the second and the third the largest, and the fourth a little larger than the first. This rule, however, does not happen to apply in the case of the litter series shown in table 6. Not only is the average number of young in the first group considerably above that which is normal for the species, but it greatly exceeds that for the fourth group; the second and third groups are fairly normal in size. Litter size is, in all probability, a factor that in the present instance does not materially affect the birth weights of the young rats, since the average size of the litters varies but little for the four groups.
Table 6 shows that the birth weights of the males increase with the ascending scale of the litter series; for the females there is a slight irregularity in the figures for the third and fourth groups, but the average weight of the young females in the fourth group is 0.3 grams greater than that in the first group. A comparison of these birth weights with those given in table 2, where the age of the mother formed the basis of the classification of the data, shows that the figures for corresponding groups are much the same, the deviation in no case being greater than 0.07 grams. No other tables given show such close agreement, and, therefore, it appears that the position of a litter in the litter series influences the birth weight of the young chiefly because it so closely involves the factor of age.
THE EFFECTS OF THE PHYSICAL CONDITION OF THE MOTHER ON THE WEIGHT OF HER YOUNG AT BIRTH
One accustomed to the care and handling of rats can quickly tell from the general appearance and weight of an animal whether or not it is in good physical condition. A hunched back, labored breathing, dark red eyes and a relatively light weight indicate internal disorders from which the rat rarely, if ever, recovers. On the other hand, a heavy weight for body size and pronounced vigor in action shows that the animal is in excellent health.
Seven of the 85 females whose litters were weighed were noted as in exceptionally good condition when theu' litters were born, and seven others showed unmistakable signs of ill health. Records for the weights of the litters from these 14 females are given in
228
HELEN DEAN KING
table 7. In the litters from females that were in excellent condition when their litters were born the average birth weight of the 3oung for both sexes is much greater than the normal, and it exceeds the average weights of the young in the litters cast by females that were in ill health by over 1.1 grams, as is shown in table 7. This result indicates that the physical condition of the mother, irrespective of her age, is a very important factor in determining the weight of her young at birth, probably through its action on the nutritive conditions to which the embryos are subjected.
TABLE 7
Shoiving the birth xveiglit data for 14 litters of stock and of inbred albino rats arranged according to the physical conditions of the mothers at the time that the litters ivere cast.
^
3
! §S
S2
£S
£2
ij
tt
o „
' £ «
5 ="
3n
H »
COXDITIO.V OF MOTHER
o
6.
2 "
i*^
K « ti
to
S
5
m
^ to
< <
w O O
z
z
S
f=,
EH
H
•<;
^
Excellent
7
66
37
29
183.2
136.6
4.95
4.71
Poor
7
39
14
25
53.4
89.5
3.81
3.58
Another way of studying the same problem is to compare the physical condition of the mothers of the litters in which the individuals were unusually heav}^ at birth with the condition of the females that cast litters having a very light weight. Data for such a comparison are given in table 8, w^here the body weights of the mothers of the litters are taken as indicative of the physical condition of the animals. There was a total of 13 litters in which the average weight of the young of both sexes was 5 grams or over at birth. The records show that in ten cases the mothers of the litters weigh considerably more than the amount normal for their age; each of the remaining females had a body weight that corresponded exactly with her age. All 13 litters, therefore, were cast by females that were in very good physical condition as far as one could judge from the general appearance and weight of the animals.
WEIGHT OP ALBINO RAT AT BIRTH 229
TABLE 8
Shoicirig the body weights of the mothers of those litters in which the birth weights of the young were above or below the normal weight
AVERAGE WEIGHT OF YOUNG IN GRAMS
NUMBER OP LITTERS BODY WEIGHT OF MOTHERS
5
or
more
/ I
10 3
above normal normal
■4
or
less
•• {
9 9
below normal normal
No definite conclusions can be drawn from the records of the 18 Htters in which the young had a birth weight of 4 grams or less since the;, results are, in many cases, complicated by the factor of litter size. In 9 cases the body weight of the mother of the litter was below the weight normal for the age indicated; in the remaining cases the weights of the females were normal for their age, but all of the litters were very large (containing from 9 to 14 young) so Utter size was doubtless a factor that had lowered the weight of the young to a certain extent. As far as the evidence from this set of records goes it seems to indicate that, in some cases at least, the small size of rats at birth is due to the poor physical condition of the mother which prevents the proper nourishment of the young and so inhibits their growth.
SUMMARY
1. Data for 113 litters show that the body weight of the young at birth differs considerably in various strains of rats. Stock and inbred albino rats weigh about the same at birth; the average weight of the males being 4.54 grams and that of the females 4.27 grams in the 85 litters that were weighed. Extracted albinos have a very low birth weight; while piebalds and extracted grays weigh much more at birth than do any of the albino strains (table 1).
2. The male rat at birth usually weighs about 0.2 grams more than does the female.
3. There is a wide range of variation in the birth weights of albino rats. The weights for the males range from 2.6 grams
THE ANATOMICAL RECORD, VOL. 9, NO. 3
230 HELEN DEAN KING
to 7.5 grams; those for the females vary between 2.7 grams and 5.9 grams.
4. In any Utter, as a rule, individuals of the same sex are practically of the same size and body weight. Alarked exceptions to this rule are probabh' due to a slight difference in the age of the embryos at the time that parturition occurs.
5. The body weight of rats at birth depends upon a number of different factors that are more or less closely related:
(a) The age of* the mother. Individuals in litters cast by older females weigh more at birth, as a rule, than do the individuals in litters belonging to very young females (table 2).
(b) The physical condition of the mother. Rats in good physical condition bear young with a birth weight considerably above that of the young cast by females in poor condition (tables 7, 8).
(c) The body weight of the mother. The body weight of a female influences the birth weight of her young chiefly because it depends on the two more important factors of age and physical condition. Rats that have a Ver\' heavy body weight are older and in better physical condition than rats with a light body weight, and their litters comprise individuals with a corresponding greater weight at birth (table 4).
(d) The size of the litter. Individuals in small litters weigh more at birth than do individuals in large litters (table 5).
(e) The position of the litter in the litter series. The birth weight of young rats increases directly with the ascending scale of the Utter series. But, as the litter series is always an age series, it is probable that the number of the pregnancy affects the birth weight onh' because it involves the factor of age (table 6).
(f) The length of the gestation period. The evidence regarding the influence of this factor is slight as yet. It is very probable that the prolongation of the gestation period for even one day materially increases the weight of the young at birth.
WEIGHT OF ALBINO RAT AT BIRTH 231
LITERATURE CITED
DoxALDSON, H. H. 1906 A comparison of the white rat with man in respect to the growth of the entire body. Boas Anniversary Volume, New York.
J.\CKSox, C. M. 1912 On the recognition of sex through external characters in the young rat. Biol. Bull., vol. 23.
1913 Postnatal growth and variability of the body and of the various organs in the albino rat. Amer. Jour. Anat., vol. 15.
KixG, Helex Deax 1913 Some anomalies in the gestation of the albino rat
(Mus norvegicus albinus). Biol. Bull., vol. 24. Miller, Newtox 1911 Reproduction in the brown rat (JNIus norvegicus). Amer. Natur., vol. 45.
MixoT, C. S. 1891 Senescence and rejuvenation. I. On the weight of guineapigs. Journ. Phys., vol. 12.
Sloxaker, J. R. 1912 a The normal activitj^ of the albino rat from birth to natural death, its rate of growth and the duration of life. Jour. Animal Behavior, vol. 2.
1912 b The effects of a strictly vegetable diet on the spontaneous activity, the rate of growth, and the longevity of the albino rat. Leland Stanford Junior Univ. Pub.
OBSERVATIONS ON THE ORIGIN OF THE MAST LEUCOCYTES OF THE ADULT RABBIT
PRELIMINARY NOTE
A. R. RINGOEX
From the Histological Laboratory of the Department of Animal Biology, University of Minnesota, Minneapolis
The investigations of Maximow have shown that in mammals the connective tissue mast cells are very different from the mast cells of the blood. _ Maximow and Weidenreich believe that the only feature which the two types of cells have in common is the presence of basophilic granules in the cytoplasm, which stain metachromaticallj^ with basic aniline dyes. The two types of cells, however, represent independent lines of leucocyte differentiation and development, with their own peculiar nuclei and granules.
Maximow ('06) found histogenous mast cells in all the mammals he investigated, even in the rabbit where most investigators have failed. He calls attention to the fact, that where there are relatively few histogenous mast cells, the deficiency is made up by increased numbers of haematogenous mast cells, and vice versa. That such a close compensatory relationship exists between the two tj^Des of cells is shown very well in the adult rabbit, there being comparatively few histogenous mast cells, but numerous mast leucocytes.
Within the past few years the origin of the haematogenous mast leucocyte has been the subject of considerable haematological investigation. The earlier mvestigators, including Ehrlich, assumed that mast leucocytes were represented in the bonemarrow by certain characteristic myelocytes and evolved like the other granular cells. Weidenreich, however, has recently shown that this is not the case with the human mast leucocyte. He believes that human mast leucocytes are formed from de 233
234 A- '^- RINGOEN
generating lymphocytes within the circulation. He derives the mast granules from the fragmenting nucleus and not from the protoplasm of the degenerating cell. Various other investigators in working on the mast leucocytes of the rabbit have come to similar conclusions with reference to the origin of these cells within the blood stream.
In 1909 Proscher concluded from his observations that the mast leucocytes of the blood of the rabbit are merely lymphoid cells of various types whose 'spongioplasm' has undergone a special form of mucoid degeneration, which results in the formation of granules which are closely related to mucin. Mast leucocytes of the rabbit are, therefore, not true granulocytes and are not derived from myelocytes of the bone-marrow.^
Pappenheim's students Benacchio ('11), .Kardos ('11), and St. Szecsi ('12) came to similar conclusions with reference to the mast leucocytes of the guinea-pig and the rabbit. . They could find no mast myelocytes in the marrow of either of these animals; they therefore concluded that the mast leucocytes are not true granulocytes. They beheve that the mast leucocytes of the guinea-pig are merely eosinophil leucocytes whose granules have remained in an unripe basophilic condition. They claim that they can find all the intermediate stages between these so-called mast leucocytes and the ripe eosinophil leucocytes whose granules have an acid staining reaction. Benacchio concluded that all of the myelocytes with basophilic granules in the marrow of the guinea-pig and rabbit were either unripe eosinophiles or special cells. In other words, he believed that all of the granulocytes with basophilic granules were destined to differentiate either mto eosinophiles or into special cells, and that mast cells are not present in the marrow of these animals. ^
1 Pappenheirn came to similar conclusions. His views, however, are based largely upon the work of Proscher.
2 Pappenheirn and St. Szecsi also believe that mast leucocytes are not represented in the marrow of the rabbit. "Die sog. Blutmastleukozyten stai-men natiirlich aus dem Knochenmark, aber z. T. sind sie keine eigentlichen Mastzellen, sondern nur unreifkornige sonstige Granulocyten, deren Granula andere chromophile Reaktion hat, z. T., soweit sie eigentlichen Blutmastzellcn sind. bilden sie sich aus Lymphoid-zellen wohl erst im Blut selbst oder untcr pathologischcr Einwirkung (Myelosc)."
ORIGIN OF MAST LEUCOCYTES 235
In a recent paper Maximow ('13) has shown that mast myelocytes are present in the bone-marrow of man, and that they are actually seen undergoing mitosis. Maximow, therefore, believes that the granules of haematogenous mast ceils cannot be products of the degenerating nucleus or spongioplasm. He could never find any evidence for the degenerative processes described by Weidenreich and Pappenheim. Maximow was able to trace the differentiation of the granules and the evolution of the cells from the tji^ical myelocytes and beUeves, therefore, that the haematogenous mast cells are true granular leucocytes which are equivalent to the other types of granulocytes of the blood and marrow. Maximow is also the chief exponent of the theory that the mast leucocyte of the rabbit is a true granular cell, which is in all respects equivalent to the human mast cell. He found nothing that would lead him to conclude with Benacchio, Kardos, and others that the mast leucocyte is not differentiated in the bone-marrow.
Maximow's observations on the origin of the mast leucocytes are of the greatest importance, but his observations should be confirmed by further studies, since he maintains that the mast leucocytes do not arise in the circulating blood from altered lymphocj^tes, but are differentiated in the marrow from certain specific, characteristic, basophilic granulocytes. Downey's* recent studies ('13) on the mast leucocytes of the guinea-pig have resulted in the complete corroboration of Maximow's findings. He finds that the granules of mast leucocytes can always be distinguished from those of eosinophil and special myelocytes, even though they are subject to shght changes in size and shape. My observations on the mast leucocytes of the rabbit, which were carried on under the direction of Professor Downej^ and to whom I wish to extend my most sincere thanks, are also a further confirmation of Alaximow's results.
It is a well known fact that the early myelocyte stages of eosinophiles and special cells have a primitive or 'prodromale' granulation which is decidedly basophilic when first differentiated.
3 A preliminary report was published in the Proceedings of the American Association of Anatomists, The Anatomical Record, 1914, vol. 8, no. 2.
236 ' A. R. RINGOEN
According to Pappenheim ('12) this primitive granulation is supposed to have nothing to do with the final eosinophilic or special granulation which is developed later. Pappenheim believes that this primitive granulation is basophilic, but that it disappears when the specific granulation develops later. The latter is also basophilic when it first forms. According to Maximow ('13) the prunitive granulation is azurophilic. Downey has shown that in the guinea-pig histogenous mast cells are derived from a type of cell similar to the clasmatocyte with a prunitive granulation. MTiether the primitive granulation disappears or becomes the final mast cell granulation is not known.
Maximow and Pappenheim have called attention to the very decided basophilic quota of j^oung eosmophil and special granules in the eosinophiles and special cells of the rabbit. Bone-marrow of the' rabbit, prepared according to Pappenheim's^ method, show the preponderance of basophilic granules in eosinophil and special myelocytes very well. The granules are seen to ^^ary in size, but are generally rounded or sUghtly irregular and show no definite arrangement withiu the cell body. All of the granules when first formed have a strong affinit}^ for basic aniline dyes, in which respect they resemble the basophilic granules of mast cells. Other cells, however, whose general character is siaiilar to these contain a few granules which are intermediate in staining reaction, having an affinity for both the acid and basic component of the staining mixtm^e which gives these granules a mixed tone. Cells can also be found in which the number of basophilic granules is greatly reduced with a corresponding increase in the number of the intermediate granules. This change of staming reaction in the basophilic granule suggests that the early myelocyte with basophilic granules is being differentiated into a cell in which the granules are acidophilic, and shows that granules of this type are not true mast granules.
Benacchio has made similar observations; however, he goes further and concludes that the myelocytes with basophilic granules, similar to those described above, are the only type of
Folia Haem, Archiv., Bd. 13.
ORIGIN OF MAST LEUCOCYTES 237
basophilic myelocyte present in the marrow of the rabbit; in other words, that all of the myelocytes with basophilic granules are destined to differentiate into eosinophiles and special cells.
Kardos, in workmg with sections of bone-marrow fixed in 100 per cent alcohol and Helly's mixtm-e, found neither mast cells, nor cells of any kind which contained basophilic granules. Paraffin sections and smears were studied in the present investigation, but with decidedly different results from those obtained by Kardos. In sections (material fixed in 100 per cent alcohol and stained in alcoholic thionin) basophilic myelocytes are just as numerous as they are in the bone-marrow smears prepared according to Pappenheim's method. The alcoholic material shows practically the same conditions as are seen in the bone-marrow smears. Sections stained in May-Giemsa show many cells which contain basophilic granules only, while others contain both basophilic and eosinophihc granules, and in still other cells all the granules are decidedly acidophilic. Furthermore, and in direct opposition to the findings of Benacchio and Kardos, it is possible to demonstrate in these same preparations and in smears also, a second type of basophihc myelocyte in the marrow of the adult rabbit. This is the mast myelocyte or the precursor of the mast leucocyte. Scattered throughout the section one sees numerous cells which contain a variable number of granules; the granules have a remarkable avidity for basic aniline dyes. These granules are metachromatic as well as basophilic, in fact, the metachromasia of the granules is so pronounced that they can neither be over-looked nor interpreted as the ordinary basophilic granule of the eosinophil and special myelocyte. The size and shape of the metachi'omatic granule, and the configuration of the nucleus are very suggestive of the mast leucocyte of the blood. On closer investigation and observation their identity is at once apparent.
In the marrow of the adult rabbit, in addition to the fully differentiated mast leucocytes with a more or less poljinorphous nucleus, all intermediate stages between them and their myelocytes can be followed out. In the early myelocyte stages the nucleus is round, but later it becomes polymorphous. A dis
238 A, R. MNGOEN
tinctive feature of the mast myelocyte, as pointed out by iVIaximow, is its very thick nuclear membrane. The writer can also add in further support of Maximow's statement, that eosinophil and special myelocytes usually occur in groups, while mast leucocytes and mast myelocytes appear more or less scattered throughout the section.
Mast mj^elocytes are well preserved in bone-marrow smears fixed in lucidol-acetone and stained in either alcoholic thionin or May-Giemsa. For fixation the solution devised by St. Szecsi^ is used. Smears of fresh bone-marrow were made by rolling a small piece of marrow over a chemically clean cover-slip. Without allowing the smears to dry in the least they were immediateh' placed into a covered dish containing the lucidol-acetone fixative. At the expiration of fifteen minutes the smears were removed from the lucidol-acetone mixture, transferred without drying to another covered dish containing a mixture of acetone and xylol, thi'ee parts of the former to two parts of the latter. St. Szecsi states that the object of using this mixture is to dissolve the lucidol crystals, and clear the preparations, ten minutes are sufficient to complete the process. Finally the smears are placed in methyl alcohol, from one-half to one minute. Bonemarrow smears, provided that the smear is not too thick, are well fixed after being subjected to the action of the lucidol-acetone.
In view of the fact that several modern haematologists have denied the presence of mast leucocytes in the bone-marrow of the rabbit, the lucidol-acetone preparations are of particular value and interest. After seeing a single preparation there can be no doubt as to the presence of mast myelocytes and mast leucocytes in the marrow of this animal. A single preparation usually shows great numbers of mast cells. In a single field I have often counted as manj^ as six fully differentiated mast leucocytes.
The mast myelocyte is such a distinctive type of cell that it is easily distinguished from eosinophil and special myelocytes. In lucidol-acetone preparations stained with May-Giemsa the granules of mast leucocytes stain an intense bluish black, while the
^ His method of procedure appeared in the Deiitschen Medizinischen Wochenschrift, no. 33, 1913.
ORIGIN OF MAST LEUCOCYTES 239
basophilic granules of eosinophiles and special cells are of a reddish black tinge. The sharp contrast in the staining reactions of the mast myelocyte as compared with the eosinophil and special myelocyte is so pronounced and so characteristic that every mast cell is easil}^ separated from the eosinophil or special myelocyte.
Of the various methods tried none gave sharper pictures for the demonstration of mast cells than did the lucidol-acetone fixation. The basophilic granules of the mast leucocyte are very well preserved. In some instances the cell body is so filled with the basophilic, metachromatic granules that the outline of the nucleus is extremely difficult to follow.*^ In cases, however, where the exact outline can be seen, I find that the nucleus is typically polymorphous and shows no similiarity to a lymphocyte nucleus, neither does it possess lymphocyte characters nor show signs of degeneration. My preparations showed nothing to support Proscher's theory that the mast leucocyte is derived from a h^mphocyte and that the nucleus remains practicalhidentical with the lymphocyte nucleus. In all probability Proscher based his theory on the early, basophilic, mononuclear myelocj^tes of eosinophiles and special cells. At any rate, the technique which he used was such that the granules of mast leucocytes would not be preserved.
The lucidol-acetone preparations show further that the basophilic granules of mast leucocytes vary in form, size, and in number as previously stated. In the rabbit the granules are fine, usually rounded or slightly irregular. As far as the mast leucocyte of the rabbit is concerned there is no evidence to show that the nucleus is concerned in the elaboration of the granules, as is claimed by Weidenreich for the mast leucocyte of man. Proscher also claims that the nucleus takes no active part in the elaboration of granules.
Maximow's method of fixing bone-marrow in 100 per cent alcohol followed by staining in alcoholic thionin or May-Griinwald was also tried. These preparations also show that the
^ A more detailed description of these cells, with figures, will appear in the final publication (Folia Haematologica). »
240 A. R. RINGOEN
mast leucocytes are present in the marrow and that the staining reactions are very characteristic. It is not the object of the -WTiter to re-describe Maximow's results with this method.
In previous work on the mast leucocytes of the rabbit, Maximow has called particular attention to the fact that the granules of these cells are extremely soluble in water, and has cautioned against using watery fixatives and watery stains. The writer found that after fixation in Helly's mixture no mast granules could be detected with any of the various stains used. This would indicate that the mast granules are soluble in water. However, after alcohol and lucidol-acetone fixation, the granules are able to resist the short exposure to water to which they are subjected while being stained in the Giemsa solution. In the material fixed in Helly's mixture, however, the granules are exposed to the action of water for a long period of time which is sufficient to dissolve them. It is obvious that only those methods of technique which preserve the granules will be of real value in determining the origin of mast leucocytes, since the cells are difficult to recognize after their granules have been dissolved out. Little heed has been given to Maximow's repeated warning as to the solubility of the granules in water, and in all probability this accounts for the fact that Benacchio and others have failed to find mast leucocytes in the marrow of the rabbit.
SIBEVIARY
The bone-marrow of the rabbit contains true mast myelocytes with basophilic granules in addition to the myelocytes of eosinophiles and special leucocytes whose granules are also basophilic. With ordinary methods all of the myelocytes with basophilic granules seem to belong to the latter two types of leucocytes, but after fixation in alcohol, or better in lucidolacetone, the granules of the true mast leucocytes are also preserved. Their distinctive characters are such that they can always be distinguished from the basophilic granules of the eosinophil and special myelocytes.
ORIGIN OF ]VL\ST LEUCOCYTES 241
The general life history of the mast leucocyte runs parallel to that of the other granular leucocytes of the bone-marrow. Their granules are differentiated gradually out of the basophihc cytoplasm of mononuclear cells. The granules are strongly basophilic from the moment of their first appearance and remain so throughout the life-history of the cell. As the number of granules increase the nucleus gradually changes shape, becoming distinctly polymorphous in the fully differentiated cell.
Fully differentiated mononuclear mast leucocytes are never found in the blood or marrow of the adult rabbit. These cells, therefore, do not show the relationships to lymphocytes of the circulation described by Prdscher and others, and they are never differentiated from the lymphocytes of the circulating blood in the normal animal.
When the proper methods of fixation have been used the mast leucocytes of the rabbit show no evidence whatever of degenerative changes. Then' granules are, therefore, not products of a mucoid degeneration of the spongioplasm of lymphocj'tes (Proscher, Pappenheim and others), but are formed by the progressive differentiation of the cytoplasm of mononuclear cells of the bone-marrow.
The haematogenous mast cells of the rabbit form a distinct and independent line of granulocytes which is in no way related to the eosinophil or special leucocytes excepting through the non-granular parent-cell of the bone-marrow.
LITERATURE CITED
Benacchio, G. 1911 Gibt es bei Meerschweinchen und Kaninchen Mastmyelocyten und stammen die basophilgekornten Blutmastzellen aus dem Knochemnark? Folia Haem., Archiv, Bd. 12. Downey, H. 1913 The development of the histogenous mast cells of adult guinea-pig and cat, and the structure of the histogenous mast cells of man. Folia Haem., Archiv, Bd. 12.
1914 Heteroplastic development of eosinophil leucocytes and haematogenous mast cells in bone marrow of guinea-pig. Anat. Rec, vol. 8, no. 2.
1914 The origin and development of eosinophil leucocytes and of haematogenous mast cells in the bone marrow of the adult guinea-pig. To be published shortly in the Folia Haem.
242 A. R. RINGOEN
Kardos, E. 1911 Uber die Entstehimg der Blutmastzellen aus dem Knochen mark. Folia Haem., Archiv Bd. 11, part 1. AIaximow, a. 1906 Uber die Zellformen des lockern Bindegewebes. Archiv,
f. mikr. Anat., Bd. 67.
1907 Experimentelle Untersuchungen zur postfotalen Histogenese
des myeloiden Gewebes. Beitr. z. path. Anat. und allg. Pathol., Bd. 41.
1910 Die embryonale Histogenese des Knochenmarks der Saugetiere.
Archiv f. mikr. Anat., Bd. 76.
1913 Untersuchungen liber Blut und Bindegewebe. Vi Uber Blutmastzellen. Archiv f. mikr. Anat., Bd. 83, Abt. 1. Pappexheim, a. 1899 Vergleichende Untersuchungen iiber die elementare
Zusammensetzung des roten Knochenmarks eineger Saugetiere. Virch.
Archiv, Bd. 157.
1904 Zusatz zu der Mitteilung von Proscher uber experimenteller
Leucocytosen. Folia Haem., Bd. 7.
1912 Zur Blutzellfarbung im Klinischen Bluttrockenpraparat und
zur histologischen Schnittpraparatfarbung der hamatopoetischen
Gewebe nach meinen Methoden. Folia Haem., Archiv, Bd. 13, 1.
Tail. Pappenheim, a. und St. Szecsi 1912 Hamoz3'tologische Beobachtung bei
experimenteller Saponinvergiftung der Kaninchen. Folia Haem., Bd.
13. Proscher, Fr. 1909 Experimentelle basophile Leukocytose beim Kaninchen.
Folia Haem., Bd. 7. St. Szecsi 1913 Lucidol, ein neues Fixiermittel. Deutsche Medizinische
Wochenschrift, no. 33. Weidenreich, F. 1908 Zur Kenntnis der Zellen mit basophilen Granulationen
im Blut und Bindegewebe. Folia Haem., Bd. 5.
1910 D.'e INIorphologie der Blutzellen und Ihre Beziehung zu Einander. Anat. Rec, vol. 4.
1911 Die Leucocyten und verwandte Zellformen. Weisbaden, J. F. Bergmann.
A NOTE ON THE COURSE AND DISTRIBUTION OF THE NERVUS TERMINALIS IN MAN
ROLLO E. .AIcCOTTER
From the Deparlment of Anatomy oj the University of Michigan
TWO FIGURES
Johnston ('13) was the first observer to determine the presence of the nervus terminahs in man. He first reported its occurrence in human embryos and later ('14) described the nerve for the adult. Brookover ('14), working independently, also observed the presence of this nerve in adult man. Apparently the material used by these authors permitted only of the examination of a portion of the intracranial course of this nerve. It is the purpose of the present paper to report observations on the intracranial course and nasal distribution of the nervus terminahs in man.
The observations about to be reported are based on gross dissections of prepared specimens of the heads of several human fetuses varying in age from ten weeks to the newborn. Two adult heads were examined. The ner\'Tis terminalis was identified in all the specimens. Drawings were made from the two most favorable dissections. Figures 1 and 2 represent such drawings. The former represents the medial sagittal dissection of the head of a six-months human fetus, the latter a similar dissection of a ten-weeks human fetus. For purposes of dissection the specimens were prepared as described by the writer ('12) in a previous communication.
The intracranial portion of the nervus terminalis, as shown in figure 1, appears on the surface of the brain in the region of the olfactory trigone and courses anteriorly over the medial surface of the olfactory tract and bulb and on to the lateral surface of the crista galli, to pass thi'ough foramina in the cribri 243
244 ROLLO E. McCOTTER
form plate well forward. In its course over the medial surface of the olfactory tract it will be seen that the nerve forms a compact bundle of nerve fibers. On the medial surface of the olfactory bulb, however, it breaks up into a close plexus of fibers intimately associated with the fila olfactoria. It forms
Fig. 1 Medial section of the head of a six-months human fetus with the nasal septum removed, showing the origin, course and distribution of the nervus terminalis. Cor. Col., corpus callosum; A.C., anterior commissure; N.T., nervus terminalis; O.B., olfactory bulb; O.N., olfactory nerves; V.N.N. , vomero-nasal nerves; V.N.O., vomero-nasal organ; Hy., hypophysis.
a loose plexus on the lateral surface of the crista galli imbedded in the layers of the dura mater. In this position the separated filaments of the nervus terminalis lie some distance dorsal to the cribriform plate of the ethmoid bone instead of lying directly on its upper surface as do the fila olfactoria. In the specimens examined the height to which the nerve attains on the lateral surface of the crista galli or the amount of arching upw^ard of
NERVUS TERMINALIS IN MAN
245
the filaments of the nervus terminahs in this region, depends apparently upon the degree of development of the crista galli. The distribution of the nervus terminalis to the nasal septal mucosa is similar to that described by Huber and Guild ('13) for the rabbit. Within the cranium filaments of the nervus terminalis join the olfactory and the vomero-nasal nerves and
Fig. 2 Medial section of the head of a 4.5 cm. human embr3^o, with the nasal septum removed to show the origin, course and distribution of the nervus terminalis. C.H., cerebral hemisphere; L.T., lamina terminalis; N.T., nervus terminalis; P.C., posterior commissure; O.B., olfactory bulb; S.P., soft palate; V.3, third ventricle; V.N.O., vomero-nasal organ.
apparently pass to the septal mucosa with them. The majority of the fibers, however, form a single strand and pass through the cribriform plate anterior to the exit of the vomeronasal nerves. Upon reaching the nasal cavity the nervus terminalis takes a path anterior to that of the vomero-nasal nerves, lying just posterior to the antero-superior border of the nasal septum. In figure 1 it is represented as breaking up into three main filaments which can be traced downward nearly to the level of the vomero-nasal organ. In the first part of its nasal course it is joined by a small filament from the mechal nasal branch of the anterior ethmoid ner\'e.
THE ANATOMICAL RECORD, VOL. 9, XO. .3
246 ROLLO E. McCOTTER
In figure 2 the long axis of the olfactory tract and bulb occupies a plane approaching the perpendicular instead of the horizontal, as is shown in figure 1. The nervus terminalis appears on the surface of the brain in relatively the same position as in figure 1 and passes directly downward to the cribriform area where it lies in close proximity to the vomero-nasal nerves. After sending a few strands to accompany the vomero-nasal nerves the larger portion of the nervus terminalis passes through the cribriform area and is distributed to the septal mucosa anterior to the path of the vomero-nasal nerves.
In conclusion it may be stated that on account of the relation of the nervus terminalis to the crista galli, where the latter is sufficiently developed to cause a stretching out, as it were, of the overlying dura mater with its contained nerve, the continuity is here usually lost in gross dissections and the fibers associated with the vomero-nasal and olfactor}' nerves alone remain to determine its distribution to the septal mucosa.
The distribution of the nervus terminalis in man as in the rabbit is mainly to the mucosa of the nasal septum anterior to the path of the vomero-nasal nerves. Their ultimate terminations could not be determined.
LITERATURE CITED
Brookover, Charles 1914 The nervus terminalis in adult man. .Jour.
Comp. Neur., vol. 24, p. 131. HuBER, G. Carl, and Guild, S. R. 191.3 Observations on the peripheral
distribution of the nervus terminalis in mammals. Anat. Rec, vol.
7, p. 253. Johnston, J. B. 1913 The nervous terminalis in reptiles and mammals.
Jour. Comp. Neur., vol. 23, p. 97.
1914 The nervus terminalis in man anrl mammals. Anat. Rec,
vol. 8, p. 185. McCoTTER, R. E. 1912 The connection of the vomero-nasal nerves with the
accessory olfactory bull) in the opossum and other mammals. Anat.
Rec, vol. C, p. 29'i.
ox WEBER'S :^1ETH0D OF RECONSTRUCTION AND ITS APPLICATION TO CURVED SURFACES
RICHARD E. Sa^IMON
Institute of Anatomij, University of Minnesota
FIVE FIGURES
In embiyological work it is sometimes a matter of importance to determine with accuracy areas of epithelial thickening which are not demonstrated satisfactorily by the ordinary methods of graphic or plastic reconstruction. Placodal thickenings of the skin ectoderm, thickened zones in the neural tube, and the thickened areas found in the early archenteron are examples of structures which are not well demonstrated by these customary methods. A method of reconstruction which brings out graphically the extent and comparative thickness of such areas was de\ised some years ago by A. Weber /and was used by him with much success in the study of the very early development of the great glands of the digestive tract. Weber first published an account of his method in 1902,' and again a shorter summary in connection with the final report on his work in the following year.- Apparently the method has not been employed elsewhere. I have found it of such interest in the study of the earlier stages of the pancreas and li^'er that it has seemed desirable to present it here with some modifications which may be found useful, particularly in its application to curved surfaces.
AVeber's method is based upon that of graphic reconstruction' from transverse sections. An outline (from either lateral or dorsal view, as desired) of the organ to be reconstructed is plotted
^ Une methode de reconstruction graphique d'opaisseurs et quelques-unes de ses applications a I'embryologie. Bibliog. Anat., T. 11.
- L'origine des glandes annexes de T intestine moyen chez les vertebres. Arch. d'Anat. Micr., T. 10.
■2A7
248 RICHARD E. SCAMMOX
out on transverse section lines in the customary manner. Instead, however, of completing the reconstruction by indicating the contour of the surface thus outlined (as is commonly done), the thickness of the wall which forms that surface is measiu'ed in the transverse sections, and the variations in this thickness are plotted out upon the reconstruction. The reconstruction will then bear a nmiiber of lines which mark the boundaries between areas of epithelium of different thicknesses. To finish the reconstruction, the areas thus outlined are filled in with different shades of a single color, the darker shades being used for the thicker areas of the epithelial wall. The finished reconstruction then will exhibit the outline of the structure and the variation in the thickness of the part of its wall which is shown in this particular view. It will give no conception of the surface modeling of this wall aside from what can be determined from the outline alone.
The picture obtained is much the same as would be secured were it possible to remove the wall of the structure, render it translucent and, magnifying it greatly, hold it before a bright light. The thicker portions of the wall would then appear to the observer as darker areas, as they do in the reconstruction.
Figure 1 is a graphic reconstruction of the left side of the archenteron of an embrj'o of Torpedo ocellata, 4.0 mm. in length (Xo. 765 of the Harvard Embryological Collection). The portion of the gut l.ying on either side of the anterior intestinal portal and included between the lines A and B has been reconstructed by Weber's method and is shown in fgure 4. The latter figure shows by means of its coloring a broad band of thickened epithelium extending dorso-^'ent rally across the gut at the level of the anterior intestinal portal. The lower and thickest part of this band represents the anlage of the gall bladder and liver, while the upper part includes the future pancreatic region. A thickened spur extends backward from this zone, and marks off the line along which the intestine will eventually separate from the yolk-stalk ventral to it.
The method of prepaiiition of such reconstructicms can best be explained in detail by following an example of the process.
ON WEBER S METHOD OF RECONSTRUCTION
249
For this purpose I will use the reconstruction shown in figure 4 which has just been described.
In preparing this reconstruction, drawings were made of each section of the portion of the gut involved, although fairly satisfactory results can be secured with drawings of every other section. These drawings should be made at a high magnification, preferably over 300 diameters. As the sections are drawn, the}' show of course the curvatures of the contour of the archenteron wall. It is desirable to eliminate these curvatures in the reconstruction and to present the gut wall as an approxi
Fig. 1 A graphic reconstruction (lateral view) of a portion of the archenteron of an embryo of Torpedo ocellata 4.0 mm. long (H.E.C. 765). X 30. The outline of the embryo is represented in broken line. The archenteron is drawn in solid line. The jiortion of the archenteron lying between the dotted vertical lines A and B is shown in a Weber's reconstruction, at higher magnification, in figure i. • ■
mately flat plane. If this is not done, areas which project sharph' from the general plane of the archenteron will be represented in the reconstruction as much narrower than they actually are in the specimen or, if small, may be lost entirely. To eliminate these curvatures, it is necessary to divide the outer margin of each section into a number of short cords or segments, each of which will be comparatively straight, and to lay off segments of an efiual length on the transverse section lines of the reconstruction. This can be done with a pair of small screw compasses. For work at a magnification of 300 diameters and over,
250 RICHARD E. SCAMMOX
segments of 1 cm. are small enough to eliminate most of the error from curvature. Figure 2 is of a section (No. 104) of the reconstruction under discussion. The short lines on the inner side of its margin represent the boundaries of these centimeter segments.
The thickness of the epithelium forming the gut wall is now to be measured in each drawing. For this purpose a unit of measurement must first be determined. Working with drawings made at a magnification of 300, it has been found that the 1.5 mm. is the smallest unit practicable for such a scale. By using this unit one is able to measure ^'ariations of less than 5 micra in the thickness of the gut wall, and errors of projection and ch"awing would probably render more refined measurements of little value. A scale of 1.5 mm. units is laid out upon a stiff card, or better, a piece of transparent celluloid. This scale or gauge is then passed over each section, care being taken to keep its graduated margin at right angles to the axis of each segment of the drawing and its zero point at the inner margin of the epithelium. This process is begun at the dorsal median line on each section and is carried laterally or ventrally from that point. At the first place where the thickness of the epithelium is found to correspond to a graduation point on the scale, a fine line is drawn out to the side of the section and the thickness noted at its end. The scale is carried along the section imtil a point is reached where the thickness of the epithelium corresponds with the next graduation (either above or below the former one) on the scale. Again a line is drawn out from the section at this point and the thickness noted. This process is continued until the entire side of the section has been gauged and marked off into segments in which the variation in thickness is not over 1.5 mm. on the drawing or 5 micra in the corresponding section. An example of a section thus gauged is shown in figure 2. As shown by the gauge lines, the epithelium immediately below the dorsal median Une is between 30 and 25 micra f6.5 mm. at X 300) in thickness. The thickness falls to 25 micra at the point indicated. This is followed by a broad zone which is less than 25, but over 20 micra thick. Ventral to this zone, the epithelium increases to
ox WEBER S METHOD OF RECONSTRUCTION
251
a thickness of between 35 and 40 micra, and then again decreases to less than 10 micra as it approaches the blastoderm. The arrows seen in the figure point towards the thinner edge of the segment and are used to avoid _confusion in mapping out the gauge lines on the reconstruction at a later time. It is impor
%H
Fig. 2 Drawing of a transverse section of the archenteron of a Torpedo embryo 4.0 mm. long (H.E.C. 765) showing the method of measuring sections for reconstruction by Weber's method. The short lines extending into the section mark the centimeter segments used in eliminating lateral curvature from the section. The longer lines extending out to the right from the section are the gauge lines. The figures at the end of the gauge lines indicate in micra the thickness of the epithelium at the points touched by them.
tant that in gauging the section drawing the scale be held at right angles to the long axis of that particular part of the wall which is being measured rather than at a similar angle to either the inner or outer surface of the epithelial strip. This holds particularly when gauging the thickness of an epithelial band which
252 RICHARD E. SCAMMON
is rapidh' changing in caliber, or when gauging a portion of a band which forms a sharp curve. Failure to observe this precaution causes noticeable error both in the gauge readings and in the position of the gauge points on the drawing. There will often be encountered considerable portions of the wall, the thickness of which corresponds exactly to the gauge unit or a multiple of it. My rule, made arbitrarily, has been to mark as the gauge point the first Ti.e., most dorsal) level at which such a zone is encountered.
A reconstruction outline is now made on section-lined paper in the usual manner, except that only the dorsal margin of the gut is mapped out. Using this margin line as a base, the transverse section lines are divided into centimeter segments corresponding to those made on the margins of the drawings of the sections. In practice it is convenient to mark the point separating each block of 5 cm. segments in a different color to aid in plotting. The ventral margin of the gut and the gauge points, which have been determined on the cross section drawings, are now plotted on the transverse lines of the reconstruction in the usual manner, except that instead of measuring the distance of each point from the dorsal margin of the section and transferring this measurement to the corres])onding section line as is customary, one measures the distance of the point from the nearest centimeter mark in each case. The ventral outline of the reconstruction and the gauge lines indicating the thickness of the epithelium are established by connecting all the points of the same order as is done in an ordinary graphic reconstruction. The reconstruction will now have the form seen in figure 3. There the base and transverse reconstruction lines are lightly drawn and the centimeter segment points are indicated as small dots on the latter. The outline of the reconstruction is indicated in heavy line. The ventral margin posterior to the anterior intestinal portal has been cut away arbitrarily in a straight line. The gaug(^ lines indicating the thickness of the epithelial wall of the organ are represented in heavy broken line. The vertical figures placed at the termination of the gauge lines indicate their values in micra.
ON WEBER S METHOD OF RECONSTRUCTION 253
90 s lOO
110
120
130
30 20 10
Fig. 3 Reconstruction plat of a Weber's reconstruction (lateral view) of the portion of the archenteron bounded by the lines A and B in figure 1. The light vertical lines represent the transverse sections. The outline of the reconstruct on is represented in heavy solid lines. The gauge lines bounding areas of epithelium of different thicknesses are represented in heavy broken lines. The dots on the transverse lines represent the boundaries of the centimeter segments described in the text and shown in figure 2. The vertical figures at the margins of the reconstruction give the value of the gauge lines in micra. The figure is reduced to one-half the size of the original reconstruction, which was made at a magnification of 350.
There remains but to color in the areas which are marked off by the gauge Imes. Weber did this with water-color and secured excellent results, as his figures show, but a much simpler and quite as satisfactory a method is to use papers of different shades of gray for the different areas. Such papers should be as near 'pure' mixtures of black and white as it is possible to secure. The only kind which I have found satisfactory is the 'Herring' series which has been prepared for color work in psychological
254 RICHARD E, SCAMMOX
laboratories. The shades numbered 3, 4, 7, 10, 16, 20, 30, 42 and 46 make a satisfactory series of marked but fairl}- equal gradations.
To build up the final colored reconstruction gray papers are selected, equal in number to the areas mapped off on the reconstruction by the gauge lines. The entire outline of the reconstruction is now traced upon the lightest colored paper. This area is cut out and pasted firmly upon a piece of compo or plaster board. Upon the paper of the next darker shade there is traced an outline similar to the preceding except that the space representing the thinnest area is cut away. This second sheet is pasted upon the first one so that the similar angles and sides correspond. In this way the area of thinnest epithelium is represented by the lighter colored paper and all thicker areas by the darker one. This process is continued, all thinner areas being cut aw^ay from each new outline until the darkest and final shade of paper will have the shape and will represent the area of the thickest epithelium only. This method of building up the papers of different colors in strata will l)e found much easier than to cut out each separately and then attempt to fit them together in a mosaic. Figure 4 is a half-tone made directly from such a colored reconstruction and based upon the plotting illustrated in figure 3.
The example just described is of a lateral view reconstruction. Reconstructions may be made by Weber's method to show dorsal or ventral views of epithelial structures as well. For this purpose the transverse sections of the structure are drawn and measured in the same manner as that described. A reconstruction i:)lat is then laid out as follows. A vertical line is drawn in the middle of the sheet to represent the dorsal median line of the structure and transverse section lines are di'awn on either side at the proper distance apart and at right angles to the median vertical one. The centimeter points, which on lateral view reconstructions must be measured off on each transverse section line, can be located on the plat in this case by simply ruling vertical lines parallel to and at centimeter intervals from the median one. The gauge i)oints and lines mai'king the boundaries
ox WEBER S METHOD OF RECONSTRUCTION
255
Fig. 4 A finished \yeber's reconstruction (lateral view) made from the plat represented in figure 3 and including that part of the archenteron lying between the lines A and B in figure 1. The thicknesses of the epithelium in several parts of the reconstruction are represented bj^ the various shades of grav. Starting with the lightest, these several shades represent the following epithelial thicknesses; (1) below IOjjl; (2) 10 to 15^; (3) 15 to 20^; (4) 20 to 2om; (5) 25 to 30^; (6) 30 to 35m; (7) 35 to 40m ; (8) above 40m. This figure is reduced to tlueo-fifths the size of the original reconstruction.
between areas of different epithelial thickness are mapped out as in the lateral-view reconstruction. Should the reconstruction be of a tubular structure, the ^'entral median line must be determined in each cross section drawing. In reconstructing, the tube is then represented as split along its ventral median line and flattened out lateralh^; i.e., the lateral margins of the reconstruction represent in fact the ventral median line of the original tube. Figm-e 5 is an example of such a reconstruction plat made from the same object employed for the lateral view reconstruction already described. The method of representation and lettering are the same as used for figure 3.
256 RICHARD E. SCAMMON
Several of the possible uses and advantages of this method of reconstruction have been mentioned. It remains to speak of a few disadvantages and sources of error. In the first place the outlines of the reconstructions, aside from the one used as a base line, are not strictly such as would be secured by the ordinary graphic or plastic methods. In eliminating the lateral curvatures from the transverse sections, the dimensions of the figure are increased dorso-ventrally or laterallj^, as the case may be, without a similar increase antero-posteriorly. Weber made no attempt to eliminate the curvature seen in cross sections, regarding it in fact as of some value in indicating contour. I have already pointed out the disadvantage in reconstructing without making this correction, which I think should be done even at the expense of some accuracy of outline, which can easily be determined by the other reconstruction methods.
While the error introduced by this method is small when dealing with structures having surfaces approximating those of cones or cjdinders, it is considerable when applied to spherical surfaces. Spherical surfaces do not admit of being spread out into planes as do those of cones and c^dinders. Cartographers, who have much this same problem in representing large areas of the globe, have developed a number of methods of projection to meet, in part, this difficulty. These are, however, too complex for our work and are all based upon representations of the perfect sj^here. For the purpose at hand surfaces which approximate spherical ones can best be treated by first compensating for the vertical curvature by the method of segmenting the outline of the section already described; and, second, by allowing for longitudinal or horizontal curvature by increasing the distance separating the cross section lines on the reconstruction plat. T\w latter can be done by determining the actual length of the outline of the structure to be reconstructed and dividing this distance by the number of sections which the structure contains. Multiplying the figure thus secured by the magnification at which the reconstruction outlines are drawn gives one the distance which should separate the section lines on the reconstruction plat. This j)ractice differs from that of ordinary reconstruction in
ON WEBER S METHOD OF RECONSTRUCTION
257
V
\
'^
■^
^
\\
V
V
\
\
\ \
\
1»
...
\
.^
1
\
1
1
..
\
1
i
V
k^
\
■•*
/
1
I
1
1
V
1
V
/
-^
/
\ (
1
h
\
1
/'
«
1 1 1
» 1
%
^,
/
1 1
x
' 1
»
\ 1 1
1
i
f
•'
1 1 1
■>
N
1
1 1
/ 1
1
1
\
1
_ JL
t
t 1
1 1
1 1
1
^
1
1
ij
^
1
1 1
1
1
^
^
.,
-.
/
1
1
1
/
/
\
\
N
.
„
'
I
,o
/
i
^'
•
^■
-.
\
"*
""
"
-'
,
'C
,' 1
'
/
\
1
^,
^
^
'
r
1
'
■'
.
.'
1
'
1
I 1
\
-■
/'
'
-'
■
/■
V, /
/
1
1
/
<
,.
•'
> '■■■
». ./'
\
'■\
V.
s
-'
>.
,.
,'
L
—
"•
>•
-'
-'
-S
^
^ :^
,
/'
-N
/
X
>x
,'
c:
«.
..
'•
~
-,
-.
»
S'"-^
».
"'■
/ 1
'\
I
1
~
St
•\
»
"^
'
/
■'
^
' N
I
\\
\
N,
\.
i
'\
^.
\
•'
"n
\
\\
V
\,
^,
^
V
y
-S
1 '
/
^
-.
^
,
-lO
\
\
\
1
1
""
'
/
/
/
/
/
N
\
1
/
,'
V
^<*
'%
,'
6
-V.'
X,
\
1
1 1
4
'
V^
'
S.
^
\
\
//
I
^
' V
V
N
1 1
1
\ \
.
/
1
1 1 f
■*■»
\
1 1 1
1
1 1
1 1
/
1 >l
/
»
V
1
\
1)
li
'1
f
>
1
1
n
1
/
1
/
\
1
/ /
/
/
/
1
1 1
r
^*
'/ /
1
,
1
f
i /
/
>, /
,'
.'
1
/
J.
1-,
_
_
o y Jl_
CO
-C
o
■v>
71
a;
43
—
O
ec
G
c
s
03
^
0/
-►^
->^
f=
i-O
^
=0
^
03
o
— ^
^
"1
^
,,
C3
^
y3
258 RICHARD E. SCAMMOX
that the longitudinal axis of the structure is used in the customary method instead of the actual length of the outline as in this case. Reconstructions made with these corrections will show approximately the area and shape of any gi\-en outline upon a curved surface, although neither will be strictly accurate. As a rule such correction will be unnecessary unless one is dealing with surfaces which curve very abruptly. INIeasurements of the thickness of epithelial plates which curve longitudinally or horizontally will always be a little exaggerated because such plates are cut somewhat obliquely by transverse sections. There seems to be no practical way of eliminating this error.
Finally, the reconstruction will represent the variations in thickness of the epithelium as occurring in definite steps and not as gradual transitions as in nature. Alost of this artificial distinction can be ehminated by using the smallest unit of measurement practicable and thus increasing the number of shades present in the reconstruction while decreasing the degi'ee of their difference. Great reduction of figures in their reproduction also aids in securing the effect of gradual transition.
ox THE STRUCTURE OF THE ERYTHROCYTE
CHARLES D. CUPP
From the Depnrlmeni of Anatomy, Tulane University of Louisiana
FOUR FIGURES
This paper was undertaken with the hope of contributing something as to the structure of the so-called stroma of erythroc3'tes and as to the presence and character of a capsule or cell-membrane belonging to them.
The fact that the mammalian red blood corpuscle upon losing its nucleus becomes biconcave, its peripheral ring remaining thicker than its central region from which the nucleus has been lost, has always suggested the existence of a supporting framework. Were the corpuscle merely a hemoglobin-carrying sac borne in the blood stream, the natural tendency would be for it to assume a spherical form, the pressure on all sides being equal. The mammalian erythrocyte after being carried through capillaries smaller than its diameter, and after being drawn around angular cur\'es of capillaries, which may stretch and modify its form considerably, always resumes its original form when returned to larger capillaries. This not only suggests a supporting frame-work for its content but also that this framework is plastic or even possesses certain elasticity.
The question of a membrane about the corpuscle is most disputed in the discussions. The most frequently advanced theory against the existence of such is that the corpuscle possesses no membrane at all but is merely co^'ered by a film of lecithin or other lipoid substance acquired from its environment, and Xorris (Physiology and pathology of blood '82) even suggests evidence that fluid droplets enclosed by lipoids (myelin) tend to assume flattened shapes, while when enclosed by films of ordinary fats they are invariably spherical. The very probable presence of a film of some lipoid or fat surrounding the entire corpuscle is
259
260 CHARLES D. CUPP
seldom disputed and is accepted here. This is indicated by and explains the phenomena of forming into rouleaux and is supported by other arguments as well. But the presence of such a film need not disprove the existence of a membrane or capsule as a structure belonging to the corpuscle itself and which may be covered without b}^ a film of lipoid substance. Fluid droplets enclosed in lipoid or fat films when shaken will become divided into smaller droplets and droplets of greatly var^'ing size, whereas the erythrocytes of normal blood are strikingly uniform in size and while shaking them may break up or burst many of them, fragments resulting do not assume the form of the original corpuscle. Further, it seems very improbable that a film of lipoids alone would result in the folded and crumpled appearances shown by the so-called 'crenated corpuscles' resulting from extraction of a portion of the content. On the contrary, it would seem that such a film, still in the blood plasma, would merely thicken instead of becoming folded. Also, crenated corpuscles have lost the biconcave form, do not have the flattened shape suggested by Xorris for fluid droplets surrounded by a film of lecithin. Crenation is suggested here as indicating the presence of a membrane intrinsic to the corpuscle which has become folded due to partial extraction of the original content by exosmosis, the crenated corpuscle becoming approximately spherical, due to a disturbance or destruction of its original internal structure by the process. The chemical compound, hemoglobin, so far as is known, has no anatomical structure. It is a complex organic compound in solution capable of various degrees of dilution. The loss of the biconcave form in crenation and the assumption of the spherical form in swelling from the action of distilled water, for example, suggest the destruction of a framework in wliich the hemoglobin was supported and })y wliich the biconcave form of the nonnucleated corpuscle was maintained.
The original cell, the typical cell, possesses a framework of spongioplasmic filaments, a cytoplasmic and karyoi)lasmic reticulum, in the meshes of which are the more fluid portions (hyaloplasm) and the various forms of granules comprising the structure and content of the cell. In the functional differentiation
STRUCTURE OF THE ERYTHROCYTE 261
of certain cells a marked increase in the evidence of a cytoplasmic reticulum is well known. The neuro-fibrillae of the nerve cell, for example, consist of anastomosing filaments more abundant and more evident than in the germinal or neuroblast stage of this element. Certain functioning gland cells show a reticulum in their cytoplasm. Hardest}' COo) and Nemiloff ('10) have shown that the emulsion comprising the medullary sheath of the nerve fiber has its component globules suspended and supi:)orted by a delicate reticulum, Hardesty showing this reticulum to be continuous into the membrane or neurilemma without and into a thinner bounding membrane (axolemma) about the axone of the ner\'e fiber, and that this framework and these two membranes, and not the fat of the sheath, maintain the shape of the sheath. He considered the membranes of the sheath as condensations of the internal framework. It is very probable that the corpuscle of adipose tissue, the fat cell, possesses an internal framework continuous with its capsule, both maintaining its general shape. Schafer (vol. 2, part 1, Quain's Anatomy, 11th edition, '12) cites the fact that an erythrocyte, that of the newt for example, may be cut into two without resulting exudation of its content. Like the medullary sheath, the muscle fiber, the nerve cell or the fat corpuscle, the erythrocyte is an example of an especially differentiated arrangement for the performance of a special function.
The findings in the literature dealing with the structure of the red blood corpuscles seem to vary according to the methods of preparation used by the various authors and according to the different interpretations of the results obtained by them. Of the very voluminous literature, many of the papers consulted indicate a lack of familiarity with histology in general and an incompleteness of investigation to an extent that they seem of no value here. Howe\'er, some authors give definite statements as to what they deem the structure of the erj'throcyte and a number have considered the subject thoroughh'.
Renant in his Histology ('89-'93) says red blood corpuscles possess no true cell-membrane, but merely a peripheral condensation of the cytoplasm such as that formed about a cell when
THE AXATOMICAL RECORD, VOL. '.), XO. 3
262 CHARLES D. CUPP
exposed to air. Von Ebner (in Kolliker's Handbook, '02) states that a membrane must be assumed on the surface of the erythrocyte, a membrane insoluble in water and which allows osmosis, but which cannot be called a membrane in the ordinary sense and may be similar to the exoplasm of other cells. Lohner ('07) believes that a true histological membrane, similar to a cell membrane, is improbable. His method of preparation of the corpuscles for study consisted in crushing mammalian corpuscles on the slide. He also tried drying them in indifferent substances. He describes them as jelly-like and elastic in character with a thinner, more compact peripheral layer and a broader less compact inner region, similar to the exoplasm and endoplasm of Protozoa. If the outer compact layer is referred to as exoplasm, he thinks the superficial part of this may be termed a physical membrane or plasma-film. Dehler ('95) studied the red corpuscles of the chick, fixing them in sublimate solution and staining with ironhematoxylin. He described the convex cells as possessing a sharp border or rind of .25 to .5 yu in thickness. Heidenhain ('96), working with the red blood corpuscles of Proteus, using the same method as Dehler, found the same appearances. Nicholas ('96), likewise using the same procedure, found the sharp border manifest on the erythrocytes of the chick, salamander, triton and viper.
Meves f'04 and '06) studied amphibian red blood corpuscles (frog and salamander). After condemning the technique used by Weidenreich and showing the results obtained by the latter were artifacts, he describes a very complicated procedure used by himself. In general this consisted of dried smears on the slide subjected to warmth for 30 minutes, then subjected to Flemming's fluid (weak formula) plus 1 per cent sodium chloride, washed in running water, stained with safranin, hematoxylin and safranin and Gentian orange, decolorized with neutral alcohol, and cleared and differentiated in clove oil. He describes the corpuscles as showing circumferential fibrils, which were disposed either in an arrangement parallel to the surface or in the foi'iu of a continuous skein, and radial fibrils running, some from the periphery toward the center of the corpuscle and some crossing each other, form
STRUCTURE OF THE ERYTHROCYTE 263
ing a net. In some cases the net-work arose only in part from the peripheral arrangement. He refers to the circumferential fibrils as a membrane which is continuous with the fibrils of the network within, and he interprets the fibrils as resulting from linear arrangements of mitochondria, granules having fused to form them. He thinks the arrangement of the fibrils is similar to that described by Heidenhain for Krause's membrane in muscle fibers, the fibrils continuing into the membrane (sarco-lemma) in a regular system. Shafer ('05) suggests that Meves' circumferential or peripheral fibrils merely represent a part of the reticulum of the corpuscle.
Ruzicka ('03 and '00 j worked with frog, guinea-pig and human blood. He stained with a dilute solution of methylen blue both without and after the action of 1 per cent pyrogallic acid, which latter he thought dissolved hemoglobin. The drop of blood was mounted in normal salt solution and the methylen blue solution (0.5 gram to 1000 cc. of water) added at the edge of the coverglass. When used, the pyrogallic acid was applied followed by the methylen blue solution. Blood from the three sources gave the same results. All showed a fine meshed reticulum with occasional knobs at the junction of its filaments. He did not think an actual membrane is present but that the corpuscle is bounded by the reticulum. The knobs at the junction of the fibrils of the reticulum being found smaller and fewer after the action of pyrogallic acid, he assumed them to represent hemoglobin. He denied the possibility that the reticulum represented a coagulation product, thinking the meshes too fine and uniform and not arranged as coagulum filaments.
In the mammalian erythrocytes, Ruzicka also observed the previously described large granules dispersed in the center of the corpuscle, the so-called "nucleus of the mammalian corpuscle." Quoting Lowit ('87) as claiming a character for these granules in the rabbit similar to nuclear chromatin, he claims they are only present in case of incompletely dissolved hemoglobin and are thus analagous to the larger of the knobs described at the junctions of the filaments of the reticulum and therefore are not of chromatin nature.
264
CHARLES D. CUPP
Brvce ('04) studied tlie erythrocytes of the larvae of Lepidosiren paradoxa. He fixed the tissues with the blood vessels containing the erythrocytes in dtu and used sections of 10 m- He Zd 'ublimate-acetic as a fixing fluid. This was tned in th.s laboratory but was found unsuited for the study of adult erythrocytes in that it precipitates hemoglobin. Bryce's dlustrataons show that at least in the erythrocytes of these larvae, not having acquired sufficient hemoglobin to color them there is a reticular structure in the cytoplasm connecting with a membrane at the periphery, and he describes a meshwork or reticulum in them which was radially arranged from the nucleus to the periphery the meshes of which in section were 3 to 4 m m size. At the nocial points or junctions of the filaments forming the nieshe. he found "strongly refractile granules of considerable size and he states that in some of the corpuscles the filaments near the nuc eus appear as arranged in parallel threads, extending from the .mcleus'a short distance toward the periphery He doe. no pa judgment on the nature of the membrane observed, but states that he was dealing with very young corpuscles.
\L\TEKI.\I. AND MKTHODS
\s is known the Amphiumae carry the largest red blood coipu^e of anv\-ertebrate as yet examined. One of the spec.es th^ animai, the Amphiuma means (the ■blind eel (being easi Iv obtainable in the ditches of New Orleans, a study o he sti~ of its corpuscle was suggested by Professor Hardesty. .. easure ments of its corpuscles in the fresh gave, measured "" 'h ^a^'^ .,n average of 72.9 m in length by 44..5 m m width. \\ ith the "me chnique as f^rally employed for those of the Amphiuma erythrocytes from the alligator, frog, snake, guinea-pig an.l human were also prepared and studied in coinparison.
Obviously, a study of the internal structure ot the e > lu" cytes of Amphiuma and those of other animals, or the study of membranes probably existing, could not be a«-omplished in a,^ detail except with very thin stained sections ;->";.«' framework existed and if a membrane p,.ssesscd v-'"'"; * Ji to observe such, it was equally obvious that the hemoglobm ..u
STRUCTURE OF THE ERYTHROCYTE 265
ried must be wholly, or at least partially remoA'ed from the specimens. Therefore any accomplishment of the purpose in mind depended largelj' upon the technique employed.
The greatest difficulty was encountered in finding a fixing fluid suitable for the purpose. A fluid was desired in which hemoglobin is dissolved rather than precipitated. Hemoglobin not removed or precipitated within the corpuscle of course obscures whate\'er other cytoplasmic structure it may possess. Further, a fixing fluid was necessary whose osmotic action results in neither appreciable shrinkage nor swelhng of the corpuscles, and one the diffusion currents resulting from whose action does not break up the structural content.
Osmic acid, bichloride of mercury, chromic acid and its potassium salts precipitate hemoglobin, and fluids in which an}- of these act of themselves were found impossible for the results desired. It is doubtful whether, according to Ruzicka ('03), pyrogallic acid is a solvent of hemoglobin. It was deemed necessary here to fix, embed and section the corpuscles, which he did not do, and no suitable fixing fluid containing pyrogallic acid has been devised. Hemoglobin is dissolved in alcohol, distilled water, acetic acid and formic acid, and formalin does not precipitate it. The action of alcohol alone distorts the corpuscle and produces shrinkage and rupture, and acetic acid, formic acid and formalin not only produce swelling but in themselves are very poor fixing agents for the structure of cells. Van Gehuchten's (Carnoy's) fluid, containing absolute alcohol, acetic acid and chloroform, was tried and found to rend the fresh corpuscles into small fragments.
Bryce obtained his results after the action of sublimate-acetic, but he was dealing with corpuscles of larvae which probably contained considerably less hemoglobin than the corpuscles of the adults whose study was here desired. After trying a number of fluids containing corrosive sublimate and acetic acid, it was decided that the action of the sublimate in all precipitated the hemoglobin. Suggestive but incomplete results w^ere obtained with corpuscles fixed in a mixture containing 5 cc. saturated aqueous solution of bichloride of mercury, 2 cc. glacial acetic acid, cc. 40 per cent formaldehyde, and 88 cc. 95 per cent alco
266 CHARLES D. CUPP
hol. Preparations of nucleated corpuscles after this fluid showed a cytoplasm more or less transparent in places and a distinct membrane bounding the periphery. In the clearer places in the c3"toplasm a fairly well marked reticulum could be discerned. In preparations of mammalian (non-nucleated) corpuscles such were much less indicated.
Of all the fixing fluids tried, the most satisfactory results were obtained after using a well ripened mixture containing the following parts :
Aqueous 3 per cent potassium bichromate 100 cc.
Commercial (40 per cent) formaldehyde 4 cc.
■ Glacial acetic acid 5 cc.
When first made, this fluid has the color of the bichromate solution, but if allowed to stand or if warmed it becomes a dark greenish brown, due chiefly to the oxidation of the bichromate in the resultant reactions. Our best results were, or maybe happened to be, obtained with a mixture which had been standing several weeks. The formation of formic acid is one of the results of the ripening process.
In using this fluid fand all those tried) a fairly large shell-vial was filled about half full of the fluid and, gently shaking it, the blood was dripped, drop by drop, directly from the animal into the fluid. Even by agitating the mixture while adding the blood, some coagulated clumps cannot be avoided. The corpuscles in these clumps were found not so good for study as those floating free in the fixing fluid. All finally settle upon the bottom of the vial and the fluid may be removed with a pipette or decanted. The fluid was allowed to act for twelve hours. Changing it once or twice was thought to result in better extraction of the hemoglobin.
This fluid does not require a preliminary washing of the material in water. Small paper boxes were made from ordinary tliin letter j)aper, labels written on the sides in pencil, and the accumulated corpuscles transferred to the boxes by i:)i])ette. The boxes nearly full of fixing fluid and corpuscles were then placed in a stender dish containing 30 per cent alcohol to a depth of about
STRUCTURE OF THE ERYTHROCYTE 267
half the depth of the boxes. In 5 to 10 minutes the corpuscles settle to the bottom of the box and some of the fixing fluid may be pipetted awa}'. Gradual dehydration is accomplished by the transfusion of the alcohol through the paper walls of the boxes. If several boxes are carried in a small stender dish, the 30 per cent alcohol should be changed during the hour. Then the 30 per cent alcohol in the stender dish was replaced with 40 per cent alcohol and so on with grades of alcohol progressively increasing by 10 per cent in strength up to 90 per cent, which was replaced with 95 per cent alcohol and this in turn by absolute. To insure complete transfusion of the different grades and the action of each upon the corpuscles, the paper boxes should remain in each alcohol at least one hour with the stender dish covered.
From the absolute alcohol, the boxes were transferred to equal parts absolute alcohol and xylol for about 30 minutes and then placed in pure xjdol to complete clearing. The boxes were next placed in a dish of melted paraffin in the thermostat for two hours. Owing to the xylol present, it was found necessary to change the paraffin or transfer the boxes to another dish of paraffin during the first hour. When first in melted paraffin, the corpuscles float about and show a tendency to adhere to the sides of the box, and to avoid much of this in the final the boxes should be gently shaken a few times during the first half-hour.
The specimens were embedded, paper box and all. The corpuscles w^ere settled in a layer at the bottom of the box, so, when the paper was pulled off, a paraffin block was obtained with them accumulated in one side.
For the sections, the ordinary Alinot rotarj' microtome was used, set at 1 ii. The large corpuscles especially were found to have settled for the most part on the flat, or with their widths parallel to the bottom of the boxes. To obtain sections cut in this plane, the paraffin block had to be arranged with its bottom surface parallel with the edge of the knife. The paraffin ribbon was of course much jmcked and crumi)led and, of course, few if any of the sections could have been of 1 ii in thickness, but setting the microtome at 1 ^ was thought to give thinnest sections
268 CHARLES D. CUPP
possible. The sections were straightened and fixed on the shde by the usual water-method, without using albumen fixative. After drying, the paraffin was removed with xylol, the slides transferred to absolute alcohol and then passed through the gradually descending grades of alcohol, 3 to 10 minutes in each, down to water.
Of the staining methods tried, including gentian violet with safranin, the best results were given b}- alizarin and toluidin blue. To distilled water was added enough of a saturated solution of the sulphalizarinate of soda in 70 per cent alcohol to make the water a straw yellow, and in this the sections were immersed for about 12 hours. The sections were then rinsed in distilled water and some slides were placed in a 0.5 per cent aqueous solution of toluidin blue to stain nuclear structures. Other slides, in order to study the region occu])ied by the nucleus in greater detail, were not stained with the toluidin blue at all. The stained sections were then rinsed with distilled water and dehydrated by passing through the gradually increasing strengths of alcohol, 3 to 5 minutes in each, up to absolute, cleared with xylol and mounted.
The action of the fixing fluid and the technique of embedding, etc., when carefully applied, seemed to have produced little change in the shape and size of the corpuscles of the Amphiuma, frog, alligator and snake. Measurements of the sections of those of the Amphiuma, judged truly sagittal by the shape and the position and size of the nuclei, gave an average of 69 m loi^g ^^.Y 38.6 }x wide as compared with the average 72.9 m long and 44.5 /i wide given b}' the fi-esh corpuscles. For the blood of all four of the animals mentioned, sections showing the evenly oval contours characteristic of the fresh corpuscles were frequent on the slides and no disturbances of interior arrangement seemed evident in these. P'ragments of corpuscles and fragments of sections of them were abundant, but most all such, from their form, were evidently })r()du('ed by crushing and breaking by the knife in cutting and by the crumpling of the paraffin ribbon. Fragments of .sections were often more favoi'able for the study desired than sections of whole corpuscles. In the sections containing the
STRUCTURE OF THE ERYTHROCYTE 269
inainmalian corpuscles ( giiinea-pio; and human), there showed much more evidence of distortion hoth as to contour and internal arrangement.
OBSERVATIONS
The nuclei of the nucleated corpuscles could be best observed as to their position, size, form, and arrangement of chromatin in corpuscles stained whole, after fixation, without embedding. The nuclei in the corpuscles of the Amphiuma and alligator especially appear to consist of a tangled coil of one or more coarse rods of chromatin supported in non-chromatin substances. The coiled and tangled chromatin rods in Amphiuma are very much larger than those in the alligator and, in Amphiuma especially, the peripheral loops of the rods produce a scalloped and lobulated contour of the nuclei quite evident in whole specimens. In the thin sections of stained nuclei, the chromatin rod or rods appear cut into short segments, as shown in figures 1 and 3. A delicate membrane about the nucleus was e\'ident in all the nucleated corpuscles examined, but could be seen only in the thin sections and best in those in which the nucleus was not stained ffigs. 2 and 3, B).
A membrane about the entire red corpuscle, after the technicjue here emplo^^ed, was distincth' present in the blood of all the animals used. In proportion to the size of the corpuscle, it appeared relatively thicker and more condensed as possessed by the mammalian corpuscles (fig. 4, human). In the thin sections of corpuscles of Am}:»hiuma. this membrane, actually thicker than that of smaller corpuscles, could be resolved under oil immersion ol> jective into an apparently parallel arrangement of very delicate threads. With the fragments of corpuscles, broken and torn by the knife in sectioning and frequently found in the preparations, the nature of the membrane could be better observed than with intact sections. Torn and broken membranes often appeared slightly frayed in the tearing and close study led to the conviction that in Amphiuma, at least, the membrane consists not of a condensation of concentrically arranged parallel threads, and certainlv not of concentric lamellae, Init instead, of a very deli
270
CHARLES D. CUPP
^^»
■}
S^ji^
Fig. 1 Drawings from very thin paraffin sections of erythrocytes of Ani])hiuma means, fixed in the bichromate-formalin-acetic acid mixture and stained with sodium sulphalizarinate and toluidin bhic, showing the capsule, reticulum, perinuclear membrane and the coiled rods of nuclear chromatin in stained section. A, corpuscle sectioned on the flat; B, corpuscle sectioned in profile.
Fig. 2 Erythrocyte of Amphiuma sectioned on the flat and slightly tangentially . Same technique as in figure 1 except that the toluidin blue was omitted. Given to show the full shape of the fixed corpuscle and, especially, the perinuclear membrane and reticulum within the nucleus.
STRUCTURE OF THE ERYTHROCYTE
271
<>■<■
^^-T: ^f^:
P'ig. 3 From sections of en^throcytes of the alligator (Alligator mississippiensis). Prepared w-ith same technique as figure 1 and drawn in same scale as figures 1 and 2 and to represent the same structures. A, section with nucleus stained by toluidin blue; B, piece of section from preparation in which tohiidin blue was omitted.
Fig. 4 From sections of human erythrocytes prepared with same technique as figure 1, except that toluidin blue was not applied to the sections used for the drawing. Drawn in somewhat larger scale than figures 1 to 3. A, two sections of erythrocytes considered as representing more nearly the normal condition of the reticulum, capsule and central knots, with hemoglobin for the most part removed. B, erythrocytes with knots more distributed. C, D and E, varying degrees of rupture of the reticulum, presumably produced by diffusion currents set up by the reagents.
979
CHARLES D. CUPP
cate reticulum so condensed or compressed that its meshes are much elongated and thus produce the impression of a parallel arrangement (fig. 1, A). Aleves, describing red blood corpuscles of frog and salamander as "showing circumferential fibrils arranged either parallel to the surface or in a continuous skein, must have obtained the same appearances.
A very distinct reticulimi is evident in the cytoplasmic areas of the sections and the threads of its meshes are continuous into the membrane. The meshes of this internal (cytoplasmic) reticulum, or stroma of the corpuscle, appear larger in the corpuscles of Amphiuma than in those of the alligator ffig. 3), frog and snake. Corpuscles of the frog and snake, not figured here, gave appearances practically identical with those given by the alligator. In the more transparent sections, those supposedly more free from hemoglobin, the fibrils of the reticulum, though very delicate and varying somewhat in size, appeared distinctly threadlike, and the meshes made by them were angular in form. That the threads of the peripheral meshes grade directly into the peI'ipheral membrane, which itself appears as a condensed reticulum, supports the conclusion that the meml^rane is nothing more than a peripheral condensation of the internal reticulum and that the membrane could better be called a 'capsule' of the corpuscle. In one of his papers, Meves referred to it as a 'feltwork membrane.' This, and likewise the statement of Ruzicka that the corpuscle j)()ssesses no actual membrane })ut is bounded by the limits of a fine-meshed reticulum, seem warranted.
In preparations of corpuscles, of Amphiuma especially, in which the hemoglobin was obviously not so completely removed, the component threads of the internal reticulum (stroma) appeared coarser and ga\'e the impression of being rod-like in form, and the meshes of the reticulum appeared o\al instead of angular. The oval form of the mesh was due in part to larger knobs or accumulations of substance at the points of junction ('nodal points') of the threads in forming the meshes. Ruzicka described these knobs in liis preparations of frog and mammalian blood as round in form and concluded they represented undissolved hemoglobin. Their larger size in certain preparations here and the
STRUCTURE OF THE ERYTHROCYTE 273
apparent!}^ larger size of the threads of the reticukim accompanying them is interpreted as cUie to imremoved hemoglobin adhering to or precipitated upon the reticulum throughout, for such preparations were always less transparent and coarser in appearance and the membrane or capsule of the corpuscle appeared dark and homogeneous as compared with those from which the hemoglobin was considered more completel}' removed.
All the figures here given are attempts to represent corpuscles considered as having been rendered most free from hemoglobin. The knobs, or junctions of the filaments making the reticulum, appear quite small in these, usually about as large as would be possible were two or more plastic filaments to cross each other in contact and fuse giving an increased amount of substance at their junctions, a knob, or knot of the mesh of the net. In the most clear of the preparations, occasional larger knobs occurred at the junctions of the threads. Several such are shown in figure 2. These larger knobs were usually angular or stellate with their points extending upon the threads joining in them and are interpreted as representing small masses of unremoved hemoglobin adhering upon junction points of a greater than usual number of threads.
In the corpuscles of the Ami^hiuma, alligator and frog, the threads which join or grade into the capsule at the periphery of the reticulum appear for the most part to join with the capsule at right angles and thus present here a somewhat radial arrangement. The meshes formed by these threads in joining the capsule average somewhat larger in size than elsewhere in the cytoplasm of the nucleated corpuscles. The greater amount of light admitted by these larger meshes compared with the greater density of the capsule gave the impression of a narrow, clear zone about the corpuscle just under the capsule. This was especially true with the corpuscles of Amphiuma (figs. 1 and 2) in which the meshes are larger throughout than in the other nucleated corpuscles examined. Aleves obser^'ed this radial arrangement of the peripheral filaments of the reticulum in the corpuscles of the frog and salamander, and thought them in regular system similar to the relation of Krause's membranes in the muscle fiber.
274 CHARLEvS D. CUPP
Bryce likewise observed it in the corpuscles of lepidosiren larvae, but he considered it a part of a general radial or parallel arrangement of the filaments extending throughout fiom the nucleus to the periphery of the corpuscle. Our preparations do not show these peripheral filaments to join the capsule in a definitely regular system but at various angles and to form meshes of varying size and shape; nor do the threads of the reticulum extend from the nucleus to the capsule in definitely radial, and certainly not in parallel, arrangements. Only in the thinnest region of the cytoplasm, in the middle of the corpuscle on the flat where the nucleus is nearest the capsule, can threads be traced directly from the nucleus to the capsule ffig. 1, B, section in profile). Here the peripheral zone comprises practically the entu'e cytoplasm. As Schafer f'05) suggests, these peripheral threads are only a part of the general reticulum of the corpuscle. Continuous into and stretched from the capsule or membrane, they happen to appear more sparse and regular than the threads of the remainder of the reticulum.
For the larva of Lepidosiren paradoxa, Bryce described in each end of j^rofile views of the corpuscles a small area free of reticulum but occupied by a number of fine dots, and he interpreted these dots as transverse sections of cu'cumferential filaments running m the plane of the flat dimension. Our preparations showed no differences in this respect betw^een sections cut on the flat and profile sections Tfig. 1).
As noted above, the chromatin in the nuclei of the corpuscles of the Amphiuma and alligator appears collected into definite coiled rods instead of being scattered in granules of varying size throughout the nucleus. Preparations in which the nuclei are stained differentially do not show a sharp, dark-staining membrane bounding the confines of the nucleus as is found in certain other tissue cells where such is probably due to chromatin material being invoh'ed in or adhering upon the nuclear membrane. On the contraiy. ilic miclear membrane appears here to be another but thinner condensation of the reticulum or general framework, being comparable in origin and structure with the capsule, or membrane, about the corpuscle. The threads of the cyto
STRUCTURE OF THE ERYTHROCYTE 275
plasmic reticulum grade directly into it and it stains in the same way as the threads of the reticulum. It could best be studied in preparations to which a nuclear stain had not been applied (fig. 2 and fig. 3, B). Figure 2 represents a slightly tangential section of a corpuscle of Amphiuma on the flat. In such corpuscles, with nuclei not differentially stained, it could be noted that the threads of the reticulum grade directly into the nuclear membrane, that the meshes become suddenly smaller or more dense to produce it and that the membrane carries the same stain-reaction as the reticulum. In other words, it is here suggested that the nuclear membrane is but a peri-nuclear condensation of the general framework of the corpuscle. Sections of corpuscles in profile (fig. 1, B) show the threads to serv^e as continuations between the capsule and the nuclear membrane.
Furthermore, our preparations suggest that the general reticulum, or framework, is continuous into and throughout the nucleus, contributing to the support of its structures. Figures 1, A, and 2 and 3, B, are attempts to show this suggestion. In such corj)uscles, the very thin sections showed the dehcate filaments extending from the nuclear membrane and forming a reticulum throughout the nucleus. The meshes of this intranuclear framework seemed somewhat larger than the average of those in the cj^toplasm, and especially large when containing a segment of the coiled chromatin rod. The threads stained just as those of the cytoplasm and the membranes. Knobs at the junction points of the threads could not be observed as so definite nor so large as in the cytoplasm, probably due to some extent to the necessarily more obscured nuclear area. The suggestion that the cytoplasmic reticulum is continuous and identical with the nuclear reticulum is somewhat supported by the frequently presented view that the spongioplasmic reticulum of the general cell structure is identical to (stains the same) and is continuous with the karyoplasmic reticulum, or nuclear linin.
The chemical composition of either or both of the reticula, including the membranes, may be that of a nucleo-proteid or lecithin or cholestrin, but in our preparations it occurs in the form of a network in whose meshes is supported the remaining cell load,
276 CHARLES D. CUPP
including the hemoglobin and the nuclear structures. Schafer states that substances dissolving lecithin or cholestrin will produce an increase in the permeabiUty of the membrane. Bryce thinks that the peripheral capsule ("peripheral ring or band," he calls it) is due to a condensation or massing of the meshes of the reticulum, and he suggests that the filaments of the reticulum are not necessarily fixed fibers but that they may be of colloidal nature. Taking all into consideration, he thinks the reticulum is not an artefact but an actual protoplasmic framework. Citing Blitchli's "Foam theory" of the structure of protoplasm, and noting that the meshes of the reticulum are larger than the limits given for the protoplasmic alveoli of this theory, and much larger than the meshes described for the cytoplasm of leucocytes, Bryce thinks that if the protoplasm of the erythrocyte may at one time have been alveolar in structure, the reticulum could be later derived from a vacuolated condition in which the hyaloplasm of the cell is greatly reduced and the alveolar arrangement lost by the breaking through of the walls of the alveoli. Ruzicka thinks the observed reticulum is not artefact but an actual structure, that its meshes and the knobs at the junctions of the threads are not coagulation products, for the meshes are too fine and uniform and the threads are not arranged as are coagulum filaments; that the hemoglobin is carried dispersed in the meshes of the net to the periphery of the corpuscle, that the net is the vegetative part of the corpuscle and the hemoglobin the functional part.
Whatever the chemical character, our preparations seem to support the conclusion that the structures observed are not artefact, that the corpuscle is pervaded by a true reticulum, a network of threads joining each other throughout and extending m all the planes of space. That the thrc^ads of the reticulum observed serve as a supporting framework of the coriniscle and possess a certain amount of elasticity is suggested by several observations on the living red blood corpuscle: (1) the manmialian corpuscle, after losing its nucleus, remains thinner in its center from which the nucleus was lost; (2) a living corpuscle may be cut in halves and neither half suffer exudation of its content; (3) corpuscles in the circulation may l)e elongated and their usual
STUrCTlTRE OF THE ERYTHROCYTE 277
shape considerably distorted in passing- through the smallest capillaries but always resume theu- shape upon reaching the larger vessels; and (4), in this laboratory, by tapping under the microscope the cover-glass upon fresh mounts of corpuscles of the Amphiuma, the nuclei could be made to shift considerably from their normal position, moving back and forth with the tapping, sometimes being slightly' spread by the pressure, but, the pressure removed, they would resume their normal position and size in the center of the corpuscle floating in the plasma.
With human corpuscles, the technique here employed was not altogether as successful as with the nucleated forms. The stained sections showed more distorted contours and internally ruptured corpuscles than those of Amphiuma, alligator and frog blood prepared in the same way. In the latter preparations, evidences of rupture produced by the reagents used were somewhat rare. Figure 4 is given to represent appearances found in the preparations of human corpuscles. Blood from the guinea-pig gave the same appearances. The two corpuscles farthest in the left (fig. 4, A) represent the form most common in the preparations and, showing less distortion in contour, were the form considered as most nearly representing the normal. The remaining four corpuscles were selected as an attempt to illustrate, progressively, appearances of injurious effects produced by the reagents.
It may be noted that the capsule, the sharp peripheral border of the corpuscle frequently mentioned in the literature, is here relatively thicker and more densely staining than that of the Amphiuma. This may be due in part to a less complete extraction of the hemoglobin considered evident in the interior. The threads of the reticulum, while grading into the capsule and continuous with each other throughout the corpuscle, do not present so nearly uniform size as in the other forms studied. Filaments much thicker than others appear to radiate from the central region, while attached to these are numerous smaller threads joining throughout and completing a general reticulum. At the junction of the larger filaments with the capsule there usually appears a visible knob, or hillock of attachment. One gets the impression in close study that the smaller, probably the
THE ANATOMICAL RECORD, VOL. 9, NO. 3
278 CHARLES D. CUPP
normal, meshes are themselves crossed by still finer threads, while a larger mesh may appear perfectly clear. Fme knobs show at the junction or nodal points of the threads, varymg in size with the size of the threads.
In the corpuscles considered more nearly normal (fig. 4, A), the frequently described large knobs dispersed in the central region of the corpuscle comprised the most promment feature m our sections. These appeared approximately spherical m shape and uniform in size and the number observed varied from 1/ to 6 a variation due no doubt in part to the planes m which the corpuscles were sectioned. In sections in which the knobs were less darklv stained, one could get the impression, under highest magnification, that these knobs themselves are fibrillar, that they are knots or condensations of very fine threads. If they carry anv nuclear material (they have been called the nucleus of the mammalian ervthrocyte), our sections suggested it very doubtfully that anv stain reaction of a chromatin character isretamed in them Vniline nuclear dves. such as gentian ^■iolet, and even hematoxvlin mav be retained in them longer than in the threads of the reticulum, but the color will wash out. and the fact that it is retained in them more deeply at times is considered due (1) to their greater compactness of mass holding more of the dye than the smaller and looser structure, but indifferently nevertheless and (2) to unremoved hemoglobin being entangled within them Thev were alwavs darker than the filaments, but this is to be expected, due to their greater density obscuring the light Occasionallv the knobs or knots appeared more scattered throughout the corpuscle and then to vary more in size, as shown
in B of figure 4. i i i • i
The larger filaments, those usually described and wJiicli appear more or less radially arranged, are here considered as artefact ('orpuscles manifestlv ruptured internally by the action of the reagents (fig. 4, E) show large, clear, vacuole-like meshes and these are always bounded by the thick filaments. It is suggested that these large clear meshes represent areas of the reticulum ruptured bv diffusion currents of the fixing agent or alcohols in preparation, and that the broken threads of these
STRUCTURE OF THE ERYTHROCYTE 279
areas have been washed together or condensed to form the thicker filaments. Often the latter give the impression that they themselves are composed of finer threads. Thej^ occur more or less radially between the centrally placed knots and the capsule probably because these knots are the firmest fixation points.
As to the knobs or knots ('nuclei'; of the central region, we beg to suggest the possibility that they may represent remains of an originally existing membrane about the nucleus and of a reticulum within the nucleus of the erytlu'oblast, or nucleated stage of the mammalian corpuscle, similar to those found in the nucleated corpuscles of the Amphiuma and alligator; that the knots represent the remains of these resulting from or after the disturbance of the central reticulum at the time the nucleus disintegrated or was extruded, and that their densit}' may be added to by hemoglobin retained in them.
The membrane or capsule of the erythrocyte, derived as a peripheral condensation or massing of the reticulum, must be the most resistant part, the structure last to rupture under stress. It is permeable, as is well known, allowing a readj^ diffusion through it. The fact that the mammalian erythrocyte assumes a spherical form before bursting, when swollen and distended by excessive endosmosis in the action of distilled water or weak acetic acid, for example, may be explained as due to its less resistant internal framework being first torn asunder and destroyed by the diffusion currents and the stress. The erythrocyte then becomes a mere turgid sac, the capsule itself rupturing at continued pressure. Both the capsule and framework are dissolved by alkalies.
The red blood corpuscle is a tissue element extremely differentiated in structure for the performance of an extremely specialized function. In the mammal it is more differentiated than in the lower vertebrates, being one of the two bodies possessed that, containing no nucleus, can no longer be called a cell. From the above studies it is concluded (1) that it normally possesses a framework in the form of a fine threaded, somewhat elastic reticulum in the meshes of which its hemoglobin is supported so intimately that its content partakes of the physical characters of
230 CHARLES D. CUPP
gelatin- (2) that in the nucleated forms, this same reticulum is continuous into the nucleus, supporting its structures; (3) that it possesses a peripheral mpni])rane or capsule into which the threads of the reticulum grade and which is deri^-ed from and consists of a peripheral condensation or massing of the reticulum; (4) that there is a similar but thinner perinuclear condensation of the reticulum l^ounding the confines of the nucleus and forming a nuclear membrane of the nucleated forms, and (5) that the central knots, or nucleus of the mammalian corpuscle," are masses of the material of the nuclear membrane and reticulum originally existing in the central part and result from or after the distm-bance of the interior produced by the dismtegration or
extrusion of the nucleus. . ^ ^^ a
Finally is due an expression of appreciation of the kindness of Professor Hardesty at whose suggestion and with whose guidance and collaboration this study was made.
BIBLIOGRAPHY
Bryce, T. H. 1903-1905 Trans. Royal Soc. Edinburgh.
Dehler, a. 1895 Archiv fiir mik. Anat., Bd. 46.
FoA 1889 Ziegler's Beitrage, Bd. 5.
Hardesty, I. 1905 km. Jour. Anat., vol. 4 Heidenhaix, M. 1912 Quoted from Schafer m Quam's Anatomy, ^ol. _, pait 1.
Lohxer, L. 1907 Archiv fiir mik. Anat., Bd. 71.
LowiT, M. 1S87 Sitzungsbr. d. K. Akad. d. ^\ len, lid. 9o.
Meves, 1"k. 1904 Anat. Anz., Bd. 24.
1906 Anat. Anz., Bd. 28 Nemiloff, a. 1910 Archiv fiir mik. Anat., Bd. 76. Nicholas 1896 Bibliographia Anatomique. RuzicKA, V. 1903 Anat. Anz., Bd. 23.
•1906 Archiv fiir mik. Anat., Bd. 67. Schafer, E. A. 1905 Anat. Anz., Bd. 26.
x%
ox the provisional arraxgeaiext of the e:\ibryoxic lymphatic system
AN ARRANGEMENT BY MEANS OF A^TilCH A CENTRIPETAL LYMPH FLOW TOWARD THE VENOUS CIRCULATION IS CONTROLLED ANT) REGULATED IN AN ORDERLY AND UNIFORM MANNER, FROM THE TIME LYMPH BEGINS TO COLLECT IN THE INTERCELLULAR SPACES, UNTIL IT IS FORWARDED TO THE VENOUS CIRCULATION
CHARLES F. W. IMcCLURE
From the Department of Comparative Anatomy, Princeton University
SIX FIGURES
One of the most interesting problems of the lymphatic system is the determination of the manner in which the continuous centripetal lymph flow is established in the embryo, in relation to the developing lymphatic vessels by which it is subsequently conveyed to the venous circulation.
It is well known that the anlagen of the lymphatic system do not normally make their appearance in the embryo until after the haemal vessels have been established. As soon, however, as the haemal vessels begin to function, lymph begins to collect in the intercellular spaces of the embryo and, as we know, is subsequently collected by a set of newly-formed vessels, the lymphatics, which convex" it to the venous circulation.
Those who maintain that the lymphatics sprout centrifugally and continuoush" from the veins, would necessarily hold that the lymph in the intercellular spaces patienth' awaits the arrival of closed and hollow outgrowths from the veins, the lymphatics, before it can be received into any portion of the lymphatic system. Such continuous outgrowths from the veins would necessarily take place in a centrifugal direction which is opposed to that of the centripetal flow of lymph they would receive.
It has always proved a difficult matter for some of us to reconcile the view that the dii^ection of the growth of the vessels
281
THE ANATOMICAL RECORD, VOL. 9, NO. 4 APRIL, 1915
2g2 CHARLES F. W. McCLURE
in which the flow takes place should be opi)osed to that of the flow One might expect a continuous centripetal lymph flow toward the venous circulation to be estabUshed m a gradual manner in the embryo, and to be regulated from the tune b^nph first made its appearance in the intercellular spaces, until it was continued on to the venous circulation. In fact, one might expect to find some provisional condition of the embryonic lymphatic svstem which should exactly accord with the maintenance and regulation of such a centripetal flow. Such a provisional condition of the embryomc lymphatic system I believe^ 1 ham been able to demonstrate in a positive manner in the living
embryo of the trout. , , • .i •, i
One of the most salient features noticeable m the development of the lymphatic system of the trout, as well as m that ot mammals, is that the main lymphatic channels are formed through a oradual concrescence of independent and discontinuous anlaoen or lymph vesicles. These independent lymph vesicles mike then- appearance in the embryo in a progressive manner along the lines subsequently followed by continuous lymphatic vessels and. in certain districts of the mammalian embryo, they utilize the static fine vacated by degeneratmg veins so that certain Ivmphatic ^'essels of the body subsequently follow the course of abandoned veins. These independent lymph vesicle'^ of the embryo first become concrescent to form continuous channels contiguous to the points at which the lymphatics establish tvpical communications with the veins. AA ith these points of Ivmphatico-venous entry, the vesicles continue to become concrescent in a progressive manner, so that the outlying or peripheral lymph vesicles are the last to establish a communication with the vein^.
The view which calls for the development ol th(« mam lymphatic channels through a confluence of independent and discontinuous anlagen was advanced by Hunthigton and Mc( lure
h chauiu'ly
. Huntington HH.l M.Clure. The development „f the mam lymph of the cat 'n their relations to the venous syste.u. Anat Eee.. ol. 07
llcad before the An.eriean Assoeia.ion of An,....nnsts a, the -neetn,, hehl .n New York in Decemlx-r. 1906.
PJMKRYONIC LYMPHATIC SYSTEM 283
in 1906. Since a recognition of this fact is a necessary corollary to the main i jsue invoh'ed in the present paper, we will first consider what constitute the main lymphatic channels in the embryo of the trout and show how, in their development, they follow this ]5lan.
Figure 1 represents a ventral view of a reconstruction of the main hiiiphatics, veins, and arteries found in the regions of the head and pharynx of a rainbow trout embryo, on the twentysecond day after fertilization. This embryo was developed at a temperature of about 10. o"" C. and its Ijanphatic system is represented, for the most part, by a continuous system of vessels which drain into the veins at typical points. The typical points of lymphatico-venous communication in the regions of the head and pharynx occur in the cardino-Cuvierian district (9): with the precardinal (jugular) vem near the caudal end of the otocyst (IS) ; and with the precardinal, near a point where the latter leaves the cranial cavity (2). The first and last mentioned points of communication appear invariably' to be retained in the adult.
The principal or main lymphatic vessels found in the regions of the head and pharynx of a twenty-two day rainbow trout embryo are as follows:
1. The Huhocuhir lymph sacs (saccus lymphaiicus subocidaris,
1 in figure l)
The subocular hmiph sacs of the trout embryo consist of two relatively huge sacs or vesicles, each of which lies ventro-medial to the 63^6. They are more or less triangular in form, with their apices directed forward. In the twenty-two day trout under consideration, they extend between the hyoidean artery (Id) and the olfactory invagination (21). as shown in hgure 2. At its postero-lateral angle, each subocular lymph sac (1) communicates directh' with the lateral pharyngeal lymphatic (3 in fig. 1) and it is solely through the latter vessel that the subocular lymph sac drains into the veins.
284
CHARLES F. W. M( CLURE
Fig. 1 Reconstruction of tiie main lymphatics, arteries and veins found in the regions of the head and pharj^nx of a twenty-two-day rainbow trout embryo; ventral view. P. K. C. series 648. Reconstructed after the method of Horn at a magnification of 200 diameters.
EMBRYONIC LYMPHATIC SYSTEM
285
REFEREXCE NUMBERS
1, Subocular lymph sac 11,
2, Medial pharyngeal communication 12,
3, Lateral pharyngeal lymphatic L\
4, Medial pharyngeal lymphatic 14,
5, Precardinal (jugular) lymphatic 15,
6, Precardinal (jugular) vein 16,
7, Communication between the pre- 17,
cardinal lymphatic and the lat- 18,
eral pharyngeal lymphatic 19,
8, Caudal end of otocyst 20,
9, Cardino-Cuvierian conununication 21, 10, Lymphatic of the lateral line of the 22,
trunk
Postcardinal vein Duct of Cuvier Otic communication Dorsal aorta Hyoidean artery First efferent aortic arch iSecond efferent aortic arch Third efferent aortic arch Fourth efferent aortic arch Caudal end of eye Olfactory invagination Carotid artery
Fig. 2 Sagittal section taken through the head and pharynx of a twentytwo-day rainbow trout embryo, showing the relations of the subocular lymph sac to the eye, the hyoidean artery and the olfactory invagination; /, subocular lymph sac; 1-', hyoidean artery; 21, olfactory invagination. P. E. C. series 816.
286 CHARLES F. W. McCLURE
2. The lateral pharyngeal lymphatic (truncus lymphaticus pharyngeus lateralis, 3 in figure l)
This vessel occupies a superficial position in the lateral wall of the pharjTix and forms, on each side of the body, the dhect anterior continuation of the lymphatic of the lateral line of the trunk (truncus lymphaticus longitudinalis lateralis, 10 in fig. 1). The lymphatic of the lateral line of the trunk is completely de\'eloped in the twent3^-two day rainbow trout and can be followed can dad to the region of the caudal Ijaiiph heart. The lateral pharyngeal Ijanphatic drains the subocular lymph sac and, at a shghtly later stage of development, also the dorsal region of the head and pharynx, the operculum and the lower jaw. It communicates with the veins in the cardino-Cuvierian district (9) in common with the lymphatic of the lateral line of the trunk (W).
S. The medial pharyngeal lymphatic {truncus lymphaticus pharyngeus medialis, 4 in figure 1)
This vessel occupies a more central position and is more deeph" situated than the lateral pharyngeal lymphatic. It follows an oblique course, in a postero-anterior direction, from about the middle of the lateral pharyngeal lymphatic with which it subsequently comnumicates, to open into the precardinal vein just caudal to the ]:)oint where this vein emei'ges from the cranial cavity (2 in fig. 1).
j!f. The precardinal or jugular lymphatics {truncus lymphaticus precardinalis vel jugularis, 5 in figure 1)
These vessels tlevelop along I lie hue of the ])i-ecardhial vehis. The}' are not completely established on the twenty-second day in the form of continuous channels and are represented by r.flch vessels only near the caudal end of the ]:)harynx, where they communicate with the lateral ])haryngeal lymphatic (7 in fig. 1).
All of the continuous lym))hati(' channels described above, as occui'iing on the t wentv-second d.-iv in the cmbi-vo of the I'liin
EMBRYONIC LY:\rPHATIC SYSTEM 287
bow trout, can be readily injected from the veins, through any one of the typical points of communication which are estal)Ushed between the lymphatics and the veins. By injecting into the subocular lymph sacs it is also possible to fill the continuous lymphatic A'essels with inject a. as well as the veins. At this particular stage of development, in the absence of lymphatico-venous valves, blood may flow freely into the lymphatics from the veins, at the tj^pical points at which the hmiphatics communicate with the \'eins. In rainbow trout embrj'os slightly older than twenty-two days, and in which lymphatico-venous valves have been formed, the application of chloretone apparently vitiates the normal function of these valves. Blood may then also flow freely from the \'eins into the lymphatics and fill up completely all of the continuous lymphatic channels, including the subocular lymph sacs. If such embryos are remo\'ed to water in which no chloretone is present, the blood will flow back from the lymphatics into the \'eins.
The age of the rainbow, steelhead or brook ti-out embryo in which a continuous system of lymphatic channels, as shown in figm'e 1, is met with for the first time, naturally varies wdth the temperature of the water in which development takes place. Much variation is also met with in the rate at which the lymphatic system of the trout develops in different embryos of the same age, as well as upon opposite sides of the same embryo. When developed at a temperature of about 10. o° ('.. a continuous lymphatic system, as shown in figure 1, is usually found in rainbow and steelhead trout embryos on the twentj-second day after fertilization, which is six or seven days after its earliest anlagen can first be observed in sections.
In all stages of development, prior to the establishment of such a continuous system of h^mphatics as shown in figure 1, a condition is invariably met with in the embryo in which the IjTQphatic system is represented by a progressively appearmg series of discontinuous anlagen, that present varying degrees of concrescence with one another to form continuous lymphatic vessels which communicate with the veins, only at the typical points at which the h^mphatics establish communications with the
288
CHARLES F. W. McCLURE
Fig. 3 Reconstruction of the Ij'mphatics, arteries and veins found in the regions of the head and pharynx of a twentj'-day rainbow trout embryo; ventral view. P. E. C. series 668. Reconstructed after the method of Born at a magnification of 200 diameters; reconstruction of an injected embryo. For reference to numbers see under figure 1.
veins. Two siioh stages of dev^elopment are herewith presented ill illustration of this fact.
Figures 3 and 4 represent, respectively, reconstructions of the lymphatics, veins, and arteries found in the regions of the head and pharynx of a rainbow trout embryo on the twentieth and
EMBRYONIC LYMPHATIC SYSTEM
289
Fig. 4 Reconstruction of the lymphatics, arteries and veins found in the regions of the head and pharynx of a twenty-one-day rainbow trout embryo; ventral view. P. E. C. series 646. Reconstructed after the method of Born at a magnification of 200 diameters; reconstruction of an injected embryo. For reference to numbers see under figure 1.
twenty-first days after fertilization. These embryos were developed at a temperature of about 10.5°C. When compared with the conditions found in the embryo on the twenty-second day (fig. 1), it is seen that the subocular lymph sacs (1 in figs. 3 and 4) are entirely independent of the veins and of other independent h'mph vesicles, and that a series of discontinuous
290 CHARLES F. W. McCLURE
and independent lymph vesicles lie exacth' in the Ime subsequenth' followed by continuous lymphatic channels on the twenty-second day (fig. 1). It is also seen that these tymph vesicles present vaiying degrees of concrescence beginning at the points at which the lymphatics establish typical communications with the veins. It may also be observed that none of these independent lymph vesicles communicate with the veins except those which lie contiguous to the points at which the lymphatics establish typical communications with the veins.
For the purpose of the present paper reconstructions of the twenty and twentj-one day trout are sufficient to illustrate the general principle of development followed b}* the main lymphatic channels.
The main question in\'olved in the j^resent issue is, however, what functional significance does the presence in the embryo of such an mdependent and discontinuous series of lymph \'esicles miply? A study of one of these lymph vesicles, the subocular lymph sacs, as observed and experimented upon in the living trout embryo gives, I believe, a positive answer to this question.
When rainbow and steelhead trout embryos are de\'elo]ied at a temperature of about lO.o^C, the de^'elopment of the anlagen of the subocular lymph sacs is initiated on about the sixteenth day after fertilization. These anlagen first appear as small clear vesicles which lie in the mesenchyme in an area between the hyoidean artery and the eye and just dorsal to the maxillary ridge. These small vesicles finalh' become confluent to form a larger \esicle which gradually increases in size. Injection experiments \\\)(m the living trout embryo, controlled by sections, have shown that the anlagen of the subocular lymph sacs never establish a communication with the veins in the neighborhood of the sacs. They have also shown that the subocular lymph sacs cannot be injected from the veins at any stage of their development, until after the independent and discontinuous anlagen of the lateral ]:)haryngeal have become concrescent with one another to form a continuous vessel and until a comnumication has been established between this vessel and the subocular
EAIBKYONIC LYMPHATIC SYSTEM 291
Ij'inph sac (compare figures 3 and 4 with figure 1). In other words, at no stage of development has it been possible to inject the subocular lym])h sac of the trout embr^^o except by way of the lateral pharyngeal h'mphatic, which signifies that the subocular lymph sac of the trout embryo is entirely independent of the veins and of other Mnphatics, until this communication has been made.
Fig. .3 Photograph of the ventral aspect of the head of a twenty-day rainbow trout embryo on which an attempt was made to inject the lymphatics and the veins by injecting into the subocular lvmi)h sacs. Spalteholz prejiaration. 1, subocular lymph sacs.
Injection experhnents have also proved that the subocular lymph sacs of the trout embryo do not grow caudad and that they are invariably bounded posteriorly by the hyoidean artery; (compare 15, hyoidean artery, with 1, subocular lymph sac, in figures 2, 3 and 4) . Figure 5 is a photograph of a Spalteholz preparation of a twenty-day rainbow trout embryo on which an attempt was made to inject the lymphatics and the veins by injecting into the subocular lymph sacs (1 in fig. 5). The figure shows the position occupied by the subocular lymph sacs
292
CHARLES F. W. McCLURE
ill the living embryo and, since the sacs alone filled with the injecta, is illustrative of an experiment which proves that the subocular lymjih sacs have not grown caudad, nor do they communicate with the lateral pharyngeal lymphatic, nor with the veins at the stage of development presented bj^ this twenty-day trout.
Fig. 6 Transverse section taken through the subocular lymph sacs of a twenty-one-day rainbow trout embryo in which, as independent structures; the sacs have reached the maximum stage of their development. P. 1^. C. series 646. Injected embryo. /, subocular lymph sac; 6, precardinal (jugular) vein; 20, eye; 22, carotid artery.
The subocular lymph sacs of the trout embryo are non-pulsatile in charactei- and are Hned by an endothelium around which no muscular coat is formed. J)uring the stage of their independence they become gradually distended with lym]ih which must necessarily enter them in a centripetal direction from the
EMBRYONIC LYMPHATIC SYSTEM 293
intercellular spaces of the head. As this lymph gradually mcreases in amount the subocular sacs of the trout can be easily observed in the living eml^ryo, and are especially prominent in the rainbow trout between the nineteenth and twentyfirst days. B}' the time the subocular h'mph sacs have attained their maximmii size as independent structures, on the twentyfirst day (figs. 4 and 6), a considerable pressure must be exerted by their lymph upon their walls, which would account for the distended appearance presented by the sacs (1) at this stage (fig. 6). As soon, however, as the subocular haiiph sacs establish tlieir communication with the lateral pharyngeal lymphatic, so that their h'mph can flow continuously and centripetalh' to the veins, this pressure against their walls is immediately released, and the sacs then appear less prominenth' in the li\-ing embryo, due to a partial collapse of their walls.
It is thus seen that, during the period of its independence, the subocular lymph sac of the trout embryo serves as a local and independent reservoir for the reception of lymph which enters it in a centripetal direction from the intercellular spaces; that it retains this lymph only temporarily, until the sac establishes a communication with the lateral pharyiigeal lymphatic, through which a continuous centripetal lymph flow may then pass from the intercellular spaces to the venous circulation.
The subocular lymph sacs of the trout embryo therefore furnish us with a striking example of the fact that, not only do independent and discontinuous anlagen of the lymphatic system actually exist, but, that they can also be observed and be experimented upon in the living embryo.
The functional role played by the subocular lymph sacs of the trout embryo, during the stage of their independence, as well as after they have estabUshed a communication with the venous circulation, undoubtedly gives us the clue to the function assumed by the independent lymph vesicles of the embryo in general. It also explains the manner in which a continuous centripetal lymph flow is estabUshed in the embryo, between the intercellular spaces and the venous circulation, in relation to the developing hmiphatic vessels.
294 CHARLES F. W. McCLURE
On the basis of the functio?ial rule played by the subocular lymph sacs — and this can be actually demo7istrated in the living trout embryo^-dt is highly probable that the independent lymph vesicles, of the embryo in general, also serve as local reservoirs for the temporary retention of lymph which enters them in a centripetal direction from the iiitercellular spaces; that these lymph vesicles become progressively concrescent with one another to form continuous channels, through which the lymph collected and temporarily retained by them is then forwarded to the venous circulation. In this manner the centripetal flow of lymph which continuously enters these independent lymph vesicles from the outlying intercellular spaces, is continued on to the venous circulation.
It may be mentioned here, incidentally, that the functional role plaj^ed b}' the subocular lymph sacs of the trout embryo during the stage of their independence, is also evidence of the fact that the lymphatics of fishes function solely in the capacity of lymphatics at the time of their inception, and that they are therefore not transformed \'eins.
In case the independent lymph vesicles of the embryo should fail to become concrescent with one another and with the veins, at the typical points of lymphatico-venous entry, an oedematous condition of the body would undoubtedly arise, aiid the ontogenetic condition, in which only independent and discontinuous anlagen are present, might then be retained in the adult. That such might actuall}- be the case seems to be borne out by the conditions observed in an oedematous human foetus alreadj^ desci'ibed by Smith and Birmingham.'- These investigators ha\'e described a case in which that peculiar and rare condition known as oedematous foetus was found to depend upon the complete absence of the thoracic duct, lymphatic glands and lymphatic trunks in general, and in which the lymph was stored in what they described as "greatly distended tissue spaces" which neither communicated with one another, nor with the veins.
- Smith iintl IJirmingham. Absent tlioracic duct ciiusiuy: ocdeina of n foetus. .]our. Anat. aiirl Physiol., vol. 23.
EMBRYONIC LYMPHATIC SYSTEM 295
Only two possible conclusions can be drawn regarding the significance of the complete absence of continuous lymphatic trunks and of the presence only of independent and discontinuous lymph-containing tissue spaces in this human foetus: Either there has been a complete failure on the part of the h^raphatic system even to initiate its development, so that these 'tissue spaces' are in no sense related to the true lymphatic system, or, the presence of these spaces signifies a condition in which the normal development of the Ij-mphatic system has been arrested at an early ontogenetic stage.
Huntington and the writer'* have repeatedly described and figured the presence of independent lymph vesicles or lymph spaces in the mammalian embryo, and Huntington has recently made a more extensive and detailed study of these structures, as theA' occur in the subclavian and primitive ulnar regions of the cat. Although one is not given the opportunity of studying these independent lymph vesicles in the living embr3^o of the mammal, as is the case in the trout, it would now seem quite a waste of time to parley further over the question of their presence in the mammalian embryo, or, that it is through the concrescence of such independent vesicles that the main lymphatic channels of mammals are formed.
The oedematous human foetus described by Smith and Birmingham appears to present us with the most striking evidence, not only of the fact that the presence of independent lymph vesicles may actualh^ be demonstrated, in certain circumstances, as functional structures in mammals, but, also, that the functional role played by them is similar to that plaj'ed b}' the subocular h'mph sacs of the trout during their independent stage. It is
•' Huntington and ]\IcC"lure. The anatomj- and development of the juguhir Iraiph sacs in the domestic cat (Felis domestica). Amer. Jour. Anat.. vol. 10, 1910. fig. 66.
' Huntington. The develoiunent of the mammalian jugular lymi)li .sac. of the tributary primitive ulnar lymphatic and the thoracic ducts from the viewpoint of recent investigations of lymphatic ontogeny, together with a consideration of the genetic relations ot lymphatic and haemal vascular channels in the embryos of amniotes. Amer. Jour. Anat., vol. 16, 1914. Also, The development of the IjTnphatic drainage of the anterior limb in embryos of the cat. Proc. Amer. .\ss. .\nat.. Anat. Rec, vol. 9, 1915.
296 CHARLES F. AY. McCLURE
highly probable, therefore, that the conditions found in this human foetus indicate that the development of the lymphatic system had been arrested at a normal ontogenetic stage; that its oedematous condition was due to the circumstance that the independent and discontinuous anlagen of the lymphatic system had failed to become concrescent with one another and with the veins, in order to establish a continuous system of channels, through which a continuous centripetal lymph flow could pass from the outlying tissue spaces to the venous circulation.
THE PYRA]\IIDAL TRACT IX THE GUINEA-PIG (CAVIA APEREA)
IDA L. REVELEY
Fro))i the Pinjsiological LohDralnnj, Mtdlcdl College, Cornell University,
Ithaca, X. Y.
TEX FKUTfES
IXTRODUCTIOX
The pyramidal tract (fasciculus cortico-spinalis) in rodents, so far as it has been examined in this order, is crossed and runs in the dorsal column of the spinal cord, but there are exceptions to this rule. In the family Leporidae, including- the rabbits and hares, it lies in the lateral columns, and in the ( 'anadian porcupine there is a dorsal column, a lateral column and a ventral column tract (Simpson '14). In view of the fact, therefore, that such wide variation exists between closely related species, it is desirable that as man^^ as possible of these be examined.
In the guinea-pig, the animal with which this paper deals, Spitzka ('86), Bechterew ('90) and Wallenberg ('03) have found that the pyramidal tract decussates into the posterior column.
Ranson ('13) states that in the albino rat the tract consists of a mixture of meduUated and non-meduUated fibers, and b}' the use of the pyridine-silver method of Cajal (modified), the non-medullated fibers are stained, so that the course of the tract can be followed by this means.
According to Linowiecki ("14), who worked in Ranson's laboratory, also with the pyridine-silver method, the same obtains in the guinea-pig. In this annual the tract lies in the posterior column, but it does not forrh such a compact uniform area when stained by this method as is found in the rat, indicating, ajoparentl}', that the proportion of medullated to nonmedullated fibers is greater in the guinea-pig.
297
THE ANATOMICAL RECORD, VOL. 9, XO. 4
298 IDA L. REVELEY
The pyridine-silver method may be regarded as the complement "^of the Marchi method since the latter stains only medullated fibers in the process of degeneration.
PRESENT INVESTIGATION
The object of the present research was to trace the fibers of the pyramidal tract in the guinea-pig from their origin in the cerebral motor cortex to their termination in the lower levels of the brain and spmal cord. The method of secondary degeneration was employed, with Marchi staining.
Eight anhnals (adults) in all were used. The cerebrum was exposed on the left side, under ether anesthesia, and the motor cortex removed. At the end of periods varying from twelve to sixteen days after the operation they were killed by ether or coal gas, when the brain and spinal cord were removed and placed in 3 per cent potassimii bichromate. After three weeks in this fluid, with frequent changing, the tissue was cut into slices 3 to 4 mm. thick and placed in ^larchi's fluid (3 per cent potassimn bichromate, 4 parts, 1 per cent osmic acid, 1 part). At the end of eighteen days the pieces were removed, washed in running tap water for twelve hours, and taken through the alcohol-xvlene-paraffin series into paraffin in which they were imbedded and cut. Sections from all levels of the bram and from most of the segments of the spinal cord were mounted and examined.
COURSE OF PYR.-OIIDAL TRACT FOLLOWED BY SECONDARY DEGENERATION
The course of the pyramidal tract through the midbrain, pons and upper part of medulla oblongata is similar to that found in the higher mammals such as the cat, dog, monkey and man, and is so well known that no detailed description need be given. In the midbrain it occupies the middle three-fifths of the crusta, more or less, and is continued downwards as the pontine bundles, which unite at the lower border of the pons to form the anterior pvramid of the medulla. Above the level of
PYRAMIDAL TRACT IN THE GUINEA-PIG
299
the general decussation, in the lower part of the medulla oblongata, there is no evidence of any crossing of fibers; all the degeneration appears to be confined to the side of the lesion.
Sections through the lower or closed half of the medulla oblongata, about 1 mm. below (caudal to) the calamus scriptorius, show the beginning of the pyramidal decussation. The
Fig. 1 Transverse section, medulla oblongata through upper extremity of pyramidal decussation. X 10.
Fig. 2 Transverse section, medulla oblongata through middle of pyramidal decussation. X 10.
pyramid, in transverse section, is triangular in outline at this level, and from the dorso-mesial angle a few fibers can be seen passing backwards along the median raphe. They cross the raphe close to the central gray matter and curving outwards, in front of the hypoglossal nucleus, turn backwards ip the gray substance. One or two small bundles reach the posterior column but most disappear in the gray matter (fig. 1).
300
.• »t , lower level, al^out the .nickUe of the decus In sections at a lo«ei le^e , ^^^ ,„ore or less
sation, the fibers cross in great ™>^b«^^ j^ ; ^^^,,,,^ intei ,,ell defined bundles ff^^ Crt tund side. After
lacing with -"-P-;^;"|^r U^rroutwards and then curve crossing the raphe the Abes 'i ^^^^ ^j ^,^^„^
backwards and inwards *'°"Sb the 8ia transversely,
passing into the funiculus '^'^^^^^l^ik- 2). Along the Ly form a distinct and ^°™Pf^\*7„'u bundles are seen dorsal margin of the gray matter a ^e.^ sm
on the mesial side of f?Xt:aTf generated fibers runs At this level a smg e small stianc| « g ^^^^^
backwards through the f^-^^-^^^Zv bend outwards; to the central canal, and *""^^Vit Caches the dorsal
^— Tis^trf:Lrwir:^^^^^^^
r:Us"r:entoii^^infour c^^^^^^^ ^,^. 3, At the junction of the medulla ^Mth tlie i ^^^^^
practically all the fibers have er - ^ « ^ he ^^.^^^
L 1,1 the ^^--^vraHetached .Zues extend from its m outline, but a le^^ -m^ matter, as de mesial angle along ^^^^^^^'^^Zll^-'^ ^^--Ale, seen near scribed in the last section, ihc "o" , ^^^^ ^,o,,_
the middle of the "at'on, is a^^^^^^^ ^ ^^^ „„ ,,, ing seems to l^e complete, no ae„e decussa same side. Between the upper ^^^^ ^,, Regeneration many fibers seem to ha^ e "^^^^^^^^ ^^^ , „„,.« tion m the anterior pyramid '^«^".^' ' ,^„,„ ^.-aet. These extensive area than in the crossed do sal oUn ^^^_^^^ ._^ have presumably terminated m the g.aj matte, this region. , „,.„„,ed pvramidal tract I, the first c«-v.ca -g™-* . * , ' Xmn of Burdach of reaches its largest size. /* "^" "\ ' „„,terior horn and «ray the opposite side, in -"*-';"* ^.^i™ line, its ventro commissure. It - — *^^ ™to and meeting its fellow mesial angle extendmg t«the middle ^^ ^^^ ^^^^,^ ^^^,^
of the homolateral side (fig. 4). All ^ .,,„l„„ in tlie
decussated and there is no evidence of ,un cU^o
PYRAMIDAL TRACT IX THE GUIXEA-PIG
Wl
crossed lateral or direct ventral columns as is the case in the Canadian ])orcupine.
Sections through the second cervical segment (fig. 5) show a considerable change in the area occupied b}' the fibers of the tract. It is crescent-shaped; the dorsal border is concave; the mesial border lies against the ])osterior medium septum, oc
Fig.
5 6
3 Transverse section, medulla oblongata through lower (caudal)
extremity of pyramidal decussation. X 10.
Fig. 4 Transverse section, first cervical segment of spinal cord. Fig. 5 Transverse section, second cervical segment. X 10. Fig. 6 Transverse section, fifth cervical segment. X 10.
X 10.
cupying about one-fourth of the distance between the posterior gray commissure and the free margin of the section. The degeneration is less dense than in the first cervical segment indicating a distinct diminution in the number of fibers.
In the third, fourth and fifth cervical segments (fig. 6) the general appearance of the tract changes little, but there is a progressive faUing off in the number of fibers which it contains.
302
IDA L. REVELEY
The degeneration seems to be densest near the gr^y/^^^*^^' the fibers becoming more and more scattered towards the dorsal
border of the area. .
Between the fifth cervical and first thoracic segments (hg. 7) a still further dhninution in the nmnber of fibers is evident. In the latter segment the tract, considerably reduced m size, occupies an oval area which is no longer in contact with the posterior median septum except at its ventro-mesial extremity.
Fie 7 Transverse section, first thoracic segment. X 10.
Fig 8 Transverse section, eighth thoracic segment. X 10.
Fig q Transverse section, first lumbar segment. X 10.
Fig. 10 Transverse section, fourth lumbar segment. X 10.
In the eighth thoracic segment the area of degeneration is still more restricted (fig. 8). It has now withdrawn from the middle line and lies in the recess formed by the narrowing of the neck of the posterior horn. Tracing it caudalwards it is found to occupy the same relative position in the succeeding segments, becoming more and more reduced m size ^ntil the fourth lumbar segment is reached, where it is represented by a very small number of scattered fibers lying against the neck of the posterior horn (figs. 9-10). Beyond this level it cannot be followed.
PYRAMIDAL TRACT IX THE GUINEA-PIG 303
It is interesting to compare the above results, obtained b}^ the ]\Iarchi method, where the medullated fibers alone are stained, with those of Linowiecki, in the same animal (guinea-pig), who used the pyridine-silver method which brings out the nonmedullated fibers. According to his description: "In the seventh cervical segment the pjTamidal tract is located in the ventral
part of the posterior funiculus The fibers of the
tract are more densely grouped ventrally and lateralh' near the grey substance and this gives the cross section of the two tracts somewhat the form of the letter V." (Compare with figures 6 and 7.)
At the level of the eighth thoracic segment, by the p,vridinesilver method, the tracts are crescentic in outline and much diminished in size; they are still further reduced at the twelfth thoracic segment where they consist of two compact groups of axons which have become separated at the posterior median septum. Proceeding caudalwards they become less distinct and at the level of the second lumbar segment the groups tend to move posteriorly and to separate from each other. From here on they narrow markedly and fade in color until at the level of the fifth lumbar segment they consist of two narrow strips, one on each side of the posterior median septum, which are hardly visible.
It will thus be seen that the descriptions of the position and outline of the pyramidal tract, as brought out by the two methods, are in close agreement. This would indicate that the mixture of medullated and non-medullated fibers, of which the tract appears to be made up, is more or less uniform throughout its entire course in the spinal cord.
In the fifth lumbar segment, according to Linowiecki, the tracts consist of two narrow strips on each side of the posterior medium septum,"^ but he does not say whether they are in contact at the septum or separated from each other. At the level of the fourth lumbar segment almost the same words might
1 Taken as it stands, this sentence would seem to indicate that the tract is represented by tico narrow strips on each side. What the author does mean, probably, is that there are two narrow strips, one on each side.
OQ^ IDA L. REVELEY
be used to describe the tract, as brought out by the degeneration method, if it be added that the narrow- strip hes close to the S aspect of the gray n.atter forming the neck of the posterior
horn (fig. lOj.
SUMMARY
The course of the pyramidal tract in the guinea-pig, from the beginning of the decussation in the medulla oblongata caudalwaids, as brought out by the method of secondary degeneration with ^larchi staining, is as follows: , c .^
The decussation begins al,out 1 mm. below the evel of he calamus scriptorius and ends near the junction of he medulla ^"h the spinal cord. All the fibers cross, between these limits, "nd most pass on into the funiculus cuneatus where they turn 'coudahvards into the spinal cord but many end in the gray matlei of the bulb in this region. As this dorsal column tract is followed downwards, from segment to segment of the cord
t outline changes considerably (see figures) and here is a progressive diminution in the number of fibers which it con ains but this loss of fibers is most marked in the upper cervical and
'X'tr c—be traced farther than the fomth lunibar segment, where it is represented by a very few degeneiated fibers Iving close to the gray matter of the posterior horn.
Iccording to Eanson the pyramidal tract consists of a mixture of meduUated and non-medullated fibers, the former o which, while un<lergoing degeneration may be ^t^ined by the Marchi method, the latter bj' the pyndine-silver method. Ihe d sc tioi of the spinal portion of the tract in the guinea-pig J^n bv Linowiecki, who used the pyridine-silver method fstaclose agreement with what I have found by the degenera ion ,ethoc°- thi' would appear to point to the fact that the mixture ofihe two varieties of fibers, within the tract, is fairly uniform throughout its course.
PYRAMIDAL TRACT IX THE GUINEA-PIG 305
BIBLIOGRAPHY
Bechterew, W. 1S90 I'eber die verschiedenen Lagen imd Diniensionen der Pyramidenbahnen beim ]\Ienschen und den Tieren iind fiber das \'orkommen von Fasern in denselben, welche sich durch eine friihere Entwickelung anszeichnen. Xeurol. Centrabl., p. 738.
LixowiECKi, A. J. 1914 The comparative anatomy of the pyramidal tract. Jour. Com]j. Xeur., vol. 2-4, p. 509.
Raxsox. S. W. 1913 The fasciculus cerebro-s])inalis in the albino rat. Amer. Jour. Anat., vol. 14. p. 411.
SiMPSox. S. 1914 The motor areas and pyramid tract in the Canadian porcupine (Erethizon dorsatus, Linn.) Quart. Jour. Exper. Phj'siol., vol. 8, p. 79.
Spit/ka, E. C. 1886 The comparative anatomy of the pyramid tract. Jour. Comp. ;Med., vol. 7. ]>. 1.
Wallexbekg, C. a. 1903 Cited by Cokistein. Zur vergleichenden Anatomie der Pyramidenl)ahn. Anat. Anz., Bd. 24, p. 454.
A STUDY OF THE AFFERENT FIBERS OF THE BODY
WALL AXD OF THE HIXD LEGS TO THE CERE BELLU:\I OF THE DOG BY THE AIETHOD
OF DEGENERATION
GILBERT HORR.VX From the Anatomical Laboratory of the Johns Hopkins Medical School
SEVEX FIGURES
This work was begun under the stimulus of the work of Bolk, who based an hypothesis of a localization of a series of coordinating centers in the cerebellum upon studies in comparative anatomy. Bolk has simplified the nomenclature of the parts of the cerebellum and, using his terms, sums up the localization in the cerebellum as follows:
The lobus anterior cerebelli contains the coordinating centers for the groups of muscles of the head, nameh' those of the ej^es and tongue, the muscles of mastication and muscles of expression beside those of the larynx and pharjaix; the IoIduIus simplex contains the coordinating centers for the neck musculature; the upper part of the lobulus medianus posterior contains the unpaired coordinating centers for the right and left extremities; the lobuli ansiformes and paramedian! contain the paired centers for the two extremities, while the rest of the cerebellum has the coordinating centers for the' trunk musculature ('07, p. 170).
In general, Bolk thinks that the coordmating centers for symmetrical muscles which act together are in the vermis, while the centers for those which act independently are in the hemispheres.
This hypothesis has been borne out by the experunental work of Rynberk ('04) from Luciani's laboratory, for example, by obtaining special movements of the neck muscles after a unilateral exthpation of the lobulus simplex. These results suggest further work on the end station of the fibers of the different regions of the body by the inethod of degeneration, though
307
308 GILBERT HOKRAX
it is of course clear that the tracing of the afferent fibers of each region to their end station does not unravel the nature of a coordinating center in the cerebellum. The results of tracing the fibers of the different regions of the cord to the cerebellum indicate that the fibers of each of the regions of the body sends fibers to almost the entire vermis.
The most recent and the most extensive work on determining the distribution of spinal fibers in the cerebellum is that of Sir Mctor Horsley in 1909. In this article he gives a complete analysis of the literature of the work on the fasciculis spinocerebellares ^'entralis and dorsalis. As far as the point of the distril)ution of the fibers to the cerebellum is concerned, the main results are as follows: In 1890 Auerbach stated that the fasciculus dorsalis (Flechsig) ended in the dorsal — that is to say, in the cephalic — part of the superior vermis, and the fasciculus ventralis (Gower's) in the ventral part. In 1892 Mott reversed this statement by showing that the fasciculus dorsalis ends in the vermis, caudal to the end station of the fasciculus ventralis, which enters the cerebellum farther cerebralwards by way of the brachium conjuncti^^um or superior cerebellar peduncle. This point is well shown in the well-known diagram of his figure 1, page 219.
Collier and Buzzard in 1903, in an anaWsis of human material, confirmed Mott's view of the relative position of the end station of the dorsal and ventral cerebellar tracts; that is, that the fasciculus spino-cerebellaris dorsalis ends in the inferior vermis, though they find that some of the fibers end in the nucleus dentatus and in the nuclei of the roof. The fasciculus spino-cerebellaris ventralis they trace b}^ way of the superior cerebellar peduncle to the superior vermis but in small part also to the lateral hemispheres.
Sir A'ictor Horsley divided the coi-d in a general way into four regions: the first, from the fii'st to the fourth cer\'ical segment, representing movements of the head and neck ; the second, from the fifth cervical to the first thoracic segment, representing movements of the arm; the third, from the second thoracic to the secfuid lumbar, representing movements of the body: and the
AFFEREXT FIBERS OF THE DOG 309
fourth, from the third kinibar to the second sacral, representing movements of the leg. He made the lesion cover the fibers representing a given region, by destroying the cells of origin for the tract rather than the fiber tract itself. Indeed, his purpose was to determine which cells of the cer\-ical and lumbar regions of the cord are homologous with the nucleus dorsalis. The lesion for the first or upper cer\-ical region was in the gray matter of the cord, taking in the cells in the homologous position to the nucleus dorsalis and the cells of the middle region of the gray matter. Within the cerebellum the fibers passed to all the vermis except the most anterior part of the lobus anterior, namely, the lingula, and the most posterior part of the lobulus medianus posterior, namely, the uvula and the nodulus. Thus the end station for the afferent fibers of the neck is not limited to the lobulus simplex but includes almost the entire anterior lobe and most of the median posterior lobe as well.
The fibers of the second region, representing fibers from the arm, he found to pass forward mainh' on the same side but in part on the opposite side. M'ithin the cord they run both in the dorsal and in the ventral cerebellar tracts, ^^'ithin the cerebellum they end in the lobulus centralis, the ventral half of the culmen and the ventral half of the pyramidalis; or in Bolk's termmology, in the lobus anterior, in all the lobulus simplex and in most of the lobulus medianus posterior. Horsley did not study the fibers of the third of the iDody regions, but the fibers from the leg region he found ended in exactly the same parts of the cerebellum as those of the arm region.
Thus from the literature it is clear that the spino-cerebellar fibers end in the vermis; that the fasciculus spino-cerebellaris ventralis fGow^er's tract) ends in the more cerebral part of the vermis, while those of the fasciculus spino-cerebellaris dorsalis end in the more caudal part of the ^'ermis. The fibers representing the four regions of the body — namely, the neck, the arms, the body and the legs — pass through both cerebellar tracts and are distributed to all parts of the vermis except the most anterior and the most posterior folia. These results are confirmed in the experiments herein reported.
310 GILBERT HORRAX
As far as the symptoms of lesions of the cerebellar tracts are concerned, our results also agree with those of Horsley, who found that there was no loss of efferent (purposeful) movement in the muscles involved ; that all the motor effects were transitory and probably due to interference with the anterior cornus, and that there was ataxia and clumsiness of mo\'ement. These results are practically the same as those of Bing.
TECHNIQUE AND METHOD OF INVESTIGATION
Three dogs, each about the size of an ordinary fox terrier, were used for the purpose of our study, and in all cases the technique of operation and preparation of material was identical. In Dog 1, experiments with regard to various sensations were carried out during the period between the operation and the killing of the anunal for histological study. As these experiments were not of a sufficiently satisfactory nature, they were omitted in Dogs 2 and 3, but are recorded in connection with Dog 1 for the sake of completeness. The artificial lesion in the cords of the dogs was made as follows : An incision was made in the median line of the back, extending between the scapulae down toward the lower dorsal region, for a distance of about 10 cm. Very little hemorrhage took place and this was quickly stopped. Laminectomy of the 4th, 5th and 6th dorsal vertebrae was performed and the cord in its dura exposed. An aneurism needle was placed under the cord very genth' and the cord in its dura was lifted slightly and rotated a little toward the left. A \'ery superficial slit was then made with a narrow scalpel, through the dura and into the substance of the cord at right angles to its long axis, and on the right side of the animal, in an effort to cut the fasciculus spino-cerebellaris dorsalis (Flechsig) and possibly the fasciculus spino-cerebellaris ventralis (Gowers) ; the cord was then slipped back. There was no hemorrhage in any case during the cutting of the cord.
The wound was closed in tighth^ by sutures of silk through the muscles, fascia, subcutaneous tissue and skin. All the dogs made prompt recoveries, but in Dog 2 the skin layer of the
AFFERENT FIBERS OF THE DOG 311
wound was opened by the dog's scratching on his cage. The skin separated, but the wound was kept swabbed out with iodine so that the lower layers did not open and were not involved in the superficial infection. The dog improved steadily and granulations formed rapidh' over the wound.
The operations were performed by Dr. Goetsch, of the Johns Hopkins Hospital, on Dog 1; by Dr. Hunnicutt, of the Johns Hopkins Hospital, on Dogs 2 and 3, under strict aseptic precautions, the total time of anesthesia being from an hour and a quarter to two hours; ether was used as an anesthetic.
The dogs were kept alive for periods of ten days to three and one-half weeks, during which time observations were made on them in order to ascertain the nature of the sj'mptoms caused by the lesion. The wounds of the operations healed perfectly within a few days. The symptoms observed were recorded each day, and in brief were as follows:
Dog 1. April 26, 1910, the da}' after the operation, 9.00 a.m. No apparent disability except in use of hind legs; dog sits at rear of cage, with left leg extended and raised at an angle of 35° to -15°; right hind leg somewhat flexed and lying on floor. ^Yhen called and coaxed by snapping the fingers, the dog responds by wagging tail, and by slight effort s at movement, but does not actually change position. Head, fore legs and fore feet, are moved in a perfectly coordinated and intelligent waj'.
April 27, 1910, 9.30 a.m. Dog still sits in about the same position as on previous day, but the left hind leg, instead of being raised, is more nearly or quite touching the floor. When called, the dog crawls to the door of the cage, locomotion being accomplished mainly by the use of the fore legs, the animal remaining in the sitting posture throughout. The hind legs are both more or less flexed, and during locomotion perfectly definite, although almost ineftectual, movements are made by them; the toes of the left hind leg are occasionally flexed. Movements of all parts of body except hind legs seem perfectly normal.
April 28, 1910. When taken out of cage to-day, the dog moves along with its hip against the wall for support, the hind legs not working nmch better than the day before. Once it sat down and scratched with the right hind leg, just posterior to the right fore leg.
April 29. 1910. When taken out on the grass, the dog took several steps in a normal manner on all four legs, but finalty the hind legs weakened, spread apart, and collapsed, throwing the rear half of the dog's body first to one side and then to the other. The dog was seen to scratch again in the same manner as j^esterday, only the
312
GILBERT HORRAX
left hind leg was used. Hot and cold water was applied to all four less bv means of dipping the latter into a beaker of water (during these tests the dog was bhndfolded). Following are the results:
Temperature of 3° to O^J. Causes no effect on any of feet.
Temperature of 20^ to 37°, 47°, 57° Causes no effect on any of feet.
Temperature of 67° Causes both hind legs to be withdrawn from the water 20 seconds after time of immersion; both front legs were withdrawn 2 to 3 seconds aftei- immersion; pinching toes with forceps causes prompt withdrawal of all four legs.
\pril 30, 1910. This morning the dog shook the fore part of the body, also got up on all four legs and stood for some time.
Temper Temper
ature of 37° ature of -47°
Temper Temper
•ature of 58°
Temperature of 65°
Gives no effect on any of feet.
Xo effect on hind legs; right front leg was withdrawn when the tips of toes came in contact with the water; left fore leg not tried.
Xo effect on hind legs; right fore leg was withdrawn on being touched to the water; no effect on left fore leg.
All legs withdrawn from 6 to 12 seconds after time of immersion, fore legs a litt'e t-ooner than hind legs; toes of hind legs were shaken in the water before withdrawal.
Same as 58°.
Tail withdrawn 6 seconds after immersion m water of 56° temperature.
^Iav 2 1910. Dog showed marked improvement in walking more
normally.' Hind legs used most of the time, but with an unsteady, swaying motion from side to side.
Xo effect on hind feet: front feet were withdrawn upon contact with water. Xote: ^^ hen the fore feet were held in the water, no signs of discomfort were apparent.
Same as at 37°.
Same as at 37°, except that discomtort was evidenced when fore feet were held in the water.
Left hind foot withdrawn in 5 seconds; right hind foot withdrawn in 15 seconds; fore Teet l)oth withdrawn ui)on contact.
May 3 1910. Improved general condition was noted to-day, with better use of hind legs, although the drunken, swaymg gait was still marked.
Temperature of 37'
Temperature of 43' Temperatui'c of 53'
Temperature of 57'
AFFERENT FIBERS OF THE DOG 313
Temperature of 28° No effect on hind feet; fore feet withdrawn upon
contact with the water. Temperature of 35° Hind feet withdrawn after 15 seconds; fore feet
withdrawn upon contact. Temperature of 47° Right hind foot withdrawn upon contact; left
hind foot not withdrawn at all; both fore feet
withdrawn upon contact; no effect on tail.
May 4, 1910. Temperature observations.
Temperature of 28° Hind feet at first withdrawn upon contact, but subsequently allowed to remain in water; fore feet withdrawn upon contact and not subsequently allowed to remain.
Temperature of 37° No effect on hind feet; fore feet withdrawn upon contact, but allowed to remain upon subsequent immersion.
Temperature of 47° All feet withdrawn almost immediately ; when beaker containing no water was touched to the hind feet there was no withdrawal nor other noticeable effect; same beaker to fore foot caused withdrawal of latter; right fore foot was not withdrawn.
\Iay 7, 1910. Temperature of 42°, Xo effect on hind feet; fore feet withdrawn upon
contact. Temperature of 45°, 47° Same as 42°. Temperature of 55° Hind feet withdrawn after 9 seconds; fore feet
withdrawn upon contact.
May 10, 1910. Again a marked improvement in the use of the hind legs was noted. Dog was very lively, running and jumping around, but there was still lack of coordination in the hind legs.
May 11, 1910. Dog was livelier than on previous day; ran about and sprang upon the observer in puppy-like fashion, playing around with very little departure from normal movements. However, the same uncertain gait of the hind legs was noticed when the dog would stop jumping and either walked or ran slowly away.
May 13, 1910. Knee jerks of hind legs present and equal on both sides.
May 14, 1910. Dog killed; brain and cord removed.
Dog 2. December 7, 1912. Operation, 10 a.m.
December 7, 1912. Dog conscious and sitting up at 2.30 p.m.
December 8, 1912. Sitting up; can move back legs.
December 9, 1912. Makes attempt at locomotion with hind legs.
December 10, 1912. Can stand up on all fours.
THE ANATOMICAL RECORD, VOL. 9. XO. 4
3][4 GILBERT HORRAX
December 11, 1912. Walks a little, with a typical extreme ataxic gairibaik feet sometimes interfering with each other) and often falls to one side or the other.
December 13. 1912. In getting into his box thei^e seems to be more uncertainty of his right than of his left hind leg; dog walks more todqv same aait but some improvements. . , i i
%ex"nLl-14,1912. 4 p.m. walks f^--\ffjr'^^^nAZ' follows one around the room, comes when ca led etc. His hnd legs however, are very ataxic. th<. right bemg noticeably more so than the
^^ecember 15, 1912. Walks alx.ut: shows umch improvement in
December l(i. 1912. Cait much improved: ataxia in right hind leg
^%eSer 17-20. 1912. Gait steadily improving; also marked gain
"' December 20, 1912. Dog killed; l)rain and cord removed.
Doqfl. December 14, 1912. Operation. 10.30 a.m. ,.,.,,,,„.
December 14 1912. 4 p.m. ; dog gets up on all fours and ^^ alks about the room his hind legs being very ataxic, but his recovery m general behio qSckei alui his ability to walk coining much sooner than was the
'%e«>mW?%2."Dog walks about the room: ataxia <.f right
^'"D(^nber Ki. 1912. Walks about : ataxia of right hind leg perfectly
'^'December 17-20, 1912. Improvement rapid and more complete than in Dogs 1 and 2; runs and jumps about the room.
January 8, 1913. By this time no ataxia can be noticed, dog > Rait and' actions are apparently normal in every way.
.January 10. 1913. Dog killed; brain and cord removed.
At periods, varying from 10 days to :U weeks from the date^ of operation, as noted above, the dogs were anesthetized with chloroform. The right femoral vein was opened, after which a cannula was inserted into the left common carotid artery and thr.mgh the latter a liter and a half of 10 per cent formalm solution was injected into the animals.
\fter waiting half an hour in order that as much hardening as possible might lake place, the brain and cord were remoAec carefully. The cord was cut inK. three approxnnately eciua lengths' and put at once into a vessel contaimng a 10 per cen solution of formalin. The material was left in this solution un .1 various parts of it were wanted for study. Th(> f-rst blocks
AFFERENT FIBERS OF THE DOG 315
were cut out of the cord three days after it had been put into the formalin and the other blocks were taken subsequenth^ for study througliout the following year; all the blocks were from 5 to 10 mm. thick. The part containing the medulla and pons, with the cerebellum, was cut into four blocks which w^ere numbered as is shown in figure 1. As will be seen, the first and second blocks contain, in Bolk's nomenclature, the lobulus medianus posterior of the cerebellum; the third block contains the lobulus simplex and the caudal part of the lobus anterior; the first block contains the rest of the lobus anterior. The blocks were all treated in the same manner; they were stained en hlocke by a modified Marchi method; they were placed for from o tf) 7 days in the following solution:
Osmic acid 1 part
XalOs 3 parts
Distilled water 300 parts
The blocks were embedded in celloidin and all the sections showed an excellent staining of the degenerated myelene.
EVIDEXC'ES OF DEGENERATIOX IX THE SECTIONS
The study of the sections of the first dog showed ab()\'e the lesion degenerated fibers scattered throughout the section on both sides. In the upper cervical region, as shown in figure 2, there is a very abundant, scattered, bi-lateral degeneration of fibers, covering tlie entire ai'ea of the fasciculus spino-cerebellaris dorsalis and ventralis. It is difficult to understand wh,v there is so extensi\'e a degeneration on the left side, inasmuch as the cord was cut only on the right; l)ut as has been seen, the symptoms in\'olved both sides and there is an almost symmetrical degeneration.
Sections through the first block containing the cerebellum, as shown in figure 3, have a concentration of the degenerated fibers in the corpus restiforme and along the lateral margins of the medulla, with a second concentration in the tractus spinocerebellaris ventralis, especially of the right side. In the cerebellum the degeneration is confined to a liand across the vermis
316
GILBERT HORRAX
in its ventral third. This degeneration does not appear until the cephahc end of the first block is approached
\s one follows the sections farther cerebralward m the second block, as seen in figures 4 and 5, there is the same concentration of degenerated fibers in the corpus restiforme and along the rioht margin, while the fibers on the left side are less nunierous but have the same general pattern. Withm the cerebe lum the degenerated fibers are found throughout the second block, that is, the lobulus medianus posterior. For the most part, the degeneration is confined to the vermis but in figure o can be seen a few fibers in the edge of the hemispheres.
From here on the course of the filbers of the fasciculus spmocerebeUaris dorsalis can be followed easily in their position m the mferior cerebellar peduncle. In the third block, as seen in figure 6, the fibers of the corpus restiforme can be seen m the foUa of the vermis of the lobulus simplex. The cephalic Innit of the ending of the fasciculus spino-cerebeUaris dorsahs withm the cerebellum is reached m the caudal half of the thhd block, as shown in figure 6. Above this level, namely, m the^ lobus anterior, only the fibers of the fasciculus spmo-cerebellans dorsalis are to be found (fig. 7). , , ^^i f .i
It will tluis be seen that we have traced those fibers of the fasciculus spino-cerebellaris dorsahs (Flechsig), which represent the legs and possibly the lower body waU, from their situaion in the cord, up through the corpus restiforme of the medulla into the vermis cerebelli, in the caudal half of which they were distributed. In the most caudal part of the verinis their distribution is confined to one or two lammae, but farther cerebralward the distribution is very diffuse throughout all the
laminae. , , , , , ii
V part of the sensory fibers from the legs and lower body ^ all pass to the cerebellum through the fasciculus spino-cerebellaris ventralis (Gowers). These fibers occupy the antero-laiera margin of the cord, being more scattered than those of the dorsal tract of Flechsig. The ventral position of the tract is plain in figures 4, 5 and 6 for the region of the medulla. In the pons the
AFFERENT FIBERS OF THP: DOG 317
ventral fibers shift to their more dorsal position in the lateral margin of the brachium eonjunctivum, as seen in figure 7. Throughout the cephalic half of the third block and a part of the fourth block — that is, in the lobus anterior — the fibers of the fasciculus spino-cerebellaris ventrahs are distributed throughout the folia of the vermis, as seen in figure 7.
It has thus been made clear that the fibers from the legs and lower body wall pass to the vermis of the cerebellum and are distributed throughout the ^'ermis, with the exception of the most anterior and the most posterior folia. A part of these fibers pass through the dorsal cerebellar tract and the inferior cerebellar peduncle to the caudal half of the vermis, while the fibers of the ventral tract pass through the superior cerebellar peduncle to the cephalic half of the vermis.
In the second dog the lesion was in the sixth segment and sections at the level of the operation showed that the entire dorsal cerebellar tract, and a part, if not all, of the ventral tract, were cut. Both above and below the lesion, degenerated fibers were very numerous all through the ventral half of the white columns and along the dorso-lateral margins. There were a few scattered degenerated fibers in the fasciculi gracilis and cuneatus. Above the lesion the degeneration was nearh' identical with that found in the first dog. The degeneration of the cerebellar tracts was again double, though the lesion was confined to one side. There were fewer degenerated fibers on the un-operated side than in the first dog. In the distribution of the degenerated fibers in the cerebellum there was a little farther extension of the fibers into the lateral hemispheres. In the third dog the results were practicalh' the same as in the other two. There was the same bilateral degeneration, less extensive on the un-operated side; the distribution of the fibers in the cerebellmii also showed a slightly greater extension into the lateral hemispheres than is shown in the figures from the first dog.
31^ GILBERT HORRAX
c()^'CLUSIOXS
1 The onlv symptoms caused by a lesion of the spmo-cerebellar tracts ill the dog are those referable to a loss of muscle sense and tone The symptoms were bilateral in all three experiments and there was almost complete recovery in three weeks
9 These symptoms, in a lesion of the tracts at the level o the sixth thoracic spinal nerve root, are confined to the hmd legs and possibly the lower portions of the trunk.
3 The fasciculus spino-cerebellaris dorsalis, so far as its distribution in the cerebellar cortex is concerned, is confined to the caudal half of the vermis, and to the medial portion oi the
lateral hemispheres. , „ • ^ i 4 The distribution of fasciculus spino-cerebellaris ventraiis in the cerebellar cortex is confined to the cephalic half ol the
vermis. „ . ^. ,
5. There is no definite cerebellar center for association regarding the hind legs.
6. The cerebellar tracts are represented fiy crossed, as well
as by direct, fibers in the dog.
.My verv earnest thanks are due to Dr. Florence R. Sabin, whose cooperation and help alone have made this paper possible- and I also wish to express my thanks and appreciation to Drs ' Emit Goetsch, Jacobson, and Hunnicutt, of the Johns Hopkins Hospital, for theh invaluable help in operating on the dogs used.
LITERATI-RE CITED
\rFKH\<-.. L. ISOO Zur Anaiounv .ler Vorderscitenstrangreste. An-l,. f. path \nat. and Phys. imd f. klm. Medicin., Bd. 121.
li.xr Hmhkht lUdC. Experimentelles zur Physiologic der Tractus spmo(•c.iol,cUares. Arch. f. Anat. u. Phys. Abth. , • , ,
li,„K 1) mr, 1907 Das Corebclhun der Saiigetierc. Erne vergleichend ' ■ anatomischo Untersuchung. Erstor und Zwoiter Toil. Potrus ( amper., Bd. 3. Dritter Toil., Bd. 4.
Bruce \ X.niax 1910 The tract of Cowers. Quart. Journ. Exp. 1 hys. Nol-i.
CoLLTKR AND Buz.AUD 1903 The degenerations resulting from lesions of posterior nerve roots and from transverse lesions of th<' spinal .-ord m man. A study of twenty cases. Brain, vol. 2G.
AFFERENT FIBERS OF THE DOG 319
LuxA 190(3 Localizzazioni cerehcUaii : Contributo sperimentale anatomofisio logieo. Ricorche fatto iicl Lah. di Anatomia della R. Univ. di Roma,
torn. 13.
1908 Einige Beobachtungen iibtn- die Lokalisationen des Kleinhirns.
Anat. Anz., Bd. 32. .MacXalty and Horsley 1909 On the cervical spino-bulbar and spino-cerc bellar tracts and on the question of topographical representation in
the cerebellum. Brain, vol. 32. ^loTT, F. A. 1892 Ascending degenerations resulting from lesions of the spinal
cord in monkeys. Brain, vol. 15.
1895 Experimental enquiry u])on the afferent tracts of the central
nervous system of the monkey. Brain, vol. 18. Pellizzi, G. B. 1895 Sur les dcgcncrescences secondaries, dans le systeme
nerveux central, a la suite de lesions de la moelle et de la section de
racines spinales. Contribution a I'anatomie et a la physiologic des
voies cerebelleuses. Archives Ital. de Biologie, torn. 24. Vax Gehuchtex, a. 1904 Le corps restiforme et les connexions bulbo-ccrc' belleuses. Le Xevraxe. tom. 6. Vax Ryxberk, Ct. 1904 Tentativi di localizzazioni funzionali nel cervelletlo.
Archivio di Fisiologia, tom. 1.
1908 Das Lokalisationsproblem in Kleinhini. Ergeb. d. Phys.,
Bd. 7.
320
GILBERT HORRAX
PLATE 1 EXPLAXATIOX OF FIGURE
1
Dorsal view of the medulla and the cerebellum of a dog, to show the blo;ks i^to .iich the cerebellum was cut. The well marked groove be ..^^^^^ the lobus anterior and the lobulus simplex falls m the third block The first and second btocks contain the lobulus medianus posterior; the thud block mXdes lobulus simplex and a part of the lobus anterior, while the fourth block
Tir>liirlp<? the rest of the lobus anterior. rE^;: IpUX a^^^ L I^er^ed ,..r. «Uea .. f,ee hand. TKe H,M
Selellar s ventralis in the medulla. In the cerebellum ,t shows degenerated fibet In one folium of the n.iddle .egion of the lobulus -'-^^J-f ° „^,^, 4 Section through the medulla and cerebellum of Dog 1. taken t"'""?" the caudal part of the second block. X 8. It shows the separation o the fibers of the' corpus re.,tifornre from those of the fasciculus sptno-cerebellarrs
"f tctSnThrough the medulla and cerebellunr of Dog 1, taken through the ceohalk end of the second block. X 8. It shows degenerated fibers of the Lft ?ltuTu"pino-cerebeUaris dorsalis, entering the lobulus med.anns posterior
°' rSectfonthrough the medulla and cerebellum of Dog 1, taken through the Cauda end of the third block. X 8. The section is so near the hne shown on figure 1 that the section of the pons is incomplete ; it show s the .obulus simple v 7 Section through the pons and cerebellum of Dog 1, taken througn the cephaTc end of the third block. X 8. It is above the level at which the o pns'restiforme enters the cerebellum and shows the f-"™'-f^--f„7,t lari^s vcntralis in the edge of the brachmm eonjunctivum and the fibers of same tract in the lobus anterior of the vermis.
AFFEHEXT FIBERS OF THE DOG
GILBERT HORRAX
PLATE 1
/
^- C\ r-)
I
^^vV,
^
r
_!>
L
I
321
SOME NEW re(;eptacles for cmda^^ers and
GROSS PREPARATIONS
RALPH EDWARD SHELDON
From the Anatomical Laboratories, the School of Medicine, C'nlrerslty of Pittsburgh
EIGHT FHiURES
RECEPTACLES FOR CADAVERS
lustituticjii.s whicli do not possess cold storage facilities usually keep cadavers in tanks of various kinds. These may be of metal (galvanized iron), wood with a metal lining (lead, zinc or coppei'). or of concrete. The fluids used ai'e such, however, that the galvanized tank rusts through in a comparatively shoi't time, since in most cases it cannot be protected by paint, on account of the solvent power of the alcohol and carbolic acid used. The stock is usually thm, necessitating the use of angle irons or planks in order to prevent bulging. Also on account of this weakness, if the tank contains material, it may not 1)6 moved without injury to the bottom. The lined tanks, which must be soldered at the corners, likewise eventually leak, the solution then making its wa}' into the spaces between the lining metal and the wooden supjiort. The lining is I'arely smooth, considerable dirt accumulating, therefore, in cracks and uneven places. On account of these factors, the tank becomes foul and undesirable in a well-kept laboratory. Concrete tanks properly built and lined with cement are satisfactory, excepting that they cannot be moved and that they take up a large amount of space.
After some experience with various types of receptacles, the tank described below was designed for the Anatomical Laboratories at Pittsl)uigh. It has now been in use for more than a year and has given perfect satisfaction. Essentially it consists of a box built of two-inch cypress with half-inch bolts running in the wood, horizontally through the bottom and vertically in the sides and ends. The individual pieces of wood are fitted together as shown in figure 1. By this method of construction a solid and exceedingly strong box is obtained. If any leakage occurs through excessive drying, it is necessary only to tigliten the nuts on the bolts. The case is further strengthened by two angle irons (d, figs. 2 and 4) running lengthwise along the bottom at the corners. Through these pass the horizontal bolts (a. figs. 2. 3 and 4), also the vei'tical bolts (c, fig. 3). The ends of
324 RALPH EDWARD SHELDON
all other bolts pass through bars of iron one-half mch by two inches; e, figure 2, for the horizontal bolts, b, figure 2; /, g and h, figures 2 and 3, for the vertical bolts c. These strengthen the case and prevent heads and nuts from being drawn into the wood through its swelling or the tightening of the nuts. Two strips of oak one inch by six inches {i, figs. 2 and 3) are placed lengthwise under the tank in order to permit the floor underneath to be cleaned. These are removable so that they may be eventually replaced. It will be noted that the}" do not come in contact at any point with the wood forming the floor of the tank, thus preventing its rotting. The top is hinged and consists of one-inch cypress, properly battened. A gate valve for emptying the tank is desirable. Hot oil should be applied to the raw wood unless it is to be painted.
While this tank is moderately hea-vy, it can be easily moved when emptj", or even when partially' full, as it is not injured by the use of levers. Even when full of solution there is no bulging of the sides. Since all parts are tightty fitted, it may be used for the storage of cadavers in alcohol fumes. In fact, such a tank has been thus used here for nearh' a 3'ear, the material keepmg perfectly. Such a receptacle, six feet, six inches long by two feet, ten inches wide and two feet, eight niches high, inside measurements, will hold fifteen cadavers of average size. It is practically indestructible and should last indefinitely. The first cost is sixty dollars.
RECEPTACLES FOR GROSS PREPARATIONS
Large gross preparations, particular!}^ dissections, are ordinarily kept in tanks smiilar in structure to those mentioned for the storage of cadavers, although often smaller in size. Such tanks are subject to the same disadvantages as when used for cadavers. In addition, they are usualh' unsightly in appearance and therefore undesirable in laboratories and museums. Glass jars of sufficient size for large human preparations are very expensive and easily broken; earthen crocks are likewise very fragile. It is not feasible, moreover, to use either of these for large preparations, such as longitudinal sections. The receptacles described below are designed to serve as museum cases for large specimens and as storage receptacles for material wliich it is desired to have constant!}'" available for students or members of the instructing staff'.
These cases are constructed on essentially the same plan as the cadaver tanks. Lighter material, however, may be used in their construction, since the cul^ic contents is considerably less. The cy]3ress used is somewhat thinner, the bolts are three-eighths inch instead of one-half inch, while the angle irons and iron bars are omitted entirely, washers being used instead. The tank proper (j, fig. 6) is placed on a removable base (/.-, fig. 6) in order to make it of the proper height for use. The top, which is largely of glass, is sloped foiwai'd in oi'der to give a good view of the contents when the case
for Cad5ver5
^t
^P
h
n— ^:^ =^
^"-^+7--^r
9 b
w
□
tt
i
1
^
JX ^
rB=3s
5 5ide Lleve^tion.
10 1
"niT
Tn"
C
iPJi
nil
"W
r
4 Pla^n of b5.5e 325
Tor Gross PrepQ.r5.tion5.
5. Tront Ll€v5,tion.
jL
8. Plo-n of b^-se. 326
-*
-^
S
NEW RECEPTACLES FOR CADAVERS 327
is used for niuseuiu preparations. The groove (/, fig. 7) runs entirely around the case: into it projects one arm of an angle iron (m). When the groove is filled with cotton the case is made practically air-tight so that there is almost no evaporation where it is desired to keep material in alcohol fumes. The catches used are Corbin tool box locks number 1217 in, fig. 5). By the use of catches of this t3'pe, it is possible always to draw the co\-er tight and hkewise lock the case if desired.
At Pittsl)urgh. we have placed cases of this type in tlie dissectingroom, where we have found them very useful for longitudinal sections of cadavers and for large gross dissections kept as museum pi-eparations. One case is used for the best dissections made each year b}the class, which are thus available the succeeding 3^ear as demonstration preparations. Cases of this type, two feet two and a half inches wide, eight feet long, one foot high in front, one foot, six niches in the rear, inside measurements, with a base, maj- be secured finished for forty-two dollars each.
Acknowledgments should be made to Mr. E. B. Lee. architect. Pittsburgh, for the prelhninary sketches. It should also be stated that the idea of using ])olts in the manner indicated above was first suggested to me b}' some small wooden cases which I saw some years ago in the Anatomical Laboratories at the L'niversity of Pennsylvania.
INCREASE IN PRICE OF JOURNALS
In order to extend and improve the journals published by The Wistar Institute, a Finance Committee, consisting of editors representing each journal, was appointed on December 30th, 1913, to consider the methods of accomphshing this object. The sudden outbreak of European misfortunes interfered seriously with the plans of this committee. It was finally decided, at a meeting held December 28th, 1914, in St. Louis, Mo., that for the present an increase in the price of these periodicals would not be unfavorably received, and that this increase would meet the needs of the journals until some more favorable provision could
be made.
This increase brings the price of these journals up to an amount more nearly equal to the cost of similar European publications and is in no sense an excessive charge.
The journals affected are as follows:
THE AMERICAN JOURNAL OF ANATOIVIY, beginning with Vol. 18, price per volume, $7.50; foreign, §8.00.
THE ANATOMICAL RECORD, beginning with \^oL 9, price per volume, $5.00; foreign, $5.50.
THE JOURNAL OF COMPARATIVE NEUROLOGY, beginning with Vol. 25, price per volume, $7.50; foreign, $8.00.
THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY
36th Street and Woodland Avenue Philadelphia, Pa.
VASCULARIZATION PHEXO^IEXA IN FRAGMENTS OF
EIMBRYONIC BODIES CO^IPLETELY ISOLATED
FROM YOLK-SAC BLASTODERM
FRAXKLIX PEARCE REAGAX
Department of Compnrallvc Anatomt/, Princeton Vnirersity
TEX FIGURES
Recent experimental work points to the possibility of a final solution of the j^roblem of the origin of intra-embrj-onic vascular endothelium. Methods of procedure affording results to which there can be no doubt of interpretation are especially desirable.
So far there ha\'e been developed two methods by which yolksac ^\ngioblast' may be kept out of communication with intraembryonic vessels: mechanical separation of the vessels of these two regions, and exposure of the de\eloping embryo to anesthetics. The former method was employed to the extent of partial separation by Graper,^ Hahn,- and Miller and McWhorter.^ These observers have obtained endothelium on both sides of chick embryos in which one side was severed from extraembryonic blastoderm. The second method has been perfected by Stockard^ who has, in cases of arrested development, been able to secure intra-embryonic endothelium ({uite independent of that in the yolk-sac.
The intra-embr^'onic vessels in the experiments of Miller and ]\IcWhorter were in part non-continuous and somewhat diminutive in size; this is ])erha])s due to the fact that heart-pressure is necessary for the normal growth of endothelium even after the latter has been established. The work of these two ob 1 Griiper, L., Archiv fiir Kntw. niecli.. Hd. 24. 1907. -Hahn, H., Archiv liir Knlw. mcch., lid. 27, 1909.
^ISIiller, A. ^\.. and .McWliorter, .J. K.. Anat. Roc, vol. 8, p. 91, 19U. ^ Stockard. (\ R., Proc. Am. Atssn. Anatomists, Anal. Rec, vol. 9, no. 1, 191.5.
^29 THE .VX.VTO>nCAI. HECOHI), \ Ol.. !l, Nil. 4.
330
FRANKLIN PEARCE REAGAN
servers was submitted as proof of the local origin of mtra-embryonic endothelium; previous toand following its publication, this work was objected to quite vigorously on the following grounds: the incision may not have l^een made sufficiently early or sufficiently close to the embryonic body; vessels may have grown into the injured side from the unoperated side or from either end. To the latter objection Miller has replied that this unusual growth would require a permeation of such sohd structures as notochord and neural tube. A later examination o Miller's material by Bremer, as well as ^filler's own careful reconstructions, failed to reveal such ingrowths.
While the work of Miller and ^IcWhorter seems m itselt to be quite conclusive, a confirmation of their results by more rigorous methods of experimentation may not be superfiuous. A most feasible method of procedure seems to be that o complete separation of an embryonic body, or a portion thereof, from the extra-embryonic blastoderm prior to a possible invasion by the so-called yolk-sac 'Angioblast.' If in such a meroplast there should develop legitimate vascular cavities possessed of good endotheUum it is confessedly futile to argue further for the necessity of ^\ngioblastic' origin of intra-embryonic en dothelium. . , u • i.- ^
The following experiments meet, I believe, the objections urged against the work of Miller andMcWhorter, supplementmg at the same time the work of Stockard. . - ,•
About forty chick embryos corresponding to stages 4, o, b and 7 of Kiebel's Normentafel (Zweites Heft) constitute so far the material studied. The operations consist in varying degrees of separation of the projecting head from the remainder of the embrvonic bodv and the blastoderm.
On; experiment which I shall designate as Type I consisted in the following incisions (fig. 1): longitudinal incisions lateral to, but close to the projecting head on each side, extending trom points slightly posterior to the anterior intestinal portal to the opaque area anteriorly; a transverse incision through the embryonic body just posterior to the anterior intestinal portal, l^ollowing this the blastoderm was entirely removed except the
VASCULARIZATION PHENOMENA
331
k
E-l
Fig 1 Diagram to illustrate the blastodermal incisions in experiments of lypes I and 11. Broken lines represent the incisions in Type T. Dotted line - to //indicates the transverse incision in Type II -n uhich incision ^ to F is omitted.
Ft^- 2 Lateral view of the head-fragment of Tvpe I pushed back to be severed trom tne proamnion at point where a broken line intersects.
Fig. .3 Section through the anterior portion of the forebrain of a head-meropiast, showing unusual head-coelo:n. Total incubation thirty-two hours. Operation at the time of the first intersomitic groove ( X 160). Experiment, Tvpe i. no. 19; b, aniage of ventral aorta; c, coelom; d, pharvnx; e, forebrain.
332 FRANKLIN PEARCE REAGAN
small strip of proamnion which suspended the head-fragment from the opaque area anteriorly. (In this way it was possible to section later the blastoderm and determine the status of vascular development in the extra-embryonic area and Jo hww certainly that the embryo had not yet been vascularized). Ihe head-fragment was then pulled backward and turned so that the proamnion could be snipped off close to the head-tissue proper (fig 2) leaving a free meroplast which would sink to the bottom of the sub-germinal cavity. The egg was then sealed and inculmted further. -n «•
\lthough manv other methods were utilized, it will suthce for the present work to describe one more experiment— Type II Longitudinal and transverse incisions were made as m Type I, except that the transverse incision extends entirely across the l3lastoderm. The blastoderm was removed to be sectioned while the head-fragment was left connected anteriorly with the opaque area bv the small strip of proamnion. Thus the headfold rested on a double membrane of ectoderm and entoderm which was also excised laterally but not anteriorly from the
opaciue area.
Complete separation of the posterior region proved to be quite unsatisfactorv owing to the circumstance that the reliet of normal surface tension induces abnormal conditions m the embrvonic body. In the projecting head-region surface tension evidentlv does not enter so extensively into the mechanics ot development. Furthermore a relatively small anKumt of m.iury in the head region serves to isolate C()m])lotely an embryonic fragnienl in which development proceeds in a surprisingly
normal manner.
In order to preclude the possihiHty of clrawmg conclusions tr„ni tissue which had already been 'invaded- by 'Angioblast the remainder of the blastoderm which had been ivmoved at the time of operation was sectioned. The region oi the embryonic bodv which would have first been vascularized was contained in ihe axial ])ortion of such blastoderms. Beiore each incision the instruments w.mv sterilized. These precautions together with that of complete isolation should render the
VASCl LAKIZATION PHENOMENA
333
procedure sufficiently rigorous to satisfy all reasonable demands of experimental jiroof.
The tissue was fixed in a picro-acetic mixture; sections were stained in a modification of :\Iann's methyl blue-eosin stain -^
I'lR. 4 Section through the forebraiu of a head-meroplast, sliowiiif. earlv stages in the formation of vasofactive cells. Total incubation, twentv-nine hours: Operation previous to the formation of the first intersomitic groove (X ■-'00). Experiment, Type I. no. 24; a. prevascular mosenchvme; r. coelom : d, pharynx; e, forehrain.
which i)roved especially \-aluable in the differentiation of endothelium.
' lieagan, V. P., Anat. Rec, vol. 8, no. 7, 1914.
334
FRANKLIN PEARCE REAGAN
When the head-fragments as above described had been incubated for a total period of from thirty to forty-eight hours and then sectioned, thev were found to possess blood vessels in varying degrees of development. In general it may be said that regardless of the amount of incubation beyond a total period of thirtythree hours, differentiation never proceeded beyond the normal stages of differentiation at that age. After forty-eight hours some signs of degeneration made their appearance. It seems that the embryonic meroplast possesses an inherent capacity for differentiation which tides it over to the time when heartpulsations would normally provide a means of tissue respiration. ^^'hile differentiation proceeds always at the same rate and to the same extent, growth varies greatly. Meroplasts equally differentiated may vary greatly in size.
Practicallv the onlv unusual condition met with in these headfragments i; the presence of a head coelom (fig. 3), the origin, fate and significance of which will be considered later.
Between the base of the coelomic pouch and the pharyngeal entoderm rounded or cuboidal cells become proliferated. Tbeir point of origin is in most instances between the base of the coelomic pouch and the pharyngeal entoderm (figs. 4, 5 and 7), where it is difF.cult to determine which of these two epitheha is ot primarv importance in such cell proliferation. C^ells of this sort are midoubtedlv proliferated to a certain extent by the pharyngeal wall and also by mesothelium. In neither of these latter cases do the cells originate by foldings and constrictions of cell aggregates, but singly or in a linear proliferation This same region is found, in somewhat later stages (fig. 4) to be occupied bv irregular stellate cells resembling mesenchyme cells. Their processes fuse forming a parenchyma-like complex which merges dorsallv into the true interstitial mesenchyme
Simultaneously with the accumulation of a plasma-hke fluid which may be detected in this parenchyma by sections of its coagulum, the loose structure becomes transf.M-med into a longitudinal tubeof endothelium (figs. 6 and?) . The tubes thus formed, though far anterior to the heart-region, may simulate heart-formation in a remarkable manner. The coelomic ].ouches may meet to
VASCULARIZATIOX PHENOMENA
335
possess a continuous ca\-ity, a condition approached in figure 6. Since the phar3^ix in this region is already tubular it does not become constricted in this process, the endothelial tubes meeting ventral to it. There is no reason to believe (though such an inference is possible) that the endothehal tubes thus formed are new or abnormal formations: they occupy the position of normal ventral aortae.
Fig. 5 Section through the forebrain oi a head-meroplast showing a loose parenchyma in a position occupied by the isolated vasofactive cells of figure 4. Total incubation thirty hours. Operation at the time of the first intersomitic groove (X 150). Experiment, Type I, no. 18; a, prevascular mesenchyme; c, coelom; d, pharynx; r, forebrain.
The conditions so far described are found in embryonic fragments of T}TDe I which were incubated not more than a total period of thkty-three hours. It will be well to consider some of the conditions found in a meroplast of Type II incubated for
tl—
7 ^>
Fio- ,i S..r.iun ll,n.U^h -1,.. lorrLran. ..1 a l.oa.l-nu-n.plast sl.nuin^ anlaRcn Of thrven.ral a-Hao as disn-eto ves.ols. The longitudinal m<.,s>on "" th. n, > side of th,- iH.a.l was n>lativoly c-los. to the neural fold Ihe cu edges he
nharynKoal<-n,„.lennhave l.een pulled apart, the ventral tissues having suung I tlUrt. Total uu-uha.ion thirty-.wn hours. ..perat.on at *»- t-e " »,c seeond intersount ,<■ groove (X 210). lX,.eruuent . I vpe 1. no. .,1, h. ^.nt,al aorta; c, eoelom; <L i)haryii\; r, forehram. .
Fig. 7 Section through the lorebrain of a head-n,eroplas, «»---"^/-;; Wolined ventral aorta. The right longitudinal ineisu.n was elose ^^^^l^l^] fold. .None of the exeised head has regenerated: the rnesenohynie s very com pact near the cut surface the- cells of which aresomewhatepahehal. 1 "t.d m^^a tion thirty-three hours. Operation at the tin.e of the u-st '"<— ^Z^^;;; (X 170). Type I, no. 17; /,. ventral aorta; c, eoelo.n ; ./. pharynx. <. loiebra.n.
33()
VASCULAKIZATIOX PHENOMENA
337
a total period of forty-eight hours (figs. 8, 9 and 10). No attempt will be made at present to set forth the processes which inter\-ene l^etween the stages of thirty-two and fortv-eight hours.
The ]:)roanmion (fig. 8), a region normally devoid of mesoderm, contains a rounded jiouch which is continuous anteriorly with the extra-embryonic coelom. The cut edges of the ectodermal and entodermal layers of the proamnion have fused forming a blind sac around the enclosed coelomic pouch; likewise the cut
Fig. S .Section throiigli the t'orebrain of a hcad-merophist .showing a proainn'otic sac containing a pouch of coelomic mesothelitna. On the ventral side of the sac is a peculiar proliferation of entodermal cells very constanth' ap])euring in experiments of this type, generally more symmeti'ically situated. Total incubation, forty-eight hours. Operation v\ the time of the second intersomitic groove (;< 55). E.xperiment, Type IT, no. 3 r, coelom; e. forebrain; C, ectoderm; f/, entoderm ; J, extra-embryonic vessels.
edges of the once continuous blastoderm have fused. On the ventral surface of the proamniotic sac will be noticed an entodermal thickening, in appearance not unlike an in\'erted nem-al groove. This structm'e is quite constant in experiments of this type. I would interpret it as a cell-complex representing potentially the floor of the fore-gut in case of normal infolding. The transverse incision in this experiment was made some distance behind the site of the anterior intestinal portal, so that there projected behind the posterior extent of the proamniotic
338
FRANKLIN PEARCE REAGAN
sac a portion of the embryonic bod}' bounded dorsally bj' ectoderm and ventrally by entoderm (fig. 9). The cut edges produced by the longitudinal incision have fused, the point of fusion being indicated by the apex of the projection on the left side in figure 9. In this figure we have a photograph of a 'projecting' head; the section passes through the anterior part of the midbrain region, presenting a rather puzzling condition in that it is entirely devoid of fore-gut. Two large dorsal aortae are present; the larger one is located on the side containing the greater amount of mesenchyme; correlated also with the fact that the
Fig. 9 Photograph of a section through the midbrain of the same meroplast as in figure S. show'ng we!! developed dorsal aortae and the absence of a tubular pharynx in a tubular head. Fusion of entoderm and ectoderm at points indicated by X (X 120). /, ectoderm; g, entoderm; /(, dorsal aorta; i, midbrain.
incision on this side was made at a greater distance from the median line. Both aortae are bounded by distinct and unmistakable endothelium. In the aortic cavities blood plasma has coagulated. No coelomic pouches are present in this section. Pleart-formation has not taken place in this particular experimental casei
The phenomena under consideration are not to be regarded as regenerative changes. The head does not regenerate the yolk-sac, neither does the yolk-sac regenerate the head. In figures 6 and 7 it will even be seen that portions of the head itself were not regenerated. There is a genesis of the first order in
VASCULARIZATION PHENOMENA
339
case of the. development of endothelium. While it is not always possible to be sure of the extent to which experimental conditions portray a truly normal process, the results here presented, together with those produced by Stockard seem to afford positive evidence in favor of the local origin of blood vessels.
It is of interest to note the statement of Bremer (Am. Jour. Anat., vol. 16, no. 4, p. 463) that mesothelial anlagen "might arise under abnormal conditions in positions where they are normally absent." The isolated cells proliferated from this
Fig. 10 Photograph of one of tlie hrst avaihxble sections of the bhistoderm behind the incision G to H of Experiment Type II, no. 3, showing the freedom of the pellucid area from endothelium (X 160).
imusual mesothelium are, according to my observation, not comparable to the gross infoldings of cell-aggregates described by Bremer, though they seem to be vasofactive in nature. To designate certain of these proliferated cells as mesothelium would be as uncalled for as to designate others of undoubted entodermal origin as pharynx. While I do not question the ability of mesothehum to proliferate pre-vascular mesoderm (indistinguishable from mesenchyme), I do wish to question the justice with which Bremer would accredit mesothelial tissue with the production of the entire vascular tissue. An assignable reason for conferring this distinction on mesothelium might be the desire to maintain for endothelium a monogenetic origin — the first requisite to the specificit}' of a tissue.
The fact that endothelium exists in the sauropsidan yolk
340 FRANKLIN PEARCE REAGAN
sac prior to the establishment of a coelom cannot be satisfactorily explained by a hypothetical ■"premesothelial stage of mesoderm" (Bremer, loc. cit., p. 4(53). It has been shown that 'Angioblast/ so far from l^eing a daughter-tissue of premesothelial mesoderm, is really a parent-tissue of the latter; in isolated blood islands Riickert*^ has shown that cell groups proliferated from this early vascular tissue cleave to form slit-like cavities which unite later with other similarly formed cavities to contribute to the extra-eml)ryonic coelom. Of logical necessity it follows that mesothelium and 'Angiol)last" (in the sense of His) must have come from a common cell-complex.
Should we ever be so fortunate as to find the ultimate font and source of all vascular tissue there would be no objection to its designation as 'Angioblast,' so long as we bear in mind the original implications of the Angioblast Theory: the essentials of this theory have ])een outlined by Minot (Human Embryology, Keibel and Mall, vol. 2, pp. 498-99) as follows:
Comparative embryology teaches that the first bloodvessels appear on the 3'olk-sac collectively and at one time. They form a unit anlage which wo call angioblast according to the suggestion of His. * * * The angioblast probably maintains its complete independence throughout life. In other words it is prol^ablc that the endothelium of the l>lood vessels (and of lymph vessels) and the blood cells at every age are direct descendants of the piimitive angioblast.
The fact that allantoic vessels may ai)pear j)rl<)r to those in the yolk-sac in early human development (Bremer, ibid.) is probably an expression of the tendency towards that sequence of local origin which is correlatf^d with the functioning of the part \'asculai"ized. A diffei'ent se(iiience is found in animals whei'e the allaiitois functions relatively later.
When it is stated that endothelium differentiates fi'oin an 'indifferent' mesenchyme the specificity or non-specificity of the immediate source of the pr(>-\'as(*ular mesenchyme is not
Ivi'ickcrt, .1., I'!tit\\ ickcliiufi dci' cxI r;i-('inl)r_\<)ii:il(ii ( id'asse dcr \'<ificl. IlaiidhiK-li der N'erfil. ii. exj). Entw-lehrc. IM. I. l". 1. I'.KKi. ("cbor die Ahstaimimiin dcr hIuth:ihig(Mi (IcfiissmiluKon Jx-ini Huliii. uiid iihcr dir iMit.stchuiig des Handsinus bcMrn llidiii uhd l>ci Toipfdo Sit zinitislxM-. dcr lia.w Akad. Wiss.. ino.i.
VASCrLARIZATIOX PHEXOMEXA 341
brought into question; the statement merely means that in their earUest stages such cells are morphologically indistinguishable from other mesenchyme cells which may or may not be capable of like de^'elopment. Whether each vasofactive cell behaves as it does in accordance with a definite teleology is at present entirely beside the question.
Like considerations apply to the study of 'indifferent' capillary plexuses, some of the meshes of which ]3ersist while others degenerate. That one should observe this process from time to time, and from the mere fact that the process takes place, be able to determine the extent to which the changes are due to heredity or mechanical influence is to be accepted with some caution; cell organization is excluded with great difficulty. In a regenerating jugular vein Clark' believes he has a case in which heredity plays no part. It may be stated that authorities on regeneration do not usually exclude considerations of heredity from the conclusions at which they arrive. Certain it is that the development of the vascular system furnishes an unproductive field for the solution of the problems of preformation and epigenesis.
In conclusion it is well to consider the following recently established facts which should share in defining our morphological interpretations. The yolk-sac is not necessarily the site of formation of the earliest blood vessels. Intra-embryonic vessels de\-elop in .situ when communication of extra-embryonic vessels with intra-embryonic tissues is prevented \)y chemical or mechanical means.
•Clark E. R.. Proc. Am. Assn. Anatomists, Anat. Ket-.. \ol. !i. no. 1. litl.T.
M E M O I R S
The publication of this series of Anatomical Monographs has been undertaken with the purpose of presenting the results of original investigation in anatomy which are too extensive for incorporation in the already over-crowded current periodicals.
No 1 The Anatomy and Development of the Systemic Lymphatic Vessels in the Domestic Gat, by George S. Huntington, Professor of Anatomv, C olumbia University, New York City. This monograph states m a few pages the various theories held in regard to lymphatic development in general and then proceeds to present the result of six years' careful investigation on mammalian lymphatic development. Part I deals with the development of the svstemic lymphatic vessels in their relation to the blood vascular system. Part II deals with the development of the preazygos and azygos segments of the thoracic duct. This memoir contains 17o pages of text, 8 text figures (two in color), 254 photomicrographs and 21 colored plates. Sent post paid to any country for S4.03
No 2. Contribution to the Study of the Hypophysis Gerebri with Especial Reference to its Comparative Histology, by Frederick Tilnev, Associate in Anatomy, Columbia University, New York City Part I contains a historical review of the literature. Part II deals with the comparative histologv of the pituitary gland and gives a report of six hypophysectomies performed upon cats. This memoir contains 72 pages of text, 2 text figures, 60 photomicrographs and plates. Sent post paid to any country for $1.50.
No 3. Early Stages of Vasculogenesis in the Cat (Felis Domestica) with Especial Reference to the Mesenchymal Origin of Endothelium, by H. Von W. Schulte, Department of Anatomy, Columbia University, New York City. This memoir contains 90 pages of text and Si figures of which 14 are in colors. Sent postpaid to any countrj. for Sl.oO.
Orders should be addressed and checks made payable to
THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY
Pliiladplpliirt, Pa.
THE IDEXTIFICATIOX OF TISSUES IN ARTIFICIAL
CULTURES
E. D. CONGDON
From the Medical Department of Leland Stanford Junior University, Calif oryiia
TEN FIGURES
In spite of the numerous studies upon tissue culture the usefulness of the method is still limited greatly by uncertainty in classification of many cell tj^pes most vigorous in their growth and of commonest occurrence. The nerve cell, the thyroid parenchyma, kidney tubules, ectoderm and a few other tissues are easily identified. But the ever varying linear, reticular and epitheloid growths are of undetermined origin. Although they are the predominating form in cultures they have only been referred provisionally to the group of embrj^onic sustentative tissues.
The structure of the growths is of only secondary importance in determining their classification. While it has by no means been established that tissues take on a more embryonic character in the plasma it is true that with few exceptions they assmne the form of cell strings, reticula or membranes. Thus even the tubules of embryonic kidney are described by Lewis and Lewis ('12 b) as growing out in the form of a loose mesh. Much can be accomplished in the wa^^ of classification by comparing the growths from various organs and drawing inferences based on their tissue composition. In this way the probability has already been established that many of the common growths are derived either from supporting tissues or endothelium. Yet the character of the growths is too variable and too little under control to arrive at a final classification b}' indirect methods. They must be traced directly to their source in the parent tissue. For this purpose sectioned cultures are necessary. Little direct
343
THE ANATOMICAL RECORD, VOL. 9, KO. 5
344 E. D. CONGDON
evidence of this character has been obtained because up to the present time whole preparations have been used to the almost complete exclusion of sectioned cultures.
In the present study sectioned and whole preparations were used to supplement one another in identifying the growths. Cultures were made from chick ventricle of ages rangmg from four to eighteen days. Lunb buds of from four- to seven-day embryos were used. A much smaller nmiiber of series were made from liver and intestme. The comparison of the growths from the younger and older organs disclosed certain differences dependent upon the histogenetic stage of the organs. These are considered briefly.
The cultures were made in plasma according to the procedure of Carrel and Burrows ('11). The description of this method has been given too frequently to require repetition m detail. The plasma was taken from young chickens varying m age from two weeks to four months. Most vigorous growths took place in clots from a mixture of two parts of plasma to one of distilled
w^flter
The preparation of cultures for sectioning presents considerable difficulty due to the marked tendency to shrink shown by the plasma clot and the very watery cells of the growths. Experience shows that care in adapting the technique of fixation and imbedding to the peculiarities of this material is especially
worth while. . ,
A brief description of the various types of heart ventricle growth will be given before considering the evidences as to their classification. The cultures from embryos of more than five days' incubation are divisible into a number of regions, four ot which have a concentric arrangement determined by the rounded surface of contact between plasma and tissue. These are not well-defined in preparations of four- and five-day tissue because of the flowing and distortion consequent on the fluidity o the early embryonic tissue. Figure 1 shows them diagrammaticaliy in a section of a culture cut parallel to the cover-slip. The central zone a is made up of tissue that remains for the larger part inactive. There is no evidence for the migration of any ot
IDENTIFICATION" OF CULTURES
345
its cells outward although it is not possible to prove that some few do not leave the region. The boundary of the inactive zone does not become well-defined until degeneration has begun and the tissue external to it has become modified by the migration outward of its cells. Degeneration is often found in eighteento thirtj'-six-hour cultures to be confined to a central area much smaller than finally occupied by the inactive zone. Evidently death takes place first at the center of the tissue because of its remoteness from the plasma. The limits of the dead region then extend gradually toward the surface of the tissue. The inactive zone shows pyknosis and chromatolysis of the nuclei
Fig. 1. Diagrammatic section of ventricle culture; a, inactive zone of implanted tissue; h, active zone of implanted tissue; c, region of reticular growth; d, cover-slip sheet; e, covering layer.
of heart-muscle cells. In two-week-old ventricle there is a clumping of the cytoplasm into large masses staining with basic as well as acid dyes. Epicardial and peritoneal endothelium have greater vitality than the heart-muscle. This also may be said for the endothelium of sinusoids except where the breaking down of the erythrocytes brings injury to the contiguous wall. Surrounding the central zone except upon the cover-slip side, is a peripheral active region from which cell migration into the plasma takes place (fig. lb). It is never many cells thick and does not necessarily include all of the living tissue if the period of incubation of the culture has been short. In cultures of pulsating heart segments the cells contained in the active zone are put on the stretch at every systole. Fixation often causes the heart tissue to contract and thus preserve them in the condition of extension. Cell debris is usually present at the line of contact between tissue and plasma as a result of cutting the tissue from the parent mass.
346 E. D. CONGDON
Aside from a very fine, often degenerate reticulum, the remaining growths lying free in the plasma can best be described under two types, one fine and the other coarse. Neither of these give evidence of being separated by cell walls when stained with iron hematoxylin and erythrosin after Zenker or osmic acid fixations. There are all intermediate forms but many of the cultures are predominated by the one or the other type. Of the two kinds the finer ismuch more abundant. The nuclei in both varieties contain the one or two chromatic masses usually to be found in embryonic chick tissue, A measurement of the nuclei shows no constant difference in diameter from growths of heart-muscle, endothelium or reticular tissue. The cytoplasm of all cells in the plasma is watery and coagulates into a loose foam structure not resembling heart-muscle substance.
The finer mesh crosses the field at all angles. It differs from the coarse reticulum primarily in the much greater independence of its cells. It is made up of two intergrading cell forms of which one is elongated, slender and cylindrical while the other is polyhedral and usually triangular or quadrilateral in optical section. The ends of the first t^^pe and the angles of the other are drawn out into longer or shorter filaments. The cylinders may not be more than a micron in diameter although they are many micra in length. Their nuclei are forced to take on a rod shape by the limited diameter of the cells. The contact of the cells is at all times slight and often made only by the most delicate of filaments. The free ends at the periphery of the mesh send out fine pseudopodial processes. The two-cell forms intermingle freely in the reticulum.
The coarse mesh in all but the four-day cultures consists of bands 2 to 6 m in width. They tend to flatten in the plane parallel to the cover-slip although they connect with each other at all levels. The growth has flowing outlines and forms loops which are characteristic in appearance and include spaces of more constant dimensions than found between the elements of the fine mesh. Nuclei are occasionally found side by side in the broader bands. The diameter of the narrower bands are no greater than usual for the fine mesh. In the nodes of the
IDEXTIFICATIOX OF CULTURES 347
coarser growth several nuclei occur all of which are in a plane parallel to the cover-slip. The triangular intersections of the strands 2 ^x in diameter contain only one nucleus.
A sparse reticular growth of very fine texture often occurs in cultures of ventricle which have been injured in handling. By the use of dog plasma as a culture medium a sunilar growth is obtained. Both elongated and polyhedral cells are present in the fine mesh. The former are frequently drawn out into filaments not more than half a micron in diameter. The polyhedral cells also may be extended into such long threads at their angles that the central portion is much reduced in volume. The filaments are often tortuous and enlarged at successive points to give a bead-like appearance. The more normal part of this growth stains faintly. In other regions there is pyknosis and chromatolysis of the nuclei. The peculiarities of the growth are plainly an expression of decreased vitality and in many cases of actual cell death. In some instances it is the action of plasma from alien species that causes the injury. When the plasma is not responsible the growth comes from the tissues injured in cutting and handling. In this case toxic substances from the degenerating cells probabl}^ not only act upon the growth before it reaches the plasma but do harm by diffusing into the culture medium. Lambert ('12) describes a similar sparse and delicate growth from cultures of chick heart in rat plasma.
The cover-slip growth (fig. 1 d) is of almost as constant occurrence in vigorous cultures as the reticulum. It is so frequently in plain continuity with the latter that no doubt of the identity of the two is possible. Sections of many cultures made perpendicularly to the cover-slip show the reticulum grading into the cover-slip membrane and frequently stained whole mounts enable one to trace the mesh into the sheet growth. The continuity' occurs most frequently close to the tissue where the growth is most crowded. It is often possible to make out a progressive flattening of the bands of the reticulum as they approach the cover-slip. Close to the tissue the cover-slip growth may be many cells in thickness. Tracing outward,
348 E. D. CONGDON
however, it soon flattens out into a single laj^er. The elements are often much flattened, especially at the border of the sheet and the nuclei may attain an area in this plane ten times as great as its usual cross section. The membrane growths agree with the reticular formation in the apparent absence of cell walls. While they have every appearance of a syncytial structure a final decision is impossible because a silver nitrate test for intercellular cement substance was not made. Aside from any differences as to cellular independence that may exist in various growths depending upon the greater or less development of cell walls it is certain that membranes show marked diff'erences in this respect depending upon the extent of the spaces between the cells. The cover-slip growth associated with the finest mesh is made up of cells separated from each other by wide intervals of plasma and seldom intercommunicating by more than a few slender threads: The finer normal reticulum when flattened upon the cover-slip shows a frequent union by broad bands but intercellular conamunication may also be only by slender processes (fig. 6). In the membrane growth from the coarse reticulum broad attachments and multinuclear sheets occur.
The difference between the cover-slip growths corresponding to the coarse and fine mesh is especially noticeable at the border of the sheet. The former sends out cell bands which either project radially or form loops. The border of the membrane associated with the other mesh has a more finely broken border although its cells may also be connected with one another by their slender processes to form loops (fig. 6).
The third region of growth (fig. 1 e) of ventricle cultures is in the plasma next to the surface of the tissue. Burrows ('11) mentions its occurrence in earl}^ embryonic chick preparations but few others have referred to it. It is not confined to the earher embyronic growths but is well developed from two weeks old ventricle. It has doubtless failed to attract atcention because it can not be seen so well in whole preparations as in sections. It consists of flattened cells which are often piled upon one another to a great depth. Tangential sections from the surface of the implanted tissue opposite to the cover-slip
IDENTIFICATION OF CULTURES 349
often show them to be as thin and sheet-hke as at the border of cover-shp membranes. They are often elongated and rounded in cross section on the sides of the cultures. These dissimilarities of form are the result of the varying conditions of pressure in different regions of the culture brought about by the shrinkage of the plasma clot. The cells are in contact by slender filaments or less commonly by wide bands. Although frequentlj^ closely packed together a thin layer of plasma always separates them on most of then- peripheries from neighboring cells. The transition of the layer into the reticulum and coverslip sheet is usually gradual. As the reticulum is traced toward the tissue and into the covering sheet it is evident that even near to the tissue where the cells are the most flattened and in closest proximity they still form a compressed reticulum. Cell associations similar in make-up to the covering layer are often formed in relation to the free surface of the plasma or around droplets of serum contained within the clots.
The cultures from four-day ventricle give rise to an extremely heavy reticulum with bands often exceeding 15 m in width (fig. 5). Their massiveness will be appreciated when it is recalled that the coarse mesh of older heart cultm-es are at the most 6 m across. The appearance of the growth is also similar to the coarse mesh with its oval spaces. Thick masses apparently made up of the same tissues as the coarse mesh flatten out upon the cover-slip to form a sheet one cell in thickness at the periphery. Sometimes a finger-like form occurs in partial contact with the cover-slip and intermediate between membrane and reticulum. The great fluidity of both reticular and membrane growi^h is shown by their flowing outlines and lack of cell independence. A finer reticulum is also common in the four-day cultures which has nothing to distinguish it from the coarse form of the older tissue. The five-day cultures are intermediate in character between the growth from four-day and older heart. Figure 4 is from one of the finest reticular growths seen at this stage.
Before considering the evidence regarding the sources of the ventricle growth the tissues which compose it may be enumer
350 E. D. CONGDON
ated. They include heart-muscle, nerve, endothelium and supporting tissue. Heart-muscle is, of course, the most abundant of these since it comprises the bulk of the myocardium. Nerve tissue is represented by a few neuroblasts. Endothelia include the endo- and peri-cardial layers and in the ventricle of the older embryos the walls of the sinusoids as well. The myocardium is separated on its inner and its outer surfaces from the endothelial covering by reticular layers. These consist of a mesh of formed substance upon whose strands are stretched cells which during embryonic development are becoming progressively more independent of their support. In the region of the atrial canal the sub-endocardial reticulum is continued into thickenings called endocardial cushions whose appearance is not unlike embryonic mesenchyme. EndotheUum and reticulum approach the heart-muscle in abundance and like it are met on every cut surface.
In considering evidence for the growth of heart-muscle the cultures from six- to eighteen-day ventricle require consideration separate from that of four- and five-day tissue. Material from the former source gives no direct evidence for the participation of heart-muscle in the culture. Many hundreds of sections were searched for myofibrillae without success. Direct evidence for heart-muscle growth was obtained however in the sections of a culture from five-day heart which happened to include not only ventricle tissue but a portion of the atrial canal. The contents of the walls of the atrial canal was seen to be flowing out into a heavy band such as made up the coarser loop-like reticulum of the four-day ventricle cultures. The wall at this point is so thin and the process growing out from it relatively so large that the conclusion is unavoidable that the whole cell-mass was moving out into the growth. At six days muscle and reticulum make up about equal parts of the wall of the canal. At five days the tissue is in large part a primitive myocardial layer. Since there is every reason for supposing that the heart-muscle cells differentiate in situ the evidence is therefore very good that primitive heart-muscle cells take part in the growth. The appearance
IDENTIFICATION OF CULTURES 351
of the coarse growths in unsectioDed four-day cultures seems to justify their classification as primitive myocardium. Their strands broaden out in such a way at the base as to appear as a projection from the whole ventricle mass rather than from small portions of its surface. If primitive heart-muscle cells grow from five-day heart, it is, of course, probable that they are more frequently represented in growths of fom-day ventricle. On the other hand, since the verj^ hea\'y reticulum is rare in fiveday cultures and entkely absent in growths of older tissue they probabl}' do not grow from embryos more than five or six days old. These conclusions are in agreement with Burrows' ('11) observation of sparse growths of heart-muscle in a small part of his cultures of two-and-a-half -day chick heart.
A growth of nerve cells was not found in any ventricle culture. The neuraxones are so characteristic in appearance that they could hardh^ have escaped detection had they been present. Their failure to appear m the growth is doubtless to be explained by the smallness of their number and the consequent imhkelihood of their coming into contact with the plasma, for the responsiveness of neuroblasts to cultivation has been amply demonstrated by Harrison ('07), Burrows ('11), Lewis and Lewis ('12) and Ingebrigtsen ('13).
The abundant growths of the older embryonic ventricle and finer reticulum of the four-day organ which are not made up of heart-muscle tissue must evidently take their origin either from the endothelium of pericardium, endocardium or sinusoids if not from the reticulum. Few of the various studies of chick ventricle growths which have appeared are especially concerned with questions of identification. Lambert ('12) describes the mesh as of connective tissue origin and Burrows ('11) as mesenchyme. Lewis and Lewis ('12 b) think that it may be mesenchymal but consider the matter of its classification unsettled. In various growths of chick organs certain of these authors have suggested that flat polygonal cells may be of endothelial origin but have not traced their connection with the tissue. It is difficult to follow the growth either to endothelium or to reticulum. There
352 E. D. CONGDON
is confusion in the histological picture because of the cell debris which results from cutting the tissue. The collapse of the ventricular ca\dty and distortion of the tissue, especially in the younger ventricles, prevents contact with the plasma on surfaces sufficiently large to allow a proof of their continuity with the growth.
Cultures in which the pericardium and its reticulum are next to the plasma are better adapted for this purpose than sections through the endocardium covering the trabeculae of the spongy ventricle. Relatively large masses of pericardimn may come in contact with the plasma while the endocardium is always intimately mingled with muscle. Figure 2 is from a photograph of a sectioned culture in which the active zone is made up of reticulum. The denser tissue internal to it is heart-muscle. Although the preparation was not killed until degenerative processes were well under way there is no difficulty in making out these tissues. The contrast between the two is seen much more clearly through the microscope than in the photograph because the muscle stains very intensely. A growth of fine mesh is plainly seen coming off from the reticulum. Xo traces of the pericardial endothehum which at first separated the reticulum from the plasma can be made out.
Figure 3 is from a photograph of a sectioned growth of the endocardial cushion of a thirteen-day heart. There has been no cutting with a knife. The free endocardial surface is in contact with the plasma. Because of the absence of the usual dead tissue resulting from cutting, strands from individual cells can be traced directly into the parent tissue. There is no room for doubt that the growth comes from the endocardial cushion. Just as the cushion is distinguishable from the reticulum by the presence of only one type of cell, so the elements of its growth are of a single form. They resemble the polyhedral cells of the fine reticulum.
There was no ventricle growth in which a connection could be traced with endothelium. The difficulties in the way of finding a region favorable for this purpose are too great to warrant concluding that endothelium does not take part. Indeed,
IDENTIFICATION OF CULTURES 353
there is considerable reason to think that it does grow into the plasma. In the two cultures which have just been described the tissues concerned which are separated by endothelium from the plasma could not have grown into it had not the covering sheets ceased to bar their way. Inasmuch as no remnants of dead endothelial cells are to be seen, they have in all probability migrated into the plasma. Of the two types of ventricle reticulum the coarser has the greater similarity to endothelium. The finer mesh has alread}^ been shown to have its origin in part at least from sub-pericardial reticulum. It must not be forgotten in this connection that since the very fine degenerate mesh grades into the normal fine mesh and this again into the coarser variety, the claim is possible that since the first transition is due to differences of the plasma the change from the fine to the coarse type has a like explanation. Such an interpretation is not in agreement, however, with the conditions found in the heart or other cultures. The so-called normal fine mesh and the coarse variety are vigorous and abundant in growth. Their cells give every indication by structure and staining reactions of being healthy tissue. The very fine growth without question stands apart from these as a type modified by its struggle with an unfavorable environment. If the coarse mesh comes from endothelium and the fine mesh from reticulum the inter-mixture of the two may well result from their close association in the ventricle itself.
Lewis and Lewis describe two types of cover-slip membrane for chick ventricle cultures. One of these which they find to be syncytial is said to develop from nearly all embryonic chick organs. They think that it arises from mesenchyme or connective tissue. Their figures correspond closely with the cover-slip membranes associated with the fine mesh. They also occasionally get from heart a non-syncytial membrane with pigment deposits around the nuclei. This variety was not encountered in my preparations.
354 E. D. CONGDON
LIMB-BUDS
It is of special interest to study limb-bud and ventricle cultures together because the growth of pre-muscle cells can be compared with primitive heart-muscle. The degree of similarity of the mesenchymal growths from limb-bud and of ventricle reticulum also has significance in view of the uncertainty as to whether reticulum makes its origin from endothelium or directly from mesenchyme.
The five-day-old embrj^onic limb-bud consists of ectoderm and closely-packed mesenchyme in which the axial scleretogenous tissue is just beginning to differentiate as an especially dense region. In the surrounding zone the pre-muscle cells can be distinguished by their elongation parallel to the axis of the bud. The vascular system is represented by sinusoids. Nerve fibers have not yet extended into the bud. The ectoderm consists of a layer two cells in thickness. At seven days a difference can be made out in the form and staining qualities of the sub-dermal and the scleretogenous tissues. The pre-muscle cells are markedly elongated. Sinusoids now form an extensive system and nerve fibers have migrated in.
The difference in appearance of the cells of scleretogenous pre-muscle and ectodermal regions combined with their separation into independent zones render the limb-bud favorable material for tracing the growth to its parent tissue. The manner of growth of the limb tissue differed greatly from heart, due to the influence of ectoderm and mesenchyme. The former tissue invariably retains its continuity as a sheet and when it grows vigorously can limit the distribution of other tissues to the region internal to it. The mesenchyme often flows out en masse taking with it sinusoids and pre-muscle cells (fig. 8).
The ectoderm can easily be traced from tissue to growth either in sections or whole preparations. Its presence in the plasma is clearly in part due to a creeping of the edge of the sheet out into the plasma. The portion retaining its contact is often stretched into a thin sheet. In the plasma the ectoderm may form masses several cells in thickness. Single cells or
IDENTIFICATION OF CULTURES 355
small groups occasionally project from the border of the sheet but there is always a considerable cellular contact.
The various mesodermal elements of the seven-day limb-bud are already sufficiently differentiated to give rise to a number of distinct growths.
Scleretogenous tissue can be identified in the plasma of many cultures close to regions of its contact with the plasma. Sometimes there is a migration of an entire region of the scleretogenous axis out into the plasma but the cells show no power of orientation and soon die. In most cultures the scleretogenous tissue is internal to mesenchyme and ectoderm and for this reason unable to reach the plasma quickly. Making due allowance for this handicap in position the vitality of the tissue from the seven-day embryo still appears to be of a low order.
The pre-muscle cells can be easily traced out into the plasma in many stained whole mounts because of their spindle form in the organ and their still greater elongation in the plasma (fig. 8). They form long strings which seldom branch. They are often long enough to reach the confines of the plasma drop and be deflected parallel to its surface. When in contact with the cover-slip the growth still maintains its elongated form thus differing from any other membrane growth (fig. 9). It is of interest that both primitive heart and skeletal muscle retain sufficient plasticity to appear in the plasma.
Another mesenchymal limb-bud growth is made up of large masses of cells formed, as in the case of the scleretogenous tissue, by the loosening up of the intercellular bonds of entire regions and a flowing out upon the cover slip (fig. 8). The cells are piled upon one another without a definite arrangement and at the outer border of the mass lie scattered upon the coverslip in a much more irregular manner than in the membrane growths of heart reticulum. In figure 8 the growth is plainly seen to be intermingled with spindle shaped pre-muscle cells. This identifies it as the little differentiated mesenchyme surrounding the scleretogenous and the pre-muscle cells. From among mesenchymal cells of cut surface a reticulum often ex
356 E. D. CONGDON
tends into the plasma which is apparently also mesenchymal in origin. The membrane associated with it is sunilar to the usual cover-slip sheet described with the fine mesh of the ventricle. The view that it is mesenchymal is strengthened by the fact that similar although somewhat more flowmg and embryonic growths occupy the chief place in cultures from five-day embryos. If the reticulum as well as the more massive growth is from mesenchyme there is need of an explanation for the development of two such different types from the same tissue. A possible clue is to be found in the fact that the reticulum only occurs in connection with the free cut surfaces while mesenchymal cellmasses are associated with the cultures in which a scant plasma drop has flattened out the border of the implanted tissue by its shrinkage. Together with the flattening of these cultures there is always an extension of the ectodermal sheet out toward the periphery of the drop thus greatly Imiiting the plasma accessible to the mesenchyme and remaining tissues.
If it be correct to regard the reticular growth of the limbbud as of mesenchymal origin, then heart reticulum and mesenchyme are closely similar in their growths and as far as this evidence goes are closely related. If, on the contrary, it should be shown later that mesenchyme gives rise only to the massive growth the arguments from tissue cultures would be in support of the origin of reticulum from endothelium.
LIVER AND INTESTINES
A few series of cultures from five- and eight-day intestine as well as from five- and ten-day liver were prepared as a basis for comparison of the growths.
A fine reticulum can be clearly traced in many sections to the mesenchyme of the intestine. Growths which arise from surfaces of the intestine with portions of the peritoneum intact often contain trabeculae of greater diameter than found where the growth is plainly of mesenchymal origin. It is not possible to trace the broader bands definitely to the peritoneum because the mesenchyme is always found to be taking part in the growth where the continuity of the peritoneum is lost. It is very prob
IDENTIFICATION OF CULTURES 357
able, however, that they arise from the peritoneum just as similar growths from heart ventricle also "probably are of endothelial origin. Whole mounts of the intestine sometimes show a flattening down of a part of the organ, accompanied by the flowing out of the mesenchjmie of the region upon the cover-slip just as seen in lunb cultures. It is usually possible in these preparations to trace the peritoneum out as an unbroken sheet into the plasma. Lewis and Lewis ('12 b) describe and figure this type of growtii and look upon it as peritoneal. The reticular growth which has just been described as present in sectioned cultures does not occur where there is such a flattening but is confined to free surfaces of the tissue.
Liver cells do not take part in growths from the ten-day organ. A coarse and a fine mesh occur with about equal frequency. Figure 7 shows the cover-slip growth associated with the fine mesh. The coarse form appears in many sections to come from the peritoneum. Elsewhere there is no proof of a deeper origin. Figure 10 showing the fine mesh growth is from a ten-day embryo and is fixed after twenty-four hours of incubation. No peritoneum is present in the culture. The growth can not come from connective tissue septa except possibly at a few restricted regions. In ten-day chick livers a reticulum is already present, as can be ascertained in part of the implanted tissue where degenerating liver cells have dropped out of a section. Endothelium is also present in large amounts in the form of sinusoids. The occurrence of mesenchyme in the embryonic liver has not been proven. It is therefore not possible to determine the source of the growth from the interior of the organ.
SUMMARY
Reticulu7n. The common reticulum with its corresponding membrane-growth is traceable to sub-pericardial reticulum in six-day ventricle.
Endothelium. There is indirect evidence for a coarse reticular growth from the peritoneum of liver (five- and ten-day) and intestine (five- and ten-day) and from the endocardium and
358 E. D, CONGDON
pericardium of ventricle (five- to fourteen-day). An unbroken sheet is found to arise from intestinal peritoneum under certain conditions (six-day).
Mesenchyme. Sectioned cultures of intestine (six-day) gave rise to a fine mesh similar to that of ventricle reticulum. Mesenchyme is sometimes given off from limb-buds (five- and ten-day) in the form of disorganized masses of cells. Under other conditions reticulum and a corresponding membrane growth apparently also take their origin from limb-bud mesenchyme.
Heart-muscle. In one sectioned five-day culture the contents of the wall of the atrial canal is found to be moving out into a strand of a very coarse reticular growth, such as is common in cultures from four-day ventricle. Prmiitive myocardium of four- and five-day heart therefore apparently grows out as a coarse mesh but it is unlikely that heart-muscle of older ventricle has this power.
Endocardial cushion of ventricle. A growth of polj^hedral cells resembling those from sub-pericardial reticulum was traced in sections to the endocardial cushion (thirteen-day).
Scleretogenous tissue of Umh-bud. The scleretogenous tissue can move out into the plasma for a short distance but has little vitality (seven-day) .
Ectoder?)! of limb-hud. There is an extension of the ectoderm out into the plasma but this is due, in part at least, to a creeping of the original layer, as is shown by its marked thinning on the surface of the limb bud (five- and ten-day).
Pre-muscle tissue of limb-hud. The spindle-shaped premuscle cells of seven-day limb buds give a characteristic linear growth in the plasma and upon the cover-slip.
IDENTIFICATION OF CULTURES 359
LITERATURE CITED
Burrows, M. T. 1911 The growth of tissues of the chick embryo outside the
animal both' with special reference to the nervous system. Jour.
Exp. Zool., vol. 10. Carrel, A., and Burrows, M. T. 1911 Cultivation of tissues in vitro and
its technique. Jour. Exp. Med., vol. 13. H.\HRisoN, R. G. 1907 Observations on living, developing nerve fibers. Proc.
Soc. Exp. Biol, and Med., vol. 4. Ingebrigtsen, R. 1913 Studies of the degeneration and regeneration of axis
cylinders in vitro. Jour. Exp. Med., vol. 17. Lambert, R. A. 1912 Variations in the character of growth in tissue cultures.
Anat. Rec, vol. 6. Lewis, M. R., and Lewis, \V. H. 1912 a The cultivation of sympathetic nerves
from the intestine of chick embiyos in saline solutions. Anat. Rec,
vol. 6. Lewis, M. R., and Lewis, W. H. 1912 b Membrane formations from tissues
transplanted into artificial media. Anat. Rec, vol. 6.
THE ANATOMIC.VL RECORD, VOL. 9, NO. 5
PLATE 1
THE IDENTIFICATION OF TISSUES IN ARTIFICIAL CULTURES
E. D. COXGDOX
^
-^135^
4
^ .^i^^r-*
IfVl* -J®
-"^J
%.
-■•J^^' ^*«
.^i^J*^*^.©^^
y/r^\
'ff V.-i,lt.
m
Fig. 2 Section of teu-day ventricle sliowing fine reticular growth arising from the reticulum. The more internally situated denser tissue is heart-muscle. X 600.
Fig. 3 Section of thirteen-day ventricle; the fine reticular growth is coming from an endocardial cushion. X 570.
360
TTrE IDENTIFICATION OF TISSUES IN ARTIFICIAL CULTURES
E. D. CON'CnON
PLATE 2
■
w
P
■"IBI
^^^H
p
1r.\
•
1|^H
i
\
';> '
xf'^^I^^H
1
~" V
\ ■
'^ '^C^'lw^B^^^^^^^^I^^^^H
hI <^
■"-vv. ^n
QtH^^^^^^HH
L
ll^^H
^Ik
^^^fijMj^UJ^H
^H^^^^H^^^H
^^H
t
>s' j-.3H|
I^^^^^^^^^H
^H
■
'^'^^Qk^^^^^l
^^^^^^^^H
um
li
^
fe*!
W'.^f^B^^^^I
■^^^^H
iKffife
u- ^^^^^^^^^IBHM
Fig. Fig.
Pine reticulum of tive-day heart; stained whole mounts. X 250. Coarse reticulum of four-day heart; stained whole mounts. X 270.
361
PLATE 3
THE IDENTIFICATIOX OF TISSUES IX ARTIFICIAL CLLTURES
E. n. rON'GDOX
Fig. 6 Cover-slip frrowth associated witli fine rcticuiuin ; nine-day ventricle; stained whole mounts. X 530.
Fig. 7 Cover-slip growth associated with fine mesh of five-daj^ liver; stained whole mount. X 350.
362
rHE IDENTIFICATION OF TISSUES IN ARTIFICIAL CULTURES
E. I). COXGDON
PLATl, 4
Fig. 8 Mesenchymal growth containing a few pre-muscle cells; spindle shaped pre-muscle cells are also seen projecting from the edge of the implanted tissue; stained whole mount. X 510.
Fig. 9 Cover-slip growth of pre-muscle cells from seven-day limb bud. X 410.
36;)
PLATE 5
THE IDENTIFICATION OF TISSUES IN ARTIFICIAL CULTURES
E. D. CONG DON
■■"£»=■
— .TiTrh-i i i i uM-.
Fig 10 Section of culture from ten-day liver, showing tine reticular growtJi. The cellular debris in the outer zone of the implanted tissue is made up of the remains of dead liver and sustentative cells. The inner zone shown at the right of the figure contains still uninjured liver cells. X 590.
364
THE ACTION OF ULTRA-VIOLET RAYS UPON THE
FROG'S EGG
I. THE ARTIFICIAL PRODUCTION OF SPINA BIFIDA
W. M. BALDWIN
From the Department of Anatomy, Cornell University Medical College, New York City
SIXTEEN FIGURES
While the method of the experimental embryological investigation detailed in the succeeding pages is of importance because of the constancy with which the condition of spina bifida may be produced in the embryo, the real value of the method and of its results lies in the interpretation which is thereby rendered possible of the physiological value of the several component portions of the fertilized ovum. The really fundamental problems to be determined by investigations of this nature are, first, whether the fertilized o^^um is to be regarded as a composite structure made up of various system- or organ-anlagen, or the chemical progenitors or 'ferments' of such, distributed in definite and, perhaps, constant positions throughout the cytoplasm. Or, second, is the ovum to be considered in the sense of a unicellular organism, differing in no great respect from the physiological structure of unicelhdar organisms in general but possessing that specific potential to elaborate 'ferments' and pro-anlagen at successive genetic stages, and, ultimately the anlagen of the later developmental stages? In the former instance it is presumed that the embryonic parts are pre-localized in the cytoplasm of the ovum and make their appearance, in the words of Lankester, as "a sequel of a differentiation already established and not visible." In the second assumption the embryonic parts are unrepresented in the ovum, the regions of the cytoplasm being
365
366 W. M. BALDWIN
then, so far as the future embryo is concerned, of equipotential vahie. It appeared to the author that a means by which a Umited area of cytoplasm could be destroyed and j^et left in its original relations to surrounding parts would afford a solution to this question. Accordingly, recourse was made to ultra-violet rays of such a degree of intensit}- as to cause the disorganization of the cj^toplasm in from one to thirt}^ seconds and of such a degree of concentration as to influence limited surface areas. Acting upon the suggestions made by Prof. E. H. Merritt, an apparatus was constructed which met these requirements fully.' This apparatus consisted of a large induction coil actuated by a 110-volt direct current reduced by an unknown resistance. The potential, moreover, was raised by means of several Leyden jars shunted between the electrode wires. The terminals were made of iron, and were spaced about 5.0 mm. The eggs used for the purpose were those of the various forms of frogs occurring in the neighborhood of Ithaca, New York. These were obtained early in the morning, as soon after laying as was possible. At the time at which they were influenced they were in the undivided stage. Development was allowed to progress in the laboratory in some instances, and the eggs influenced at several later developmental stages, but no egg further along in its cycle than about the 64-cell stage was used. Furthermore, care was taken to reject such eggs as were collected late in the laying season for that particular species of frog, and particularly those located near the center of the egg bunches, specifically to avoid dealing with those possessing a tendency towards abnormal development. In preparation for exposure to the rays, the eggs were freed from their jelly, which had been found impervious to the light, and placed under a i)erforated tinfoil diaphragm. The perforations differed in size in different experiments. After the egg had been rotated so that a predetermined part had been brought directly under the center of a circular ])erforatioii in the dia 1 At this ])oint I desire to express my indebtedness to Professor Merritt for helpful suggestions and to the Department of Pln'sics of the University for the use of the apparatus with which the experiments mentioned in this paper were conducted.
ARTIFICIAL PRODUCTION OF SPINA BIFIDA 367
pliragm, both were then brought under the electrodes of the apparatus and the circuit closed.
While in the numerous experiments conducted, the various portions of the white and of the black hemisphere and of the equator of the frog's eggs were influenced in order, the author decided to limit the scope of this present communication to those effects produced by the rays when influencing the white hemisphere and the equator of the egg. Indeed, in addition to the significance of the findings of the investigation in the interpretation of the larger problem of ovum structure, the immediate purpose in jiresenting this paper is to establish the fact that the condition of spina bifida may be produced at will l)v this method.
Figure 1
The aperture in tlie diaphragm used for this experiment was 0,4 mm. in diameter. The eggs averaged 1.7 mm. in diameter. Consequently but a small surface area proportionately of the total area of the egg was influenced. The latter amounted to 425.0 sq. mm. whereas but about 5.0 sq. mm. of this surface could be influenced by the rays. The relative sizes of these areas is brought out more clearly by reference to figure 1, which represents by a broken-line circle the portion of the surface area of the egg sphere ilhuninated. Further, it was found that the depth of penetration of the 0.4 mm. pencil of rays depended upon the length of exposure to the light. Uniform exposures of 30
368 W. M. BALDWIN
seconds were employed in this series. A section through an egg so influenced is shown in figure 2, in which the depth to which the ra^'s had penetrated is represented by the shaded portion on the right of the sketch. In this instance the rays passed in the plane of the section and at riglit angles to a tangent at the center of the surface of the affected area. The direct results of the illumination were corroborative of those previously observed by other investigators using violet rays, such as granulation of the chromatin and certain degenerative changes noted in the cytoplasm. Xo attempt was made, however, to study this aspect of the influence of the rays. It was noted in extremeh' long exposures of from 1 to 10 minutes, that masses of protoplasm
Figure 2 Figure 3
were in some instances extruded upon the egg surface, retaining, however, a slender connection with the main mass of the egg. In instances of such exovation, the egg died early after having made but little developmental progress. Such an exovate is shown in flgure 3. It is of importance to note the fact demonstrated by the sketch that the mass of the exovate was approximately equal to that of the influenced area of the egg (compare with figure 2). The most plausible inference to be drawn from this phenomenon, in the terms of tlic interpi-etation of the o\".ini as an organism, seems to be that of an effort on the part of the ovum to rid itself of the chemically altered or dead protoplasm which can only act as a hindrance to its further developmental progress.
ARTIFICIAL PRODUCTION OF SPINA BIFIDA 369
Reference to various series of experiments selected at random bring out the value of the method in the constancy of production of the condition of spina bifida. Tn one series of thirty-one 16-cell eggs used at the beginning of the experiments and exposed to the 0.4 mm. ra^' for 30 seconds each — various regions of the equator and of the white hemisphere being influenced — twenty-one developed abnormally, and but ten normally. Of the abnormal embryos, eight presented the condition of sphia bifida. In another and later series of fifteen eggs hi the 4-cell stage, influenced in the same manner, none developed normally. Most of these died during the early stages. Four, however, lived to swunming forms with two tails. Later in the spring, after the technique had been still further perfected and the eggs of the green frog were available, from which it was possible to remove the enveloping jelly more readily and more completely, the percentage of spina bifida embryos rose. In one set of five undivided eggs influenced in a suiiilar manner, one died about twelve hours after the experiment, having made no developmental progress, and the four others grew to swimming forms presenting the condition of spina bifida, each having two tails. This last instance is merely representative of the high percentage of these forms of malformation obtained when the white hemisphere is influenced by the ultra-violet rays.
As has been mentioned above, the most eflectual barrier to the penetration of the rays was the investing, jell}^ ^^ hen this was completely removed an exposure of 10 seconds was sufficient to influence the egg. The presence of a very thin layer, however, completely blocked the passage of the raj^s even during an exposure of as much as 10 minutes. The author attributes most, if not all, of the irregularities in percentage production of spina bifida embrj^os to the presence of this jelly. The later results of the expermients \^ere sufficiently assuring to warrant the conclusion that, when it could be positively known that the rays under the above conditions had actuall}^ penetrated the ovum in the regions above mentioned, the condition of spina bifida could be invariably brought about in the developing embryo. Taking these difficulties into consideration, however, the per
370
W. M. BALDWIN
centage production of the condition ranged between 85 and 90 per cent of the total number of eggs used.
An observation that recurred repeatedly was to the effect that the developmental period required by the eggs was lengthened as a result of the rays' influence. Under laboratory conditions ordinarily from 3 to 4 days were sufficient for the appearance
6
Fig. 4 A specimen of cauda bifida demonstrating the asj-mmetry of the tails. In this egg the rays had struck a portion of yolk farthest removed from the equator. The tails are provided, as is shown, with peculiar toe-like processes on their free extremities.
Fig. 5 In this tadpole the right tail encountered the body axis at an acute angle directed anteriorly. Two yolk plugs are to be seen, and the same splitting of the extremity of the left tail as was noted in figure 4.
Fig. 6 The lines 7 and 8 on this cauda bifida tadpole indicate the level of the cross-sections shown in the respective figures. In figure 7 the notochord lies ventral to the well differentiated neural tube. In figure 8 the asymmetry of the halved neural tube is shown, more particularly on the right side. Ventral to this lies the notochord, all traces of which are absent from the left halt of the sketch. This right neural tube-half lay in the more activeh' used tail.
of the free-swimming tadpole-forms of the green frog; in the case of the experimented eggs, howe\'er, 5 or 6 days were required and in some instances 8 days. Furthermore, it was noted that during the 12 hours immediatel}^ ensuing upon the experiment the eggs seemed to have entered into a condition of temporary suspension of development, later resuming that process but with greatly lengthened tempo.
ARTIFICIAL PRODUCTIOX OF SPINA BIFIDA
371
The further observation was made that free-swiinming forms, such as are represented in figures 4, 5, and 6, seemed to be able to move about by the use of either tail, but that the swimming movements were more vigorous in one than in the other. This is an interesting fact in connection with the results obtained by study of the microscopical sections of the same specimens. In these it was learned that in the more favored of the two tails the neural tube was greater in diameter and extended a longer distance towards the tip of the tail. In some of the specimens,
'<:^
c
♦
^.
«%
V^ .
«
.",
•
•<
'^'.V.
»
>
'//
'Vii^i
.-u.v
.^
Figure 7
Figure 8
as is to be seen in figure 8, the notochord was limited to one tail. Figures 7 and 8 are cross-sections of the tadpole represented by figure (5. The right was the more active of the two tails during life. In figure 7 the neural tube, cut just cephalic to its bifurcation, is seen to be well differentiated, with the notochord b^ng ventral and adjacent to it. The section of figure 8 was taken immediately caudal to the bifurcation and shows the notochord confined to one (the right) tail.
Several specimens presented a peculiar relation of one tail to the longitudinal axis of the body; such are figured in 5 and 9. In the latter figure the main axis of one tail joined that of the trunk at almost right angles, whereas the other tail coincided fairly well with the main body axis. In figure 5 the right tail
372
W. M. BALDWIN
met the body axis at an acute angle, looking forward. Such tails were, of course, useless from the functional standpoint; but their unportance cannot be o^'erestimated in furnishing exaggerated examples of the asymmetry of some of the types of spina bifida, such as were observed above in the cross-sections.
The study of the serial cross-sections demonstrated, furthermore, as these were followed in order caudally, that just posterior to the level of bifurcation of the neural tube each half tube des
lO
Fig. 9 This figure illustrates a marked instance of asymmetry with a division of the cord well anterior on the embryo. Here again the extremity of each tail is broken up into toe-like processes.
Fig. 10 In this embryo the rays had encountered an area well up on the equator in the median plane; hence, the bifurcation of the neural cord immediately posterior to the optic anlagen. The lines 11, 12 and 13 indicate the levels of the respective cross-sections illustrated by the succeeding figures.
tined for each tail presented an asymmetrical outline, the lateral wall being considerably thicker than the median. This is shown in part by the right neural tube in figure 8. As the series was followed farther caudally, however, a readjustment of the tube cells was observable, each half now becoming either a solid rod or a tube entirely synunetrical so far as the thickness of its walls was concerned; (see also figures 12 and 18).
The neural tube was caused to bifurcate at various levels, dependent upon tlie portion of the hemisphere influenced.
•
1.
- ■ ^
t '. ■
» ^
• * .*
e *
• V: *
' •>
> -9 '
• *
' ,
♦^*>
» ? '
%^2
■*
, • » *
%
•
•
< •>
«
->%^
f » •
1 r
" • ■
>
»»*»^
«^ ■, »
«
♦ z' v»
^%^ *.'jr .
^igs. 11, 12, 13 In figure 12 the marked asjTiimetry of the neural tube halves immediately posterior to the level of bifurcation of the cord is shown. The neuroblasts become adjusted farther posteriorlv, however (fig. 13) to form a more symmetrical tube or column of cells.
373
374
W. M. BALDWIN
Where the rays struck close to or on the equator in the median plane of the eg^ the subsequent bifurcation was noted well forward towards the head region; figure 10 well demonstrates this fact. In figure 9 the point encountered was much lower down on the hemisphere, while in figures 6 and 4 it was still farther posterior. Such are reallj^ instances of cauda bifida. Another fact of the utmost significance was gained from studies of the histological sections. Since it was definitely ascertained that the rays had killed the portion of the egg illuminated, in no specimen was there to be observed, however, a deficiency or falling out of a ]:)ortion of the neural tube, as one would expect
if the ])roanlagen or anlagen had been encountered by the rays. Notwithstanding the possibility of post-generation, with a replacement of the anlagen thus rendered inactive, the conclusion might be justified, tentatively at least, that the proanlagen are not located in the early stages of division of the ovum either in the yolk hemisphere or along the equator, but are confined wholly to a region well up on the black hemisphere. Before referring in detail to the results of other investigations dealing with spina bifida it will be well to emphasize some general considerations which must be taken into account in connection with the production of abnormalities.
It is taken for granted that there are many linking factors associating the develoi)mental processes concerned in the pro
ARTIFICIAL PRODUCTION OF SPINA BIFIDA 375
duction of spina bifida with tliose of other forms of malformation occurring in nature and produced experimentallj'-. Accordingly, one cannot well studj^ the one without taking into consideration those general fundamental phj^sico-chemical factors of development underlying both, upon the disturbance of which the production of anomalous conditions is dependent. The fact is well supported that the processes of differentiation in the tadpole, as in some other forms, appears to be dependent upon a series of complex and progressive chemical reactions which are of the nature of oxidations. In the later stages of development we are dealing with an organism whose chemical constitution differs considerably if not completely from that of the undivided ovum. The results of such reactions and chem.ical changes become apparent in the differentiation of the various anlagen. Eggs placed in pure water free from chemical compounds which might enter into the chemical structure of the organism, but supplied with a liberal amount of oxygen, can and do undergo the several processes of differentiation, finallj^ hatching and swinuning about. From that time on, however, when grovv^th of the differentiated parts is initiated the organism requires a supply of various chemical substances for its api^ropriation. In this connection, it is significant that the great majority of defects produced experimentally ai"e referable to changes in the differentiative stages of the embryo and only .secondarily to defective growth phenomena. Bearing in mind, then, the chemical modifications occurring during the differentiating cycle of the embryo, it is fair to assume that at certain specific times in the genesis of the anlagen their chemical composition is such as to render them more ready participants in chemical reaction with the chemical agent employed. Results so obtained cannot be cited without reservation as applying to the state of the undivided ovum, consequently much of the chemical work conducted along this line is subject to considerable qualification and limited in the insight which it furnishes into the constitution, chemical or physiological, of the ovum. Before the full truth can be established on this point, it must be demonstrated that the toxic action of the chemicals, employed was exerted upon the egg during the undivided state
THE ANATOMICAL KECORD, VOL. 9, NO. 5
376 W. M. BALDWIN
and not afterwards; or, in other words, upon the chemical 'ferments,' or proanlagen, and not upon the anlagen cells when they have attained what might be termed a condition of chemical completeness. Only by this means can we decide between the two conceptions of the ovmn; as an organism elaborating its organ anlagen at succeeding developmental periods, or as a composite, mosaic-like structure.
While the artificial production of defects has been known for a long while to be possible, but very few have succeeded in advancing any plausible explanations covering the instance of spina bifida embryos. There are two aspects to the problem of the artificial production of spina bifida which are brought out in a review of the literature dealing with this subject. The first consideration is whether we are dealing with the action of an external agent upon some specific substance in the egg; and the second, whether the nature of this reaction is specific, referable to the agent alone^ which possibly reacts upon the egg as a whole.
Morgan in 1894 and O. Hertwig in the year following were both successful in the production by chemical means of a large percentage of embryos showing the defect of spina bifida. That 0.625 per cent solutions of sodium chloride should produce so high as 50 per cent of this form, of malformation was an argument in favor of a definite specific chemical or physical property of the compound. Prior to this date observers had recorded only occasional instances of this defect and had failed to give a convincing indication of the nature of the upset in the physicochemical factors concerned. Roux was the first to call attention to the occurrence of spina bifida among frog's eggs, owing, apparently, to conditions found in nature. Panum recorded 38 instances of spina bifida in chicks, among 404 monsters produced. He, with Dareste and Fere, obtained monsters of various kinds by the employncient of variations in temperature (as did Hertwig), by varnishing the egg shells, by shifting the long, axis of the egg to the vertical, by traumatic injuries, shaking, magnetism, electrical means, various gases, vapors of lavender and by injecting different toxines and chemicals, such as turpentine, an
ARTIFICIAL PRODUCTION OF SPINA BIFIDA 377
iseed, absinthe, and cloves into the white of the egg. Their inabihty to associate any given deformity with a known and controllable cause led to a failure in the analysis of the normal developmental factors of the embryo. Richter found three instances of spina bifida among several hundred hen's eggs upon which he had experimented. Spemann, however, produced twotailed embryos by simply tying a ligature between the two blastomeres, demonstrating the bilaterality of the anlagen but throwing no light on the nature or antero-posterior extent of the organ-building substances. Fol, Rauber, Born, and O. Hertwig attributed the duplicity to double fertilization. This explanation was too compromising regarding the anterior portion of the embryo, and later was found to be unnecessary. Godlewski's experiments with reduced pressure, and Herbst's with lithium salts, Morgan's with the centrifuge, Samossa's with atmospheres of nitrogen and of hydrogen, and Wilson's with Ringer's solution and with sodium chloride, furnish additional evidence of the diversity of ways by which this abnormal condition may be produced.
In this connection, it is interesting to note that Mall has reported 12 instances of spina bifida among 163 pathological human embryos, attributing as a possible cause of the condition, faulty unplantation of the embryo. Analysed still further, however, by analogy to the conditions found among lower vertebrates, it seems possible that the hmnan ovum, too, requires but little else than a good supply of oxygen for its differentiation during the early stages of development. At this time the causal forces are operative for the production of spina bifida. Undoubtedly, a deficiency in the supply of oxygen could be brought about by the imperfect unbedding of the ovum in the uterine mucosa. Bearing this fact in mind in connection with the features of differentiation of the ovmii given on page 375, it seems superfluous to seek an explanation of the condition in man through the action of chemical substances or of altered temperature. Though the possibility of the direct or indirect dependence of the processes of oxidation upon the action of the latter agents must be
378 W, M. BALDWIN
admitted, reasoning from conditions as we find them in the frog, a sufficient and probably a more primal cause, at least, is referable to faulty oxidation.
Guthrie produced these defects by the use of strychnine, caffeine, and nicotine, as had Hertwig, but with concentration far below that of 0.625 sodium chloride. Jenkinson, however, tested out this question of the osmotic pressure of the salt solutions by employing a great variety of isotonic solutions of various salts, such as chlorides, bromides, iodides, nitrates, and sulphates of ammonium, lithium, sodium, potassium, calcium, barimn, strontium, and magnesium, and, in addition, solutions of cane sugar, dextrose, urea, and gmn arable. He obtained spina bifida with especial success in his sodium chloride and sodium nitrate solutions. His conclusions are best given in his own words: "There is very little room for doubt, that the malformations in question may be due to some property of a salt other than its osmotic pressure." Bataillon had previously come to the conclusion that malformations were not specific to the means employed. Gurwitsch's belief was that halogens affected the position and development of the blastopore and of the brain, sodium chloride acting upon both, and sodium bromide upon the brain alone, whereas lithium chloride seemed selective on archenteron and blastopore.
It would appear, considering the production of this malformation by the diverse methods outlined above in connection with that detailed by this paper, that we were justified in concluding that in the question of specificit}^ of reaction in the production of spina bifida the weight of argmiient at present refers the causative forces more particularly to an upset of a specific substance in the egg, rather than a specific action of the agent. It is in the conception of the changes which occur in the chemical composition of the ovum during its differentiation, as previously outlined, that we find support for this statement. It follows tiiat the composition of any particular proanlage or anlage may be such at different stages of its chemical elaboration as to possess a marked affinity for widely varying chemical reagents. The
ARTIFICIAL PRODUCTION OF SPINA BIFIDA 379
developmental end-product of the reactions so brought about would be the same, e.g., spina bifida, notwithstanding the wide diversity of character of the chemical reagents employed.
Furthermore, since we cannot deny, we must take into account the possibility of a two-fold manner of production of spina bifida; the one, owing to an upset in the contents of the cells of the unpigmented hemisphere whose yolk is intended for the nutrition or elaboration of the other component, viz., the proanlagen or chemical ferments restricted to the cells of the pigmented hemisphere. For the other, we can assume the possibility^ of an interference in the function of these proanlagen as the result of the chemical reactions experimentally induced, apart from an upset referable to the composition of the nutritive yolk particles. The author's work, however, points out clearly that in the white hemisphere alone are resident sufficient causes for the production of the malformation, so that, while the possibility of an involvement of the proanlagen exists, the weight of experimental evidence points to the yolk hemisphere as the more vulnerable of the two. Jenkinson observed, for instance, that as the result of chemical action the yolk cells were primarily affected, and Godlewski, employing reduced pressure, came to the same conclusion. The disturbing influences of insufficient aeration and cold, as ascertained by Morgan and others, were noted first in the yolk cells, and to this same region Morgan attributes the causative factors in the results of the centrifuge, while Hertwig drew the same conclusions from studies of overripe eggs.
The production of an area of altered protoplasm, which serves as a mechanical check to the approxunation of the lips of the blastopore during the backward progression of the latter, emphasizes very naturally the importance of synchronized tempo in the two processes directly concerned with the elaboration of the neural cord. Under normal conditions the differentiation of the neural anlagen (consequent upon the backward migration of the proanlagen) occurs apparently synchronously with the backward migration of the blastopore and fusion of its dorsal lips. These two processes are approxmiately cooperative in point of
380 W. M. BALDWIN
time, i.e., the anlagen of each half-tube become progressively differentiated in a backward direction at about the time when the half of the dorsal lip in which it is localized meets and fuses with its corresponding fellow of the opposite side. The two processes are not causally dependent upon each other, however, since differentiation takes place in the experimented eggs at about its former rate but now along the equator and not, as usual, parallel to the median plane.
The absence of the proanlagen and anlagen of the egg along the equator in the earliest stages of development of the ovum is sufficiently attested for by the ultra-violet method . Incidentally, it should be remarked that the restriction of these proanlagen at all times to the pigmented hemisphere seems to the author's mind a very significant fact. In this connection, it should be stated that ultimately the yolk mass is wholly drawn into the body of the embrj^o. Even though by this later process the neural tube halves may be approximated, subsequent fusion does not take place, however, since each half tube has postgenerated into a whole tube.
The conclusions reached by this method of experimentation upon the fertilized ovum, are, therefore; first, that the killing of a small localized area of the yolk hemisphere or of the region of the equator of the frog's egg produces invariabl}^ the condition of spina bifida in the embryo; and second, that the neural tube proanlagen, or formative substances, do not lie either in the yolk hemisphere or along the equator of the frog's egg, but are wholly restricted to the pigmented half of the egg. These proanlagen attain their definitive positions bj^ a process of backward migration, the rate of which is synchronous with that of the backward progression of the dorsal lip of the blastopore. The action of the ultra-violet rays in destroying a small localized area of the yolk hemisphere or equator results from mechanical causes in an upset of the synchronism of the two factors, i.e., differentiation of the neural anlagen and approximation of the lips. The former proceeds at its normal tempo, while the latter is retarded. Consequently, the former, always restricted to the pigmented hemi
ARTIFICIAL PRODUCTION OF SPINA BIFIDA 381
sphere, come to lie along the equator and are later carried towards the median plane by the subsequent approximation of the lips, but the half tubes, having already differentiated into whole tubes, do not subsequently fuse.
BIBLIOGRAPHY
Bataillon 1901 Arch. Entw. Mech., Bd. 12.
Born 1887 Bresl. arztl. Zeitschr. fur 1882, Bd. 14, Bd. 15.
Dareste 1891 Recherches experimentales sur la production artificielle des
monstruosites, Paris. F^RE 1893-1901 Compt. Rend. Soc. Biol. FoL 1879 Mem. de la Soc. de Phys. et d'hist. Nat., Genevd. GoDLEWSKi 1901 Arch. Entw. Alech., Bd. 11. GuRwiTSCH 1896 Arch. Entw. Alech., Bd. 3; Zeitschr. f. wissensch. Zool., Bd.
55, 1893. Herbst 1897 Mitt. d. Zool. Station zu Xeapel, 11, 1895; Arch. Entw. Mech.,
Bd. 5. Hertwig, O. 1896 Festschr. f. Karl Gegenbaur, Leipzig (Arch. f. mikr. Anat.,
Bd. 44, 1805). Jenkinson 1906 Arch. Entw. Mech., Bd. 21. Knower, H. McE. 1907 Anat. Rec, Amer. Jour. Anat., vol. 7. Lankester 1877 Xotes on embryology and classification. LoEB 1893 Pfliiger's Archiv... Bd. 54; Amer. Jour. Physiol., Ill, 1900. Mall, F. P. 1908 Jour, of Morph., vol. 19. Morgan, T. H. 1894 Anat. Anz., Bd. 9, Quart Journ. Micr. Sci., vol. 35, no. 5
1897 The development of the frog's egg.
1909 Anat. Rec. vol. 3. Pantjm 1878 L'ntersuchungen liber die EntstehungderMissbildungen, zunachst
in den Eiern der Vogel, 1860, and Virchow's Arch., Bd. 52. Ratjber 1878 Virchow s Archiv., Bd. 71, Bd. 73 and 74; Bd. 81, 1883. RiCHTER 1888 Verhand. der Anat. Gesellsch. Roux 1895 Gesammelte Abhandlungen, Leipzig. Spemann 1903 Zool. Jahrb. Suppl. Wilson, C. B. 1897 Arch. Entw. Mech. V.
ON THE PRESENCE OF INTERSTITIAL CELLS IN THE
CHICKEN'S TESTIS
t
T. B. REEVES
Anatomical Laboratory of the University of Virginia
THREE FIGURES
Since there is a difference of opinion as to whether interstitial cells are present in the testis of the domestic cock, and because of the obvious bearing of the question upon the theory which attributes to these cells an important influence upon secondary sex characters, it has seemed worth while to investigate the matter. To illustrate the difference of opinion I shall abstract briefly two very contradictory reports on the subject.
Alice M. Boring^ reports observations on testes of roosters from one day to twelve months of age. In the young as well as in the older testes she fails to differentiate any cells of the intertubular tissue from ordinary connective tissue cells. The variation in size, shape and character of the nuclei is attributed to mechanical conditions of pressure. The fat observed in the intertubular tissue was not found inside of the cell bodies, hence it was thought to be brought there by the circulation and deposited. Her conclusion is that there are no interstitial cells present at anj^ time.
In the summarj^ of the work done by J. des Cilleuls,- it is stated that the external differentiation of the rooster from the pullet begins to be apparent at about the thirteenth day; and that at this time interstitial cells first make their appearance in the testis. Des CiUeuls says the interstitial cells and cock characters increase pari passu and the cock characters are accentuated
1 Alice M. Boring. The interstitial cells and the supposed internal secretion of the chicken testis. Biological Bulletin, vol. 23, no. 3, August, 1912.
^ J. des Cilleuls. Interstitial testicular cells and secondary sex-character. Summary in Journal of the Royal Microscopical Society, December, 1912.
383
384 T. B. REEVES
while the seminal tubes still remain in an embryonic condition, until after the sixtieth day. The explanation offered is that the internal secretion of the interstitial cells serves as a stimulus for the development of the secondary sex characters.
This report is made after the study of testes from cocks three, five-and-a-half, nine and eighteen months old. T^he tissue was removed inmiediately after killing the fowl and fixed in the following solutions: formalin, Zenker's, Bouin's and Ciaccio's fixative :
5% potassium bichromate 20 cc.
formalin 4 cc.
acetic acid 1 cc.
Fix in the above forty-eight hours, then in 3 per cent potassium bichromate one week. Sections were stained chiefly with hematoxyhn, and congo red, iron hematoxylin, and Mallory's connective tissue stain. The Ciaccio fixative and Mallory's stain gave the best results for the study of the intertubular tissue, although the other preparations showed up fairly well.
For the study of fat I used frozen sections of tissue fixed in formalin. These were stained with Sudan III and hematoxyhn.
Microscopic examination of the sections from the eighteen months testis shows the seminiferous tubules in a state of active spermatogenesis. The intertubular tissue is small in amount and compact, allowing the tubules to he close together. Where three or more tubules come in juxtaposition small triangular or irregular areas are formed. In most of these areas there is a small blood vessel surrounded by connective tissue which contains both spindle- and oval-shaped nuclei. In other areas there are, in addition to the above structures, typical interstitial cells also, as shown in figure 1 . The nuclei are round or oval in outUne, rather rich in chromatin and contain an evident nucleolus. The cytoplasm is granular and in certain areas, especially around the nucleus, it is conden-sed, while near the periphery of the cell it is much less condensed or even vacuolated.
Sections from a five-and-a-half months' testis show the tubules in an inactive state, without spermatozoa, and the intertubular connective tissue shghtly greater in amount than in the preceding chicken. The intertubular areas are somewhat larger, but
INTERSTITIAL CELLS IN CHICKEN S TESTIS
385
in other respects the appearances are quite similar to those observed in the cock of eighteen months.
In the sections of the three months' testis, the tubules are much smaller and lined with SertoU cells, imbedded iq which are numerous young sex cells. Relative to the tubules the intertubular spaces are far larger than in any of the older testes. The connective tissue with its spindle-shaped nuclei is readily differentiated from the interstitial cells. The former surrounds the tubules very closely, while the interstitial cells are usually located in the irregular areas formed by three or more tubules
Fig. 1 C. t., connective tissue; s. I., seminiferous tubule; i. c, interstitial cell; 6. c, blood cell.
coming close together. In most of these areas there are several interstitial cells; often they form large groups (fig. 2). The cell boundaries are more distinct than in any of the older testes and the cytoplasm is very much more vacuolated. Indeed some of the cell bodies appear ahnost clear, containing only the nucleus and a small amount of granular cytoplasm.
On examination of the sections stained for fat the interstitial cells in the three months' testis appear to be almost completely filled with fatty material (fig. 3). Thus the vacuolated appearance of the cells in figure 2 is explained. Most of the fat is within the interstitial cells, though there is a good deal free in the intertubular tissue and also a very small amount in the tub
386
T. B. REE^'ES
i. C.
|>
Fig. 2 S. <., seminiferous tubule; i. c, interstitial cell; c.<., connectivetissue. Fig. 3 .S. i., seminiferous tubule;/., fat; i. c, interstitial cell.
ules. In the older testes the fat is very much less in amount and appears as very small particles both in and outside of the interstitial cells. No attempt was made to determine the nature of the fatty material; as it was not rendered insoluble by Ciaccio's fixative, it probably does not consist of phosphatid lipoids to any large extent.
The primary object of this short study was merely to determine the presence or absence of interstitial cells in the testis of the domestic cock; there can be no doubt but that they are present in all the stages examined.
A SIMPLE METHOD OF BRAIN DISSECTION
PAUL E. LINEBACK
Harvard Medical School
FIVE F13URES
Every instructor realizes how hard it is to make clear to students the deep or inner structures of the brain. It is difficult to give them a lucid description of even the simpler and more superficial parts, but when it comes to explaining the intricate mechanism, it is an almost hopeless task.
Efforts have been made to disclose the regions, parts, tracts, and nuclear masses by means of a series of cross sections or fiber tract dissections. To all but those well trained in technique and familiar with the general make-up of the brain, these methods are confusing and difficult. Tract dissection necessitates a general understanding of how and where a tract runs, and a series of cross sections presents to a student a mass of labyrinthian vagaries. Most students remember an important structure in a cross section series as it appears in a few well-defined segments, but do not have a clear mental picture or distinct understanding of its extent and relationship. With the following method a student, being guided l)}'^ a few easily located landmarks, can get the greatest degree of clearness and satisfaction from his work, and have the least amount of cutting and mutilating of tissue.
The procedure is as follows: Using one-half of the brain, clear away all pia mater from the regions to be cut; sylvian fissure, central fissure (Rolandi), post central fissure, superior frontal sulcus, and about the uncus and temporal pole. This is important to make the field of operation perfectly clear and prevent blocking the knife. It is also important to use a long scalpel the blade of which should be about 7 or 8 cm. long and not more than 1 cm. in width; 0.5 cm. is still better. Place the hemisphere with frontal region upward or toward the student and depress the temporal pole sufficiently clearly to expose the uncus. Now cut across the upper part of this convolution, going from within outward and slightly downward, extending the cut about 2 cm. lateralward and the same backward (fig. 1). Make further depression of the temporal lobe there by widening the sylvian fissure, and cut at nearly right angle to the first incision along the lower border of the island (fig. 1). When this cut is extended 2 or 3 cm. directly backward, the tip of a cavity can be exposed, the anterior extremity of the inferior horn of the lateral ventricle.
387
388
PAUL E. LINEBACK
J.Arej?/rj_
-^^-e
Fig. 1 Showing hemisphere in position, frontal pole forward and temporal pole depressed to make first and second cuts.
Fig. 2 Showing median surface with third and fourth cuts made and knife in position making fifth cut.
SIMPLE METHOD OF BRAIN DISSECTION
389
Fig. 3 Showing 'removable' portion, with hippocampus, fornix, and inferior horn of lateral ventricle.
Fig. 4 Showing 'basal' portion with lateral ventricle and its floor and lateral boundary structure in clear viev/.
390
PAUL E. LINEBACK
ust back °f th^ ff ^™2Lhere begtaning he incision at a point on 1? t: callosum (genu, thus -•^■>« -,° ^ f? n•taScutS:>^h1
Fig. 5 Showing lateral surface and dotted line marking, on the surface, the course of the shoulder of the knife in making the fifth cut.
and make the following incision with the shoulder of the knife^^^^^^^^^^ excised portion to give clear view of the he 1 ^^^^P;^2;^^-3hrrp curve
rn=„rbtcf.rtrttpii? n^^^^^^^^^^^^^^
at Us lateral-most boun,la.y where that cav.ty turns downward. When
SIMPLE METHOD OF BRAIN DISSECTION 391
the shoulder comes to the sylvian fissure, lift the handle to about 60° with the horizontal plane, thus forcing the point downward, at the same time turning the sharp edge forward.
Now make a little more bold depression of the temporal lobe through the S3'lvian fissure, and continue the incision forward along the lower border of the island to meet the original cut at the uncus, all the while holding the knife at about 60° with the horizontal plane.
The incisions have now been completed. With the median surface facing upward, grasp the frontal lobe in the right hand and the occipital lobe in the left hand (this is for the left hemisphere; if the right hemisphere is used, the hands will be reversed), and carefully separate the two portions to about 3 cm. This wall stretch out the choroid membrane which can be easily followed almost throughout its entire extent of attachment. When this is carefully studied, the removable portion containing the hippocampal lobe with its fornix can be entirely withdrawn, and a clear view of the lateral ventricle will be had (fig. 3). The complete separation of the two parts will rupture the choroid membrane, but the ragged edges will still give a clear view of the line of its attachment. On the basal portion, in plain view, will be the structures forming the floor and lateral Isoundary of the lateral ventricle, caudate nucleus, taenia semicircularis, thalamus, etc. The optic tract, geniculate bodies, quadrigeminal bodies, and pes pedunculus can also be easily seen (fig. 4). The two segments can be easily, quickly, and repeatedly separated with no harm whatever to continuity of tissue. When the sections are in place, the cuts are scarcely perceptible; when removed, there is the greatest amount of exposure of hidden structures. A special advantage in this method is that specimens too soft for filler tract dissection or cross sectioning, or hardened after being mashed or pressed out of shape, can still be used with a good degree of satisfaction when cut as outlined above. There is to be offered this last and most important point, that the removable segment comes off from the basal portion in approximately the same course the hemisphere pursued in its early stages of development. Following this course of development as displayed by such a method of removing part of the hemisphere, it is much easier for the student to see how the velum interpositum was at one time a part of the wall, the roof portion of the forebrain, of the neural tube, and its presence in the fully developed specimen, attached to the sharp edge of the fornix on the one side and the taenia semicircularis and thalamus on the other, makes a closed cavity of the lateral ventricle and its horn.
This method is not to be used for complete work; the nuclear masses and fiber tracts demand deeper dissection. But using the method on one hemisphere and cross section on the other, the student will have far greater and more gratifying results, and will have good material, easily kept, for future reference.
Finally, I wish to express mj- appreciation to Dr. Bremer and Mr. IMiller for their kindness in reviewing this paper and making valuable suggestions.
THE AXATO.MICAL RECORD, VOL. 9, NO. O
BOOKS RECEIVED
The receipt of publications that may be sent to any of the five biological journals published by The Wistar Institute will be acknowledged under this heading. Short reviews of books that are of special interest to a large number of biologists will be published in this journal from time to time.
THE ANATOMY OF THE DOMESTIC ANIMALS. By Septimus Sisson, S.B., V.S., Professor of Comparative Anatomy, Ohio State University, College of Veterinary Medicine. Second edition entirely reset. Octavo of 930 pages, 724 illustrations. Philadelphia and London: W. B. Saunders Company, 1914. Cloth, $7.00 net; Half INIorocco, $8.50 net.
Preface to first edition. The lack of a modern and well-illustrated book on the structure of the principal domestic animals has been acutely felt for a long time by teachers, students, and practitioners of veterinary medicine. The work here offered is the expression of a desire to close this gap in our literature.
The study of frozen sections and of material which has been hardened by intravascular injection of formalin has profoundly modified our views concerning the natural shape of many of the viscera and has rendered possible much greater precision in topographic statements. The experience of the author during the last ten years, in which almost all of the material used for dissection and for frozen sections in the anatomical laboratory of this University has been hardened with formalin, has demonstrated that many of the current descriptions of the organs in animals contain the same sort of errors as those which prevailed in regard to similar structures in man previous to theadopt ion of modern methods of preparation.
While the method of treatment of the subject is essentially systematic, topography is not by any means neglected either in text or illustrations; it is hoped tha^ this will render the book ol value to the student in his clinical courses and to the practitioner. Embryological and histological data have been almost entirely excluded, since it was desired to offer a text-book of convenient size for the student and a work of ready reference for the practitioner. * * *
Preface to second edition. This book supersedes the author's Text-book of Veterinary Anatomy. A comparison of the two will show the new title to be justified by the extent and character of the changes which have been made.
Continued observations of well-hardened material and frozen sections have led to a considerable number of modifications of statement. It is scarcely necessary to say that the recent literature, so far as available, has been utilized.
Many changes in nomenclature have been made. Most of the synonyms have been dropped or relegated to foot-notes. Exceedingly few new names have been introduced. Nearly all eponyms have been eliminated, on the ground that they are not designative and are usually incorrect historically. The changes made in this respect are in confomiity with the report of the Committee on Revision of Anatomical Nomenclature which was adopted by the American Veterinary Medical Association two years ago. Progress in the direction of a simplified and uniform nomenclature is much impeded by the archaic terminology which persists to a large extent in clinical literature and instruction. * * *
Septimi^s Sis.son.
The Ohio State University, Columbus, Ohio, September, 1914.
392
THE ROLES OF NUCLEUS AND CYTOPLASM IN MELANIN ELABORATION
DAVENPORT HOOKER
From the Anatomical Laboratories of the Schools of Medicine of Yale University and the University of Pittsburgh
ONE FIGURE
The more recent work on the formation of melanin seeks to derive this pigment from chromatin elements. In 1889 Mertsching attributed the formation of melanin to the breaking down of the cell and especiallj^ to the destruction of the nucleus. Jarisch ('92) makes the statement, "Das Oberhautpigment entwickelt sich ^us einer Kernsubstanz, dem Chromatin, oder einem diesem chemisch oder wenigstens raumlich nahe stehenden Korper." Rossle ('04), in the discussion of pigment formation in melanosarcomas, recognizes fives stages in the process, based on the appearance of the nucleus. He believes the melanin granules to be small particles of chromatin which are extruded from the nucleus, impoverishing the latter in the process.
Aurel von Szily ("11) has made a notable contribution to the theor}^ of pigment formation from chromatic elements. He worked on the elaboration of melanin in developing eyes of a variety of vertebrates and in melanotic tumors of the human eye. According to his results, the melanin granules arise as colorless rod-like bodies (Pigmenttrager) extruded from the nucleus, being derived directly from chromatic elements. These Pigmenttrager are typical for species and locality of production. They also correspond exactly with the size and form of the melanin particles met with in that species and location. After being freed from the nucleus and wandering to a more or less peripheral position in the cell, the colorless Pigmenttrager become colored probably by the action of cell ferments. The pigmentation of the Pigmenttrager begins at one end of it and proceeds to the other. The origin of the Pigmenttrager from the chromatin and their transition into the cell cytoplasm
393
THE AKATOMICAL RECORD, VOL. 9, NO. 5 JCXE, 1915
394 DAVENPORT HOOKER
may be followed step by step. The nucleus which gives rise to them may be either productive or degenerative. In the former case, no impoverishment of the nucleus takes place; in the latter, pigment formation is accompanied by marked degenerative changes in the nucleus.
In 1910 Harrison presented evidence of a new type upon the question of pigment formation. In his paper on nerve growth in vitro, he mentioned and figured certain cells which became pigmented during the life of the cultures. To quote, Harrison found that the pigment first arose as a round mass of granules lying just to one side of the nucleus. This (mass) gradually increased in size and then the pigment granules became scattered through the cytoplasm" ('10, p. 812).
In 1912, while working on the reactions to light of embryonic connective tissue melanophores, material for a careful study of the actual elaboration of the melanin granules themselves was found in the developing connective tissue and epithelial cells of embryos of Rana pipiens. For this study, Harrison plasma cultures, living embryos and serial sections of carefully fixed embryos were used. The development of the melanin was followed in embryos varying in length from 3 to 10 mm.
Though the melanophoric cells found in the epidermis of older frog larvae are certainly mesodermal in origin, many ectodermal cells of young embryos elaborate this pigment within themselves. The pigment, however, after existing for a time in the cells, gradually disappears. That connective tissue cells elaborate melanin has long been an established fact. The mode of its development in these two types of cells is the same.
Plasma cultures. Small pieces of mesenchyme and epithelium from Rana pipiens embryos 3 to 4 mm. long were implanted in the plasma of frogs of varying species. Seventeen primary cultures were made. These lived in good condition over periods varying from three to forty days. In the case of the older cultures, the plasma was changed at frequent intervals. From these primary cultures, secondary were made, to the number of sixteen, by removing fragments of tissue and reimplanting in new plasma, fti all, thirty-three cultures were studied.
At first (fig. 1, A) the cells were clear and translucent, without
MELANIN ELABORATION
395
C D E
Fig. 1 Diagrammatic sketch to show the different stages of melanin eh^boration. The heavily outlined circle in the center represents the nucleus of the cell, the smaller ovals represent oil droplets. The cell is represented in median section.
pigment granules, but contained fat droplets scattered throughout the cytoplasm and a more or less centrally placed nucleus. Then a few small, spherical, brownish granules became visible in the cytoplasm of the cell immediately adjacent to the nucleus, appearing simultaneously, or nearly so, on all sides of it. Their number gradually increased until a well defined hollow sphere surrounded the nucleus (fig. 1, B). On their first appearance, these melanin granules presented all the characteristics which are normal for them. No positive evidence of any increase in size of the individual particles of pigment was obtained. Owing, however, to their minuteness, an accurate determination of any such change presents alinost insurmountable difficulties.
No granules made their appearance at any time inside the nuclear membrane nor in the cytoplasm of the cell away from the nucleus. The region of production was limited to that zone of the cytoplasm which is in contact with the nucleus. Nor was there any sign of the presence of colorless granules, which, by a process of pigmentation, could be transformed into the particles of melanin. The most careful search for any morphological structure within the cell that might serve as a Pigmenttrager in von Szily's sense, was unavailing.
With each succeeding day, the number of melanin granules
396 DAVENPORT HOOKER
increased. During the earlier stages of this process, they remained concentrated in the center of the cell, the periphery of the cloud of densely packed granules becoming larger and further away from the nucleus (fig. 1, C). Then the granules began to spread throughout the cell. This process was gradual and was accomplished by the slow separation of the individual elements of the dense cloud about the nucleus. At first, but a few granules, widely separated from one another, detached themselves from the central mass and wandered among the oil droplets nearest at hand. As this process of distribution, accompanied by further production of new granules around the nucleus, continued, more and more of the pigment particles spread toward the periphery (fig. 1, D), in such a manner that the number of granules per unit area steadily decreased from within outward. The periphery of the cell always contained fewer granules than the center until the final stage of melanin elaboration was reached (fig. 1, E). At this time, the cytoplasm of the cell was filled, one might say, to 'saturation.' Through this cloud of granules, the fat droplets and the nucleus were visible as clear, translucent areas, absolutely free from pigment. The entire process here described is completed in a period of eight to fourteen days, on an average.
Living embryos. By carefully dissecting out scraps of mesenchyme and epidermis from embryos of different ages and mounting them in plasma or isotonic saline (0.4 per cent), the different stages of melanin elaboration may be observed. That the steps thus seen are not continuous, but isolated from one another, is true. This objection to the nlethod is obviated, however, by the check provided by the cultures. Every step in the elaboration of melanin observed in the cultures is exactly duplicated in the normal body. The light brown granules of melanin are found, at first, only in the region of the nucleus and then spread through the cytoplasm of the cell. They are colored on their first appearance, a fact which seems clearly to do away with the colorless Pigmenttriiger idea.
Serial sections. Like the study of fresh material from li\'ing embryos, that of serial sections serves principally as a check upon the cultures. A description of the findings here would
MELANIN ELABORATION 397
be but a repetition of facts already stated, with one important exception. The most minute examination of series fixed in all stages of melanin formation fails to show the slightest change in character and content or the least sign of degeneration or depletion in the nucleus. Nor can any evidence be found to show that the nucleus plays a part in the formation of melanin by a process of extrusion of any of its elements into the cj^toplasm. All the granules of pigment are found in the cytoplasm near the nucleus, but they have no visible, structural connection with the nucleus or any of its contents.
Certain very important objections to the chromatin idea of the origin of melanin are evident from the results set forth above.
The nuclei of the pigment-forming cells suffer no depletion during the process Though von Szily claims that certain nuclei which he terms 'active,' develop pigment without loss to their content, the actual depletion of many others was also seen by him. Rossle describes minutelj^ many changes which occur in the nucleus and states that after the extrusion of chromatin to form melanin, the nucleus is bladderlike, with a reduced amount of chromatin.
No colorless anlagen for the melanin granules are to be found in the form of 'Pigmenttrager' (v. Szily, '11) or Tigmentbildner' (Fischel, '96). That a chemical anlage in the form of a chromogen is present is almost certain in view of the work of Bertrand ('08) but that the melanogen exists in the frog as a definite morphological structure, which, without any other change than in its coloration, becomes the pigment granule itself, may be denied.
The process of pigmentation of a melanophore in the frog begins in the area nearest the nucleus and spreads from that point throughout the cell, that is to say, it progresses from the center of the cell to the periphery. This is in direct opposition to the observations of von Szily, who states that the pigmentation of the colorless Pigmenttrager takes place gradually while they are wandering about in the cytoplasm. Indeed, his figure 4 (plate 4) shows the process going on irregularly throughout the cell and figure 8 of the same plate illustrates a process directly the reverse of that noted in this paper, namely, the appearance
398 DAVENPORT HOOKER
of pigmented granules at the periphery of the cell before they are present near the nucleus.
The last, and probably the most important, objection to the supposed chromatin origin of melanin granules is the fact that no process of extrusion of chromatin nor any of the steps of such a process are to be observed. The pigment granules appear near the nucleus, in fact, in almost direct apposition to the nucleus, but no evidence was found in this work which even suggests a morphological relationship to either the nucleus or its contents.
It may safel}^ be concluded that, in the normal ontogenetic origin of melanin in the frog, the chromatin plays no direct role. On the contrary, all the evidence obtained goes to demonstrate that the melanin granules are formed in the cytoplasm, from elements already present in solution in it, through some action of the nucleus.
Bertrand ('96) isolated an enzyme in plants (Russula and Dahlia) which, by its oxidizing action on tyrosin, was named tyrosinase. Von Fiirth and Schneider ('01) found this same ferment in the haemolymph of Lepidopteran larvae and noted its occurrence in many animal forms. The action of this enzyme on tyrosin gives a melanin and von Fiirth suggested,
. . . . dass die physio logische Bildung melaninartiger Pigmente in den tierischen Geweben auf das Zusammenvvirken von zweierlei Fermenten zuruckzufiihren sei: durch ein autolytisches Ferment konnte ein aromatischer Komplex aus dem Eiweismaterial abgespalten und dieser sodann durch einc^ Tyrosinase in ein Melanin iibergeftihrt werden. (1901, p. 242). 1
The remarkable results of Bertrand's ('08) more recent work demonstrate the manner in which tyrosin and its derivatives may form the various types of the melanins usually met with in the animal body. He determined that many substances may be transformed into melanin by the oxidizing action of tyrosinase, each giving a characteristic color. During oxidation, a play of colors results, the earlier stages of the process giving lighter colors than the more advanced. The essential constitu ^ It is not within the province of this paper to review in detail the literature on the melanins. The reader is referred to the excellent Sammelreferat of von Fiirth ('04).
MELANIN ELABORATION 399
ent seems to be a benzene ring with an hydroxyl radicle. Tyrosin itself gives a black pigment, while paraoxyphenylacetic and paraoxyphenylpropionic acids give browns. It should be remembered, however, that the individual granules of ^black' melanin are brown; those of 'brown' melanins, yellowish in color. Gessard ('03) has given the strongest evidence yet adduced for the actual formation of melanin in the animal body from tyrosin. Working on melanotic tumors in horses, he determined the presence of free tyrosin and adds: "La tyrosine est done le chromogene dont I'oxidation par la tyrosinase determine la formation du pigment noir commun a divers produits physiologiques et pathologiques de I'economie animal." (p. 1088).
The recent work on protein digestion demonstrates that aminoacids are absorbed, unchanged, by the blood stream from the ahmentary canal and are distributed to the tissues (Folin, '14). Van Slyke and Meyer ('13) have shown that "the disappearance of intravenously injected amino-acids from the circulation is the result of neither their destruction, synthesis nor chemical incorporation into cell proteins. The acids are merely absorbed from the blood by the tissues, without undergoing any inmiediate chemical change." They have also demonstrated that there is a limit to the amourlt of amino-acids that may be absorbed by the tissues, so that a certain equilibrium exists between the blood and the tissues so far as the amino-acids are concerned. Further, Osborne and Alendel ('12) have shown that, while certain amino-acid groups will sustain life if fed as an exclusive diet, on the other hand it is clear that when certain aminoacid groups are lacking, nutritive equilibrium is impossible. The cyclic derivatives, tyrosin and trj^ptophane, appear to be included here" (p. 326).
Several of the protein putrefaction products mentioned by Bertrand as sources of melanin, as paracresol, paraoxyphenylacetic acid and others, all derivatives of tyrosin, also are known to be absorbed as such by the organism. There can be but little doubt that sufficient quantities of melanin-forming substances occur normally in the body.
" H. Eppinger ('10) isolated a melanogen from the urine of patients suffering from melanosarcoma which turned black on oxidation. This he believes to be derived, not from a tyrosin base, but from tryptophane.
400 DAVENPORT HOOKER
J. Loeb has repeatedly made the statement that oxidation is a prominent function of the nucleus in normal development and regeneration. Perhaps the most striking proof of this fact in specific tissues in the adult has been given by the work of R. S. Lillie ('02). By soaking thin slices of living tissues in solutions of substances which, colorless in the unoxidized condition, give brilliant color reactions on oxidation, he was able to determine the exact location in the individual cell where this reaction proceeds to the greatest extent. His findings furnish, it is believed, conclusive evidence that in many tissues the nucleus is the chief agency in the intracellular activation of oxygen; and, further, that the active or atomic oxygen is in general most abundantly freed at the surface of contact between nucleus and cytoplasm" ('02, p. 420).
The findings in the normal development of melanin in the embryonic frog furnish strong histological evidence that the nucleus of the cells elaborating this pigment provides something vitally necessary for its production. The melanin granules appear, not in haphazard manner throughout the cell, but in the cytoplasm immediately about the nucleus or, in Lillie's words, ^'at the surface of contact between nucleus and cytoplasm." Lillie's work seems to indicate clearly that 'the vitally necessary element for melanin elaboration provided by the nucleus is an oxidizer. Jaquet in 1892 demonstrated that the oxidizing action of the cell was not alone a property of living tissue, but was also evinced by broken-down cells which were no longer living. Nevertheless, the oxidases present in dead cells were originally elaborated by the nucleus. Lillie's work demonstrates this.
The particular form which the oxidizing action of the nucleus takes in melanin elaboration is that of an oxidase, perhaps of a type of tyrosinase. A host of investigators, following in the footsteps of Bertrand's ('96) original discovery of the presence of tyrosinase in plants, have isolated this enzyme in many animal forms and in such bodily positions as to serve normally for the manufacture of melanin.
The data derived from these various sources may be briefly summed up as follows:
1. Tyrosin, or its derivatives, acted upon by an oxidizing
MELAXIX ELABORATION 401
agent, tyrosinase, gives a melanin. (Bertrand, '96 and '08; von Ftirth and Schneider, '01, etc.)
2. Free tyrosin was discovered by Gessard ('03) in horses with melanotic tumors and it is now a well krfown fact that derivatives of tyrosin are absorbed by the animal body.
3. Lillie ('02) gave definite proof of the role played by the nucleus as a producer of oxygen or of an oxidase.
4. The normal presence of tyrosinase discovered in many parts of the body by Gessard ('01, '02, '03,) Przibram ('01), ^ Dewitz ('02), Durham ('04), Weindl ('07), etc.
When the histological data presented in this paper are considered in connection with the facts just reviewed, it will be seen that they are in full accord with one another. While no evidence has been obtained from this work that tyrosin is present in the cells under consideration, it is shown that the base from which the melanin granules are formed probably exists in a soluble condition in the cytoplasm. The role of the cytoplasm, then, is that of a carrier of the chromogen. That the nucleus plays an all important role is evident.
CONCLUSIONS
It is felt that the evidence here brought forward demonstrates conclusively the following points:
1. That the theory of the origin of melanin from chromatin elements extruded from the nucleus into the cytoplasm is untenable, at least in the frog.
2. That, however, the nucleus plays an essential part in pigment formation by some activity which greatly resembles an oxidizing action.
3. That melanin is formed in the cytoplasm of the cell at the point of known greatest efficiency of the nucleus as an oxidizing agent.
The following general conclusion from these facts seems justified: that, in the cells of embryo frogs, melanin is formed from some substance (probably tyrosin or its derivatives) in solution in the cytoplasm when acted upon by the nucleus (perhaps an oxidase reaction).
Anatomical Laboratory, University of Pittsburgh
^ Evidence given by von Fiirth and Schneider ('01, p. 241).
402 DAVENPORT HOOKER
LITERATURE CITED
Bertrand, G. 1896 Sur une nouvelle oxydase ou ferment soluble oxydant
d'origine vegetale. Compt. rend, de I'Acad. d. Sci., torn. 122, p. 1215.
1908 Recherche§ sur la melanogenesis : Action de la tyrosinase sur
la tyrosine. Ann. de I'lnst. Pasteur, torn. 22, p, 381. Dewitz, J. 1902 Recherches experimentales sur la metamorphose des insectes.
Compt. rend. d. 1. Soc. Biol., torn. 54, p. 44. Durham, F. 1904 On the presence of tyrosinase in the skins of some pigmented
vertebrates. Proc. R. S. London, vol. 74. Eppinger, H. 1910 Uber Melanurie. Biochem. Zeitschr., Bd. 28, p. 181. FiscHEL, A. 1896 Uber Beeinflussung und Entwicklung des Pigmentes. Arch.
f. mikr. Anat., Bd. 47, p. 719. FoLiN, O. 1914 Intermediary protein metabolism. Jour. A. M. A., Vol. 63,
p. 823. VON FtJRTH, O. 1904 Physiologische und chemische Untersuchungen liber
melanotische Pigmente. (Sammelreferat). Centralbl. f. allg. Path.
u. path. Anat., Bd. 15, p. 617. VON FtJRTH, O. UND ScHNEiDER, H. 1901 Uber tierische Tyrosinasen und
ihre Beziehung zur Pigmentbildung. Hofmeister's Beitrage z. chem.
Physiol, u. Path., Bd. 1, p, 229, 1901-02. Gessard, C. 1901 Etudes sur la tyrosinase. Ann. de I'lnst. Pasteur, torn.
15, p. 593.
1902 Tyrosinase animale. Compt. rend. d. 1. Soc. Biol., tom. 54, p. 1304.
1903 Sur la formation du pigment melanique dans les tumeurs du cheval. Compt. rend. d. 1. Soc. Biol., tom. 136, p. 1086.
Harrison, R. G. 1910 The outgrowth of the nerve fiber as a form of protoplasmic movement. Jour. Exp. Zool., vol. 9, p. 787.
Jarisch 1892 Uber die Bildung des Pigmentes in den Oberhautzelien. Arch, f. Dermat. u. Syphlis, Bd. 23, p. 223.
LiLLiE, R. S. 1902 On the oxidative properties of the cell nucleus. Am. Journ. Physiol., vol. 7, p. 412.
Mertsching 1889 Histologische Studien iiber Keratohyalin und Pigment. Virchow's Arch., Bd. 116, p. 484.
Osborne, T. B. and Mendel, L. B. 1914 Amino-acids in nutrition and growth. Jour. Biol. Chem., Vol. 17, p 325.
RossLE, R. 1904 Die Pigmentierungvorgang in Melanosarkom. Zeitschr. f. Krebsforschung, Bd. 2, p. 291.
Van Slyke, D. D. and Meyer, G. M. 1913 The fate of protein digestion products in the body. III. The absorption of amino-acids from the blood by the tissues. Jour. Biol. Chem., Vol 16, p. 197.
VON SziLY, A. 1911 Uber die Entstchung des melanotischen Pigmentes im Auge der Wirbelticrembryonen und in Chorioidealsarkomen. Arch. f. mikr. Anat., Bd. 77, p. 87.
Weindl, T. 1907 Pigmententstehung auf Grund vorgebildcter Tyrosinasen. Arch. f. Entw-mech., Bd. 23, p. 632.
ON THE NORMAL SEX RATIO AND THE SIZE OF
THE LITTER IN THE ALBINO RAT (MUS
NORVEGICUS ALBINUS)
HELEN DEAN KING AND J. M. STOTSENBURG
From the Wistar Institute of Anatomy and Biology
ONK FIGURE
Literature dealing with the early development of the albino rat contains references to but two papers that give information regarding the normal sex ratio and litter size in this animal (Cuenot '99; King '11). Marked differences in the results of these two sets of investigations, which were made on relatively small numbers of individuals, render it necessary that a large series of observations should be recorded in order to furnish adequate standards by which one can judge the effects of experiments aiming to modify the sex ratio or to alter the fertility of the albino rat. To supply the material for such standards the data given in the present paper were collected.
All of the records given are of litters cast by stock albino rats kept in the animal colony of The Wistar Institute. During the period when the data were being collected (1911-1914) all of the animals used for breeding were subjected to similar environmental conditions, and they all were fed on a mixed diet that experience has shown is necessary if rats are to be kept in good condition for any length of time.
THE NORMAL SEX RATIO IN THE ALBINO RAT
Practically all of the data were obtained by examining litters at or very shortly after their birth, since the sexes can readily be distinguished at this time as Jackson ('12) has shown. The removal of the young rats from the nest entails some risk that the mother will not care for them after they are replaced, but it is necessary that the records be taken at this time if one wishes
403
404
HELEN DEAN KING AND J, M. STOTSENBURG
an accurate determination of the sex ratio or of the Utter size. Not infrequently Utters contain one or more stiUborn young which are usuaUy eaten by the mother within a few hours after their birth. Often, too, some individuals in the litter, particularly^ if the Utter is large, will be killed by the mother when they are several days old, or if one or more of the young rats in a large litter are constitutionally weak they will die from lack of nourishment, being unable to cope with their stronger brothers in their efforts to obtain food.
Xo attempt was made to obtain the sex records for all of the litters of stock albino rats that were born in the colony during the years 1911-1913. The data that were collected during this period have been grouped together, according to the months when the Utters were cast, and are given in table 1.
TABLE 1
Showing the sex raiios and the average number of young in litters of stock albino rats born during 1911-1913. Data arranged according to the months when the litters were cast
January. . . February. .
March
April
May
June
July
August. . . . September. October. . . November. December.
NUMBER
NUMBER
AVERAGE
OF LITTERS
OF INDIVIDUALS
M.U-ES
FEMALES
MALES TO
100 FEMALES
NO. YOUNG
PER
LITTER
28
194
103
91
113.2
6.9
18
123
65
58
112.1
6.8
32
236 !
113
123
92.2
7.3
16
101 j
47
54
87.0
6.3
21
135
69
66
104.5
6.4
27
194
108
86
125.6
7.1
22
160
87
73
119.2
7.2
12
77
40
37
108.1
6.4
11
80
38
42
90.5
7.2
46
316
159
157
101.3
6.8
16
117
65
52
125.0
7.3
26
195 1
102
93
109.7
7.5
275
1928
996
932
106.9
7.01
One fact clearly brought out in the above table is that there is no restricted breeding season for the albino rat. Litters are cast during every month of the year, but, as the records for many thousands of litters show, relatively more litters are pro
NORMAL SEX RATIO AND LITTER SIZE IN RAT
405
duced in the spring than during other seasons of the year. In table 1 the sex ratios for the different groups of Utters do not show a very great range of variation considering the small number of litters involved. The highest sex ratio is that for the 27 litters cast during the month of June; the lowest sex ratio is found in the litters of the April group. For the entire series of 275 litters the sex ratio is 106.9 males to 100 females.
During the year 1914 an attempt was made to obtain the sex data for as many as possible of the litters of stock albino rats born in the colony. The cages containing the breeding animals were examined nearly every day throughout the year and practically all of the litters cast were recorded. The data obtained, arranged according to the months when the litters were cast, are given in table 2.
TABLE 2
Showing the sex ratios and the average number of young in litters of stock albino rats born during 191 4. Data arranged according to the months when the litters were cast
NUMBER
NUMBER
AVERAGE
MONTHS
OP
LITTEKS
OF IXDIVID. U.\I.S
MALEiS
FEM-\LES
MALES TO 100
FEMALES
NO. YOUNG
PER
LITTER
January
57
407
202
205
98.5
7.1
February
56
410
197
213
92.5
7.3
March
58
432
210
222
94.6
7.4
April
51
367
199
168
118.4
7.1
May
60
430
217
213
101.9
7.1
June
101
744
387
357
108.4
7.3
July
116
821
430
391
109.9
7.0
August
109
776
438
338
129.6
7.1
September
111
751
377
374
100.8
6 7
October
31
187
104
83
125.3
6.0
November
33
188
99
89
111.2
5.7
December
31
178
96
82
117.1
5.7
SI 4
5691
2956
2735
108.1
6.99
Although the number of records taken during the year 1914 is about three times greater than that collected during the period 1911-1913, the range of variation in the sex ratios of the Utters cast during the various months is only slightly greater than that given in table 1 . The lowest sex ratio in this series
406
HELEN DEAN KING AND J. M. STOTSENBURG
of records is found among the litters cast in February; the highest sex ratio occurs in the Utters born in August. The sex ratio of the 814 Utters examined during the entire year is 108.1 males to 100 females. This sex ratio is remarkably close to that found in the 275 litters previously recorded (table 1).
A summary for all of the data collected is given in table 3. In order to give equal value to the two sets of records the sex ratios in this table, and also the averages for the size of the litters cast in the various months, represent the arithmetical mean of the records as given in table 1 and in table 2; they have not been computed in any instance on a litter basis.
TABLE 3 A combination of the data given in table 1 and in table 2
NUMBER
OF LITTERS
.January 85
P'ebruary 74
March 90
April 67
May 81 June 128
July ; 138
August '■• 121
September 122
October 77
November. . . '. 49
December .57
1089
NUMBER
NUMBER
AVERAGE
OF INDIVID
M.VLES
FEM.'K.LES
MALES TO
100
NO. YOUNG PER
UALS
FEMALES
LITTER
601
305
296
105.7
7.0
533 •
262
271
102.3
7.0
668
323
345
93.4
7.3
468
246
222
102.7
6.7
565
286
279
103.2
6.7
938
495
443
116.9
7.2
981
517
464
114.6
7.1
853
478
375
118.8
6.7
831
415
416
95.6
6.9
503
263
240
113.3
6.4
305
164
141
118.1
6.5
373
198
175
113.4
6.6
7619
3952
3667
107.5
7.0
As arranged in table 3, the data show that the sex ratios are somewhat higher in the litters cast during the latter part of the year than in those cast in the early part of the year. With the exception of the record for. March the sex ratios for the litter groups from January to May show a variation of less than three points; and the sex ratios for the litters cast from June to December, omitting the record for September, vary less than four points. The pronounced drop in the sex ratio for the litters produced during September is found in both sets of records, and at present there is no satisfactory explanation for it.
NORMAL SEX RATIO AND LITTER SIZE IN RAT
407
In the total of 1089 litters examined there were 3952 males and 3667 females, giving a sex ratio for the series of 107.5 males to 100 females. This sex ratio is somewhat higher than that given by Cuenot, who found in 30 litters of albino rats a sex ratio of 105.6 males to 100 females, but it is practically the same as that given by King ('11) for 80 litters of albino rats (107.3 males to 100 females). The sex ratio found among adult rats is doubtless considerably lower than that given above, as growth experiments with the albino rat at present under way seem to indicate that female rats, as a general thing, live longer than male rats and show somewhat less susceptibility to disease at all stages of their growth.
It would be futile to make a comparison between the sex ratios of the various litter groups owing to the inequality in the number of litters recorded for the different months. For the purpose of a somewhat closer analysis than that given above, the two sets of records have been grouped in table 4 according to the season of the year when the litters were cast. The averages given for the two sets of records were obtained in the same manner as were the averages in table 3.
TABLE 4
Showing the data for sex ratios and size of the litters in the albino rat arranged
according to the season of the year when the litters were cast
1911-1913
1914
1911-1911
SEASONS
NUMBER OF LITTER3
NUMBER MALES TO
100 FEM.^LES
a S a ■^ a.
NUMBER
OF LITTERS
NUMBER MALES TO
100 FEMALES
la ^ a
NUMBER
MALES TO
100 .
FEMALES
a z a
March to May
June to August . . .
September to November
December to February
69 61
73
72
94.2 119.9
104.4
111.6
6.8 7.0
7.0
7.1
169 326
175.
144
103.8 115.6
106.2
99.0
7.2 7.1
6.4
6.9
99.0 117.7
105.3
105.3
7.0 7.05
6.7
7.0
275.
106.9
7.01
814
108.1
6.99
107.5
7.0
There is a very striking agreement between the corresponding sex ratios for the two sets of records, as is shown in table 4. In each case the sex ratio for the litters cast in the spring
408 HELEN DEAN KING AND J. M. STOTSENBURG
1 1 1 1 1 1 1 1 1 1 1 1
I
Number males to 100 females,
/
\
^
^/
V
\
/,
/
^^^
^
li
^
■s
N
\
110
r
"^
^
—A
/■
s
s
y>.
<
5n"
<^\
/
'/,
N
N
^
-^
'V9^
y
/
/
"^
■>
<:
^
—
C
/
/
/
■^
^
100
/
/
%
'^
^
/
/r
?
V.
— B
M
f
90
Spring
Summer
Autumn
Winter
_
_
_
1 i i
1 1 1
L
1 1 1
1 1
Mar. -Way
June-Au9.
Sept -Nov.
Dec. -Feb.
Fig. 1 Graphs showing variations in the sex ratio of the albino rat at different seasons of the year. A, graph constructed from data for litters cast during 1911-1913; B, graph constructed from data for litters cast during the year 1914; C, graph constructed from the averages for the two sets of data.
is considerably below the normal sex ratio of 107 males to 100 females; the average for the two groups giving a sex ratio of only 99.0 males to 100 females. Each set of data shows likewise a sharp rise in the sex ratios of the litters born during the summer months and then a drop to below the normal ratio for the litters born in the fall. The two sets of records for the litters cast in the winter months do not, for some reason, show the same agreement as those for the litters produced in other seasons of the year, as in one case the sex ratio is somewhat above the normal and in the other case it is below the normal.
Figure 1 shows graphs, constructed from the data given in table 4, which bring out very clearly the changes in the sex ratio that are found to occur among rats born at different seasons of the year.
Judging from observations and from the records for several thousand litters cast in our colony during the past six years.
NORMAL SEX RATIO AND LITTER SIZE IN RAT
409
the rat breeds more readily in the spring than in any other season of the year, and there is a second, less pronounced, period of sexual activity in the early fall. The lowest points in the graphs shown in figure 1 are found to coincide with the period in which the greatest sexual activity occurs. Lacking adequate means for heat regulation the rat suffers greatly from heat during the summer months, and for years the highest mortality among the animals in our colony has occurred in July and in August while relatively fewer litters are produced at this time than at other seasons of the year. It is during the hot weather when the breeding animals are not in the best physical condition that the litters produced show the highest sex ratio, as is indicated by the graphs in figure 1.
The seasonal variation in the sex ratios that is shown by these records cannot be ascribed to environmental conditions other than temperature, since the routine of caring for the animals in our colony is the same throughout the year and there is no change in the character of the food.
That the sex ratios in various mammals seem to show a pronounced variation at different seasons of the year has long been known. From the large body of statistics examined by Diising ('83) it appears that relatively more boys are born during the winter than during the summer months. Table 5, compiled from data collected by Wilckens ('86) and by Heape('08), shows the apparent seasonal variation in the sex ratio that occurs in the young of various kinds of domestic animals.
TABLE 5
Showing seaso7ial variations iti the sex ratios of some dornes tic animals. Data
collected by Wilckens and by Heape
NUMBER IXDIVIDU.\LS
NUMBER MALES TO 100 FEMALES
Birth in warm Birth in cold
mos. mos.
1
Birth during entire year
Horses
Cattle
Sheep
Swine
Greyhounds
16,091
4,900 6,751 2,357*
17,838
96.6 97.3 114.1 103.0 102.9 94.0 115.0 109.3 116.3 122.1
97.0 107.3
97.4 111.8 118.5
THE ANATOMICAL RECORD, VOL. 9, NO. 6
410 HELEN DEAN KING AND J. M. STOTSENBURG
Except in the dog, and in the horse where these statistics are at variance with those collected by Schlechter ('84), the sex ratios as given in table 5 are relatively high among the animals born in the warm months and correspondingly low where the births occurred during the cold months. To be available for analysis by any current theory of sex-determination, however, these records would have to be arranged according to the time when conception occurred, since it seems most probable that sex is determined at or before the time of the fertilization of the ovum and cannot be altered by the nutritive or other environmental conditions to which the embryo is subjected. The gestation period in the rat is so short, only 21 days, that the time of conception and the time of birth may be said to take place in the same season of the year. Since the gestation periods in the various animals for which sex ratios are given in table 5 vary so greatly, the sex records cannot be arranged on any basis except that of the time of birth, and they are of value, therefore, merel}^ as indicating that there is apparently a seasonal variation in the sex ratio of other animals as well as in that of the albino rat.
If it can be shown bj^ a sufficiently large body of statistics that the sex ratio in various animals changes in a definite direction with the time of year at which conception occurs it will indicate that some metabolic process occurs in one or the other or in both of the parent organisms at stated periods which tends to swing the sex ratio in one direction rather than in the other. Assuming that sex is determined by the chromatin constitution of the spermatozoan that fertilizes the egg, we must add to this theory the probability that some form of chemical attraction or repulsion exists between each ovum and one kind of spermatozoan in order to account for the constantly increasing mass of evidence that under changed environmental conditions sex ratios in various animals can be altered in a definite direction. Chance, therefore, cannot play as important a role in the process of sex-determination as some investigators, have maintained, and any egg is not fertilized by any spermatozoan that happens to come in contact with it. The laws of chance, according to our
NORMAL SEX RATIO AND LITTER SIZE IX RAT 411
present conception, are not subject to periodic changes in their action, and while they offer a very attractive explanation for the existence of an equality of the sex in certain species, they utterly fail to explain sex ratios that vary in a definite direction, whether as the result of seasonal changes or as the outcome of experimental attempts to modify the sex ratio.
THE EFFECTS OF THE AGE OF THE MOTHER OX THE SEX RATIO
OF HER YOUXG
It has been stated by many investigators that the age of the mother has a pronounced influence in determining the sex of her young. According to a considerable body of statistics collected by Punnett ('03) the sex ratio among the first children in a family is 140 boys to 100 girls. This ratio falls to 117 boys to 100 girls for the second births among the children of these same mothers, and it then declines steadily until, at the ninth birth, the chances for the two sexes are about even. In a compilation of birth statistics for the first born of women of various ages, Bidder (78) found that the sex ratio was 122.2 boys to 100 girls when the mothers were under 19 years of age; this ratio falls to 104.6 boys to 100 girls for the children of women between 20-30 years of age and it then rises to 131 boys to 100 girls when the first conception occurs after the woman has reached 40 years of age. Conditions closely paralleling these for man are found in the horse according to "Wilckens, but this investigator states that heifers predominate among the first offspring of cattle.
Data given by Copeman and Parsons ('04) from their inbreeding experiments with mice show the relation between the age of the mother and the sex of the offspring as given in table 6. Normally there is about an equal proportion of the sexes in mice as is shown by the investigations of Schultze ('03) and of Welden ('06).
The sex records for the mouse, as given in table 6, agree with those for man and for the horse in that they show that the sex ratio in the young is at its lowest point when the mother is at
412
HELEN DEAN KING AND J. M. STOTSENBURG
the height of her reproductive powers. Schultze, on the other hand, states that young female mice tend to produce a slight excess of females among their young, and he concludes that the age of the mother has no effect whatever on the sex of her offspring.
TABLE 6
Showing the effects of the age of the mother on the sex ratio of mice. Data collected by Copeman and Parsons
AGE OF FEMALE AT NUMBER OF NUMBER OF M.VLES
CONCEPTION LITTERS TO 100 FEMALES
2mos 21 103.7
3-5 mos 27 96.5
6mos 21 123.3
For comparison with the records given by Copeman and Parsons and by others we have the sex data for 75 litters cast by 21 stock albino rats. These data, arranged according to the location of the litter in the litter series, are given in table 7.
TABLE 7
Showing the sex ratios and average number of young in 75 litters of stock albino rats. Data arranged according to the position of the litters in the litter series
LITTER SERIES
NUMBER
OF LITTERS
NUMBER
OF INDIVIDUALS
M.\LES
FE.MALES
NUMBER M.\LES TO
100 FEMALES
AVERAGE
NO. YOUNG
PER
LITTER
1
2
3
21 21 18 15
131 '
162
127
96
72 85 64 41
! 59 77 63 55
122.0
110.4
101.6
74.5
6.2
7.7 7.0
4
6.4
75
516
262
254
103.1
6.8
At the time that the first litter was cast each of the 21 females was about three months old. As shown in table 7, the sex ratio in the j^oung rats belonging to the first litters is 122.0 males to 100 females. For the individuals in the second litters the sex ratio drops to 110.4 males to 100 females, and it goes down to 101.6 males to 100 females for the rats belonging to the third litters. At the time that the females cast their fourth litters the majority of them were seven to nine months old.
NORMAL SEX RATIO AXD LITTER SIZE IN RAT 413
The female albino rat, if she is in good physical condition, will continue to bear young until she is about fifteen months old. The third and the fourth litters of an albino female, therefore, are usually cast during the period when the female is at the height of her reproductive power. In the above table the sex ratio for the fourth Utters is much lower than that for the first three litters, being only 74.5 males to 100 females.
The records given in table 7 are, of course, too few to furnish evidence from which very definite conclusions can be drawn. As far as they go, however, these records indicate that the sex ratio among the first offspring of very young females is higher than that found among the offspring of the same females at a period of life when they are at the height of their reproductive power. The results, therefore, are in agreement with those obtained by Piinnett, by Bidder and by Copeman and Parsons. In what way the age of the mother can affect the sex of her offspring is not known as yet. The fact that female rats at the height of their sexual activity in the spring and fall and also at the zenith of their reproductive powder tend to produce relatively more female than male young would seem to indicate that the physical condition of the female, either as the result of age or of environment, produces changes of metabolism that tend to affect the sex of the young. It is possible that anabolic processes predominating in the female at certain periods might affect the ova in such a way as to cause them to be more easily fertilized by a female-producing than by a male-producing spermatozoan. In very young females, on the other hand, and in females not in good physical condition, katabohc processes that would give the male-producing spermatozoa an advantage over the female-producing spermatozoa in the fertilization of the ova, might be assumed to occur. Until, however, our knowledge of the mechanism of sex determination rests on a more secure foundation than it does at the present time, it seems useless to offer even tentative suggestions as to the manner in which this mechanism can be influenced.
414 HELEN DEAX KING AND J. M. STOTSENBURG
THE RELATION BETWEEN THE SIZE OF A LITTER AND THE SEX
OF ITS MEMBERS
E\TLdence for man as to whether one sex or the other tends to predominate in large f amihes is conflicting. According to Nichols (/07), it has been shown by several investigators, particularly bj' Geissler ('89), that in large f amihes there is a greater proportion of sons than in small famihes. Geissler's statistics show that in 159,042 families containing more than seven children the sex ratio was 106.8 boj's to 100 girls, while in 839,719 families ha\'ing from two to seven children each there were only 105.8 boys to 100 girls. From the statistics of a very much smaller number of families, Punnett ('03) comes to the opposite conclusion that girls tend to predominate more in large families than in small ones.
Copeman and Parsons's breeding experiments with mice show that the percentage of males is shghtly less in large litters (containing more than 6 young) than it is in small litters. Welden, on the contrary, states that in a given generation of mice there seems to be a positive tendency for large litters to contain more males than females.
The sex data for 1089 litters of albino rats have been arranged on the basis of litter size in order to ascertain if, in this animal, there is any relation between the sex of the individuals and the size of the Utters to which they belong. For the purpose of this analysis the litters have been arbitrarily divided into three groups: large litters containing nine or more young; medium litters with six to eight young; small Utters having Uve or less members. The records collected during the year 1914 are sufficiently numerous to warrant their separation into groups according to the months when the litters were cast; the data obtained during 1911-1913, being too few to be divided in a similar way, have been grouped together. The results of this arrangement of data are given in table 8.
As shown in table 8, the results obtained bj^ this analysis are so conflicting that no definite conclusions can be drawn from them. The data for the year 1914, arranged according to the months when the litters were cast, show that the highest sex
NORMAL SEX RATIO AND LITTER SIZE IN RAT
415
TABLE 8
Showing the sex ratio in different sized litters of albino rats. Data collected during 1914 arranged according to the months when the litters were cast
9 OR MORE YOUNG
January. . . February. .
March
April
May
June
July
August. . . . September. October. . . November. December.
Number litters
14
19
24
14
23
34
36
29
28
1
3
3
Number males to
100 females
108.7
98.9
87.1
112.3
104.6
100.6
97.7
129.9
100.0
125.0
93.8
130.8
6 TO 8 YOUNG
O OR LESS YOUNG
Number [ Number
Number 1 males to Number I males to
litters : 100 litters 100
females ; females
27 25 22 24 19 42 56 57 46 18 15 15
93.1
87.6
98.7
144.3
85.9
109.9
120.6
134.1
98.1
119.3
123.4
106.3
16 12 12 13 18 25 24 23 37 12 15
91.7 87.5 125.0 75.8 126.5 131.3 116.7 111.4 107.8 140.9 100.0
Data for 1911-1913.
228
65
Total.
293
103.7 109.9 106.8
366
110.7
142 508
110.5
110.6
13 220
133.3
111.7
68
90.2
288 ! 100.7
ratio occurs in the members of the largest htters in onl}^ two cases, while in six cases it is found in the individuals comprising the smallest litters. In the records for the entire year the highest sex ratio, 111.7 males to 100 females, occurs in the indi\dduals composing the smallest litters; the lowest sex ratio, 103.7 males to 100 females, being found in the rats belonging to the largest litters. The records fol- 1911-1913, on the other hand, give the highest sex ratio, 110.5 males to 100 females, in the indi\dduals belonging to htters of medium size; the records for the small litters show a sex ratio of only 90.2 males to 100 females. For the entire series of data, litters of medium size show the highest sex ratio, 110.6 males to 100 females, and the lowest sex ratio occurs in the individuals of the small litters.
The lack of uniformitj' in the results of this arrangement of data indicate that apparently there is no well defined relation between litter size and sex in the albino rat.
416 HELEN DEAN KING AND J. M. STOTSENBURG
THE NORMAL SIZE OF THE LITTER IN ALBINO RATS
Available data concerning litter size in the rat indicate that the average number of young in a litter varies considerably in different species. Miller ('11) finds for the common gray rat (Mus norvegicus) that there is a range of 7 to 12 young in the litter and that, on the average, a Htter contains 10.5 young. Data recorded by Lantz ('10) give 8.1 as the average number of young in a large series of pregnant females of this species killed in India. Litters of the black rat (Mus rattus) are apparently much smaller than those of the gray rat. Lantz states that 5.2 young is the average for the Utters of this species. This average is practically the same as that given by Lloyd ('09).
But few observations have been recorded regarding litter size in the albino rat. Crampe ('84) states that the average size of a litter of albino rats is 5.6 young, which is exactly the result obtained by one of us (King '11) from an examination of 80 litters of stock albino rats. Cuenot records 8.5 as the average number of young in 30 litters of albino rats, but this is undoubtedly a higher average than would be found in a larger series of litters.
In addition to the sex ratios tables 1-4 give the average number of young in the various litters examined during the years 1911-1914. The records for the period from 1911-1913, as given in table 1, show that there is very little variation in litter size in the various groups of litters cast during the different months of the year; the range being from 6.3 young, the average size of the litters cast during April, to 7.5 young, the average of the litters produced during December. The largest litter examined contained 14 young, the smallest contained only two individuals. For the series of 275 litters the average size of the litter was 7.01 young.
A similar analysis of the data collected during the year 1914, as given in table 2, shows for the entire series of 814 litters an average of 6.99 young per litter, which is remarkably close to the average for the litters examined in 1911-1913. While the records for 1914, as a whole, show a great uniformity in the
NORMAL SEX RATIO AND LITTER SIZE IN RA.T 417
average size of the litters cast in the various months, there seems to be a tendency for the Utters cast during the first part of the year to be sUghtly larger than those produced during the latter half of the year. A similar tendency, however, is not noted in the records of table 1, so that it can have little, if any, significance.
Records for the entire series of 1089 litters give 7.0 young as the average number of indi\iduals in a Utter. According to these observations, therefore, the size of a litter of albino rats is, on the average, greater than that of the black rat, but it is smaller than that in the gray rat of which it is the domesticated variety.
The data for litter size, arranged according to the season of the year when the litters were cast, are given in table 4. A marked uniformity in the various series of records is again e\'ident. In the final averages the Utters cast during the fall of the year show a relatively small size. This result probably has Uttle, if any, meaning, since it is due entirely to the low average size of many of the Utters cast during the fall of 1914. Records for the litters cast in corresponding months of the years 19111913 give 7.0 young as the average number of individuals per litter. It is evident, from these results, that there is no pronounced seasonal variation in the size of the litters at all comparable to the evident change that occurs in the sex ratio at stated periods in the year. Seasonal changes in the sex ratio are independent of litter size just as the normal sex ratio is independent of litter size.
Crampe ('83) states that the first litter of an albino rat is not as large as the second and that the second litter is an index of the size of subsequent litters. The first part of this statement can be corroborated by our records, but the latter part of it needs to be modified. A large second litter gives no indication whatever as to the size of the following litters, as the records for litters from many hundreds of females coUected by one of us shows. In many cases a large second litter is followed by an unusually small Utter, and there are marked individual differences in females regarding the size of the litters they pro
418 HELEN DEAN KING AND J. M, STOTSENBURG
duce. Some females never have over five or six young in a litter; other females invariably cast litters containing eight or more j^oung.
The average size of 75 Utters cast by 21 stock albino rats is given, with other data, in table 7. In these records the average size of the first litter is found to be considerably less than that of the second, whUe the second of the four litters is the largest of the group, containing an average of 7.7 young per litter. In this particular series of records the average size of the third litters is considerably below that for the second litters, but in a larger series of data it would probably be found that the third litter is nearly, if not equal, to the second in size. The fourth litters are, as shown in table 7, only a little larger than the first, as a rule.
For the entire series of 75 litters the sex ratio is below normal, and the average size of the litters is somewhat small, being only 6.8 young per litter. The number of young in a given litter is dependent to a marked extent on the age and physical condition of the female (King '15), and it is not improbable, as previouslj^ stated, that these factors also have an effect on metabolic processes that play an important role in determining the sex of the embryo.
SUMMARY
1. Albino rats breed throughout the entire year, but the periods of greatest sexual activity are in the spring and autumn.
2. The sex ratio in the 1089 litters of albino rats examined was 107.5 males to 100 females.
3. There is, apparently, a seasonal variation in the sex ratio of the albino rat. Litters cast in the spring and early fall show a relatively low sex ratio; those cast in summer have a much higher sex ratio (fig. 1).
4. Data for 75 litters produced by 21 albino females indicate that the sex ratio among the first offspring of young females is higher than that found among the offspring of the same females when they are at the height of their reproductive power.
NORMAL SEX RATIO AND LITTER SIZE IN RAT 419
5. There is apparently no relation between the size of a litter of albino rats and the sex of its members.
6. The 1089 litters examined contained an average of 7.0 young per litter. Litters of albino rats, therefore, are smaller than those of the gray rat and larger than the litters of the black rat.
7. There is no pronounced seasonal variation in the Utter size comparable to the seasonal variation noted in the sex ratios.
8. As a rule the first of an albino female's four litters is the smallest; the second and the third litters are the largest; the fourth litter is a little larger than the first.
420 HELEN DEAN KING AND J. M. STOTSENBURG
LITERATURE CITED
Bidder, F. 1878 Ueber den Einfluss des Alters der IMutter auf das Geschlecht
des Kindes. Zeitschr. Geburtshiilfe und Gjiiakologie, Bd. 11. CoPEMAN, S. M., and Parsons, F. G. 1904 Observations on the sex in mice.
Proc. Royal Soc. London, vol. 73. Crampe, H. 1883 Zucht-Versuche mit zahmen Wanderratten. I. Resultate
der Zucht in Verwandtschaft. Landwirthschaftliche Jahrb., Bd. 12.
1884 Zucht-Versuche mit zahmen Wanderratten. II. Resultate der
Kreuzung der zahmen Ratten mit wilden. Landwirthschaftliche
Jahrb., Bd. 13. Cu^NOT, L. 1899 Sur la determination du sexe chez les animaux. Bull. Sci.
de la France et de la Belgique, t. 32. DusiNG, K. 1884 Die Regulierung des Geschlechtsverhaltnisses bei den Ver mehrung der Menschen, Tiere und Pflanzen. Jen. Zeitschr. Natur.
Wiss., Bd. 17. Geissler, a. 1889 Beitriige zur Frage des Geschlechtsverhaltnisses der Gebo renen. Zeitschr. d. k. sachsischen statistischen Bureaus, Dresden,
Bd. 35. Heape, W. 1908 Notes on the proportion of the sexes in dogs. Proc. Cambridge Phil. Soc, vol. 14. Jackson, C. M. 1912 On the recognition of sex through external characters in
the young rat. Biol. Bull., vol. 23. King, Helen Dean 1911 The sex ratio in hybrid rats. Biol. Bull., vol. 21.
1915 On the weight of the albino rat at birth and the factors that
influence it. Anat. Rec, vol. 9. Lantz, D. E. 1910 Natural history of the rat. U. S. Bull. Public Health and
Marine Hospital Service. Lloyd, R. E. 1909 Relation between fertility and normality in rats. Report
6f the Indian Museum, vol. 3. Miller, N. 1911 Reproduction in the brown rat (Mus norvegicus). Amer.
Nat., vol. 45. Nichols, J. B. 1907 The numerical proportions of the sexes at birth. Mem.
Amer. Anthropological Assoc, vol. 1. Ptjnnett, R. C. 1903 On nutrition and sex-determination in man. Proc.
Cambridge Phil. Soc, vol. 12. Schlechter, J. 1884 L^eber die Ursachen welche das Geschlecht bestimmen.
Biol. Centralbl., Bd. 4. Schultze, O. 1903 Zur Frage von den Geschlechtsbildenden Ursachen. Arch.
mikr. Anat., Bd. 43. Wklden, W. F. R. 1906 On heredity in mice. I. On the inheritance of the sexratio and of the size of the litter. Biometrika, vol. 5. Wilckens, M. 1886 Untersuchung ueber das Geschlechtsverhiiltniss und die
Ursachen der Geschlechtsbildung bei Haustieren. Biol. Centralbl.,
Bd. 6.
AN INSTANCE OF ACIDOPHILIC CHROMOSOMES AND CHROMATIN PARTICLES
IVAN E. WALLIN
Anatomical Laboratory, Cornell University Medical College, New York City
ONE PLATE (twelve FIGURES)
Studying the cytological literature, I have been unable to find a record of acid staining chromosomes in a normally dividing cell. In an investigation of sections of petromyzon larvae numerous mesenchyma and blood cells are seen which contain nuclei staining a uniform and brilliant red. These cells are scattered among other cells having nuclei of exactly the same structure, yet staining the usual deep blue color with the hemotoxilyn eosin stain. After studying these cells more closely, I found that the cells with the red nuclei were able to undergo mitotic division in the same manner as the cells with blue nuclei, the chromosomes in such cases staining red rather than the characteristic deep blue or black seen in the neighboring cells.
Such cells have been found as free blood cells in the blood vessels and also as mesenchyma cells in the pharyngeal and head regions of the larvae. Wherever found, aside from the peculiar property the nuclei show in the absorption of the acid dye, these cells are exactly similar to others in the region in which the nuclei stain characteristically with the basic dye (hematoxylin). The accompanying plate shows the two types of cells in the resting condition and in different stages of mitosis.
It is of interest that these cells have been found only in the 5 mm. larvae of my collection. These particular 5 mm. larvae were procured at Naples by Professor Stockard in the spring of 1910. They were fixed at the time of collection in picroacetic and preserved in 80 per cent alcohol. I received them
421
422 IVAN E. WALLIN
in the fall of 1914, sectioned and stained them in hematoxylin and eosin. The remaining specimens of my collection, which comprise developmental stages ranging from the segmentation sphere up to an including a transforming larva and the adult, have been kindh' supplied to me by Professor Gage. They were collected from the waters in the neighborhood of Ithaca, X. Y. Unfortunately, I do not have any 5 mm, larvae of the American species so am unable to say whether these cells are peculiar to the European species and to this age of larvae. Several specimens of these 5 mm, larvae show cells with nuclei taking the acid stain. They are not equally numerous nor do they take the stain equally well in all cases. In one specimen, for instance, such cells are difficult to find while in others they are distinct and plentiful. Embryos in which these cells are prominent may show more than half of the blood cells containing nuclei in which the chromatin has absorbed the acid dye.
Cells have been found in which both kinds of chromatin are present. Figure 7 represents such a cell in the resting state. Two lumps of chromatin in the central part of the nucleus have taken the basic stain while the chromatin at the periphery is stained with the acid dye, Heidenhain ('07) has shown a chromatolytic nucleus which somewhat resembles this, but in which the acid-staining chromatin was in the central part of the nucleus while the basic-staining chromatin was at the periphery. Figures 5 and 6 represent cells with two kinds of chromatin in the process of mitosis. Figure 6 shows that a small part of. the acid-staining chromatin has apparently been taken over with the basic-staining group. Figure 5 represents the separation of the two kinds of chromosomes in the daughter nuclei which appears to have been complete. The great majority of these cells with acid-staining chromatin, however, are pure in regard to their staining reaction. Figures 1, 4, 8, 9 and 10 show nuclei containing no granule or any other part which absorbs the basic dye.
Stockard ('06) confirmed the observations of Schniewind-Theis ('97) in which it was shown that some nuclei in the deeper layers of actively secreting nectar glands of Vicia faba take the plasma
AN INSTANCE OF ACIDOPHILIC CHROMOSOMES 423
stain. In the living gland some rows of cells have a blue while others have a red coloration. By introducing alkaline and acid fluids to sections of the living gland, Stockard found that these cells responded to the fluids in the same way that litmus does to alkalies and acids. This experiment shows quite conclusively that the chemical reaction of the glandular plant cell is not constant during ^'arious physiological phases. The staining reaction also indicates that the nuclei apparently respond to the stain according to their physiological state. The stain used was Auerbach's (methyl green and acid fuchsin) which gives a delicate differentiation of the acid and basic qualities. It was determined in these investigations that materials were formed by the nucleus and passed out into the cytoplasm, the cytoplasm in such cases finally accumulating enough of the nuclear products to stain with the nuclear dye. Further, when the secreting activities of the nucleus had apparently been spent the nucleus stained with the plasma stain. These reactions occurred only in vegetative cells, the dividing cell always contained chromatin which stained in the normal way with the basic dye. In these studies the tissues had been carefullj^ fixed with neutral fluids so as to preserve the chemical reaction of the living cell. The fixation used in my specimens, as was pointed out above, was an acid fluid, yet the differences in the reactions of the cells and portions of some nuclei were sufficient to maintain their character and to respond to the ordinary hematoxylin-eosin stain in the pecuhar ways shown in plate 1.
The presence in the same section of the lamprej' larva of cells with acidophilic nuclei together with cells which stain in the normal way, and the fact that both kinds of chromatin are present in a single cell, make it difficult to give any other interpretation of these reactions than that they represent the result of physiological changes which occurred during life.
As far as I can ascertain this is the first account of a case where the chromosomes in a dividing cell have definitely taken the acid stain.
424 IVAN E. WALLIN
LITERATURE CITED
Heidenhain 1907 In Bardeleben's Handbuch der Anatomie Des Menschen.
Schniewind-Thies, J. 1897 Beitrage zur Kenntniss der Septalnectarien, Jena (cited from Stockard (1906) ) .
Stockard, C. R. 1906 Cytological changes accompanying secretion in the nectar-glands of Vicia faba. Bull, of Torrey Bot. Club, vol. 33, pp. 247262.
PLATE 1
EXPLANATION OF FIGURES
All figures were drawn with the aid of the camera lucida to the same scale of magnification (1/12 oil inmaersion objective, compensating occular No. 12). Higgins' carmine and true blue inks were used in reproducing the colors of the stained specimens.
Figures 1, 2 and 5 are mesenchjTna cells; all other figures represent blood cells.
AN IXSTAXCE OF ACIDOPHILIC CHROMOSOMES
IVAN E. WALLIN
PLATE 1
r.
(<?VJ
a
1
10
.^
%
12
425
THE ( ONXECTING SYSTEMS OF THE REPTILE HEART
HENRY LAURENS
From the Osborn Zoological Laboratory, Yale University
EIGHT FIGURES (tWO PLATES)
THE SINO-VENTRICULAR CONNECTION
A sino-^'entriclllar bundle, or dorsal ligament, has already been described by me in the hearts of Lacerta agihs, and L. viridis, of Clemmys lutaria and Chelopus insculptus. By observing this ligament in living hearts under the binocular microscope, and from the study of transverse, frontal and sagittal sections, it was seen to be a band of connective tissue, containing nerves and blood vessels, running between the sinus and the ventricle well over to the right side of the heart, (Laurens '13 a and '13 b). In my first paper it was further shown that this ligament had no significance for the coordination of the heart beat. This band of tissue had previously been described and experimented with b^' several investigators, and a discussion of their various views as to its structure and physiological importance will be found in my papers and in a recent publication by Mangold (;14).
Mackenzie ('13) has recently described in the heart of the salempenter, a South American lizard, a ' ' sinu-auricular bundle" of specialised muscle, connecting the sinus with the specialised tissue lying in the floor of the auricle. Although from my earlier studies of the lizard and tortoise hearts it was certain that such a bundle did not exist in the hearts of the reptiles examined by me, this publication of Mackenzie's induced me again to go over my preparations, to which had been added in the meantime sections of the heart of the fence lizard (Sceloporus undulatus) and of the spotted tortoise (Chelopus guttatus). A part of this later material was fixed and stained by the same methods as were earlier used ('13 b) — fixation in strong Flem 427
THE AXATOMICAL RECOn'>, VOL. >J, N'O. 6
428 HEXRY LAURENS
ming or in concentrated corrosive sublimate, and staining with iron hematoxylin and picric acid fuchsin. The remainder consisted of sections of hearts treated with methylen blue according to various methods, by CajaFs double impregnation method, as given by Hofmann r02), and by the silver reduction method, as given by ^leikeljohn CIS).
From this further study of these lizard and tortoise hearts no doubt has been thrown on the truth of the statement that the dorsal ligament is here a sino-ventricular bundle. But from Mackenzie's descriptions and from his figures it can also not be doubted that in the heart of the salempenter conditions are diiTerent. As he has himself pointed out, the conditions in this respect shown bj^ the hearts of different reptiles are not the same and various stages can be recognized. We shall see that the conditions found in the hearts of the reptiles Usted above represent still another stage in addition to those whi'ch he has described.
According to Mackenzie fp. 129) the sinu-auricular-ring" of the fish heart is represented in the heart of the salempenter by a bundle or leash of fibres which hes in the groove between the left venous valve and the spatium intersepto-vahailare." The sinu-auricular bundle courses round the posterior and under aspect of the sinus venosus just where the left duct of Cuvier enters the sinus and runs ... a short distance as a free bundle to become continuous with the .speciahsed tissue h'ing in the floor of the auricle, this tissue becoming in turn continuous with the auricular canal." In the heart of the crocodile (p. 130) the sinu-auricular muscle" is present at the base of the left venous valve at the junction of the sinus with the spatium. There is no direct continuity between the sinu-auricular bundle and the auriculo-ventricular bundle in the crocodile. The interruption takes place in the region of the sinus septum where the left duct of Cuvier enters the sinus." Later ^p. 135) he goes on to say:
The .sinu-auricular nodal tissue appears to become lost in this septum. It would appear that there are reptiles which in respect of this point exhibit an intermediate stage between the lizard (salempenter) and the crocodile. An example of this is the iguana, in which the sinuauricular bundle is interrupted by a cord of fibrous tissue with iso
COXXECTING SYSTEMS OF THE HEART 429
lated muscle fibres and large nerve trunks. This cord occupies a corresponding position to the continuous muscle structure in salempenter and in front appears again as a short isolated muscle bundle which in turn becomes continuous ^^1th the muscle of the auricular canal.
The conditions found in the lizard and tortoise hearts listed above represent another intermediate stage between that found in the iguana and that found in the crocodile, as described bj' Mackenzie. In these hearts there is no isolated muscle bundle" which becomes continuous with the auricular cajial. At the left venous valve, near its upper portion, there is, in the fibrous tissue a large group of nerve cells, which represents the endings of a branch of the right vagus nerve. The nerve fibers connected ^^dth these nerve cells can be followed for quite a distance along the right vein, as far as it is present in the sections, being connected with other large ganglia here and there along the vein. From this portion of the left valve there runs a band of connective tissue, a fold of the pericardium, which bending under the left vein becomes free from the dorsal surface. From this point it is continued downward as a free band, superficially over the dorsal surface of the right auricle to the ventricle, over the anterior dorsal surface of wliich it spreads, being wider at its point of attachment than elsewhere. Sometimes the bundle, before it reaches the ventricle, divides into two, or even three, parts, and often under these circumstances a fine branch can be seen bending still further to the right and running in the auricular-ventricular groo\'e to the ventral side.
In the bundle there are numerous blood vessels and large nerve trunks with several groups of nerve cells. Sometimes these nerve cells are single and scattered, but there are also many large ganglia. By studying ^Mackenzie's figures of sagittal sections (plate 2, figs. 1-3) one sees on the dorsal surface of the hearts a mass of tissue which extends from the sinus region to the anterior dorsal portion of the ventricle. This tissue Mackenzie has not labelled, but it has a position very similar to the continuous sino-ventricular bundle m the hearts of the Uzards and tortoises which I have studied. From figure 3 of this same plate of ^lackenzie's, however, it is seen that the sinu-auricu
430 HENRY LAURENS
lar bundle" does go over direct I3' into the auricular funnel musculature, the latter being, according- to his representation, on the dorsal side a continuation of the " sinu-auricular bundle." A glance at the figures which are presented with this article will show that this is not the case in the animals with which we are dealing. The figures are untouched photographs of sagittal sections of the heart of the tortoise Chelopus guttatus. The hearts of the other tortoises and of the lizards show the same conditions. Drawings of the lizard heart have already been given (Laurens '13 b) and for that reason the tortoise is selected for the illustrations here.
After reaching the ventricle, the sino-ventricular bundle spreads out over its anterior dorsal surface. A portion of it runs to the back of the ventricle, while another portion goes down into the space between the funnel musculature and the inner wall of the ventricle. This space is filled \vith connective tissue containing blood vessels, nerves and ganglia (Laurens '13 b, fig. 4), and the portion of the sino-ventricular bundle which goes down into this space becomes continuous with this connective tissue, which is also, of course, a portion of the pericardium. The nerves in the sino-ventricular bundle, two of which are shown in figure 6, are also distributed, some of them to the anterior dorsal surface of the ventricle, and some to the connective tissue filled space between the funnel musculature and the inner wall of the ventricle. The latter innervate the auriculo-ventricular funnel and also supply the inner wall of the ventricle. Quite often the nerves in the sino-ventricular bundle are insignificant, and even entirely lacking, a fact which was also noted by Gaskell.
There is no muscle tissue, either continuous or isolated, in the sino-ventricular bundle. At its beginning (sinus end) and ending (ventricular end) a few isolated striated muscle fibers can sometimes be seen in the connective tissue (as in the sections from which figures 3 and 4 are taken). But it is clear that these muscle fibers have been pulled into this position by the knife tearing them away from the walls of the sinus and of the ventricle. In its free course there is only fibrous tissue in the bundle, in which nerve fibers, ganglion cells and blood vessels are found.
CONNECTING SYSTEMS OF THE HEART 431
As the sino-ventricular bundle is followed in transverse and sagittal sections it is seen to be nothing more than a portion of the pericardium which is folded off as a free band to run between the sinus and the ventricle. In some places it is even connected with the pericardium proper over the right auricle by fine strands (fig. 5). By studying the figures, which represent sections in order from left to right, the manner in which the sino-ventricular bundle is folded off from the continuous pericardium can be made out. Figures 1 and 2 are sections to the left of the median line, and the dorsal ligament does not show at all. In figure 2, however, one of the large nerve trunks is seen running from the sinus, under the pericardium on the dorsal side of the auricle, to the ventricle across the auriculo-ventricular groove (Laurens '13 b and Dogiel and Archangelsky '06) to end in the auriculo-ventricular funnel musculature, after it has gone through the connective tissue in the space between the funnel and the ventricle. In figure 3 we see the beginning of the free portion of the sino-ventricular bundle in an out-folding of the pericardium on the anterior dorsal surface of the ventricle, and opposite to it a corresponding outfolding on the sinus wall. In figure 4 these folds have advanced further and in figure 5 they have met to form the continuous band of connective tissue, which can here be followed up along the sinus until bending over to the right it disappears from the section. Figure 6 gives another view of the sino-ventricular bundle further to the right. In this section a large nerve trunk is seen in the Hgament coming from the sinus and going dowTi into the connective tissue filling the space between the funnel and the ventricle. It also shows to the extreme right a portion of a smaller nerve going over to the outer wall of the ventricle. Figures 7 and 8 serve to illustrate the appearance of the ligament further over to the right hand side of the heart. From a study of these figures it will be clear, I believe, that the dorsal hgament is simply a fold of the pericardium, which, retaining its connection with the sinus and with the ventricle, runs free over the dorsal surface of the right auricle from the sinus to the ventricle, and is therefore strictly a sino-ventricular bundle.
432 HENRY LAURENS
THE SINO-AURICULAR COXXECTION
The connection between the smus and the right auricle is a direct muscular one in the reptile hearts described in this paper. Gaskell pointed out the fact that this was the case in the tortoise with which he was working, though the details concerning the manner in which the connection was actually brought about do not hold here. Kiilbs and Lange TIO) also describe a direct muscular connection between the sinus and the right auricle in the lizard. But in the heart of the salempenter (Mackenzie) the ring of specialized muscle with numerous nerve cells and fibers at the ' sinu-auricular junction' in the fish is represented by a bundle or leash of fibers which lies in the groove between the left venous valve and the spatium intersepto-valvulare." In the reptile hearts that I have examined there is a complete muscular ring. Nerve cells, in larger and smaller gangha, and nerve fibers are all around this ring in the connective tissue. The musculature of the sinus goes over into that of the right auricle in much the same way that the musculature of the auriculo-ventricular funnel goes over into that of the ventricle. At the junction of the sinus with the auricle there are the two valves which completely close the oval shaped opening which runs obhquely from the upper right hand side to the lower left, as Mackenzie shows in his figure on plate 3. At the right, or lower, valve, the musculature of the sinus goes over into that of the auricle at the free edge, the two kinds of musculature being here continuous. At the left, or upper, valve the conditions are somewhat different. In its upper portion the valve is a continuation of the wall of that portion of the sinus, and the musculature of the sinus joins directlj^ with that of the auricle along the valve (fig. 7). But the extreme lower portion of the left valve, which, in sagittal sections taken from the left to the right comes first into view, is formed from a portion of the auricular septum, being really a continuation of it (fig. 3), and the wall of the sinus here goes directly over into this portion of the septum, the left valve being here separated from the wall of the auricle by a layer of fibrous tissue, a condition particularly clearly shown in transverse sections.
CONNECTING SYSTEMS OF THE HEART 433
THE AURICULO-VENTRICULAR CONNECTION
Attention may be here again called to this connection in the hearts of lizards and tortoises, since, from Mackenzie's description, there appear to be slight differences between the conditions found in the salempenter and those found in other Uzards and in tortoises. In the salempenter, the auricular canal, according to Mackenzie, is specialized, an assumption also made by Kiilbs and Lange ('10) for the lizard (L. viridis and L. muraUs) and by Kiilbs ('12 and '13) for the lizard and tortoise. It has already been pointed out (Laurens '13 b) that the musculature of the auriculoventricular funnel is not very different from the musculature of other portions of the auricles. The fibers and nuclei are similar to those of the auricles, though there is more sarcoplasm and fewer fibrillae. However, the funnel musculature is richly supplied with nerves and contains numerous capillaries, and in between its fibers, which are arranged circularly, there is a considerable amount of connective tissue. The striation of the fibers is distinct but fine, and is quite similar to that of the auricles. The striation of the ventricular fibers is coarser and the nuclei are much elongated and narrower than are those of the auricles and of the funnel.
The function of the auriculo-ventricular funnel in co-ordinating the contractions of the auricles and of the ventricle was very carefully worked out in the hzards, L. viridis and L. agilis, and in the tortoise, Clemmys lutaria. From this work (Laurens '13 a) it is evident that there is here a physiological differentiation in that certain portions of the funnel are more efficient than others in allowing the passage of the contraction wave from the auricles to* the ventricle, and furthermore, in preserving the co-ordination of ventricular with auricular beat, when other portions of the connection between these parts of the heart are cut away. The portions showing this greater efficiency are the right and left sides of the funnel. Later (Laurens '13 b) it was shown that this physiological speciaUzation had an anatomical basis in that, at these two portions, there was a more intimate connection between the funnel musculature and the musculature of the ventricle.
434 HENRY LAURENS
In the salempenter (Mackenzie, p. 129) the auricular canal is described as a tube invaginated into the ventricle, becoming at its lower end continuous with the ventricular musculature in the region of the papillary muscles to which the auriculoventricular valves are attached. This invagination does not of course occur at that part of the orifice where the auricular canal is continued on to the bulbus musculature." In all the lizard and tortoise hearts studied by me the invagination does take place around the whole circumference of the orifice, the funnel being only broken through at its entrance into the ventricle, by the bulbus with the musculature of which the funnel musculature becomes continuous. The direct continuity between the musculature of the funnel and that of the ventricle does not, of course, occur at this place.
Mangold ('14) points out that on the dorsal side the funnel musculature in the salempenter, as described by Mackenzie, extends further into the cavity of the ventricle, before the fusion of the two kinds of musculature takes place, than it does in the hearts of the reptiles which were described by me, where it very soon becomes broken through by its fusion with the ventricle. This difference, however, is I think, very slight. By comparing Mackenzie's figures with mine it will readily be seen that the length of the ventricle of the salempenter is relatively, when compared with its dorso-ventral thickness, less than that of the lizards and of the tortoises here described. Moreover, that the attachment of the auriculo-ventricular valves as represented in the salempenter is much nearer the apex, and that the auriculoventricular funnel extends further into the cavity of the ventricle, than in the other lizards and tortoises. From a glance at figure 4 (Laurens '13 b) it will be apparent that on the dorsal side the funnel musculature is continued, although quite thin, almost to the attachment of the auriculo-ventricular valves to the papillary muscles, and the figures presented with the present article show this quite plainly. On the right and left sides, however, the funnel musculature is continued further into the ventricle, before the fusion between the two kinds of musculature finally takes place, the connection at these parts being
CONNECTING SYSTEMS OF THE HEART 435
therefore more intimate than at other portions. Mackenzie does not mention whether the final continuity between the auriculoventricular funnel and the ventricle takes place at all portions at the same level.
There is one other matter concerning the auriculo-ventricular connection, and that is its innervation. Nerve fibers can be seen extending downward from the sinus in the pericardium and can be followed across the auriculo-ventricular groove. For the past two years I have been studying the innervation of the reptile heart, and particularly of the auriculo-ventricular funnel muscle, by means of the special methods mentioned earlier. Although perfect results have not yet been obtained, it has been seen that the auriculo-ventricular funnel is richly supplied with nerves, in the form of a net-work of fine fibers, which come to it from branches of nerves descending along the back of the auricles in the way described earlier by me ('13 b), and by Dogiel and Archangelsky ('06). Nerve fibers and cells are also found on the inside of the auricles, running along the inner edge of the walls and along the septum, and which come into the heart along or near the entrance of the pulmonary vein and of the left duct of Cuvier. These nerves also give off branches which run down between the auriculo-ventricular valves and the inner side of the funnel to finally become distributed to the latter. In the funnel musculature itself, nerve cells are scarce, only a few scattered ones being found here and there. But in the connective tissue of the groove, and of the space between the funnel and the inner wall of the ventricle, ganglia are numerous, particularly on the dorsal side, though in sections of some hearts the number of ganglia found on the " ventral side, especially near the bulbus, and on the right and left sides of the funnel is also quite large.
The nerves which run over the dorsal surface of the auricles and of the auriculo-ventricular groove to be continued into the connective tissue between the funnel and the ventricle can also be easily seen in sagittal sections of material fixed in Flemming and in corrosive sublimate. In figure 2 one of these nerves is shown running down to finally become distributed to the auriculoventricular funnel musculature (see also fig. 6).
436 HENRY LAURENS
THE PROBABLE FUNCTION AND FATE OF THE SINO-VENTRICULAR
BUNDLE
The sino-ventricular bundle, contrary to the view of Imchanitzky ('09), has nothing to do with the coordination of ventricular and auricular contraction (Gaskell '84 and Laurens '13 a). Furthermore it has no function in bringing about a possible sino-ventricular rhythm. In the salempenter, however, according to Mackenzie, 'the sinu-auricular bundle' is made up of specialized muscle. In support of which statement, we have no physiological evidence.
The contraction wave begins in the walls of the sinus and spreads to the auricles along the sino-auricular junction, and from there along the auriculo-ventricular funnel to the ventricle. Can the 'sinu-auricular bundle' in the salempenter be a pathway for impulses passing direct from the sinus to the auriculoventricular connection and to the ventricle, and if so, what is the nature of these impulses? That they have anything to do with coordination can hardly be claimed owing to the physiological evidence against such an assumption. Gaskell (p. 83) showed that when the peripheral end of the 'coronary nerve' was stimulated that the rate of beat of the heart was not changed, although when the central end was stimulated a decided slowing of auricular rate took place. By the stimulation of either end of the nerve the force of the auricular contractions was diminished (p. 92). Further, Gaskell showed (p. 85) that, when the connection between the sinus and the auricles is severed so that the 'coronary nerve' remains intact and as the only connection between the auricle and ventricle and the body of the animal and an independent auriculo-ventricular rhythm is set up, stimulation of the right vagus brings about a decrease in the force of the auricular contractions alone. In one experiment, out of several, Gaskell obtained an inhibition of this independent auriculo-ventricular rhythm, when the right vagus nerve was stimulated. His explanation of this was that "when the coronary nerve happens to contain fibers which supply the particular muscles which originate the independent rhythm.
CONNECTING SYSTEMS OF THE HEART 437
then stimulation of the vagus can inhibit that rhythm." Moreover, as Gaskell also pointed out, the 'coronary nerve' is only one of several nerve trunks, branches of the vagus nerves, which pass from tlie sinus to the ventricle and to the auriculo-ventricular funnel musculature, and it is therefore but natural that the auriculo-ventricular rhythm should be controlled by it when it happens to innervate the muscles which originate this rhythm, just as the other branches of the vagus nerves do.
The 'coronary nerve,' when the vagus is stimulated, is also no more able to conduct an impulse to the ventricle by which its force of contraction can be diminished, than are any of the other nerve trunks passing from the sinus to the ventricle (Gaskell, p. 91). It has also another function in common with the other branches of the vagus nerves in that it carries impulses, upon stimulation of the vagus which improve the conduction power of a strip between the auricle and the ventricle during a condition of partial block, though it may also have a function occasionally in increasing such a block by diminishing the conducting power of the connecting strip (Gaskell, p. 96).
The 'coronary nerve' is not always present, and is often very insignificant. From what has just been said it apparently has no particular function that the other nerve trunks do not have, and we are forced to the conclusion that its function is relatively unimportant, or perhaps better, no more important than that of the other branches of the vagus nerves.
From physiological evidence then, the effect of the 'coronary nerve,' like the other branches of the vagus nerves, are on the auricles, and it may be assumed to have an additional regulatory effect when, for any reason, the rhythm producing power of the auriculo-ventricular funnel is brought into prominence above that of the sinus. As we have seen, some of the nerve trunks which run in the sino-ventricular bundle go into the connective tissue between the funnel and the ventricle, which is a condition that was not reahsed at the time of my earher publication. It is highly probable that the nerves of the ' sinu-auricular bundle' of the salempenter have the same function as the 'coro
438 HENRY LAURENS
nary nerve' in the heart of the tortoise described by Gaskell. But what the function of the 'speciahzed muscle' of this bundle is, must remain for the present an unanswered question.
As to the probable fate of the sino-ventricular bundle, Mackenzie has given the steps leading up to the condition found in the reptile hearts described in the present article. In the salempenter there is a ' ' sinu-auricular bundle of specialised muscle;" in the iguana the bundle is interrupted by a cord of fibrous tissue, a short isolated muscle bundle, which is continuous with the muscle of the auricular canal, being all that is left. In the hearts of the lizards and tortoises here described, the sino-auricular junction is a ring of muscle completely surrounding the opening of the sinus into the right auricle, and running from this region there is a band of connective tissue which goes to the anterior dorsal surface of the ventricle. In this band of connective tissue there are large nerve trunks and blood vessels some of which go to the ventricle, others into the connective tissue between the funnel and inner wall of the ventricle to become distributed to the musculature of both. In the crocodile, all that there is of the ' sinu-auricular muscle' according to Mackenzie, is at the base of the left venous valve at the junction of the sinus with the spatium.
Mackenzie speculates on the probable fate of the fibrous cord and of the distal isolated muscle bundle found in the iguana. He considers it not improbable that it becomes incorporated in the auricular floor, and that it is represented in the higher reptilian and mammalian hearts by a bundle of fibrous tissue which runs in the basal wall of the auricle at the line of attachment oftthe septum between the region of the coronary sinus and the septum fibrosum." As we have seen, however, in the hearts of other lizards and tortoises; the 'distal muscular bundle' has disappeared, while the 'fibrous cord' is still present as a sinoventricular bundle.
In this connection attention may be called to the recent publication of Stanley Kent in which it has been shown that in man "the auriculo-ventricular bundle is not the only path by which the functional connection between the auricle and ven
CONNECTING SYSTEMS OF THE HEART 439
tricle may be established/' (Kent '14 a). In a series of papers Kent describes and figures a specialized tissue on the right lateral aspect of the heart, as a neuro-muscular connection between the auricle and the ventricle, with large nerve trunks in the fat and connective tissue of the groove, and also in the muscular tissue. He also shows ('14 c) that, when all other "connections between the auricles and ventricles are severed, contractions of the auricles are still carried over to the ventricles. This connection, according to Kent ('14 a), may constitute a "local reflex arc which may perhaps exhibit only an occasional activity," and "which may be capable of controlhng . . . coordination when the bundle is no longer perfect."
Whether the ' sinu-auricular bundle' of the salempenter, and the sino-ventricular bundle of Lacerta agilis, L. viridis, Sceloporus undulatus, and of Clemmys lutaria, Chelopus inscn,lptus and C. guttatus have anything in common with this bundle described by Kent is of course a question for the future.
SUMMARY
1. The dorsal ligament in the hearts of Lacerta viridis, L. agilis, Sceloporus undulatus, and of Clemmys lutaria, Chelopus insculptus and C. guttatus is a sino-ventricular bundle of connective tissue, a fold of the pericardium, which, extending from the sino-auricular junction near the upper portion of the left venous valve, runs under the left vein and is continued as a free bundle to the ventricle on the anterior dorsal surface of which it ends; there becoming continuous again with the pericardium. In this bundle there are large nerve trunks and blood vessels. Some of the nerves go to the dorsal surface of the ventricle, others to the auriculo-ventricular funnel, after they have gone down into the connective tissue filling the space between the musculature of the auriculo-ventricular funnel and the inner wall of the ventricle, and still others to the inner wall of the ventricle.
2. The sino-auricular junction is in the form of an ovalshaped muscular ring. At the right valve the musculature of
440 HENRY LAURENS
the sinus and that of the auricle is continuous. At the left valve in its upper portion the musculature of the two chambers is also continuous, but in its lower portion the musculature of the sinus goes over into that of the auricular septum which forms this portion of the valve, there being here a layer of connective tissue between the left valve and the wall of the auricle.
3. The auriculo-ventricular junction has in cross section the* form of a ring; in sagittal section it is in the form of a funnel which extends down into the orifice of the ventricle, with the musculature of which it gradually becomes continuous. On the right ventral side the ring is interrupted by the bulbus with the walls of which it also becomes continuous. The fusion between the funnel and ventricle musculature takes place at the lower level of the aurico-ventricular valves near their attachment to the papillary muscles. The dorsal portion of the funnel is the first to become continuous with the ventricle, while the right and left sides are the last.
4. The auriculo-ventricular funnel is richly innervated bjbranches of the right and left vagus nerves which, coursing along the dorsal surface of the auricles, under the pericardium, cross the auriculo-ventricular groove, in the connective tissue of which there are many ganglion cells, go down into the connective tissue filled space. between the funnel and the inner wall of the ventricle, and finally end in the funnel musculature and in the musculature of the ventricle. There are also a few nerves on the ventral side, and a few which course along the inner walls of the auricles and the auricular septum to end in the musculature of the funnel.
CONNECTING SYSTEMS OF THE HEART 441
LITERATURE CITED
DoGiEL, J., und Archaxgelsky, K. 1906 Der Bewegungshemmende und der
motorische Xervenapparat des Herzens. Pfliiger's Archiv, Bd. 113,
S. 1-96. Gaskell, W. H. 1884 On the innervation of the heart with especial reference
to the heart of the tortoise. Jour. Physiol., vol. 4, pp. 43-127. HoFMANN, F. B. 1902 Das intracardiale Nervensystem des Froschherzens.
Arch f. Anat. (u. Physiol), S. 54-114. Kent, A. F. Stanley 1913 a The structure of the cardiac tissues at the auric ulo-ventricular junction. Jour. Physiol., vol. 47, p. xvii.
1913 b Observations on the auriculo-ventricular junction of the mammalian heart. Q. J. Exper. Physiol., vol. 7, pp. 193-195.
1914 a Xe\iro-muscular structures in the heart. Proc. Roy. Soc, Ser. B, vol. 87, pp. 198-204.
1914 b The right lateral auriculo-ventricular junction of the heart.
Jour. Physiol., vol. 48, p. xxii.
1914 c A conducting path between the right auricle and the external
wall of the right ventricle in the heart of the mammal. Jour. Physiol.,
vol. 48, p. Lvii.
1914 d Illustrations of the right lateral auriculo-ventricular junction in the heart. Jour. Phj'siol., vol. 48, p. lxiii. KtJLBS, F. 1912 tJber das Reizleitungssystem bei Amphibien, Reptilien und
Vogeln. Zeitschr. f. exper. Pathol, u. Ther., Bd. 11, S. 51-68.
1913 Das Reizleitungssystem im Herzen. Berlin, 28 pages. KtJLBS, F., und Lange, W. 1910 Anatomische und experimentelle Untersuch ungen iiber das Reizleitungssj'stem im Eidechsenherzen. Zeitschr. f.
exper. Pathol, u. Ther., Bd. 8, S. 313-322. Laxjrens, H. 1913 a Die atrioventrikulare Erregungsleitung im Reptilien herzen und ihre Storungen. Pfliiger's Archiv, Bd. 150, S. 139-207.
1913 b The atrio-ventricular connection in the reptiles. Anat. Rec,
vol. 7, pp. 273-285. Mackenzie, Ivy 1913 The excitatorj- and connecting muscular system of the
heart. 17th Internal. Cong, of Med., Sec. Ill,, Gen. Pathol, and
Patholog. Anat., Pt. 1, pp. 121-150. Mangold, E. 1914 Die Erregungsleitung im Wirbeltierherzen. Sammlung
Anatom. u. Physiolog. Vortrage u. Aufsiitze, Heft 25, Bd. 3, S. 1-36. Meikeljohn, J. 1913 On the innervation of the nodal tissue of the mammalian
heart. Jour. Anat. and Physiol., vol. 48, pp. 7-18.
EXPLANATION OF PLATES
All the figures are untouched photographs of sagittal sections of the heart of the tortoise Chelopus guttatus. They are arranged in order from the left to the right hand side of the heart. In all of them the auriculo-ventricular funnel is shown invaginated into the cavity of the ventricle to become continuous with its musculature near the attachment of the auriculo-ventricular valves.
PLATE 1
EXPLAXATION OF FIGURES
1 A section through the left auricle passing through the right side of the opening of the pulmonary vein into the left auricle; a portion of the auricular septum is seen.
2 A section to the right of the opening of the pulmonary vein. On the dorsal side a nerve trunk can be seen running in the pericardium to end in the musculature of the auricular-ventricular funnel.
3 A section to the left of the opening of the sinus into the right auricle. On the dorsal side the beginnings of the folding of the pericardium, which from the sino-ventricular bundle, are shown.
4 A section just to the left of the opening of the sinus; the folds of the pericardium have advanced further; the two venous valves are seen.
442
CONNECTING SYSTEMS OF THE HEART
HENRY LAURENS
PLATE 1
443
PLATE 2
EXPLANATION OF FIGURES
5 A section through the opening of the sinus into the right auricle. The sino-ventricular bundle is here seen extending as a free band over the auricle from the sinus to the anterior dorsal surface of the ventricle. Just above and to the right of where it joins the sinus there is a group of nerve cells and fibers, and further anterior and to the right, another group.
6 A section further to the right than figure 6. In the sino-ventricular bundle two nerve trunks can be seen; a stronger one running down into the connective tissue of the space between the funnel and the inner wall ( f the ventricle, and a smaller one, to the right, going to the outer surface of the ventricle.
7 and 8 Sections far over to the right of the auriculo-ventricular connection to show the fold of the pericardium which forms the sino-ventricular bundle becoming continuous with the pericardium over the sinus, the auricle, and the ventricle.
444
CONXECTIXG SYSTEMS OF THE HEART
HENRV LAURENS
PLATE 2
44.5
THE ANATOMICAL RECORD, VOL. 9, NO. 6
THE ADULT ANATOMY OF THE LYMPHATIC SYSTEM IN THE COAIZVION RAT (EPIMYS NORVEGICUS)i
THESLE T. JOB
From the Laboratories of Animal Biology, State University of Iowa
FOLK FIGURES
In the fall of 1914 the writer undertook the study of the origin and development of the lymphatic system in the common rat. Before the work had proceeded verj^ far it was evident that a knowledge of the adult anatomy would be of no small amount of help in guiding and interpreting the work on the origin and development of this system. So the plan '^f as changed to a study of the adult anatomj^ of the lymphatic system. This paper gives the chief results of the work.
Fifty rats have been examined in the course of the stud}'. The lymphatic system was injected from the soles of the feet, the tip of the tongue, the lips, the waUs of the intestines, the spleen, the lumbar, thoracic, inguinal, axial, intestinal and cer\dcal lymph nodes. The first nineteen specimens were injected with India ink, using the hjiDodermic sATinge for pressure. In the next six specimens, a solution of soluble blue was used, with a glass cannula and pressure bulb apparatus. The remainder of the work was done with Berhn blue gelatin mass, a small crystal each of thymol and potassium iodide being added to preserve and lower the melting point of the mass. The supply of gelatin mass w^as kept in a warm water bath, and the cannula occasionally warmed. Injections could be made through a much smaller aperture and in a much more satisfactory way by this method.
1 An abstract of a thesis presented to the Graduate Faculty of the State University of Iowa for the degree of [Master of Science.
447
THE ANATOMICAL RECORD, VOL. 9, SO. 6
448
THESLE T. JOB
RESULTS
Injections made through the soles of the feet, showed a variable number of superficial lymph vessels joining to form^ larger lymph vessels which followed the main course of the radial and ulnar veins, or dorsal and plantar branches of the saphenous vein, to the elbow or knee lymph node {1, 2, 3, 4, fig- !)• From the single node, 3 and 4, the femoral lymph vessel continued to the lumbar node {5-6). A branch is given off from this vessel at the juncture of the superficial epigastric vein and the femoral vein, which leads to the inguinal nodes (7-5).
Just below the posterior end of the rectum is a small single node {30) which receives the lymph from the caudal region and appendages. From this node a vessel leads to the left lumbar node {6), or to the left femoral lymph vessel, joining it at about the juncture of the iliac and inferior vena cava.
The lumbar nodes {5-6), which are normally double, lie just caudad of the iho-lumbar veins, on either side of the vena cava. There is a great range of variation, however, in the position of these nodes, due, in the main, to the variabihty of the ilio-lumbar veins. If the veins are well caudad, the lymph nodes may be crowded back opposite to each other; or, if the veins are well
FIGURE 1
1-2, right and left elbow nodes S-4, right and left knee nodes 5-6, lumbar nodes 7-8, inguinal nodes 9-10, renal nodes
11, cisterna chyli
12, cisterna group of lymph nodes
13, intestinal node
14, thoracic duct
15 to 20, axial nodes
23-24, jugulo-subclavian taps
25, thoracic group
26, posterior cervical nodes
27, anterior cervical nodes
28, submaxillary nodes
29, tongue and lip plexus SO, caudal lymph node
FIGURES 2, 3
a, renal lymph vessel
ci, cisterna chyli; 12, cisterna group
13, intestinal node; 5-6, lumbar nodes
9-10, renal nodes
p.c.v., posterior vena cava
i-l., ilio-lumbar vein
FIGURE 4
1, spleen; 2, fundus lymph vessel
3, pyloric lymph vessel
4, intestinal lymph vessel
5, connection with portal vein
6, intestinal branch to cisterna chyli
7, appendix; 8, mesenteric nodes
9, intestinal vessel
10, mesenteric branches of lymph vessel 13, intestinal node
ANATOMY OF THE LYMPHATIC SYSTEM
449
Fig. 1 Venous system shown diagrammatically in solid lines, the Ijonphatic sj'stem stippled. The lateral veins and lymph vessels are continuous, although not shown so in the figure. Lymph node 13, the intestinal node, so ■marked in all drawings.
450
THESLE T. JOB
k%
i! V
Tiny, .-.■',——■.-3? S«!l
'^"»'^
-J
\
.1^"
^
AXATOMY OF THE LYMPHATIC SYSTEM
451
'% ^12
■*^- ifi
s ^g w,
iW
3 C
Figs. 2, 3 Exceptional and type variations of the lumbar region
Fig. 4 The intestinal tract and lymphatics
452 THESLE T. JOB
forward, the left lumbar node may be a considerable distance in advance of the right node. In a few instances the double nodes were divided into two single nodes, w^hich lie some distance apart, and in two specimens an additional double node was found at the bifurcation of the ilio-lumbar vein on the right side.
The lymph vessel leading from the right lymph node (5) is so variable that a definite description cannot be given. It usually joins the left lymph vessel in some way between the lumbar node and the cisterna chyli. Figure 2, D, E, F, shows some variations of a general type. Figure 3, A and B, shows two additional variations, which are more common, in the general plan, than any of the others. In only four specimens did the right lumbar l3anph vessel lead to the right renal node (9) directly. In these cases a small lymph vessel tapped the renal vein from the renal node, and another vessel connected with the cisterna chyli (11), or the cisterna chyli group {12). In two specimens the right lumbar lymph vessel gave off a branch, which connected with the ilio-lumbar vein (fig. 3, C). This branch did not follow the veins to its tap, but went directly from the node across the ilio-psoas and psoas muscles to its juncture with the ilio-lumbar vein about 1 cm. from the vena cava.
On the left side the number of lumbar lymph vessels leading from the lumbar node (6) may vary from one to four, or form a network, depending somewhat on the mode of attachment with the right lymph vessel. However, all the vessels lead along the left side of the vena cava, in any case. If there is only one vessel, it will open into a single node (10), just anterior to the left renal vein, from which a branch is given to the renal vein and one to the group of single nodes (12) lying to the left of the cisterna chyli. If there be more than one lymph vessel leaving the lumbar node, some one of them will enter the renal node, the rest may join the cisterna group, the cisterna directly, ihe renal vein directly, or any combination thereof. The latter conditions are fewer than the single vessel method.
The number of nodes in the cisterna group (12) vary from one large one to six small ones. From this group one or more
ANATOMY OF THE LYMPHATIC SYSTEM 453
vessels enter the cisterna chyli. Two special exceptions to this method are shown in A and B of figure 2. Drawing A shows both of the lumbar lymph vessels entering the cisterna chyli directly. The nodes of the cisterna group being closely collected about the periphery of the cisterna chyli, did not fill with the injection mass. The connection with the left renal vein was made from the intestinal Ijinph vessel (13) through a small branch vessel (a). The specimen from which drawing B was made, had no cisterna chyli, only a short vessel shunted off from the main thoracic duct marked its normal position (ci). The renal branch (a) came from the main lumbar lymph vessel. In both of these specimens the connection between the intestinal h'mphatics and the portal vein was very prominent.
In addition to the connections with the cisterna chyli already mentioned, there is another lymph vessel received from the intestinal region past the intestinal node (13). The intestinal lymph does not pass through the intestinal node, but that the node is connected with the lymphatic sj^stem is shown by the fact that injections made from it fill the main lymph vessels. Also, the mass from the lumbar injections frequently pass into this node but not bej^ond it. From the cisterna chyli the thoracic duct (14) leads dorso-lateraUy along the superior vena cava and left innominate vein to its juncture with the venous system in the jugulo-subclavian district. Not a single specimen in the fifty rats showed the thoracic, duct branching or entering the right jugulo-subclavian district, as pointed out by McClure and Silvester,- Silvester,^ and Davis,^ in other forms.
A double lymph node is found in the groin, the inguinal nodes {7-8), located in the bifurcations of the superficial epigastric vein, and so closely attached to the lower dermis that in removing the skin the node remains imbedded in the subcutaneous
2 McClure and Silvester. A comparative study of the lymphatico-venous communications in adult mammals. Anat. Rec, vol. 3, 1909.
F. Silvester. On the presence of permanent communications between the lymphatic and venous sj'stem at the level of the renal vein in adult South American monkeys. Am. Jour. Anat., vol. 12, 1912.
Henry K. Davis. A statical study of the thoracic duct in man. Am. Jour. Anat., vol. 17, 1915.
454 THESLE T. JOB
tissue. While this node is connected with the femoral lymph vessel, in only two injections did the mass run backward into it. The vessel leaving anteriorly from the inguinal node soon divides into two branches. These branches lead through the lower layers of the dermis to the axial region where they again join to enter a double node, which is usually found in the bifurcation of the lateral thoracic vein (15-16). From here a lymph vessel leads to another double node (17-18) located in the same general region onh^ a little anteriorly: then to the third double node of this group (19-20) found by the intersection of the lateral thoracic and the axillary vein. This node also receives the lymph vessels from the elbow nodes (1-2). On the left side a lymph vessel leads from node 20 to the communication with the venous system, either through the thoracic duct tap (2Ji), or through a separate tap in the immediate district. On the right side the communication i^ in the same relative position (23).
In the anterior part of the thorax, between the two innominate veins, are two groups of single nodes, the thoracic groups (25). Each group varies in the number of nodes it contains, from four to eight. When one node in the group is injected all the other nodes of that group fill up with the mass and show a vessel leacUng to the venous connection in the jugulo-subclavian district. The right group connects with the right district and the left with the left district. Injections from other parts of the lymphatic system do not show these nodes.
The tongue and lips have man}' small vessels which collect in larger main vessels on each side of the head. These vessels lead to the submaxillary lymph nodes (28), which lie at the branching of the external jugular vein into the anterior facial and transverse vein. From these nodes, vessels lead almost directly inward, dorsally, to single nodes lying on each side of the trachea. Further down the trachea, on each side, are two nodes lying in close proximity, one a very small single node and the other a large double node. A lymph vessel leads from the larger node to the jugulo-subclavian tap on its respective side.
ANATOMY OF THE LYMPHATIC SYSTEM 455
From injections in the intestinal walls a great number of small vessels are shown collecting into larger and larger vessels until the}^ finally join the great intestinal lymph vessel, which follows the large intestine from the appendix to the intestinal lymph node in the anterior part of the abdominal cavity. In the region of the appendix there are several lymph nodes, all of which are single. In some cases they are so united as to form one large compound node, 2 or 3 cm. long. There is usually a single node at the juncture of each mesenteric lymph vessel (8, fig. 4) with the main intestinal lymph vessel (9, fig. 4). Just a short way from the cisterna chjdi is the large, single, intestinal node (13, figs. 1, 2, 4), which marks the branching of the large intestinal lymph vessel. From here one branch goes to the cisterna chyli (6) and the other to the portal vein (5). The injections made from the nodes in the region of the appendix did not always show the portal branch of the A^essel, but there were a sufficient number of cases to demonstrate that such a connection is fairly common. There were a varying number of single nodes in the region of the large intestinal node which did not fill up with the mass from the intestinal injection. However, injections made from them showed that they were connected with the intestinal lymph vessels.
Injections made from the spleen (1, fig. 4) show four lymph vessels leaving at various places along the hilus, accompanied by the splenic veins. Near the cardiac end of the stomach the four splenic lymph vessels unite with a small lymph vessel from the fundus (2) to form the main splenic lymph vessel which now proceeds dorsally toward the pyloric end of the stomach, from which it receives another small lymph vessel (S) . Almost immediately thereafter, the branch from the intestinal lymphatics, when present (4), and the splenic vessel unite to join the portal vein (5).
456 THESLE T. JOB
REMARKS
On the nodes. There seem to be two types of nodes; (a) a single node in which the lymph enters at the periphery, passes through the body of the node to the hilus, where a lymph vessel is formed; (b) the double node, referred to above, in which two single nodes are bound together in the same capsule. Instances showing this double nature are seen when injections made distal^ to a double node fill the posterior half of the node and pass on well into the vessels beyotid before the anterior half of the node begins to fill; injections made in the posterior half of these nodes will not fill the anterior half for some time after the vessel leading from the node is filled; injections made in the anterior half seldom enters the posterior half; in some instances where normally a double node is found, two single nodes were found only slightly separated; and finally, dissections can usually be made, to show the double nature of the nodes. The significance of the double node has not jQt been determined, but it is possible that the two parts perform separate functions. The posterior part of the node responds to the injection mass, just as the single nodes of the head, knee, elbow and caudal region do, while the anterior part responds as the intestinal and thoracic nodes. It is possible, therefore, that the posterior part and certain single nodes may act as filters primarily, while the anterior part and certain other single nodes may act as elaborators only.« A histological study is to follow.
On the vessels. No attempt was made in this study to determine the nature of the origin of the lymph vessels in the tissues. The injections do show, however, a definite tubed system beyond the origin, carrying lymph in only one direction and that toward the venous connection. The valves in the vessels are shown by the knotted appearance of those vessels filled with the mass. In general, the lymph vessels follow the blood vessels very closely the lumbar region being the main exception. The walls are
The venous connection is considered the anterior end of the lymph vessel. ' Sabin, in Morris's "Human anatomy," Part II, page 702, recognizes one type of h-mph node with two functions.
ANATOMY OF THE LYMPHATIC SYSTEM 457
very thin and easily ruptured by mechanical devices, but are capable of considerable extention in situ without injury. Where a lymph vessel passes through a node it always enters at the periphery and leaves from the hilus.
On the connections. In addition to the venous connections pointed out bj McClure," Silvester,^ and others, there appears to be two additional connections in the rat, the portal vein connection and the ilio-lumbar connection.
The portal vein connection receives the lymph from the spleen and stomach, and part of the intestinal lymph (fig. 4, 5) .
The ilio-lumbar connection receives some lymph from the right lumbar lymph vessel (fig. 3, C).
The left jugulo-subcla\ian tap, studied by AlcClure and Silvester in several forms,- in the rat receives the lymph from the left side of the head and neck, the left thoracic group, the left fore limb, side and hind limb, the deep seated vessels of the right hind limb and lumbar region.
The right jugulo-subclavian tap^ receives the lymph from the right side of the head and neck, the right thoracic group, the right fore limb and the superficial vessels of the right side and hind hmb (fig. 1,23).
The right renal vein communications, studied by Silvester,^ when present in the rat, receives part of the lymph from the deep seated vessels of the right hind hmb (fig. I, 9).
The left renal vein communication, ^ receives part of the lymph from the hind quarters (fig. 1, 10).
The portal and ilio-lumbar vein connections have not been reported previously, to the knowledge of the writer.
On general arrangement. The lymphatic system of the trunk is sinistral in the rat. The irregular occurrence and the extremely variable connections of the right lumbar lymphatics, as contrasted with the comparatively stable left lumbar system, is evidence of this fact. In all of the appendages of the body the lymph vessels follow very closely the various branches of the venous system. When the lymphatic system is injected it is very easy to distinguish the arterial, venous and lymphatic systems, when they occur together.
458 THESLE T. JOB
In conclusion, I take this opportunity to thank Dr. Frank A. Stromsten for his many helpful suggestions in the course of this work, and for his human interest, which has made this work possible and a pleasure.
SOME ANATOMICAL DEDUCTIONS FROM A PATHOLOGICAL TEMPORO-MANDIBULAR ARTICULATION
FREDERIC POMEROY LORD
From the Department of Anatomy and Histology, Dartmouth Medical School,
Hanover, N. H.
THREE FIGURES
In a previous paper, giving some observations on the jawjoint,^ I stated my belief that ordinarily the jaw is opened solely by the contraction of the two external pterygoid muscles; that the apparent barrier to the forward movement of the condyle, offered by the articular eminence, is almost entirely removed in life by the presence of the meniscus, or interarticular fibrocartilage, which, by its peculiar shape and movements largely obliterates the working depth of the glenoid fossa; and that thus during this action the path described by the moving condyle is almost a straight, rather than a curved, line, the direction of w^hich is nearly parallel to the plane of pull of the lower heads of the two external pterygoid muscles.
A few months ago, through the courtesy of Prof. F. W. Putnam, Director of the Peabody Museum at Harvard College, I was given access to its splendid collection of several thousand skulls. In the course of my work on this material I came across a particular specimen, offering very direct, corroborative testimony on the points just noted, which I had demonstrated in my 'physiological' model two years ago, as well as to hint at one or two facts not mentioned before.
This skull was that of an adult negro from Algoa Bay; it was in good condition, and w^as normal except for the bony surfaces
1 Observations on the temporo-mandibular articulation. Anat. Rec, vol. 7, no. 10, pp. 355-367.
459
460 FREDERIC POMEROY LORD
at the right temporo-mandibular articulation, and for a very slight modification in the left joint as well. The teeth were practically all present, were well w^orn and equally so on both sides, indicating that the two sides were used interchangeably in chewing. At the left jaw-joint there was a slight roughening of the posterior part of the articular eminence, as if its covering of hyaline cartilage had been eroded at that point, but elsewhere the bony surfaces appeared normal.
At the right joint, however, very extensive changes had taken place. The condyle was elongated anteriorly to double the antero-posterior diameter of that on the other side. This made the altered condylar process about as long as it was wide, being roughly the size of a twenty-five-cent piece, though not circular in outline. It was made up of two flat surfaces, one sloping down and laterally, the other down and mesially, from a central liixo^ running in a sagittal plane the whole length of the process, it thus resembled an inverted trough. A transverse section gave nn obtuse angle of about 120 degrees, and a longitudinal, or ani,,»o_pQgtej.ior, section was a straight line. Its shape was then vei^ different from that of the normal condyle as found on the othei ^i^e (fig. 1).
The opposing surface .^ the temporal bone presented an appearance equally altered y^^ ^^le normal. Over the region of the original articular emin^^jg ^^^ ^j^^ anterior half of the glenoid fossa was a rough surfa.^^ ^j^^ reverse of that noted on the condyle, being a wide trough ,^-^j^ -^^ ^^^ surfaces meeting also in a line running antero-postei ^j.iy^ ^^le whole being somewhat longer in that direction than x^^ corresponding condylar surface. Behind it was a part of t^ glenoid fossa, but evidently not utilized in any way as an a^jp^jg^j. g^rface (fig. 2).
Putting the jaw in its proper positic^^ ^^ determined by the perfect occlusion of the teeth, it was at ^^^ ^^^^ ^j^^^ ^^le peculiar right condyle fitted exactly into tht ^j.^ygj^_i-i^g surface on the temporal bone, anterior to the remt.^^ ^^ ^^le old glenoid fossa, and that the left condyle was simil^j^ advanced a short distance in front of its usual position. T^ motion on the left side, although restricted by the slightly jyanced position of
TEMPORO-MANDIBULAR ARTICULATION 461
the left condyle, was otherwise probably almost normal. On the right side the condyle moved in its socket, much as the carriage of a sliding microtome knife slides freely back and forth m its bearings, with no possible lateral deviation.
Obviously the right meniscus was either entirely lacking, or if present, did not serve its ordinary function of adjusting the bones to each other, as shown by the shape of the articulating surfaces and the uniform thickness of the intervening space
Nevertheless, as shown by the wear of the teeth, and the shape and relative size of the opposing surfaces of the enlarged condyle and its reverse, the former glenoid fossa and articular eminence, the condyle must have moved forward and backward during the various motions of the jaw very much as it did on the left side or as it does in a normal case. If there had been no forward and backward motion of the condyle, such as is necessary for proper trituration, the teeth could not have been so worn, and the articular surfaces would have had an entirely entirely different shape. With only a pure hinge movement the jaw-jomt would undoubtedly have been more like that of the Carnivora, a cylinder rotating about its transverse axis, fixed in a cylindrical fossa above.
It is easy to see how the muscles of mastication in moving the jaw during the activity of the osteo-arthritis, which was evidently the cause of this abnormality, while the diseased surfaces were still plastic, formed joint surfaces exactly adapted to their proper sort of motion and to no other. The meniscus apparently had to be sacrificed, due to the disease, and certain advantages pertaining to this mechanism were lost, such as the lessened degree of friction in this moving hinge-joint and the presence of an elastic cushion to take up the shock of sudden and hard biting. But a very satisfactory substitute for a normal temporo-mandibular articulation was thus manufactured, a substitute that gave the same kinds of motion, though less in amplitude, as ordinarily found, in opening the jaw and in trituration, and yet one that accomplished this without the presence of a glenoid fossa, articular eminence, functionating meniscus, or cylindrical condyle.
462 FREDERIC POMEROY LORD
Due to the peculiar conditions obtaining in this unusual specimen we have here crystallized evidence of what kind of motion was going on in that joint, a motion which must have been similar to that of the other, practically normal, side. The evidence is much better than that offered in the ordinary case where the presence of an active meniscus makes more difficult an appreciation of the actual path of the advancing condyle. Here the path itself is molded for us, not in moving cartilage, but in the enduring material of bone.
Study of this 'ossified testimony' gives us the following: The path of the advancing condyle was in a straight line, as noted in the working model described in my previous paper. There is no glenoid fossa or articular eminence in this case. So in the ordinary skull these irregularities are nearly smoothed out by the presence of the meniscus; and the condyle, there also, moves in a nearly straight line (fig. 3) .
If the lines of the two advancing condyles be joined by a plane, that plane is seen to be almost identical with the plane formed similarly by the lines of pull of the lower heads of the two external pterygoids. The latter plane makes a slight angle with the former, about five degrees above its level, just enough to enable the opening muscles, the two external pterygoids, to apply the condyle not too firmly against the skull above, rather than to pull it away from its bearings, keeping the joint surfaces in close contact ready for the instant action of the closing muscles to take effect at any stage of the process of opening. It is interesting to note, however, that in closing the mouth the muscle which retracts the condyle, the posterior fibers of the temporal, does so at not less than an angle of twenty-five degrees above the plane of the condylar path. Thus it always keeps the bony surfaces in very firm contact.
The path of the condylar advance lies in the plane of the temporal muscle, very naturally, it seems, as the strongest of all the closing muscles, and also as the only retracting muscle. Very likely it is this muscle that had most to do with the molding of the plastic bony surfaces during the active stage of the disease. The condylar path, in being parallel to the sagittal
TEMPORO-MANDIBULAR ARTICULATION
463
plane, is thus parallel to the resultant of the combined pull of the two external pterygoids, the main muscles utilized in opening the mouth.
Examining the condyle of the diseased joint bj^ an imaginary coronal section its two surfaces are seen to slant as follows: the slope running laterally from the ridge at its summit makes an angle of about twenty" degrees with the horizontal, while the
Fig. 1 Drawing of condyles of mandible. The upper figures show the condyles as seen from behind; the line below the upper part of the drawing gives the edge of the articulating surface. The lower figures sliow a view of the condyles from above. In each case the right hand figure represents a view of the pathological right side, showing its enlarged and altered articulating surface. The dotted lines show the approximate number of degrees of slope from the horizontal of each side of the joint surface.
surface running mesially from the ridge makes an angle of about forty degrees — being about twice as steep (fig. 1). The reason for this appears when we examine the planes in which the closing muscles act. The temporal muscle pulls practically vertically, not tending to dislocate the condyle either mesially or laterally. The masseter pulls laterally at an angle of twenty degrees from the vertical, tending to pull the condyle outward, the internal pterygoid tends to pull the condyle inward, as it acts at an angle
THE ANATO^^CAL RECORD, VOL. 9, NO. 6
464
FREDERIC POMEROY LORD
2 3
Fig. 2 Drawing of base of skull to show articulating surfaces for mandible. The broad oval area on the left of the figure is the pathological troughlike surface, articulating with the right condyle, shown in figure 1. The narrower oval of the opposite side of the skull is almost normal, although neither articulating surface extends as far back as the Glaserian fissure (shaded).
Fig. 3 Drawing of right side of skull. This shows the flattened articular surfaces of the pathological right jaw-joint, which allow the condyle to move forward and backward only in a straight line. The line of pull of the lower head of the external pterygoid muscle makes an angle of about five degrees with that of the joint surfaces, and is so directed as to pull the mandible against the skull during contraction.
of forty degrees to the other side of the sagittal plane. These lateral and mesial strains are exactly met, and in proper proportion, by the two sloping surfaces of the condyle as they are applied against the sides of the trough in which thej^ run.
I further noticed that, although the surfaces of the condyle of an ordinary skull did not show these suggestive angles, if I placed an artificial meniscus above it — one molded in wax, fitting exactly that particular specimen — then the same significant slopes were seen in a transverse section of the condyle plus its meniscus.
While the study of this specimen perhaps adds little that is entirely new to our knowledge of the jaw-joint, it certainly gives very tangible proof of certain conditions, ordinarily not readily seen.
LABORATORY AND TECHNICAL MISCELLANY
ARTHUR W. MEYER From the Division of Anatomy of the Stanford Medical School
SIX FIGURES
AN ODORLESS DISSECTING ROOM
To attain this desirable end we make use of the ordinary chemicals, namel}: glycerine, carbohc, alcohol, formaline, and find that it is largely a matter of quahty and proportion. Since the impurities in the cruder carbolic acid cling to, if not penetrate, the hving as well as the dead, we have avoided its use altogether and substituted a much better grade of acid with a melting point of 35 to 38° C. Moreover, since carbohc acid of 33 per cent strength usually crystallizes especially in the viscera of the cadavers and in this strength always is an obstacle to efficient work because of its anesthetic effects, we reduce the proportion of carbolic acid to about 10 per cent. Tliis reduction in quantity also compensates considerably for the increased cost of the better grade of acid used, avoids the strong odors of the impurities and numbing of the fingers during dissection.
In addition to the carbohc we use from 2 to 2j per cent of formahne, 20 per cent alcohol, and 20 per cent commercial glycerine. Since the formahne effectively fixes the material a larger quantity of alcohol than needed to get the carbolic to dissolve and the glycerine to flow seems wholly unnecessary. It evaporates quickly upon exposure carrying odors with it and also materially increases the cost. The large quantity of glycerine used also adds considerably to the cost, but I know of nothing that can replace it.
The total quantity of preservative used per cadaver should, to be sure, vary with the condition and size of the cadaver. From twelve to twenty hters is what we use. This keeps the cost of the preservative per body down to about two to three dollars, which is a sum sufficiently small to be within the reach of all laboratories and especially of those that can afford lead-lined storage tanks. Although we use from 12 to 20 liters of preservatives per cadaver we depend on gravity for pressure.
All our anatomical material is kept in a saturated atmosphere, in ordinary wooden plank tanks each tank requiring only about a gallon of methyl alcohol per year in spite of the ch'}' climate. A very weak solution (i per cent) of formahne will, of course, accomphsh the same result as the methyl alcohol and for some time I have immersed some cadavers
465
^gg ARTHUR W. MEYER
?K S;t p^Je^^:;trrwa*ltS^^ .«.- direction, we avoid the "-o/™-lij'^« because of -ts mes^-.^^ ^,^„ ^ half a year
sations resulting from its use are so man^' and >ts d.smtecta q
so desirable that .h-:^^_°ot, cared to omt.t^ St.eete, ^^^^^^^^ ^^^^^ ^^^
their continuation. ^ f J?^,f J.f^iSent in sufficient amounts to assui-e
Trtatf to^r:^;f -SrJcStai^tSt ^l^™ hit ceived should determine how ?f ^^^^^f ^^ ^^.^^ '"^e preferences
In one laboratory with ^\nlcn i was ^ui mormng
sary in the spring of t^e year to scrape off 0^^^^^^ J
before beginmng dissections, bmce "^«^^ ^wb .^ ^ • ^^
strong carbohc and formaline ^^f ^^^^^^f ^J^^^f 'Jontan^^ seems
Slo tT:^.tf S\*l,t"T a^ :|v?SIhrL use of arsenic
is for this reason alonejery ^^^J^f ^^"^f,^^^;. ^_,„ as our best friend and
No doubt, someone has suggested ouic^^^^^^^^^^^ most efficient helper ^ gladly admit its Ixn^^^^^^^ ^^^^^^^^^^
can and also have been accomplished ese^^eleoy^^^
As preservative to moisten the f ^^t^;';^/^ ^^^f. ff^l^ we use essentially the lotion used by f ^^^^ ^^^^^ :^^iX?h It prevents But we cover the cadavers and tables AMth ^^h^U oucioin. i
LABORATORY MISCELLANY 467
evaporation very well, looks neat, is cheap, can be discarded as soon as soiled and is therefore far preferable to oiled muslin used perennially.
Anatomists need not be reminded that in an absolute sense the above caption is, to be sure, a euphemism. Yet visitors to our laboratory both from at home and abroad have used that expression and lest there be skeptics I shall add that a medical \dsitor from Berlin, for example, repeated the words 'Geruchlos' and 'Sonderbar' on going tlii'ough the laboratory. Likewise, an eminent English physiologist who was compelled to see the laboratory literally on the run, ejaculated as he hastily went through the dissecting room, "Well, I declare! Quite-unlike-the-old-dissecting room! Smells sweet!" Other instances could be given, but enough, and this much not as a toot from our own horn but merely for the benefit of the skeptic.
Architecturally our laboratories are not ideal; but we as others have hopes. Although the ceilings of our laboratory are sixteen feet high there is no ventilation except by means of a few thi'ee-foot-square sashes placed between six to nine feet above the floor. Steam heat is used. In spite of these things our laboratory is practically odorless. We don't deoderize the cadavers simply because we don't know how but we do not add to the natural odor through the use of preservatives from which there is less escape.
THE STAINING OF ELASTIC TISSUES
While using the various elastic tissue stains on the fetal vessels I noticed that sections stained in a watery solution of orcein and mounted in glycerine alwaj^s showed a much more brilhant stain. In fact fine elastic fibers, which for some reason were scarcely noticeable by other methods, stood out ver}^ distinctly by the use of this method.
^lORDANTING WITH IODINE FOR MALLORY'S CONNECTIVE TISSUE
STAIN
For staining the reticulum of lymph and hemal nodes by Mallory's method a special fixation as recommended by him is desirable. That this is preferable is admitted but it was found that by a preliminary mordanting of sections in a very weak iodine solution, before staining, almost any kind of fixation would jdeld a fair result -^dth Mallory's stain. The advantage of this procedure lies, to be sure, onlj^ in the fact that it is an emergency measure which often enables one to use this special stain on material fixed by other methods than the preferred.
MESH DISSECTING TABLE TOPS
The hypostatic accumulation of the preservative and of the body fluids in some bodies and especially in connection with the use of certain methods of preservation suggested the use of a wire-mesh, false table-top. Such a table-top also prevents the accumulation of fat and the preservative which is applied externally during dissection on the
468
ARTHUR W. MEYER
table and keeps the cloth wrapping from becoming soaked and unsightly. Moreover, it makes the use of water for cleansmg the material easy and agreeable. .
Figm-e 1 shows the wire mesh top with the zmc binding. We use Xo 14 2 X 2 galvanized iron wire mesh but would far prefer an aluminum mesh were it not so expensive and so flexible. Such a removable
Fig. 1 IMesh table top Fig. 2 Dissecting table ready for mesh top
mesh top can be used on any kind of table for, as shown in figure 2, it is merely supported by removable rods. Since this mesh and the zmc table tops can quickly be cleaned with a hose when desired without removing the body, the accumulation of small pieces of material can be prevented easily and the tables kept clean throughout the progress of the dissection. The meshes are about one centimeter square.
If desired, a cork stopper can be placed in the drainage pipe or a valve can l)e put on it so as to permit preservative or water to stand on the table beneath the mesh. This effectively prevents drying even when a body is used for a long time for demonstration purposes.
LABORATORY MISCELLANY 469
mesh-bottom and felt-sealed specimen boxes One of the disagreeable features usually connected with the study of dissected and injected preparations and frozen sections is the fact that it is customary to keep them in boxes or jars containing a solution of carbolic acid. Indeed, in some laboratories such specimens are kept wholly immersed in a weak (5 to 10 per cent) solution of crude carbolic acid although as a matter of fact crude carbolic will dissolve in water only to the extent of about 3 per cent. Hence it is necessary to roll up the sleeves sometimes as far as the elbow, in order to remove the specimens for study. When removed the specimens also run and drip with the crude carbohc solution. In order to ob\'iate these objections we have placed false bottoms in the usual galvanized iron boxes and avoided the use of crude carbohc. These false bottoms are made exactly hke the mesh table tops except that a hghter grade of wire can be used. They lie upon triangular rests made of narrow strips of galvanized sheet-iron. These rests are about two inches high and run across the \Nidth of the boxes at such intervals as may be necessary properly to support the specimens (fig. 3).
The mesh is covered with oilcloth or mushn, and a piece of muslin, or cheesecloth, long enough to pass completely around in the box is put with its midpoint on the true bottom beneath the triangular rests, the free ends of the cloth extending up each side of the box. A small amount of a weak preservative containing but little or no alcohol is then poured into each box and the free ends of the muslin put over the specimens which lie on the false bottom. Capillarity keeps the cloth moist and the specimens can be removed at any time without the possibility of unnecessary soihng. They are always moist but never wet, dripping or macerated. Sections of frozen infants kept in such boxes for a period of four yem's are still in excellent condition in spite of frequent handling.
In order to prevent evaporation these boxes are provided with a gutter for the flange of the cover, as shown in figure 3. This gutter is fined with felt one inch thick, which can be fastened with a shellac or paraffine paint and which seals the boxes so effectively that suction is very evident when the fids are fifted. If desired, the gutters can also be filled with glycerine but this has never been necessary even in our long dry summers.
Coating the interior of the boxes with parafine wiH delay corrosion from formaline-preserved material but in the course of j'ears the boxes nevertheless get rather unsighth^ Hence, I am at present arranging to have similar boxes made in white enamel. These wiU be permanent and far preferable esthetically.
DRAWING AND BOOK STANDS
The need of convenient and stable individual book and drawing stands has been met variously in different laboratories. Although individual stands require more space and often make a room look
470
ARTHUR W. MEYER
Fig. 3 End view of specimen box; triangle supports and mesh bottom in place.
Fig. 4 Drawing and book stand. Fig. 5 Adjustable wire dissecting stool.
somewhat disorderly yet ever since my student days I have been so thoroughly convinced of their advantages that I have gone to some trouble regarding the matter.
The stand shown in figure 4 has been in use in our laboratory for over five years. It is exceedingly stable and not one stand so far has
LABORATORY MISCELLANY
471
required repairs. The heavy base has four rather than three legs, for stabihty. The legs are not provided with castors, for the same reason, but castors can be added in a moment, for holes are provided. The book box, which measures 12 X 9^ X 6 inches, is rotarj^ on the stand. Hence the student can always easilj^ reach the books. The oak drawing board, which cannot warp, is 12 X 6 inches in size. It is adjustable for slant and height and the rod supporting it can also be rotated. It is provided with a metal retaining edge which can also be adjusted for height, and wliich can be dropped to the level of the board.
Fig. 6 Adjustable wire laboratory stool
The whole stand, which is easily dismountable, is finished in white enamel except the metal supporting rods, which are nickeled, and the drawing board, which is finished in dark oak. The metal retaining edge is antique copper. This stand, which has answered our needs completely, can be obtained from Frank S. Betz, Hammond, Indiana, at prices from five to six dollars each, varying somewhat with the size of the order.
STOOLS FOR DISSECTING ROOMS AND HISTOLOGICAL LABORATORIES
While looking about for a durable and comfortable stool for the dissecting room, over half a decade since, I had the good fortune to be referred to the Chicago Wire Chair Company. I have not become a stockholder since. This company was then making a stool which answered our purposes provided we could be supplied with a longer screw assuring a greater range of adjustment. Through the courtesy of the manager of the company this was easily accompUshed and we have equipped all our laboratories with the stools represented in figures 5 and 6. The latter is a stock model and can be purchased ahnost
472 ARTHUR W. MEYER
anywhere but the former is modified to suit our needs. It is adjustable for a height of 18 to 26 inches and can easily be provided with metal ring foot-rests placed higher up than the rectangular rodbraces between the legs, if desired.
These stools, which are practically indestructible, are not at all expensive. Only one stool was damaged in six years and the parts that wear out or may break can easily be replaced.
The metal parts are finished in antique copper, although other finishes can be obtained. Both stools can be obtained with or without backs. From an experience of six years I am inclined to prefer both without backs, since an occasional careless student uses the back to knock a stool over or to tip it back while sitting on it, thus eventually loosening the screws in the oak top. The tops or seats are so built as practically to preclude warping. The address of the makers is La Salle Avenue and Ontario Street, Chicago. The only justification for this note and the accompanying illustrations lies in the frequent comments and inquiries of visitors and their dura])ility and ver}" reasonable cost.
LETHAL CHA^IBER
The essential thing in the construction of this chamber for killing small animals (dogs, cats, etc.) with illuminating gas, is a galvanized sheetiron false bottom which rests and slides on galvanized sheetiron right-angled strips attached to the walls of the chamber. In order to protect the zinc binding of the mesh bottom from soihng, an apron of wood, or preferably of sheetiron, is fastened to all sides of the box above the border of the mesh. Beneath the mesh bottom is a galvanized iron pan which is easily removable.
The animals are placed in the chamber, the front of which is provided with a well-fitting door, and the gas turned on graduall}^ If this is properly done dogs seldom utter a sound. When the animal is dead the removable mesh bottom and pan can be hosed off, thus keeping the chamber clean. Soiling of the animals is also made much less likely by the use of the mesh false bottom.
ANIMAL CAGES WITH HINGED BOTTOMS
In handling dogs it is often inconvenient to lean far into the cages in order to reach the retreating animal. To obviate this difficulty a shghtly raised and inclined board floor two feet wide can be provided on the near or door end of the cage. The rest of the floor can be made of a galvanized iron nu^sh which is hinged near the wooden floor on a galvanized iron rod, so that the farther end of the mesh floor of the cage can be raised by means of a rope or wire from the door end. This compels the animal to come forward to the door and makes inspection and removal very easy for both animal and caretaker.
LA^ORATOEY MISCELLANY 473
SECTIONAL PORTABLE LABORATORY CASES
It is often a great convenience to be able to move cases from one laboratory to another to meet some temporary need. Hence a special type of case with two sliding glass doors each in the base and top w^as designed. These cases can be built in one piece or the base and top can be built separately, and then fastened to each other by a few screws when placed in position.
Our cases are approximately 30 inches wide and 6 feet high, inside measure. The base is 2 feet deep but the top only 16 inches, leaving an offset of 8 inches which serves as a sheK. Since the doors are not hinged, anj^hing standing on the offset need not be removed before the doors are opened and there is no danger of sweeping things off onto the floor when the doors are opened hastily. Directly beneath the offset are two drawers 5 inches deep, placed side bj^ side. The central position of the drawers makes them very accessible. The size of the drawers and that of the unit itself can be made to suit personal preferences. The sliding doors can be provided with showcase locks or they can be locked with a metal peg (a sawed off 20-penny spike) and a hasp and padlock. The peg is placed in a hole which passes through the overlapping sashes in the middle and the hasp is brought over it and locked in place.
DRAWING FROZEN SECTION
Few students of anatomy attain sufficient command over pen, pencil and brush to be able to make rapid satisfactor}- freehand drawings of frozen sections. Aloreover, even if they had this command of drawing, the time consumed would be entirely too great when it is desired to make a considerable series of life-size drawings. We have obviated this difficulty by making rapid tracings on glass with India ink directly from the specimens, as is commonly done. These plates are then put on an inclined frame and illuminated from below by an electric hght. By placing drawing paper over the ink tracing an accurate cop}' can quickly be made in pencil or ink and the details filled in from the specimen. The latter are handled on a wooden tray and are always turned over between two trays or a tray and a glass pane, to prevent damage. Pencil drawings can be made permanent in the customary way with shellac. The ease with which drawings can be made in this way encourages students to get as complete a series as possible for later reference, as well as for immediate use.
NEUTRAL STAINS AS APPLIED TO THE GRANULES OF THE PANCREATIC ISLET CELLS
W. B. IMARTIN
From the Anatomical Laboratory of the Johns Hopkins Medical School
Through the investigations of Bensley and his pupils^ we are aware that two types of cytoplasmic granules are present in the cells composing the islets of Langerhans in the pancreas of most mammals. These granules are distinct from those found in normal pancreatic parenchyma cells and this enables us to identify islet tissue. Bensley's valuable contribution rests on the application of a neutral dye (e.g., Reinke's neutral gentian) to pancreatic tissue which has been fi.xed in a particular way. At the suggestion of Dr. H. ]\I; Evans, therefore, this work was taken up with the view of determining the best method of preparation of the neutral dyes, the concentration of the staining solution necessaiy for the best results and further to investigate certain of the dyes allied to gentian violet and orange G, in the hope of obtaining a neutral stain more efficient than Reinke's neutral gentian.
G#itian violet is a mixture of two dyes of the triphenylmethane series, hexamethylpararosaniline and pentamethylpararosaniline. Therefore, when gentian violet is combined with orange G the resulting neutral stain is also a mixture of two dj^es and it is this mixture that is known as neutral gentian. As the two components of gentian violet differ somewhat in their staining properties and as the relative proportion of these constituents vary in different samples of gentian violet on the market it would seem advantageous to substitute the pm-e hexa or penta compound for the mixture. This at once suggests the substitution of other dyes of the triphenylmethane group on the basic side of the reaction and also the replacement of orange G by other acid dyes of the azo series. This idea has been carried out and a number of neutral stains prepared. These dyes have been applied to pancreatic tissue fixed by the Benslej- method. The result of the study of these neutral dyes is set forth below with a brief description of each dye.
The acid dyes and basic dyes combine in molecular proportion. In some cases the ratio is a simple one of 1 : 1 and in other cases the ratio may be 1 : 2 or 1 : 4. For example, ethyl violet combines ^\^th ponceau
1 IM. A. Lane, The c3'tological characters of the areas of Langerhans. Am. Jour. Anat., voL 7, 1907; R. R. Bensle}% Studies on the pancreas of the guineapig. Am. Jour. Anat., vol. 12, 1912.
475
476 W. B. MARTIN
4 G B in the ratio of 1 : 1, with orange G in the ratio of 2 : 1 and with trypan blue in the ratio of 4 : 1.
The method of preparation of the neutral dye is the same in each case. A concentrated aqueous solution of the acid stain is added to a similar solution of the basic substance. This should be done slowl}^ and the mixture stirred thoroughly. The neutral point may be determined in the following manner: After each addition of the acid substance the mixture is stirred and a drop is taken on a glass rod and placed on a piece of ordinary filter paper. The neutralized portion of the mixture, being in the form of a precipitate, settles at once on the paper, while the liquid portion containing the unneutrahzed stain spreads in a circle around the deposit. By the color of the outer ring can be determined whether the solution contains an excess of the acid or the basic dye, and by the change in the degree of colorization one can readily perceive the approach of the neutral point. When this is reached the outer ring is entirel}^ colorless. The end point is thus made as exact as in any other chemical reaction and the filtrate in such a case is either clear or only slightly colored. One thus avoids an excess of either stain and the residue is practically free from either of its single constituents. This is of practical importance, for the failure to free the neutral stain of one of its components may account for some of the difficulties that have been encountered in the use of neutral gentian. The residue obtained above is filtered, washed with distilled water and allowed to dr}- either in the air or in an oven at a low temperature.
In order to determine the concentration of the staining solution necessarj^ to obtain the best results, solutions of varying streng-th were used, the staining time remaining fixed. A solution of known strength was made up in absolute alcohol and this was kept as a stock solution. From the stock, staining solutions in 20 per cent alcohol ranging in concentration from one of 8 mgm. of the solid dye in 50 cc. of alcohol to one of 0.25 mgm. of the dye in 50 cc. alcohol were prepared. The strength of these various staining solutions is given in table 1. Positive or negative results are indicated by the corresponding mathematical signs. To obtain the best results fairly dilute solutions should be used. A solution of neutral ethyl violet orange G containing approximately 2 mgm. of the cr3^stal dj^e to 50 cc. of 20 per cent alcohol is satisfactory. A staining solution of neutral azo fuchsine should contain 0.5 to 1.0 mgm. of dye to the same amount of alcohol.
AVhen brought together in molecular proportions these dyes react with the precipitation of a neutral dye. On drying, this is a dark green powder giving a deep purple \'iolet color in alcohol:
Crystal violet + Orange G
fthe sodium salt of benzcueHydrochloride of hexamethylpararosanilino + { azo-/8-naphthol-disulphonic
[ acid y
NEUTRAL STAINS
477
TABLE 1
Shoiriiig relative staining power of different neutral dyes. The numbers refer to Schultz's Farbsfofftabellen, 19U
Milligrams of neutral dye added to 50 cc. of 20% alcoliol| for staining solution
Orange G(38) (A) + gentian violet + + penta methyl violet (olo) (C.J.)- ■ . + — + hexa methyl violet (516) (C.J.) .... + +
+ crystal violet (517) (B) +
+ ethyl violet (518) (B) +
+ victoria blue B (.559) (B) + + +' + +\
4 2 11 0.5 0.25
+ + +
Ethyl violet + azo eosin (94) (B7) + brilliant croceine 3B (227) (B7) .... + + croceine orange G* (B7) + +
+ Ponceau 4 G. B. (37) (A) + +
+ diphenyl brown (347) (Sch) + +
+ diamine brown (344) (Sch) + +
+ azo fuchsine G (146) (B7) + +
+ trypan blue (391) (C) + +
-f heliotrope B** (Sch)
+
+
_
+
+
+
+
+
+
+
+
+
+ +
From Fabrenfabrik of Elberteld Company. Although given by Schixltz in his last edition (V)
as identical with ponceau 4 G. B. the dye is undoubtedly different.
• Not given in the last edition of Schultz under this name.
Methyl violet +
Hydi'ochloride of pentamethylpararosaniline +
Orange G sodium salt of benzene-azo/8-naphthol-disulphonic acid y
Hexa methyl violet, wliich is given in the table, has the same composition as crj'stal violet but is a purer product and gives slightly better results. Both of these dyes react with orange G in the same manner, giving a coarsely crystalline neutral dj^e of a green color and wdth a a fine metaUic luster. Sections of pancreatic tissue stained in anyone of the above d^-es present much the same picture, though the hexa methyl violet is decidedly superior to the penta methyl ^^olet. In each case the zj'inogen granules of the parenchyma are stained a ^-ivid heliotrope on a light orange background. The nuclei of all the cells are blue or \dolet, and in the islets of Langerhans the granules in A and B cells of Lane are differentiallj' stained in the same manner as when stained in neutral gentian. This is to be expected, as gentian ^nolet is a mixture of methyl violet and cr^^stal ^aolet and in staining power lies intermediate between the two:
Ethyl violet +
Hydrochloride of he.xaethylpararosaniline
+
Orange G the sodium salt of benzene-azo'/3-naphthol-disulphonic acid y
478 W. B. MARTIN
Ethyl violet and orange G react in the proportion of 2 : 1 in the usual way, giving rise to a neutral stain. This forms lustrous green cry^stals having a slightlj^ bronze cast and gives a violet solution in alcohol. This stain is much superior to any of the above. While the general picture is the same, it is much more intense in its action and the three types of granules are more sharply differentiated. It has the advantage also of resisting the action of the dehydrating and differentiating agents better:
Victoria blue B + Orange G
Hydrochloride of phenyltetramethyltri- 1 , fthe sodium salt of benzene-azoamidodiphenyl-a-naphthyl-carbinol + / \ /3-naphthol-disulphonic acid 7
The neutral dye is obtained in the same way and forms lustrous crystals of a deep brownish-red color. The alcoholic solution is blue without the red cast of the stains so far mentioned.
Sections stained in this dye present a somewhat different picture. The zymogen granules are stained blue on a pale yellow background. The granules in the islet cells are stained a more intense blue and retain their color when differentiated from a dilute solution longer than the zymogen granules. The nuclei of all the cells are stained a beautiful blue green, the nucleolus and chromatin network standing out clearly against the rest of the nucleus. This dye wall stain in high dilutions, but on account of the deficiency in color contrast between the granules and protoplasm is not as desirable a stain as the neutral ethyl violet.
The series of neutral stains so far given have been combinations of orange G with various basic dyes of the triphenylmethane series. Of these ethyl violet is the most valuable. In the following experiments this dye has been joined to a number of acid dyes of the azo series and the staining power of the resulting compound studied:
Ethyl violet — Azo eosin
,- , , , . , „ , ,, , .,. , fthe sodium salt of anisol-azo-a rlvdrocnloride 01 hexaetnylpararosaniline + < 111 11 • -j
(^ naphthol-p-sulphomc acid
The neutral stain crystallizes, forming bronze green crystals, and gives a deep violet red solution in alcohol. It is, however, lacldng in staining power and in concentration as high as 32 mgm. of dj'e to 50 cc. of 20 per cent alcohol stains the zymogen granules very faintly and the nuclei not at all:
Ethyl violet + Brilliant croceine
sodium salt of benzene-azo
Hydrochloride of hexaethylpararosaniline + \ benzene-azo-/3-naphthol disul phonic acid 7
Two parts of ethyl violet combine with one part of brilliant croceine. The neutral stain is a lustrous crystalline sub.stance of a fine green
NEUTKAL STAINS 479
color forming a violet red solution in alcohol. The protoplasm of the cells stains a pale pink while the zymogen granules take a slightly darker shade. The granules in the islet cells are stained faintly and the two types can be made out but the differentiation is poor:
Ethj'I violet + Croceine orange G
Fine bronze crystaUine substance giving a violet red solution in alcohol. Sections stained in this dye are a bright yellow. The zymogen granules are a reddish brown. The granules of the islet cells of one type are stained a deep blue while the other type takes the same j^ellow color as the background. In higher dilutions the zymogen granules still take the stain but the islet granules are not stained. The nuclei are shown fairly well:
Ethyl violet + Ponceau 4 G. B.
TTj ,,.,,, ,,, ... , fsodium salt of benzene-azo-/3 rlyarochlonde of hexaetnyipararosaniline +< i xi i ^ i i • -j
[ naphthol-/3-sulpnonic acid
These dyes combine in the proportion of 1:1, forming a gummy residue which crystallizes out on standing and gives a deep red solution in alcohol. The staining action of this dye is very much like that of the neutral ethyl violet orange G compound except that it is dissolved out much more rapidly by differentiating agents and on this account is entirely unsatisfactory:
Ethyl violet -f- Diphenyl brown
sodium salt of
^salicylic acid benzidin<f
Tnonomethylamidonaphthol-sulphonic acid
Hydrochloride of hexaethylpararosaniline + <
Ethyl violet and diphenyl brown combine in the ratio 2 : 1 to form a neutral stain. This is a dull black amorphous powder giving a violet alcoholic solution. The zymogen granules appear heliotrope on an orange background. The granules in the islet cells take the stain but they are not sharply differentiated:
Ethyl violet + Diamine brown
the sodium salt of
Hydrochloride of hexaethylpararosaniline + *
/Salicylic acid benzidinx"
^amidonaphtholsulphonic acid
THE ANATOMICAL RECORD, VOL. 9, XO. 6
W B. MARTIN 480
Ethyl violet combines with dia.^ne b.o.™ in }^^ ^^
dXrentiated and take a shade of le^^ not ^ei3 ^^^^.^^^^ .Gained
, Heliotrope B
Ethyl violet "^ [the sodium salt of diansidme • r 4- di-monoethylamidonaphtha Hydrochloride of hexaethylpararosanilme -t- | ^^^^ ^^^ptonic acid
1 ^^.nno-T deep violet alcoholic
This is a dark brown amorphns P^^^^^JTilht heSotrope and the
soEn. Zy-ogen granules are stamd^^^il^^ ,,, „ot well stamed^
i°a::ir^Jgrb:-f^nS^^^^^^^^^^^
, Trypan blue
Ethyl violet + .^^ sodium salt of tolidme n.line + diamidonaphtholdisulphonic
Hydrochloride of hexaethylpararosanabne +| ^^.^ ^
.ou. parts of ethyl violet a. ^^^^r ^^SlSL^I^^^^
KcSic :ssi:?st3^ a SAioiet - -i^y^
Iround. Thenuclejstamfair^^^^ Both types appear to take the
Ls in th\i«l^S^^^l^,,^,tnh?on^ghout is not sharp enough: stain but the color contiast tnioag
, Azo fuchsme , , , „
Ethyl violet + . ^ ^i^^ salt of p-sulphoben .,. 4. zene-azo-dioxynaphthyalene
Hydrochloride of hexaethylpararosanilme + , ^^^^^^.^ ,,id
f 9 • \ The Ethyl violet and -» /"^^s^r^^" ^th"^^^
?S"™Ms'l' aTer7?n\e„t B^a^ and n.ay be used m Ingh ddufon.
NEUTRAL STAINS 481
It is resistant to the action of acetone and alcohol thus making possible a more careful differentiation than with any of the other stains used. Sections stained in it have little tendency to fade and preparations made nearly a j^ear ago are as brilliant as when first made.
From a consideration of the group of dyes just described it is e\ddent that two of them stand out as distinctly superior to any of the others as a stain for pancreatic tissue. These are the compounds formed by the union of ethyl \'iolet with orange G and with azo fuchsine. These are both powerful dyes, staining cell granules very intensely. The color contrast between the different types of granules and between the granules and the cell protoplasm is very sharp. The 3' are efficient in high dilutions and gross precipitation of stain on the tissue is avoided. Finally it may be said the granules and the nuclear chromatin retain these stains well in the presence of acetone and absolute alcohol, thus rendering differentiation easy.
On referring to the table given above it is seen that the neutral stains formed by combining orange G with different basic dyes vary in staining strength and that as the basic substance used becomes more complexed the staining power of the neutral dye increases. Thus, the hexamethyl \dolet gives a more efficient stain than the penta methyl violet, the hexa ethyl compound surpasses the hexa methyl combinations while victoria blue joined to orange G gives a neutral dye that ^dll stain efficiently in higher dilutions t an an}' of the others
The dyes tested in the above study were not secured through dealers but in each instance from the firm concerned in its manufacture. I am indebted to Dr. Evans, who placed his collection at my disposal, and we wish to thank the following houses for cooperation both in the supply of dj^e samples and in the confirmation of the precise chemical make-up of the dyes used: Farbenfabrike of Elberfeld Company; the Badische Company; the Berline Aniline Works; Kalle and Company; Leopold Cassella & Co.; Carl Jager; and Schoellkoff, Hartford & Hanna Co.
INCREASE IN PRICE OF JOURNALS
In order to extend and improve the journals published by The Wistar Institute, a Finance Committee, consisting of editors representing each journal, was appointed on December 30th, 1913, to consider the methods of accomplishing this object. The sudden outbreak of European misfortunes interfered seriously with the plans of this committee. It was finally decided, at a meeting held December 28th, 1914, in St. Louis, Mo., that for the present an increase in the price of these periodicals would not be unfavorably received, and that this increase would meet the needs of the journals until some more favorable provision could be made.
This increase brings the price of these journals up to an amount more nearly equal to the cost of sunilar European publications and is in no sense an excessive charge.
The journals affected are as follows:
THE AMERICAN JOURNAL OF ANATOMY, beginning with Vol. 18, price per volume, $7.50; foreign, $8.00.
THE ANATOMICAL RECORD, beginning with Vol. 9, price per volume, $5.00; foreign, $5.50.
THE JOURNAL OF COMPARATIVE NEUROLOGY, beginning with Vol. 25, price per volume, $7.50; foreign, $8.00.
THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY
36th Street and Woodland Avenue
Philadelphia, Pa.
SPOLIA ANATOMICA ADDENDA I
ARTHUR WILLIAAl MEYER
From the Division of Anatomy of the Stanford Medical School
TWENTY-SEVEN FIGURES
CONTENTS
Perforate sphenoidal sinuses with sub-dural diverticula 484
Eight cervical vertebrae associated with a cervical rib 490
Unilateral absence of a vertebrae associated with other anomalies 495
Fusion of three cervical vertebrae in a gorilla 500
A peculiar double thoracic duct 501
A large phrenico-hepatic artery 502
Corpora libera abdominalis vera et potentialis 502
Bizarre lymph follicles 504
Intercalated (?) hemal nodes 506
The genesis of supernumerary spleens 507
Depleted hemal nodes 507
1. Parathymic hemal nodes from a lamb of seven months 507
2. Paraparotid hemal node from a newborn pig 508
3. Omental node from a newborn colt 508
Pancreatic spleens in rabbits 51 1
Pigmented 'hemolymph' nodes 512
Pigmented 'mixed' hemoljTnph nodes 514
Diaphragmatic lymph node in a rabbit 517
Intra-gluteal bursae 517
An anomalous vago-sympathetic plexus 518
Trajectorial structure in the sacrum 519
The molding effect of muscle pressure 521
Freely ending capillaries in a hemal node 524
Hearts with bifid apices 524
Literature cited 52G
483
THE AXATOSnCAL RECORD, VOL. 9, NO. 7 JtTLT, 1915
484
ARTHUR 'W. MEYER
i
PERFORATE SPHENOIDAL SINUSES WITH SUBDURAL DIVERTICULA
Specimen a is from the bod}^ of a white man, twenty-eight years of age, who died of tuberculosis.
There are two sphenoidal sinuses, as usual, in this specimen; the right sinus extends several millimeters across the median line. The septum is located in almost a sagittal plane lying 2 to 3 mm. to the left. The form of the right sinus is oval with its long axis — which is practically parallel to the upper surface of the basilar process of the occipital bone — ^making an angle of
Fig. 1 Developmental defect in the lateral wall of the sphenoid and sphenoidal diverticulum; side view.
approximately forty-five degrees with the vertical. This diameter measures 25.5 mm., the vertical one 15 mm. and the transverse (right to left) 18 mm. The lining membranes of the sinuses are thin and they are not abnormally adherent anywhere. The roof of the right sinus is very thin, measuring only 0.35 mm. (by micrometer caliper) directly beneath the sella and a little to the left of the median line ; its minimum thickness is only 0.5 mm.
Although the configuration of the left sinus is similar to that of the right it is much smaller, measuring only 11 mm. in a transverse Tright to left) diameter and 22 mm. in the longest
SPOLIA ANATOMICA ADDENDA I
485
oblique direction; it too is normal in appearance. The combined sinuses extended only about half way beneath the sella. The ostia are normal in size and position.
The nasal cavity is capacious but normal in appearance and the conchae, except the superior, are small. There are four conchae on the left side, the posterior ethmoidal cell opening into the supreme meatus.
At about the midpoint of the ventral (anterior) portion of the lateral wall of the right sinus immediately beneath its roof there is an oval defect (fig. Id) The longest diameter of this oval defect measures 7 mm. and extends antero-laterally, lying a' most in a horizontal plane and making forty-five degrees with
Fig. 2 The same as figure 1 ; cranial view.
a saggital plane. The short diam^eter measm-es only 4.5 mm. Through this opening a diverticulum of the sinal lining, which is 6 mm. long, protrudes into the subdural space. The wall of this diverticulum is verj^ thin, nowhere adherent to the exceedingly thin (0.5 mm.) but regular margin of the defect and can be inverted with entire ease. It extends slightly forward and upward into a triangular space bounded by the optic nerve anteromedially, the carotid artery postero-medially and the reflection of the dura laterally (fig. 2) The dura, which surrounds the base of the diverticulum on all sides, does not envelop or cover the defect but merely comes into contact with the encephahc surface of the bony margin bounding the defect. Hence it is evident that the mucous diverticulum extended directly into
4g6 ARTHUR W. MEYER
the subdural space and must have been in direct contact with the
arachnoid. . n r -u
There is no corresponding or other defect m the wall ot the left '^inus but the corresponding region is marked by a depression which lies in the base of the posterior root of the lesser wmg. This root is absent on the right side. The anterior clmoid process on the left side joins with the middle, formmg a complete foramen for the carotid artery. That on the right side unfortunately had been partly removed bUt the condition of the middle chnoid process, which is wholly intact, shows clearly that it was not joined to the anterior on this side, and the lateral wall with its dural reflection shows that the posterior root of the clinoid process was absent on this side.
The anterior and middle ethmoid cells and the frontal sinus were large but the posterior ethmoid cells were extremely small, being 3 to 4 mm. in size.
Specimen h is from the cadaver of a man thirty years old, who also died of tuberculosis.
The left sphenoidal sinus in this specimen is somewhat larger than the right and extends completely beneath the hypophyseal fossa, being separated from the pons by an exceedingly thm • bony wall only 0.37 mm. thick. The hypophyseal fossa is large and long, in a dorso-ventral direction. The dorsum sella is T-shaped' in section, low (5 mm. high) and composed of a thm plate of bone which bears the large (by comparison) posterior clinoid processes. The floor of the sella formed the dorsal (posterior) half of the roof of the sinuses.
The ventral (anterior) half of the lateral wall of the left sinus contains a defect similar in position and character to that m the preceding specimen. This defect is oval also but some^^^at larger than the preceding for it measures 6.5 X 4.5 mm. ihe bony margin bounding it protudes slightly intra-cramally, forming a small cufi" around the defect; this margin is only 0.2.)
mm. thick. i. j .
The diverticulum of mucous membrane \yhich protrudes
through this defect extends fully 3.5 mm. beyond the plane,
extending across the dural reflections over the optic and oculo
SPOLIA ANATOMICA ADDENDA I 487
motor nerves; it is slightlj^ enlarged distally. Although the lining membrane of the sinus is somewhat thicker than that of the preceding case, it is nowhere adherent and could be easily inverted. The relations of the diverticulum to the surrounding structures are exactly the same as in the previous specimen. It has a total length of 6 mm. and a width of 7 mm., in a line parallel to the optic nerve, and a width of 4.5 mm., in a craniocaudal direction. Since a thin bony rim extends intracranially around the defect, for several millimeters the optic nerve lies more above than medial to the defect. This bony rim also covers the ventral knee of the carotid artery.
The dorso-ventral diameter of the left sinus is 29 mm. and it extends several millimeters beyond the median line, which it intersects somewhat obliquely, the ventral half of the septum lying a trifle to the left of the median line. This sinus does not extend beneath the hypophyseal fossa on the right side except at its most caudal portion, although laterally to the right it extends beneath the middle cerebral fossa.
Just as on the left side of the previous case, there is a depression in the lateral wall of the right sinus in a position corresponding to the defect on the left side. As before, this depression was covered by the posterior root and in part also by a bony extension from the anterior to the middle and to the posterior clinoid process, representing the ossified ligamenta intercHnoidea not infrequently present. On the left side, . on the contrary, there is no such extension from the anterior to the middle, but only from the middle to the posterior, clinoid process.
The finding of these defects in two subjects, the skull of neither of which was damaged by disease or injury, among only eight cadavers simultaneously under dissection, remotely suggests John Burroughs' dictum that "the number of birds one sees depends on the number one looks for," for I cannot believe that these instances are isolated cases, especially not after I have examined a small series of skulls with especial reference to the shape, form and position of the posterior root of the lesser wing and the osteology of the surrounding region. It is, for example, comparatively common to find a depression — i.e., an evagination
488 ARTHUR W. MEYER
of the sphenoidal sinuses into the base of the posterior root — ■ and in one cleaned remnant of a skull found in the laboratory one of the sphenoidal sinuses communicated with the cranial cavity at exactly the same place as in the preceding specimens. Although the lesser wing of the sphenoid had been removed in this specimen it was evident from the character of the margin of the defect in the lateral wall, and from the character of the wall itself, that it is not improbable that the defect was present in this specimen also before it was cleaned for onl}' a very slender posterior root could have been present. However, the mere presence of a very slender posterior root, or perhaps even its absence, is not necessarily an indication that the sphenoidal sinus is not completely separated by bone from the cranial cavity. The absence of this root is probablj^ of significance only if the sphenoidal sinus is as large or larger than normal, or perhaps still more accurately, only when there is a tendency to extend the sinus laterally in the region of the base of the root. Just why absorption should be especiallj^ active here at the junction of the pre- and basi-spenoids, I do not know, and it is possible, of course, that tension exerted through the root in consequence of unequal growth after its fusion with the lateral wall may cause evagination of the sinus into the base of the root and its absorption.
The absence of the posterior root in both the above speci-mens must have left a weak point in the wall at this region. Since the reflections of the dura o\'er the carotid artery, the second nerve, and over the third, fourth and sixth nerves in the adult he at a higher intracranial level than the wall of the sinus, it is evident that the dura at one time must have been depressed over this region in order to clothe the portion of the lateral wall later occupied by the defect. In pre- and early post-natal life, to be sure, there could have been no such dural depression, but with the change in contour and in the relation of the different portions of the sphenoid bone and the extension of the air sinuses, with approaching maturity of the bone such a condition was bound to appear. Since in these cases the lateral wall of the sphenoid was not reinforced by the posterior root, as is normally
SPOLIA ANATOMICA ADDENDA I 489
the case, it might be assumed that this region formed a point of least resistance to the developing and encroaching air sinus, were it not for the fact that other portions of the sinal wall are not reinforced and yet no perforations or defects result. That the increased intra-sphenoidal air pressure associated with such occasional phenomena as sneezing, coughing or forcible and obstructed expirations of any character, could be responsible for local atrophy of the bony wall and the underlying dura seems decidedly unlikely, although the form of the diverticula and the bony margin of the defects seems to suggest this. Indeed, no other explanation seems to fit the anatomical findings. But until we know more about the cause of the development of the air sinuses and the factors which control their development in the different directions, it seems quite futile to speculate on the genesis of these pecuHar defects.
It seemed highly probable to me, at first sight, that the dura must have covered these diverticula, but that it did not do so except at the ver}- beginning of their extension through the bone is beyond question. Indeed, the perforation of the dura by the diverticula is the most interesting and unexpected thing. ]\Ioreover, it would have seemed Hkely that the cerebrospinal fluid, the other meninges and the brain substance might have forced these diverticula of mucous membrane back into the sinuses, or rather prevented their protrusion. It would also seem as if one might have expected a diverticulmn of the arachnoid, accompanied or unaccompanied by brain, to extend into the sinus as a result of the unopposed intracranial pressure. Moreover, since a defect was produced in the dura, it also seems as though the arachnoid should also have yielded to the same influences. Unfortunately, the brains had been removed from both these skulls, but the dui-ae were undisturbed in these regions. The presence of the protruding diverticula is conclusive evidence, however, of the fact that practically no intracranial pressure was exerted upon them, for their extremities were entirely unattached to the durae and their inversion into the sinuses was prevented only by atmospheric pressure. Besides, had they been at all firmly attached to the arachnoid it is more than
490 ARTHUR W. MEYER
likely that a tag of arachnoid would have remained attached to their distal extremities when the brains were removed.
The only investigators who mention any defects whatsoever in the lateral wall of the sphenoid are Zuckerkandl ('82), Spee f'96) and Onodi ('03). The first spoke of having noticed dehiscences and small defects in the lateral walls, and the latter of small defects in the region of the sulci carotici in a juvenile skull. Onodi, who examined 4000 entire and several hundred cut skulls, found vascular sulci and foramina in the lateral wall and also larger or elongated dehiscences in these Avascular sulci. But neither he nor Gibson, who recently examined 85 specimens, found any defects comparable to those here reported. Onodi also emphasized the fact that such dehiscences as he found may result from traumata, pathological conditions and senile atrophy, as well as be artifacts or developmental anomalies.
Since small defects in the lateral walls in the region of the carotid sulci — and in other places, for that matter — are seen not ver}' infrequently in cleaned and dried skulls, it is evident that such skulls do not furnish proper evidence regarding the actual frequency of these abnormalities. If the walls of the sinuses are exceedingly^ thin they are easily injured when the dura is stripped and still more easily eroded in the cleaning. Hence, skulls in which the durae have not been removed can alone be regarded as furnishing rehable evidence.
Since the clinical significance of such defects in the lateral osseous wall, accompanied or unaccompanied by protrusions of the lining mucosa into the subdural space, must be evident to everyone emphasis on this matter is unnecessary.
CERVICAL RIB ASSOCIATED WITH EIGHT CERVICAL VERTEBRAlC
The body from which this specimen was obtained was that of a female Swede 38 years old. There was nothing especially peculiar about the cervical rib which was present on the left side (fig. 3) except that it was .well formed. It reached to within 2 cm. of the sternum and measured 7 cm. in length, along its concave border from head to tip, and the same dis
SPOLIA ANATOMICA ADDENDA I
491
tance along its convex dorsal border from the prominent articular tuberosity of the eighth cervical vertebra. The distal extremity of the rib was thickened and articulated with the upper surface of the first thoracic rib, about 2 cm. from the costo-chondral junction of the latter. A true articular capsule and an articular surface were present. The one on the first thoracic rib was about 3.5 mm. across. Both articular surfaces were covered by cartilage, and from the lateral surface or the distal extremity of the cervical rib a narrow but strong ligament, almost tendin
Fig?.3-4 Cervical rib and brachial plexvis.
ous in character, extended to the sternum. It ran parallel and directly cranial to the border of the first thoracic rib. Only slight mobility was possible between these two ribs, which mobility w^as greatlj^ increased by section of the articular capsule at the distal extremity of the rib.
The head of the rib is small, the neck broad and flat, the tubercle comparatively large, and the shaft looks decidedly twisted because of the presence of a prominent groove formed
492 ARTHUR W. MEYER
by the medial fasciculus of the brachial plexus as it crosses the shaft obHquely. The distal extremity of this rib is somewhat thickened.
The lower surface of the broad thm neck and the upper surface of the distal third of the shaft are grooved by the ni7ith cervical nerve; and the medial fasciculus of the brachial plexus and the cranial surface of the neck and the medial half of the bod\' by the eighth nerve. The very thin and sharp medial margin of the broad neck lies between these two constituents of the brachial plexus, which join directly distal to this sharp border, 3 cm. distal from the head. This sharp medial margin also bears a slight impression from the subclavian artery. The communication from the first thoracic intercostal nerve did not cross this rib but ran along its lower border (the plexus was drawn cranially in figure 4) joining with the ninth nerve before the latter crossed the cervical rib to reach its upper surface, upon which it lay nearly to its extremity. If the subclavian vein also made a slight impression it was in common with the artery and not separate from it.
There is no separate extra rib on the right side, but a thin broad fibrous band extends from the prominent transverse process of the eighth \'ertebra to the mid-point of the first thoracic rib.
The cervical vertebrae are all normally formed; there is no sudden transition from the spinous process of the sixth to the seventh, or from the latter to the eighth, but merely a gradual increase in length of the spinous processes from the second to the eighth, and from the latter to the first thoracic. The extra vertebra had the character of a thoracic vertebra on both sides, although the two hah'es were not symmetrical. This also agrees with the statement of Bardeen ('00) who found in an examination of 59 spines that ' ' There was some variation in the form of the vertebrae on the two sides of the body but in none of the bodies which we examined for the purposes of the present study was a given vertebra of different type on the left side of the body from what it was on the right side." The asymmetry was due to the presence of the rib, and hence of an articular facet on the
SPOLIA ANATOMICA ADDENDA I
493
body and the transverse process on the left side, and the presence of a long transverse process with a foramen on the right. The dense band of fascia mentioned above arose from the extremity of this transverse process and extended to the upper border of the first right rib. The unusual length of the transverse process, as well as the contained foramen, confirm Todd's conclusion that all elongated transverse processes in this region are
Fig. 5 Cervical and thoracic curvatures in a spine with eight cervical vertebrae.
to be regarded as resulting from the attempted formation of rudimentary ribs.
The vertebral artery entered the transverse foramen of the fifth vertebrae on the right and that of the sixth on the left side.
As is usual, the cervical curvature as represented in figure 5 was markedly accentuated. As judged by the alignment of the
494 ARTHUR W. MEYER
spines from the dorsum, there is no more than the usual amount of scoUosis, but there is a rather decided curvature to the right in the regions of the bodies of the 4th to 8th dorsal vertebrae. Xhe maximum curvature is located opposite the body of the sixth cervical vertebra and the whole curvature is accentuated by the presence of flattening of the bodies of the 4th to 8th vertebrae on the left side. Since there is no such flattening on the right side, the convexity of the scoliotic curve is less evident than its concavity. The flattening of the bodies is so extensive that it can scarcely be attributed to the aorta alone. The bodies and the discs of the cervical vertebrae are only very slightly reduced in thickness and the cervical spine is hence longer than usual, especially if compared with individuals of corresponding stature. The increase in cervical curvature seemed to be compensated for, at least partly, by an accentuation of the dorsal curvature but no reduction in length compensatory to the cervical increase was evident; nor was there any reduction in the number of vertebrae in other regions. Twelve dorsal, five lumbar, five sacral and three coccygeal vertebrae were present.
The brachial plexus was normal in formation and distribution. It was not shifted cranially but caudally with the extra vertebra. This was the case of the cervical plexus also. Hence it is evident that in this case the presence of the plexus did not have any inliibitory effect upon the formation of an extra rib, as Patterson suggested. Nor was there any pre-fixing of the plexus, as emphasized especially by Jones ('13). The contribution from the first thoracic nerve was fully as large as normal, and this case then stands in striking contradiction of Jones' statement that
Just as varying grades of imperfection of development of the first thoracic rib are the outcomes of varjdng degrees of post fixation of the plexus, so the varying grades of perfection in the development of a cervical rib are the outcomes of varying degrees of prefixation of the plexus. And just as the post fixation may readjust itself with the rib elements at a lower level, so may the prefixation readjust itself at a higher level.
This generalization would also seem to be contradicted by the case reported by Frank ('14).
SPOLIA ANATOMICA ADDENDA I ♦ 495
The very intimate relation of the ninth and first thoracic nerves to the cervical rib would seem to imply that considerable nervous disturbances probably resulted from pressure upon them, as emphasized by Goodhart ('09), Todd (12), Thorburn ('05) and others.
Since such an excellent discussion and a summary of the literature on cervical ribs is given by Le Double ('12) and also by Streissler ('13) I shall only add that my own experience confirms Le Double's statement that variations in number of the cervical vertebrae are exceedingly rare. Xor does Streissler mention a case of supernumerary cervical vertebra accompanied by a cervical rib, although his summary covers 80 pages and includes 297 citations from the literature.
THE EFFECT OF UNILATERAL ABSENCE OF A \^ERTEBRA ON THE PRODUCTION OF SCOLIOSIS
• A priori, one would be likely to conclude that the failure of development of the right or left half of a vertebra would result in a very evident, even if not in a marked, deformity of the spine. That this is not an inevitable result, however, was shown in the body of an adult male from which the specimen represented in figure 6 was taken. In this case only a slight deviation in the alignment of the spinous processes was noticeable after skin, fascia and some of the .dorsal musculature had been removed, so that the true nature of the anomaly was not noticed until the thorax had been opened and the relations and the marked ventral scoliosis were revealed.
There were also a number of other abnormalities in this cadaver. The twelfth ribs were exceedingly^ small, as shown in figure 7, and the tenth and eleventh were long floating .ibs. The xiphoid was long, curved, narrow, and completely ossir^d. The twelfth ribs, which measured 2 mm. in length, are syn, metrical. Each is provided with a small articular facet 3 mm. in diameter on the body of the vertebra, and with a true articular capsule; they were slightly movable. Since the head and neck of the cadaver had been removed before the bod}- was received
496
ARTHUR W. MEYER
at the laboratory it is impossible to speak with certainty regarding the number of cervical vertebrae but even assuming that the normal number was present — according to Le Double six cervical vertebrae occurred in only 0.14 per cent of 1420 fetal, infantile
Fig. 6 Unilateral absence of half a vertebra and oblique fusion of the fourth to eighth thoracic vertebrae.
A
Fig. 7 Twelfth thoracic vertebra and rudimentary rib.
and adult vertebral columns reported in the literature — the total number of pre-sacral vertebrae is only twenty-two.
There are twelve vertebrae in the thoracic but only three in the lumbar region, the fourth lumbar being completely incor
SPOLIA ANATOMICA ADDENDA I 497
porated in the sacrum, which is composed of six vertebrae. The spinous process and the lamina of the first sacral are entirely distinct and the inferior intervertebral articular facets are preserved completely on the right, and almost completely on the left, side of the first sacral vertebra. But the sacrum differs in no essential respects from many similar sacra seen in every laboratory. As figure 7 shows, the thoracic vertebra — the twelfth — which bears the very small twelfth ribs has the characteristics of a lumbar rather than of a thoracic vertebra. This applies especially to the character of its spinous process and the direction of the surfaces of the superior articular facets. The transverse processes of the tenth and eleventh dorsal vertebrae bear no articular facets and have the characteristics of the normal ele\'enth and twelfth dorsal vertebrae. The corresponding ribs also have the characteristics of the normal eleventh and twelfth ribs. Hence if this, the twelfth dorsal numerically, be regarded as a lumbar vertebra in spite of the presence of rudimentary ribs, and the fifth lumbar be regarded as having fused with the sacrum, as is evidently the case, the normal number of five lumbar vertebrae is accounted for. This, however, leaves only eleven dorsal vertebrae represented by only eight distinct bony elements. The first three and the last four of these eleven dorsal vertebrae are separate bony elements quite normal in form and size. The eighth independent element is a bony complex represented in figure 6, and — as indicated by the number of ribs and transverse processes — should be composed of four vertebrae. There are five independent spinous processes, however, as shown in figure 6. On closer inspection, however, it will be seen that the second and third spinous processes are really only half processes of two different vertebrae each of which is represented by half a vertebra located on opposite sides of the body. The second spine belongs to a half vertebra on the left and the third spine of the complex belongs with a haK right vertebra which is fused with the body of the succeeding vertebra. This relationship becomes clearer upon examination of the bodies. As seen in figure 6, a and 6, showing right and left sideviews respectively, there are only three bodies in this complex. ^Moreover, the right portion
498 ARTHUR W. MEYER
of the body of the first vertebra in the complex is reduced decidedly in thickness and the left portion is decidedly enlarged. The former bears only a small articular demi-facet and the latter a large and small demi-facet, and one large entire articular facet on its body. Hence it would seem that this portion of the complex represents a Httle less than one vertebra on the right and two vertebrae on the left side. The same thing is indicated by the presence of two transverse processes on the left and one on the right side and by the presence of a dorsal ridge, which clearly marks the line of fusion of the two lamina on the left side, and by a sulcus between the large middle articular facet on the left side of the body, which divides this large facet into two parts one of which resembles a demi-facet. The other no doubt represents a demi-facet also, although it looks Uke a whole facet and is large enough for one. Since a single rib — the fifth articulated here — this composite articular surface evidently represents only two demi-facets of adjoining vertebrae.
As viewed from the front, the upper surface of the body of the first vertebra in the complex slopes decidedly from left to right and although this great inequaUty in the thickness of the body was compensated for largely by difference in the thickness of the intervertebral disc and very shghtly also by a slight inequality in the body of the third dorsal vertebra, it nevertheless caused a short but marked scolicosis, noticeable only ventrally.
A reference to figure 6, a, will also show that there was a shifting caudally of the lamina, transverse processes and pedicles on the right side, as a consequence of which the last lamina and transverse process, although normal in size and shape, are left without a pedicle. All the transverse processes, lamina and pedicles on the right have been shifted one vertebra caudally, and actually belong to the bodies one vertebra farther cranially. This asymmetrical bilateral error in segmentation also explams the presence of the two half spinous processes and is further confirmed by the presence of two half and one whole articular facets on the right side of the body of the last vertebra of the complex, as shown in figure 6. Hence it is evident that the right
SPOLIA ANATOMICA ADDENDA I 499
half of the last composite vertebra, in this complex of four, articulated with three ribs, as did the left side of the bod}^ of the first. Since the lamina and intervertebral articular facets of the first two vertebrae are fused on the left side of the complex, only two normal articulations remain on this side, although three are preserved on the right. One of the articular facets on the left side is also on a pedicle.
The particular interest attaching to this error in segmentation lies, I take it, in the fact that the absence of one half segment was accompanied by the oblique fusion of dissimilar halves of succeeding segments, with consequent production of a floating laminum and transverse process, both of which have a normal form and size.
Aside from a few minor abnormalities present in this cadaver, the most interesting thing was the position and blood supply of the kidneys. The lower pole of the left kidney reached only to the level of the tenth ribs. This kidney was small (9.3 X 4.7 cm.) and received its main blood supply from a renal artery which arose at the level of the origin of the superior mesenteric. This artery bifurcated se\'eral centimeters from the medial border of the kidney, which it penetrated directly cranial to the pelvis. An additional large artery, which entered the medial surface about midway between the hilum and the lower pole, arose in common with the inferior mesenteric and a right renal 5 cm. cranial to the bifurcation of the aorta. Directly from the apex of the lower pole an accessory vein emerged and emptied into the vena cava.
The right kidney, which was considerably larger, measured 11.7 + by 5 + cm. Its lower pole reached the level of the lower border of the lumbar \'ertebrae, almost half of the renal mass lying caudal to the l^ifurcation of the aorta. The superior pole reached the middle of the second lumbar vertebra. This kidney had three renal arteries. The main renal took its origin from a trunk common to the inferior mesenteric and accessory left renal, and bifurcated about 2 cm. from the renal mass. A second renal arterj^ arose from the lateral border of the right common iliac about 1 cm. below the bifurcation of the aorta. From here it took a shghtly upward course, passed dorsal to the
THE ANATOMICAL RECORD, VOL. 9, NO. 7
500 ARTHUR W. MEYER
kidney and bifurcated about 2 cm. before reaching the lateral renal border. One branch ended on the lateral border and the other on the ventral surface.
A third renal artery arose from a trunk common to the middle sacral and fifth lumbar arteries. This vessel emerged between the two common iliac arteries, then curved abruptly to the right, crossed the right common iliac ventrally and bifurcated at the border of the vena cava, the branches entering the ventral surface of the kidney 1 cm. lateral to the medial border.
Two large renal veins emptied directly into the vena cava. The largest, which was located farthest cranially, arose near the lateral border of the ventral surface betw^een the lower and middle thirds by two trunks, which soon joined and emptied into the vena cava about 2 cm. cranial to the superior pole. The second vein arose by two branches, coming from the dorsal and ventral surfaces respectively^, near the medial border of the superior pole.
The ureter arose mainly from a larger pelvis lying on the ventral surface between the upper and middle thirds and from three small accessory pelves located on the ventral surface of the middle and lower thirds. Each of these pelves had a ureter which joined the main ureter separately.
FUSION OF THREE C1:R\1CAL VERTJ-^BKAE
The specimen found in a skeleton of Troglodytes savagei is conii)osed of the 3rd to 6tli cervical vertebra, inclusive. As the illustration (fig. 8) shows, almost complete fusion has occurred. The three transverse processes on the left side, although very different in form, are quite separate, however. The articular processes are fused completely but the location of the individual processes can still be determined because of the press iCv of SI i in the inter-articulai- regions.
Tl. ^ extr ^ • of the right trans\erse processes of tlie third
and fou Ih \ are fused in such a way as to form an almost
comnlete boi th-ough which the fifth nerve passed ven +^a The - >n of the transverse in'ocess of the third
>
\
"■SPOLIA ANATOMICA ADDENDA I
501
is absent on this side but the base containing the transverse foramen is preserved. The bodies of these vertebra are fused so completely that the exact limits of the individual corpora can nowhere be detected.
The character of this splendidly preserved specimen shows clearly that the fusion did not result from injury or disease but is a simple case of embryonic fusion. The remaining cervical vertebra are normal in size and number, there being 13 dorsal, 3 lumbar, 2 sacral and 3 coccygeal vertebra. A marked dorsal inclination or retro-flexion of the dens epistropheus is present
Fig. 8 Fused third to sixth cervical vertebrae of a gorilla.
and ma}^ be due to the lessened mobilit}^ of the cervical spine resulting from the fusion of the 3rd to 6th vertebra.
AN UNUSUAL THORACIC DUCT
The careful statistical study made by Davis '15 has supplied us with a better basis for grouping anomalous ducts and for judging the frequency of the different types. Since f ase \± a aible duct, observed by me some years since, differs " se rec jrded
by Davis, Svitzer, von Patruban, Wutidr i ers, it seems
worth while to give a short description o"" it le sketc? and
notes at hand.
502 ARTHUR W. MEYER
A cysterna chyli formed by the confluence of three lymph vessels was present in the usual location. From it a large left thoracic duct extended cranially and emptied into the subclavian vein after being joined by the left cervical trunk. But in addition to the left thoracic duct a smaller duct ran to the right from the cysterna chyli and then extended cranially parallel to the left duct, continued directly with the left cervical and joined the left thoracic duct by a transverse branch in the bronchial region after receiving a large bronchial trunk. In addition to being double, this specimen of the thoracic duct was interesting in the fact that the right cervical trunk joined the right thoracic in the thorax in the retro-bronchial region and sent a transverse communicating branch, which joined the left thoracic duct after receiving a large bronchial trunk.
A LARGE PHRENICO-HEPATIC ARTERY
Although the occurrence of small hepatic branches from the phrenic arteries is mentioned in all anatomies as normal I have been unable to find a description of a large vessel such as noted here. The vessel in question, which was fully 2.5 mm. in caliber, arose from the right phrenic and entered the fossa ductus venosi near its cranial extremity. It then divided, sending one branch cranially into the left lobe the other branch running caudally along the fossa. A few centimeters from the point of bifurcation a second branch was given off to the left lobe at al)()ut the midpoint of the vessel but the main trunk extended onward to the porta hepatis, where it divided into three branches which entered the quadrate lobe. The phrenic itself arose directly from the aorta opposite the left renal artery and gave off a supra renal branch before reaching the diaphragm.
CORPORA LIBERA ABDOMINALIS VERA ET POTENTIALIA
In a former article I have reported the finding of true appendices epiploicae (not coli) and discussed their genesis and significance. Since then I have found two more cases of appendices and one case of corpus liberum.
SPOLIA ANATOMICA ADDENDA I
503
One of these appendices was found in the ventral surface of the great omentum- of a cat; it is shown in figure 9. Its actual length a-b was 8.5 mm. and the cylindrical vascular tip, which very closel}^ simulated a henial node, was 2 mm. long and 1.3 mm. in caliber. The vessels going to it were not visible in the gross.
In figure 10 a much larger fatty bovine omental appendage is represented. Its greatest width was 14.5 mm. and its length 22.2 mm. The capsule is thin and the whole appendage has the appearance of normal fat.
Fig. 9 Omental appendage from the cat; actual length, o~b, is S.5 mm.; length of hemal nodule at tip is 2 mm. and its diameter is 1.3 mm.
Fig. 10 Omental appendage from a steer; actual size 14.5 X 22.2 mm.
The third specimen was found in a small cavity on the internal surface of the ventral wall of the greater omentum of the body of an adult male. It is oval|^in form and measured 2x1 cm. It is smooth and soft and was contained within a small sac about twice its size, which was located on the internal surface of the \entral reflection of the great omentum. It could easily be rolled between the fingers and the character of its surface, as well as its form, suggest that it had often been rolled about by the action of the peristaltic movements on the great omentum.
On section this bodj^ is found to have a well-defined connective tissue capsule, 0.5 mm. thick, which contains a partially degenerated fatty material. Although no histological examination
504 ARTHUR W. MEYER
was made there is little likelihood that this body was other than a fatty omental appendage, such as that shown in figure 10. The fact that it was contained in a small omental bursa is likely accounted for b}^ a slight inflanunatory reaction which probably accompanied its detachment from the omentum. The mere process of torsion of the pedicle which precedes and accompanies separation would alone favor the exudation of a sufficient amount of serum to cause omental adhesions about the appendage. However, since these adhesions would be unlikely to form over the whole surface of the appendage, the continuous effects of peristalsis could later free it and convert it into a corpus liberum within its own small sac, instead of within the omental bursa.
Upon the genesis of a thickened connective tissue capsule from a very thin peritoneal layer I have no information to offer. That matter has been discussed often since Virchow called attention to the thick cartilaginous capsules which frequently surround loose bodies. However, since seeing the report of the extraordinary large free body found in the abdominal cavity by Campbell and Owen ('14) I am prompted to add that no evidence whatever for the idea that free bodies continue to grow in size after detachment was found in any of the cases which came under my observation. Nor can I saj^ I should have expected to find any, for the evidences of degeneration present even in very small loose bodies, as well as the evidences obtained from tissue culture and from corpora libera articulationes — corpora oryzoidea, joint mice — speak very strongl}^ against such a supposition. JVIoreover, it is well known that the presence of comparatively small loose bodies often necessitates operative interference and the larger the uncalcified body the greater the danger from degeneration if detachment has been sudden.
lUZAKRE LYMPH FOLLICLES
During my investigations on hemal nodes, especially of the sheep, I was frequently impressed with the singular form of the lymph follicles in some lymph nodes. We are so in the habit of regarding the follicles as s))hej'ic;il Ixxlies th.-it a form such as
SPOLIA ANATOMICA ADDENDA I
505
is represented in figure 11 seems odd indeed. Strangely enough, I never noticed anj' condensation of the reticulum surrounding these follicles, such as in commonly present in the spherical forms, nor could I relate their form with any peculiarity of the node or the surrounding parenchyma. They seem to result from a diffuse rather than a strictl} locahzed proliferation of lymphocytes. I am aware, to be sure, that the least-resistance theory can also be applied here and it might perhaps offer a satisfactory explanation were it not so difficult to see why the
Fig. 11 A strangely-shaped lymph follicle from a hinph node of a sheep.
obstruction to growth in the various directions around an ordinary germinal center should be so nearh' equal in an over-whelming number of cases of follicle formation.
The occurrence of a more or less laminated appearance due to an arrangement of the cells in more or less orderly concentric rows is also a fairly common occurrence in spherical follicles and not infrequently one neets with folhcles which simulate Malpighian corpuscles very closely because of the presence of a central artery.
506 ARTHUR W. MEYER
INTERCALATED (?) HEMAL NODES
The hemal node shown in outhne in figure 12 was foinul in the pre-vertebral fat of a sheep. It is a very small flattened node measuring only 2.5 X 3 X 1.5 mm., but is peculiar in being transfixed, as it were, by a vein. Although the specimen has not been examined microscopically^ from what I know of hemal nodes I feel certain that it also has arterial relations and is therefore not intercalated in a vein, as the figure might suggest and as has been assumed b}'- some. Indeed, I have never seen a hemal node solely intercalated in the course of veins, although I have seen several nodes which were completely crossed by both veins and arteries (figs. 10 and 11, Hemolymph nodes of the sheep," Standford University, 1913).
Fig. 12 Hemal uotle with two veins; actual t^ize aljuut 3 mm.
Since this node is so small it might seem that one of these vessels is a vein and the other a distended artery had not injections frequently shown that it is not extremely rare to find two veins leaving a single nod'e in opposite directions. I am cntain this is the case here.
THE GENESIS OF SFPEHXTMI-HAK V SPLEF:NS
One of the theories of the genesis of some accessory spleens has long held that they not infrequently result from the isolation of small processes or lobules from the main mass. The main and
SPOLIA ANATOMICA ADDENDA I
507
supernumerary spleens shown in figure 13 very likely had such an origin. The specimen taken from a cat is interesting in that it shows a small narrow splenic process extending out from the parent mass, with a vessel running from the latter to the small nodular supernumerary spleen lying several millimeters distant. The main spleen was entirely normal in size and appearance and one can scarcely doubt that in this case mechanical factors were responsible for the isolation of the small splenic nodule and that hence it represents the former distal extremity of the process and is consequently a true daughter spleen. Serial sections of the latter show a wholly normal and typical splenic structure.
Fig. 13 Mother and (laughter spleens from the cat; actual size of daughter spleen 2 X 1.5 mm.
DEPLETED HEMAL NODES
1. Parathymic node of Ovis aries
Some years since, while incidently interested in the occurrence of accessory parathyroid and parathymic glands in sheep, I observed that small hemal nodes were especially frequent near the parathymus glands (The parathymus glands of the sheep. Anatomical Record, volume 1, 1905). These nodes, which measured only a few millimeters in diameter, could not be distinguished from the isolated parathymus glands with the unaided eye, and were frequently found upon microscopic examination to contain very little lymphatic tissue. Some of them were indeed mere sacs of blood containing small islands or chains of
508 ARTHUR W. MEYER
islands of lynii)hatic tissue. Sometimes, as in the case of the node a cross-section of which is shown in figure 14, the capsule was extremely thin. This particular node was taken from a young sheep only seven months old, but similar specimens were ouiicl in much younger and older animals.
2. Para parotid node from a iieirhorn pig, Sus domestica
My attention was called to this specimen (fig. 15) some years since bj^ Professor Sabin, through whose courtesy I am enabled to refer to it here. The structure of this node, which measured only a fraction of a millimeter, is entirel}' comparable to that above, obtained from the sheep, except that it is still more depleted of lymphatic tissue. However a careful scrutiny of the wholly depleted areas of the node from the sheep, by means of high magnification, reveals only very slight difTerences. Indeed, except for the specific and age differences there rasiy be said to be distinctions between these nodes.
3. Node from the gastro-hepatic omentum of a colt, Equus caballus
This specimen was the only hemal node found during a careful inspection of the thoracic and abdominal cavities, of the cervical axillary and abdomino-inguinal regions of three approximately full-term colts, which material I owe to the courtesy of my colleague, Professor Jenkins. The node was 5x3x2 mm. in size, intensely dark red and lay among pink lymph nodes in the gastro-hepatic omentum. As figure 10 (giving the ai:)pearance of a portion under high magnification) shows, there is far greater depletion of the lym})hatic parenchyma in this node than in that from the sheep. The node is in fact merely a sac of blood-cells like that from the parotid region of the pig.
It seems decidedly interesting to me that nodes from such different species and regions, of such different sizes, and from animals of different ages, should have so similar a structure. A comparison of these with hemal and splenic nodes, represented and described elsewhere, will also show that small areas from
'/' T' *
a*!^.
Fig. 14 Depleted hemal node from a lamb. X 54. Fig. 15 Depleted hemal node from a newborn pig. X 42. Fig. 16 Depleted hemal node from a colt. X 275.
509
510 ARTHUR W. MEYER
such haemal and splenic nodules from cats, dogs, sheep, bovines, goats, horses and pigs are found to have a very similar and wholly comparable .structure. Indeed, were it not for specific cellular differences, we could with considerable justice speak of an identical structure.
The node from the sheep has a very thin capsule, while those from the pig and horse have relatively thick capsules, in some places at least. But since the variations in thickness of the capsules of hemal nodes and sujDernumerary spleens from the same animal are subject to wider fluctuations, this matter is evidently of no great significance in the differentiation of nodes. The node from the sheep also contains lacunae and vessels within the lymphatic tissue which contain disintegration products of erythrocytes, but these may be present merely because this node represents an earlier stage in a process of disappearance or appearance. Personally, I feel quite convinced that whether found in fetal or adult material, these are decadent nodes, although I am wholly open to conviction and do not urge my own conclusion.
The mention of splenic nodules in this connection is heterodox, no doubt, at least if the term splenic is to imply the origin rather than the characteristics of nodes. I would not expect a spleen to develop anywhere, but neither would I conclude that all spleen-hke nodules had their origin in the splenic anlage or the main spleen itself and are hence daughter spleens. Moreover, until we are better informed upon the origin of hemal nodes and supernumerary spleens a definite conclusion would seem to be quite im justifiable.
PAXCREATIC SPLKF.N8 I\ RABBITS
The occurrence of spk^en-like nodules in the jiancreas of cats and rabbits was recorded elsewhere (Meyer '14). Some of these nodules are c()mi)letely and others only partly imbedded in the pancreatic tissue and vary in size from those barely visible to the naked eye to others three or more millimeters in diameter. Some of these intra-pancreatic spleens have no capsule what
SPOLIA ANATOMICA ADDENDA I 511
soe\'er and are surrounded by pancreatic tissue, except where in contact with the capsule of the spleen itself. Others have a distinct capsule of their own. Some of these nodules contain \'ery small amounts of erythrocytes, while others at first sight look like mere hemorrhages. They may contain lymph follicles or typical Malpighian corpuscles or ma}' be composed mainly of a diffuse mass of lymphocytes. There may be no trabeculae, or large ones may be present. Phagocytosis was never noticed and so far I have found these pancreatic splenic nodes in cats and rabbits only.
Figure 17 gives the appearance of one of the smaller splenic islands found in the pancreas of a rabbit. A capsule is absent in this portion; only one lymph follicle is present and islands of pancreatic alveoli surround and extend into the splenic island. The latter contains little supporting tissue and is composed almost exclusively of lymphocytes and erythrocytes, including capillaries and vessels.
A small section of another nodule is shown inider higher magnification in figure 18. The definite capsule, with its serous envelope and a capillary near the periphery of the node, are very evident.
The fate of the imcapsulated splenic islands is a matter upon which I can offer no evidence, although it hardly seems that they can become and remain permanent constituents of the mature pancreas. Their presence to be sure, is easily accounted for from a de\'elopmental standpoint and it is possible that their apparent greater frequency in some animals is due to specific differences in the times, or rates even, of development of the spleen and pancreas. These cases of supernumerary spleens recall the case of Rokitansky of a spleen in the head and of Kolb in the tail of the pancreas. The extremely interesting cases of Sneath f '13) of a scrotal spleen and of de Tyssieu ('14) of an intrahepatic spleen also suggest that supernumerary spleens maj^ apparently occur anywhere in the peritoneal cavity, although I presume one would be justified in a certain amount of scepticism regarding the identit}- of apparently splenic structures.
512 ARTHUR W. MEYER
PIGMENTED 'HEAIOLYMPH' NODES
As stated previously, the occurrence of pigment in henial nodes is a comparatively rare thing. It is common, however, in hemorrhagic lymph nodes. Some of the latter are literally crammed full of light golden and brassy pigment and some darker pigment may also be present. Most of the pigment in hemal nodes is usually extra-cellular, although phagocytosis is very evident.
Figure 19 shows a highl}^ magnified portion of a node, the identity of which may be open to question. This node, which was 4 to 5 mm. in size, was removed from the abdominal cavity of a sheep. It is characterized especially by the presence of sinuous masses of intensely black pigment which suggests India ink. In fact, upon cursor}- examination one might suppose this section to come from a hemal node which had been injected with India ink (cf. fig. 5, The hemolymph nodes of the sheep," Stanford University, 1914). But this is not an injected specimen at all, and the great mass of the black pigment granules in certain areas lie in capillaries in the lymphatic parenchyma. It is this gorging of the capillaries with black pigment granules which makes it simulate an injection. Nevertheless, large quantities of pigment granules are scattered about among the erythrocytes and accumulations of bright yellow pigment are also seen outside of the capillaries. \"ery little phagocytosis is visible in the sections of this node and the pigment gives one the impression that it is adherent to or even imbedded in the walls of the capillaries.
'MIXED' HEMOLYMPH NODES
The portion of a section of the node in figure 20 shows an intra-capillary arrangement of black pigment entirely comparable to that in the previous s{)ecimen. In this specimen, however, no yellowish pigment is present and the capillaries are much more perfectly outlined by the contained pigment. Since this node, which was taken from the axillary region of a lamb, was also injected, it is evident that very large quantities
-,.^#IW-
■M^
>%
a.^
i?*"
(t«*?"C
^ --vi;*';
><*
5fi
i^
«^<-a^
a .
n "=31
.•sri
/<;;-\
'«f'.
r
|/
i "0'
Fig. 17 Pancreatic accessory spleen from a rabbit. Camera liicida
Fig. 18 Section of a ijancreatic spleen from a rabbit. X 410.
Fig. 19 Pigmented hemal (?) node. X 515.
Fig. 20 Pigmented mixed node. X 515.
X 79.
513
514 ARTHUR W. MEYER
of pigment must either have been formed within the node or been transported there. The capillaries are outlined in this manner by the black pigment throughout practically this whole large node. There are also loose pigment granules in abundance in the parenchyma but there is little evidence of phagocytosis. This node was obtained from an animal five and a half weeks old, which was killed by bleeding. Forty to fifty small hemal nodes were found in the abdominal cavity. None of these nodes were in connection with the mesenteric lymphatics although many of the lacteals which were engorged with chyle — the lamb had nursed before killing — ran about and over hemal nodes both within and at the root of the mesentery.
Man J' pigmented and pink nodes that were injected were found to be lymph nodes. But in the left subscapular region a large dark red node, 1 cm. in size and flattened latero-medially, was found. Except for the disposition of the vessels this node looked like a hemal node and seemed to be a mere sac of blood. It was located at the bifurcation of the subscapular vein lying between the dorsal and ventral branches. Since it was so large it was injected with India ink and as I had never found a true hemal node in this region the result of the injection did not wholly surprise me. The India ink quickly entered typical afferent lymphatic vessels, which ran to reddish nodes which lay along the course of the axillary vessels. The nearest of these axillary nodes was 4 to o cm. distant and the injection could be followed from node to node to the jugulo-subclavian junction. The result of the injection then would seem to indicate that this dark red node was, after all, a typical lymph node.
An anomalous result of the injection was the fact that some of the India ink easily and quickly found its waj^ into the dorsal branch of the subscapular vein. Indeed, so much of the ink entered the axillary vein from here that some of it was later recovered from this vein by means of filter paper and identi^ed as such. This ink had entered the axillary vein from the branch of the subscapular, which was connected with the node by a small vessel which clasped the node after the manner of the afferent lymphatics but which was filled with blood before
SPOLIA ANATOMICA ADDENDA I 515
injection. The efferent vessel, which was also filled with blood, Hkewise arose by numerous branches which clasped the opposite sides of the node as they emerged from it so that the branches from these two vessels practically clasped the whole node from opposite sides.
Upon examining the right axillary and subscapular regions, a node apparently identical in all respects with that found on the left side was found in exactly the same location. Careful inspection showed that the large vesselleaving the node, which was also engorged with blood, had the typically beaded appearance of lymphatic vessels. Had it contained lymph instead of blood no one would have questioned its lymphatic character.
The short trunk (0.5 cm. long), which was formed by a number of clasping branches which emerged from the node and entered the dorsal branch of the subscapular vein, was filled with blood and also had the appearance of the short vessel in the previous specimen. This node was excised and imbedded in celloidin; figure 20 was made from a portion of a section.
The gross appearance of both nodes and the results of the injection of that on the left side seem to indicate that both these nodes were undoubtedly in communication with the subscapular vein by means of a short vessel, the branches — or radicles — of which embraced the node, and that they were also in communication with axillary and cervical Ij^mph nodes by means of a vessel which had the form and characteristic of a lymph vessel.
That the India ink entered the subscapular vein without the least obstruction and without the least distension of the node is good evidence that the internal architecture of the node was not unduly disturbed. Those familiar with the history of the injection of the lymphatic system will no doubt be reminded immediately of the similar results obtained by the early experimenters — especially by those using mercury. Nevertheless, even the authority of such great names as those of Haller, Mascagin and Sommering, not to mention Cruikshank, regarding lymphatico-venous communications, has been compelled in the course of time to yield to the higher authority of facts.
THE ANATOMICAL RECORD, VOL. 9, NO. 7
516 ARTHUR W. MEYER
Even if the observations of Fohmann, Lauth and Panizza, on the occasional termination of the lymphatics in the femoral and iliac veins of birds, and the similar observations of J. jNIiiller on amphibia and of Fohmann on the latter and on swine, remained unconfirmed, no one could disregard the recent observations of Baum, Huntington, jMcClure and Sylvester, made in connection with modern methods. Besides, there are the early anomalous cases in human cadavers of Conring, Duvernoy, Kaaw, Kulmus, Hebenstreit, Mertrud, and also the interesting observation of Wutzer, which was confirmed at the time by J. Miiller.
Hence, it seems to me that one can hardly doubt that these two nodes really were connected with the subscapular veins in such a manner that the venous current was shunted through them and then passed through the efferent lymphatics, thus coloring the intermediate lymph vessels and nodes pink. Had there not been two nodes on opposite sides of the body, with wholly similar characteristics, one might have assumed that the point of the needle had damaged the internal architecture in such a way as to let the blood enter both a lymphatic space and a venous radicle. But even such an assumption does not explain the characteristics of the nodes before injection.
The specimen which was sectioned is completely filled with blood and lymphatic tissue. Lymph sinuses are nowhere visible, and erythrocytes and lymphocytes are intermingled except in those portions of the lymphatic parenchyma which are extended throughout the node like large trabecula, or in areas in which there are many follicles and very little blood.
Designating these nodes as 'lusus naturae' does not, to be sure, explain their presence, nor does the rare occurrence of such specimens justify one in regarding them as being representatives of a separate type of node. Nor are they the exceptions which prove the rule, although they help very materially in establishing the occurrence of atypical lymph and hemal nodes and in explaining occasional anomalous results obtained by injections of Ij^mph nodes, as already emphasized.
SPOLIA AXATOMICA ADDENDA I 517
DIAPHRAGMATIC LYMPH NODE IX A RABBIT
This node, which measured 4 X 3 X 2.5 cm., was located on the central tendon of the disphragm, ventral to the vena cava. It la}^ partly on the tendon and partly on the diaphragmatic musculature and projected into the abdominal cavity. It had the appearance of a lymph node but no lymph vessels were seen in its neighborhood. Upon microscopical examination it was found to be provided with a very evident but ill-defined capsule.
It is a very vascular node, contains no e\ddent lymph sinuses but some coagulum which looks Uke lymph and also several degenerated areas of considerable size. Some of the parench^Tua is quite definitely arranged in cords. Degenerated erythrocytes and a few pseudo-polykaryocytes are also present. Since the rabbit had been treated "«dth goat serum it is of course possible that the areas of degeneration were partly or wholly, due to these injections. There was also an excess of serous fluid in the peritoneal cavity.
IXTRA-GLUTEAL BURSA
Two large subcutaneous bursae were found symmetrically placed in the gluteal region of the female cadaver. They were empty and measured 5 cm. in depth and 2 cm. in wddth. Both were located in the medial margins of the glutei maximi muscles, directly beyond the lateral margins of the ischio-rectal fossae. The skin over them was wholly normal in appearance and they were partly contained in the subcutaneous fat of these regions. The far larger portion of each, however, was contained in the glutei. These bursae were in no sense comparable in location to the ordinary subcutaneous bursae and could with entire justice be designated as intra-muscular. Their walls were very thin and smooth and contained no evidences whatever of having had an inflammatory origin. Since their genesis in this location would be a matter of pure surmise it seems futile to speculate.
518 ARTHUR W. MEYER
AX axo:malous vago-sy^ipathetic plexus
Since comparatively few students do a sufficiently careful dissection adequately to reveal the sympathetic system, our knowledge of variations and developmental defects of the autonomic system lags far behind that of other systems. Furthermore, the descriptions in our textbooks are rather inadequate. In addition to greater skill and time on part of the student, a searching supervision on part of the instructor is also necessary to reveal the many variations. Besides, not everyone can be specially interested in the sympathetic system; I, for one, confess to having given it scant attention.
Both the right vagus and sympathetic are normal in the cervical region of this subject except that the vagus passes dorsal to the subclavian artery. The superior cervical ganglion is large and the middle and inferior ganglia are 1 cm. apart and are joined by two trunks. The lateral trunk of this joining loop gives off a branch 1 cm. long, which forms an annulus vertebralis instead of an annulus subclavii, as is not infrequently the case.^ A fairly well-defined ganglion marks the point where the annulus vertebralis begins to form. From the latter several branches are given off and from the dorsal trunk of the annulus numerous branches extend to the common carotid, subclavian and innominate arteries. The two subdivisions of the sympathetic join directly after forming the annulus, and join the first thoracic ganglion. The latter is well-defined, although somewhat elongated, and the rest of the right sympathetic chain is normal.
About 1.5 cm. caudal to the inferior cervical ganglion of the sjanpathetic another somewhat stellate ganglion, about 4 mm. in size, is located. This ganglion was joined by the recurrent laryngeal branch of the vagus, which was rather small. The recurrent branch is not merely enclosed in the same sheath but definitely forms a ganglion with the sympathetic at the point of union. A somewhat larger branch than the pre-ganglionic branch of the vagus leaves this ganglion to form the inferior
^ I have seen an annulus innominatus but once.
SPOLIA ANATOMICA ADDENDA I
519
laryngeal nerve, which is accompanied by a tracheal and several esophageal branches.
Three parallel branches about 1 mm. in caliber leave the caudal portion of the ganglion and join with several similar though somewhat smaller branches, which arise from the caudal end of the dorsal arm of the annulus vertebralis to form the right pulmonary plexus; these branches are about 3 cm. long. The accompanying diagram, figure 21, illustrates these anatomical relations as found in the cadaver of a white female.
TRA.JECTORIAL STRUCTURE IX THE SACRUM
Since it will always remain an impossibility mathematically to prove the existence of trajectorial structures in the spongiosa of
Fig. 21 An unusual vago-sjanpathetic plexus.
bones, their presence here and there must always remain a matter of opinion, even if not of conjecture. Although mathematical proof of the existence of such an architecture must remain impossible because of the indeterminable character and the complexity of the acting forces, a more careful examination will probably reveal the presence of such an architecture in some locations heretofore undescribed. Indeed, until the whole human skeleton has received as careful an examination as some of the bones, it is highly probable that we shall also remain ignorant of manj^ decidedly interesting variations in the spongiosa architecture, even if not of the prevailing fundamental structural types.
520 ARTHUR W. MEYER
During an inspection of a series of sections of laboratory remnants, my attention was attracted to some sagittal sections of sacra made in the ordinary dissecting-room routine. One of the things noticed was a prominent spur of compacta, which extended dorsally from a mid-point on the ventral surface of some of the bodies of the sacral vertebra (fig. 22). These spurs or bars, which reinforced the ventral portions of the individual sacral vertebra, were not infrequently porous and looked as though they were in a process of dorsal extension. The portions of the spurs which were not porous were extremely firm and hard, and in fact looked eburnated. None was ever found in the first sacral vertebra. In short sacra they seemed to be commonest in the bodies of the second and third, and in long sacra in the third and fourth, sacral vertebrae; they were never observed in straight sacra. Not infrequently the compacta was slightly thickened opposite the mid-points of these vertebrae and when the spurs were absent the thickening often was located at the bottom of a depression in the middle of the ventral surface.
But a more striking thing still was the existence of a ^'ery evident trabeculated disposition of the spongiosa in the third and fourth and less frequently also in the more distal sacral vertebrae. This architecture was very definite and very similar in all cases where it was evident. As indicated in figures 22 and 23, the main trabeculae were perpendicular to the ventral surface at its mid-point but there were also oblique trabeculae on either side. Other smaller and less evident trabeculae extended approximately at right angles to these.
Since such a disposition of the spongiosa was not present in all sacra examined, one is naturally tempted to speculate. But it is likely wise to defer any possible explanation until a more extensive survey can be made. It is clear, however, that the thickening of the ventral compacta, the occurrence of spurs and the peculiar disposition of the spongiosa here described, are all admirably adapted to resist the tendency of body weight in sitting and of muscle pull from breaking particularly the middle and also the lower sacral vertebrae by ventral flexion.
SPOLIA ANATOMICA ADDENDA I
521
The illustrations are a faithful representation of the originals, though they show the characteristic architecture somewhat inadequately. From an examination of these drawings it will be evident how well-adapted such an architecture is to withstand the effect of the forces that must be active in every sacrum, even if they are not equal in magnitude or in direction.
Figs. 22-23 Sagittal sections of portion and entire human sacra, sliowing spurs and trajectories.
THE MOLDING EFFECT OF MUSCLE PRESSURE
Attention was called in another connection, to the effect of tendon pressures in affecting, even if not in determining, the rehef of bones. The following three specimens of humeri — two of which I owe to the courtesy of my colleague. Professor Ophiils — seem to show the effects of muscle pressure and tension
522
ARTHUR W. MEYER
in a striking way. In the left humerus, shown in figure 24 a, the whole greater tuberosity, which is decidedly enlarged by the deposit of much additional bone as a result of periostitis, is literally drawn out in the direction of pull of the spinatus and teres minor muscles. In fact, the whole bony mass looks as though it had been pulled in this direction while in a viscous condition.
c'
Fig. 24 Humeri sliowing the molding effect of muscle pull and pressure.
The lesser tuberosity and the rest of the bone are practically normal save for very slight evidences of arthritis. Aside from the slight arthritis there was nothing about the rest of the skeleton or the cadaver which suggested an explanation for this peculiarity of the greater tuberosity.
The other two humeri, shown in figure 24 h and c, were isolated specimens. The greater tuberosity of the right humerus, 6, shows a peculiarity similar to a, but the deposit of new bone is
SPOLIA ANATOMICA ADDENDA I 523
far greater in extent, and the proximal portion of the shaft of the humerus immediately distal to the head is bent medially. The humeral head and the whole lateral surface of the proximal part of the shaft are also flattened and new bone has been deposited along both tubercular ridges and on the medial side of the surgical neck; the rest of the bone is normal.
In addition to the dragging dorsally of the greater tuberosity, distinct flattening and molding of the lateral surface of the proximal portion of the shaft and of the new bone are plainly evident. This molded surface is also marked by very fine sulci, which suggest arterial impressions. Even delicate anastomoses of fine sulci are present and quite a complete network is evident under slight magnification with a reading-glass.
The tuberosities of the third specimen — the right humerus, c and c' — are in the normal location and the greater tuberosity was but slightly affected by the divsease. The intertubercular sulcus is quite normal but is roofed over \evy largely by a thin layer of bone, which is part of the large thin mass of new bone deposited on the proximal portion of the humeral shaft. This thin mantle of new bone rises above the lesser tuberosity and hoods it and the humeral head. The outer surface of the new bone shows a canalization similar to that in the previous specimen, but it is more complete and shows sulci made by larger vessels. The molding effect of the deltoid on this new bone, is so unmistakable that a mere glance suffices to reveal it.
Since the lateral portion of the humeral head is capped by the new bone it could have articulated with a portion of the glenoid fossa onl}^ The rest of the humeral shaft is practically normal.
An explanation of the observations recorded here seems quite impossible in the entire absence of a clinical history. Arthritic deposits rarely show such very evident molding and since the dragging dorsally of the greater tuberosity, especially in the first specimen, is not limited to the new bone, one is tempted to assume that these humeri must have been fairly plastic at some time. But the condition of the rest of the bones does not confirm such a supposition and I do not see how a greater plasticity could be confined to such a small area.
524 ARTHUR W. MEYER
FREELY ENDING CAPILLARIES IN HEMAL NODES
I have elsewhere expressed the opinion that capillaries in nondepleted hemal nodes communicate directly with the venous lacunae. The main reasons for coming to this conclusion were that the results of injections into the vena cava and aorta in lambs gave entirely comparable results and that capillaries could not be seen ending freeh^ within the parenchyma of the nodes. In case of the injections the injected mass of ink could never be seen in a freely ending capillary and was always found in the venous lacunae.
In figure 25 a section of a capillary is shown, containing a number of leucocytes arranged in a row within the capillary and also several beyond the mouth of the capillar}^, similarly arranged. This drawing was made with, an oil immersion lens and the specimen would seem to represent leucocytes either entering the free end of the capillary or passing through a gap in its wall.
It is the only specimen I have found in a careful study of many many sections indeed and has been in my possession for some years. I report it here with some indifference since it is of questionable value. A somewhat similar arrangement of erythrocytes near a real gap in the wall of a venous lacuna, is a much more frequent thing, however, and this and other observations seem to suggest that such gaps in the walls of the lacunae occur normally in hemal nodes.
HEARTS \MTH BIFID APICES
As reported and fully set forth by Mall ('12) and the literature cited by him, hearts showing more or less of an indentation at the apex are not very rare. The two hearts shown in outline in figures 26 and 27 possess this characteristic to a varying extent, the bifurcation in that shown in figure 27, being very marked. This notch between the apices of the right and left ventricles was also much more evident at autopsy and is deeper than an outline of the exterior indicates. This heart, which was obtained from the body of a white man who had fasted absolutely for sixty days, was quite flabby and probably for
SPOLIA ANATOMICA ADDENDA I
525
if
Fig. 25 Freely ending capillary in a hemal node of the sheep. X 1050. Figs. 26-27 Human hearts with bifid apices; half natural size.
526 ARTHrR W. MEYER
this reason the tip of the right ventricle was turned laterally so that the gap between the two apices was wide. The specimen is otherwise normal. Since Mall ('12) explained this anomaly fully, further comment is unnecessary here.
LITERATURE CITED
Bardeex. C. R. 1900 Costo-vertebral variation in man. Anat. Anz., Bd. 18.
Baum. 1911 Konnen Lymphgefasse direkt in Venen einmlinden? Anat. Anz., Bd. .39.
Campbell, R. P., a.vd Owen, J. J. 1914 An unattached mass found in the abdominal cavity of a male. Amer. Jour. Med. Sci, vol. 148.
Cou.six", GusTAVE 1898 Anomalies du canal thoracique. Bull, de la Soc. Anat., tome 7-3.
Davis, Hexry K. 1915 A statistical study of the thoracic duct in man. Am. .Jour. Anat., vol. 17.
DuPRE, B. G., axd Todd, T. \\'. 1914 A transitional t\'pe of cervical rib, with a commentar}'. Anat. Rec. vol. 8.
Fraxk, J. 1914 Ein Fall von Halsrippe mit abnormen Xervenverlauf. Anat. Anz., Bd. 47.
GooDHART, S. P. 1909 Cervical rib and its relation to the neuropathies. Amer. Jour. Med. Sci., vol. 138.
HuxTiXGTOx, Geo. S. 1910 Ueber die Histogenese des Ij-mphatischen Systems beim Saugerembryo. Anat. Anz., Erganzungsheft, Bd. 37.
.JoxES, F. \^'ooD 1913 The anatomy of cervical ribs. Proceed. Royal Soc. Med., vol. G.
LR Double, A. F. 1912 Traite des variations de la colonne vertebrale de I'homme. Paris.
Mall, F. P. 1912 Bifid apex of the human heart. Anat. Rec, vol. 6.
-McClure, C. F. W., AND Silvester, C. F. 1909 A comparative study of the lymphatico venous communications in adult nuunmals. Anat. Rec, vol. 3.
Meyer. A. W. 1907 The parathjTnus glands in the sheep. Anat. Rec, vol. 1.
1914 .Spolia anatomica. .lour. Anat. and Physiol., vol. 48.
1914 The hemolymph nodes of the sheep. Stanford Univer.-ity, California.
1914 The occurrence of supernumerary spleens in dogs and cats with observations on tlie corpora libera abdominalis. Anat. Rec, vol. S.
SPOLIA ANATOMICA ADDENDA I 527
Meyer, A. W. 1914 The supposed experimental production of hemolympli nodes and accessory spleens. Jour. Exp. ZooL, vol. 16.
OxoDi, 1904 Die Dehiszenzen derNebenhohlenderNase. Archivfl'irLaryngol. und Rhinol., Bd. 15.
vox Patrubax, Carl Edlex 1845 Ueber die Einmiindung eines Lj^inphaderstammes in die linke Vena anonyma. Archiv. fiir Anat. und Physiol.
SxEATH, \V. A. 1912 An apparent third testicle consisting of a scrotal spleen. Jour. Anat. and Physiol., London, vol. 47.
Silvester, C. F. 1911 On the presence of permanent communications between the hanphatic and venous systems at the level of the renal veins in South American monkeys. Am. Jour. Anat., vol. 12.
vox Spee 1896 Skeletlehre Abteilung II Koff. Handbuch der Anatomie des Menschen. Bardeleben, Jena.
Streissler, E. 1913 Die Halsrippen. Ergebnisseder Chir. und Orthop., Bd. 5.
SviTZER 1845 Beobachtung einer Theilung des Ductus thoracicus. Archiv fiir Anat. und Physiol.
Thorburx, William 1905 The seventh cervical rib and its effects upon brachial plexus. Medico-Chirurg. Trans., vol. 88.
1908 The symptoms of cervical ribs. Dreschfeld Alemorial Volume, Manchester.
Todd, T. W. 1912 The vascular symptoms in the cervical rib. Lancet, vol. 2.
DE Tysieu, 1914 Rate surnumeraire incluse dans le foie. Jour, de Med. de Bordeaux, tome 44.
^YuTZER, C. W. 1834 Einmiindung des Ductus thoracicus in die Vena azj'gos. Archiv. fiir Anat. und Phj^siol.
ZucKERKANDL, Emil 1882 Anatomie der Xasenholile, Wien.
EXPERIMENTAL STUDIES AIMING AT THE CONTROL
OF DEFECTIVE AND .MONSTROUS
DEVELOPMENT^
A SURVEY OF RECORDED MONSTROSITIES "V\T:TH SPECIAL ATTENTION TO THE OPHTHALMIC DEFECTS
E. I. WERBER
From the Zoological Laboratory of Princeton University
TWENTY-NINE FIGURES
INTRODUCTION
The problem of the causes underlying defective development has recently received very exhaustive and illuminating treatment by Mall ('08). Basing his conclusions on his own study of 163 pathological human ova and on recent results of work in experimental embryology, he suggests that the human monsters are — with the exception of the hereditary 'merosomatous' terata — due to injurious influences of atypical environmental factors. He makes the specific suggestion — which seems justified in the Ught of evidence brought forth by him as well as by cUnical data — that the monstrous development of some ova may be due to their inadequate nutrition o^Aing to the imperfect implantation in a diseased uterus. It would seem that this hypothesis may hold good at least for some pathological embryos aborted during the first two months of pregnancy. It is obvious, however, that Mall's interpretation could not be extended to monstrous fetuses of the later months of pregnancy or to monsters after full-term birth; for an ovum suffering from lack
1 This contribution is based on a paper read by title ("Is defective and monstrous development due to parental metabolic toxemia?") at the meeting of the American Association of Anatomists held at St. Louis, December 29, 30, 1914. Anat. Rec, vol. 9, no. 1, pp. 133-137.
529
530 E. I. WERBER
of nutrition could not well be imagined to live on to these later stages of development.
Some environmental factors must be looked for other than faulty implantation of the ovum, to account for the occurrence of such cases. The results of investigations in experimental embryology and teratology by Dareste ('91), Roux ('95), Hertwig ('96, '10), Fere ('94), Morgan ('02, '04) Stockard ('07, '09) and others, who obtained monstrous development of ova which had been subjected to the action of physical or chemical modifications of the environment, suggested to me that the human as well as the other mammalian monsters may be due to the physical or chemical action of some substances in the blood of one of the parents on either one of the sex cells or on the fertilized ovum, respectively.
Stockard ('09) has expressed the unsupported view that cyclopia in man may possibly be due to an excess of magnesium salts in the blood of the mother, and in later publications ('10 b), concluding from his experiments with alcohol solutions, he suggests that the Cyclopean defect might perhaps be attributed to alcoholism of either one of the parents. It is rather evident that this latter assumption may be regarded as justified only for a negligibly small percentage of monsters. For such defects are not infrequently found in mammals as well as in other vertebrates where the possibility of alcoholism is entirely eliminated. The solution of this problem appeared to me to be elsewhere. Since the metabolism is the main source to which chemical modification of the body might be traced, I concluded that the toxic substances found in the blood of individuals afflicted with some metabolic disturbances might be the ones which could be made responsible for the origin of monstrous development.
MATERIAL AND METHODS
To test the validity of this hypothesis it would be necessary to produce experimentally in mammals such disturbances of metabolism as diabetes, nephritis, jaundice, etc., to mate thus diseased animals and eventually to study the heredity of such offspring as they may beget. Such experiments, however, are
EXPERIMENTS OX MOXSTROUS DEVELOPMENT 531
difficult to perform in the absence of certain facilities, which are not usually accessible to the biologist. On the other hand, the spontaneous occurrence of these diseases in animals is too rare to permit of conclusive breeding experiments. It was necessary, therefore, to confine myself to a preliminary step in the investigation. This consisted in subjecting eggs of an oviparous vertebrate to the action of some toxic substances found in the urine under certain pathological conditions of metabolism.
The eggs of Fundulus heteroclitus were chosen, and after being fertihzed they were exposed in early (1 to 2, 2 to 4, or 4 to 8, and 8 to 16, cells respectively) cleavage stages to the influence of solutions in sea-water of such substances as urea, butyric acid, lactic acid, acetone, sodium glycocholate, ammonium hydroxide, etc. Only with two of these substances, viz., butyric acid and acetone, have so far definite and positive results been obtained.
In the case of butyric acid, 10 cc. of a tV to ^t gram molecular solution in 50 cc. of sea- water was found to be approximately the optimal solution, i.e., causing the greatest number of eggs to develop in a defective maim.er. If fertilized Fundulus eggs were left in this solution for fifteen to twenty hours and afterwards transferred to pure sea-water a great number of them gave rise to monsters. Similar results were obtained with solutions of acetone. Here the eggs were exposed in a number of dishes to the action of solutions of 20, 25, 30, 35, 40, 45 and 50 cc. of a gram molecular solution of acetone added to 50 cc. of sea-water. A varying relative number of deformed embryos was found in all dishes, increasing with the strength of the solution up to 40 cc. of acetone and decreasing in solutions still stronger, which caused an increase in the death-rate of the eggs. The length of exposure to the action of acetone was forty-eight hours in most and twenty-four and seventy-two hours respectively in some experiments. It was found that while long exposures increased the mortality of the eggs the difference in effect on the surviving eggs was very shght between exposures of twenty-four and those of forty-eight or even seventy-two hours. This points to the probability that it is mainly the initial effect
THE ANATOMICAL RECORD, VOL. 9, XO. 7
532 E. I. WERBER
of the toxic solution on the ovum that causes it to develop in an atypical manner. The eggs were fixed and preserved after Child's subHmate-acetic-formaline method and were left in 4 per cent formaldehyde until the time of their imbedding. They were imbedded in celloidin-paraffin, chloroform being used as a clearing agent. Sections were cut 6 or 7/i thick, stained with Delafield's hematoxyhne and counterstained with erythrosin.
SURVEY OF THE MORPHOLOGICAL RESULTS 1. Deformities of the sense organs and the mouth
The morphological results obtained in l^oth series of experiments, with butyric acid and acetone, being very much alike, it will suffice to state that the deformities enumerated here were found to be common to both. Great numbers of cyclopean embryos were found in both these series of experiments. I have recorded in my observations the occurrence of transition from tw^o normal eyes in the typical position in the head all the way down through the more or less closel}' approximated eyes or eyes of a double composition and true cyclopia to complete anophthalmia (figs. 1-7) as described by Stockard ('09) in his experhnents with magnesium chloride and alcohol, and by Lewis ('09) in his pricking experiments. Other defects in the development of the eyes such as asymmetric monophthalmia (fig. 8) and microphthalmia ffig. 9) were also found to occur abundanth'. In some eml^rj^os all that could be detected of the eyes were lenses only or a rudiment of the choroid coat with or without a rudimentary lens. Coloboma, by which is understood a patency of the embryonic fissures of some parts of the eye, was found in some cyclopean as w^ell as in two eyed but otherwise defective embryos.
Stockard f'09) has also described a jieculiar change in the form and position of the mouth of the cyclopean embryo. The mouth in such embryos has the appearance of a snout, a proboscis-like structure, and is pushed do\ATi below the c\'clopean eye. This displacement into the ventrolateral position Stock
Figs. 1-29 Sketches of monstrous Fundulus embrj'os.
Fig. 1 Normal Fundulus embryo, twelve days old, to show the position of the eyes; h., heart.
Fig. 2 Synophthalmia bilentica, from ^ gram molecular butyric acid, eighteen days old.
Fig. 3 SjTiophthalmia bilentica, from tj gram molecular butyric acid, twentyeight days old, with 'proboscis' — mouth, m.
Fig. 4 Sjmophthalmia bilentica, from iV gram molecular butj-ric acid, twenty-eight days old.
Fig. 5 Sjmophthalmia unilentica, from tV gram molecular butyric acid, twenty-four days old.
Fig. 6 Cyclopean embryo, showing one unusually large median eye with fused olfactory pits, o.p., and proboscis-like mouth, m. From xV gram molecular butyric acid, eighteen days old.
Fig. 7 Anophthalmic embryo with club-tail and distended ear vesicles, from acetone solution (40 cc. gram molec. sol. to 50 cc. sea-water), thirteen days old.
Fig. 8 Asymmetrically monophthalmic embryo, with greatly distended ear vesicles, e.v., one pectoral fin only and club-tail. From acetone solution (35 cc. gram molec. sol. to 50 cc. sea-water), thirty-two days old.
533
534 E. I. WERBER
ard attributes to the circumstance that the cyclopean eye being frontally located has caused the mouth to move downward. This interpretation, however, seems to be insufficient in view of the fact that I have observed in my experiments this occurrence not only in cyclopean monsters (fig. 6) but also in some cases of synophthalmia (fig. 3), asymmetric monophthalmia (fig. 8) and in many cases of dorsal microphthalmia (fig. 9). In some asymmetrically monophthalmic embryos (fig. 29) the mouth, while being apparently normal in shape, occupied the position in which normally the lacking eye would have to be located. Abnormalities of the olfactory pits were also almost invariably found to occur in embryos exhibiting various degrees of median 'cyclopia' (fig. 6) as well as in asym,emtrically monophthalmic embryos (fig. 20). The}^ usuall}^ corresponded to the anomalies of the eyes of a given embryo, that is, were either blended into one median pit or exhibited various degrees of approximation or fusion respectively in the cyclopean embryos. In the asymmetrically monophthalmic embryos where the mouth had taken the position of the missing eye, the nasal pit of the side possessing the eye was usually found to be in the normal position, while the pit belonging to the side lacking the eye has sometimes been found to be located posterior to the mouth. This unilateral ectopia of the nasal pit in the monophthalmic embryo is probably secondary to the ectopia of its mouth. These changes in shape and position of the mouth as well as of the olfacotry pits are apparently due to processes of regulation after an blastolytic destruction of a certain area at the anterior end of the earh^ embryo's body.
In a great many embryos the auditory vesicles reached enormous size, which on microscopic examination seemed to be due
In a previous note ("The influence of products of pathologic metabolism of the developing teleost ovum," Biol. Bull., vol. 28, no. 1, pp. 51-57), it \vas staied (p. 54) that in some cases of asymmetric monophthalmia an open orbit was found on the side lacking the eye. Ihis error was made owing to the transparency of the living specimens. It was the mouth in the exact position of the eye that was mistaken for an 'open orbit.' 'I he error was found when the drawings of the living embryos were comi)aro(] with llic fixed six-cinicns.
Fig. 9 Microphthalmic embryo (small eyes dorsally located), with distended ear vesicles, e.v., From ^ butyric acid, twenty-nine days old.
Fig. 10 Greatly malformed cyclopean embryo with dorsallj' located eye, rudimentarj^ pectoral fins and club-tail, from tV gram molecular solution butryic acid, thirteen days old.
Fig. 11 Greatly malformed embryo from acetone solution (3.5 cc. gram molec. sol. to 50 cc. sea-water) with one rudimentary lateral eye, without fins, clubtail; p.c, distended pericardial vesicle.
Fig. 12 Extremely malformed anophthalmic embryo from acetone solution (20 cc. gram, molec. sol. to 50 cc. sea-water), fourteen days old.
Fig. 13 Greatly deformed anophthalmic embrj^o, with head partly constricted off from the rest of the body. From acetone solution (30 cc. gram molec. sol. to 50 cc. sea-water), sixteen days old.
Fig. 14 Extremely malformed, oedamatous embryo, with one rudimentary lateral eye, distended ear vesicles, e.v., with club-tail and without pectoral fins. From T2 gram molecular butyric acid, fourteen days old.
Fig. 15 Amorphous embryo from acetone solution (40 cc. gram molec. sol. to 50 cc. sea-water), sixteen days old.
Fig. 16 Egg with amorphous tissue fragments on j-olk-sac, from acetone solution (30 cc. gram molec. sol. to 50 cc. sea-water), thirteen days old.
Fig. 17 Meroplastic embryo with rudimentary eyes, from acetone solution (35 cc. gram molec. sol. to 50 cc. sea-water), twelve daj^s old.
535
536 E. I. WERBER
to an oedematous distension (figs. 7, 8, 9, 14). In some monophthalmic embryos which had hatched a similar observation was made to the one recorded by Stockard ('09) in his experiments. These embryos "swam in circles, often whirling around with great rapidity, much as Japanese waltzing mice do. Others swam in irregular spirals and only progressed in a straight direction with difficulty. " In most of them I observed that when forced to swim in a straight forward direction they would immediately drop to the bottom of the dish, while they were able to swim for some distance along the wall of the finger-bowl in which they were kept. Stockard attributes this functional anomaly to a defective muscular arrangement, the animal's body being slightly bent or twisted so that it is unable to straighten perfectly." From my own observations and Stockard's ('10 a) microscopic findings I am rather inclined to think that in these embryos — at least in my own experiments, if not also in those of Stockard's — the locomotor anomaly is due to defects in the semi-circular canals.
2. General dejects; ' meroplastic embryos^
Besides the already referred to deformities of the sense organs there were found in both butyric acid and acetone solutions great numbers of embryos which developed in a much more defective manner, the malformation involving practically the entire bodies. Thus curiously deformed dwarfs with vestigial eyes or blind, or slender, elongate, greatly malformed embryos (fig. 11) often with waistlike constrictions (figs. 12 and 13) were of not infrequent occurrence. A considerable number of eggs were recorded with amorphous embryos (fig. 15) or only amorphous fragments of tissue (fig. 16) on the yolk-sac. The amorphous embryos would correspond in homology to those found in man and described by ]\Iall (I.e.) and others. On micoscopic examination of several amorphous embryos it was found that the nervous system, the notochord and the viscera while having developed, are highly abnormal and rudimentary in structure. The sections present pictures similar to those of
EXPERIMENTS ON MONSTROUS DEVELOPMENT
537
Fig. IS Egg with 'solitary eye,' s.e., and three very small clastolytic tissue fragments, t.f., from acetone solution (35 cc. gram molec. to 50 cc. sea-water), twelve days old.
Fig. 19 AsjTnmetrically monophthalmic embryo, with club-tail, without pectoral fins. On the j-olk-sac at a distance from the embrj-o is seen an 'isolated eye,' i.e. From acetone solution (40 cc. gram molec. sol. to 50 cc. sea-water), 12 days old; p.c, pericardial vesicle, y.s., yolk-sac.
Fig. 20 Asymmetrically monophthalmic embryo with 'proboscis' — mouth, without pectoral fins. From yV gram molecular butjTic acid, twenty-eight days old.
Fig. 21 Duplicitas anterior; both components are anophthalmic. From acetone solution (35 cc. gram molec. sol. to 50 cc. sea-water), si.xteen days old.
sectioned amorphous fetuses of man. The amorphous fragments may be compared to the 'nodular forms' of ]\Iall which he found in some aborted human ova.
By far the most numerous were found to be eggs in which only a part of the body (fig. 17) had developed — 'meroplastic embryos' (Roux, I.e.). In these the hind parts of the bodies were missing to a greater or lesser extent, while the anterior parts were often extremely deformed, their shape being much distorted and only more or less rudimentary eyes being present. ^Maii}' of them m
538 E. I. WERBER
which an anterior half of the body had remained would belong to the class designated by Roux (I.e.) as 'hemiembryones anteriores/ which he and subsequently other investigators found to develop from the frog's egg if one of its first two blastomeres was punctured with a hot needle.
The other cases of meroplastic development concern embryos in which either more than the anterior half of the body had developed, or less. They range all the way from those in which hardly much more than the tail was missing to those in which only a very small anterior part of the head (usually with one or sometimes with two vestigial eyes) was present.
3. Independent development of an eye from a fragment of the
medullary plate
By far the most curious and most significant of all the meroplastic ova recorded in these experiments were some in which all that was left of the embryo was a fragment of brain tissue with a solitary eye. The fragment of brain tissue was in some cases somewhat larger than the solitary eye to which it has given rise, while in others it was smaller. In one of these eggs (fig. 18) in which the solitar}' eye was rather defective — the chorioid fissure being patent — several small amorphous fragments of tissue could be observed at different points of the yolk-sac at a considerable distance from each other. Since I have observed such fragments of tissue on many other eggs in which either a defective embryo or nothing else besides the amorphous fragments had developed, I am inclined to think that such cases give us a clue as to what processes may be involved in bringing about the effects recorded in this work.
The 'solitary eyes' when observed in the living egg had the typical appearance of eyes and no doubt could be felt that the interpretation of these sporadic cases was correct. However, as no other similar case was known, it seemed rather improbable that a small fragment of the medullar}^ plate would be able to go on developing independently so far as to give rise to such a complex organ as the eye. The possibility suggested itself that the 'solitary eye' of such an egg may be connected with an
EXPERIMENTS OX MONSTROUS DEVELOPMENT 539
embryo which had sunken in the yolk-sac leaving the eye on the surface. The main objection to this would, of course, be that such an egg would die in a very short time, for the embryo could not possibly receive a sufficient oxygen supply; and the death of such a sunken embryo would soon cause autolysis of the ovum. More convincing proof seemed to be furnished by the fact that the yolk-sacs of the eggs, owing to the method of fixation, were translucent, while the embryo had turned white and opaque. If an embryo had sunken into the yolk-sac, it could thus easily have been detected. But, in spite of very careful examination of the ova with solitary eyes, not a trace of an embryo could be seen within their yolk-sacs. Moreover, in order to establish this fact of the independent development of the eye on the firm basis of unmistaka])le evidence, I sectioned one of these eggs. I have purposely chosen the egg in which besides the solitar}" eye several (some three or four) very small fragments of tissue could be observed on the 3'olk-sac, thinking that the latter ones might possibly offer a basis for the interpretation of the morphogenesis of the solitary eye.
The microscopic examination (fig. 22) of the sectioned egg proved that my interpretation of the case as that of a 'solitary eye' was perfectlj^ correct. In the sections it can be plainly seen that the blastoderm had overgrown the entire yolk-sac just as this takes place in normal development. But no embryo can be seen in the yolk of an}^ of the sections. On the yolk-sac besides the solitary eye there can be seen on some sections the fragments of tissue mentioned above, one of which makes the impression of a defective spmal cord and when followed out in successive sections is seen to give rise to the eye in question. Some blood vessels have also developed in a very abnormal fashion and no heart is present. Of the other amorphous tissue fragments one proved to be another eye, although very rudimentary in structure. The nervous elements of this second eye are obviously greatly inhibited in development, neither a choroid nor an iris are present, but the cornea and the lens have developed to a degree permitting of safe identification. The distance between the position on the yolk-sac of this eye
540 E, I. WERBER
rudiment and that of the well developed sohtary eye being as much as 262/i it seems evident that a blastolytic fragmentation of the early embryonic material and a subsequent shifting of the fragments to distant parts of the j^olk-sac's surface has occurred.
Two more eggs wdth solitary eyes were cut into sections and a microscopic examination again confirmed the correctness of their being interpreted as such.
Several eggs were also found with asymmetrically monophthalmic and otherwise malformed embrj^os in which a completely isolated tissue fragment T\dth a well developed eye could be seen on the yolk-sac at a distance from the embryo which made its connection with the same appear impossible. One of these eggs was sectioned and the microscopic examination revealed the fact that the isolated eye was in no connection whatsoever with the embryo.
In view" of these findings no doubt can now be felt that the 'sohtar}^ eye' (embryone absente) and the 'isolated eye' (embryone praesente) offer ample evidence of the fact that a very small fragment of the medullary plate maj^ be able to develop independently of the rest of the embryo's anlage far enough to give rise to such a complex organ as the eye. This is, as far as I am aware, the first case on record of the independent development ('self-differentiation,' Roux) of the eye.
4. Defects of the brain
The destructive influence of butyric acid and acetone on the developing egg of Fundulus manifests itself also in the severe injuries sustained by the central nervous system. In teratophthalmic embr^'os the brain was in most cases found to be abnormal to a greater or lesser degree. The forebrain may often be unpaired while the mid- and hind-brain of the same embryo are bilateral in symmetr3\ The hemispheres of one brain may often differ considerably in size and shape as well as in other respects. Thus, e.g., there may be seen a striking developmental inhibition in one hemisphere where most of the
Fig. 22 Photomicrograph of section of egg with 'solitary eye,' (Cf . fig. 18) ; y.s., yolk-sac, o.g., oil globule space. X 80.
Fig. 23 Photomicrograph of transverse section through the eye region of a normal Fundulus embryo, fourteen days old, one day after hatching. X 80.
541
542 E. I. WERBER
nervous elements had not proceeded beyond the neuroblast stage, while the other hemisphere may be made up of apparently well developed nerve cells and fibers. It was also often observed both in the brains and the spinal cords of teratophthalmic or otherwise defective embryos that wherever the neuroblasts failed to differentiate into ner\'e cells a great many large, clear and empty tissue spaces (figs. 26 and 27) were present. The probability suggests itself that these spaces in the hving embryos may have been filled with body fluid. For, the heads of some defective embryos when observed in the living or even the preserved specimen were distended to a degree suggesting oedema. As was mentioned before, the same is true also for the ear vesicles. On microscopic examination it can invariably be seen that the usual size of the latter ones is due to (an apparently oedematous) distension of the semi-circular canals. A condition of hydrops is thus produced in the embryo which (since the fish brain has neither lateral ventricles nor a choroid plexus) might perhaps be considered as homologous with the internal hydrocephalus of man. Furthermore, the cranial (figs. 25 and 26) cavity may be unusually large, often containing some fibrin. Both this condition and the intracerebral oedema may often be present in the same embryo. Genetically these dropsical conditions probably are due to an arrest in the development of the blood or lymph vascular system.
5. Defects of the blood vascular system
There is a very wide range of variation in the deformities to which this system is subject. The heart is almost perfect in some embryos in which cyclopia is the only superficiall}' noticeable defect. In more extremely maKormed embryos, however, the heart may be only an exceedingly delicate, straight tube in some embryos, while it may be absent altogether in others. It is interesting to note in this connection that while acardia was frequently observed in eggs in which a whole although malformed embryo had developed, a heart, tliougli usually more or less of a rudimentary structure, but functioning, could be
EXPERIMEXT8 OX MOXSTROUS DEVELOPMEXT 543
found in some eggs in which only meroplastic embryos or no embryo at all had developed.
^'^llile in some embryos the heart can be plainh' seen to be connected with the great vessels of the embryo and indirectly \yith. the larger vessels of the extraembryonic area, no such continuity exists in most of the deformed embr^-os which develop without a complete circulation, or, often without any circulation whatsoever. To J. Loeb ('93) belongs the credit for this remarkable discover}^, which is now corroborated by Stockard's ('15) and my own findmgs. The blood vessels of the yolk-sac may sometimes be present in the form of irregular, dense, apparently continuous networks, or in some cyclopean embryos they may approach in pattern and size the normal blood vessels, while in cases of more extremely malformed embryos only blood islands may be seen scattered in the yolk-sac. Yet blood vessels are usually found in such cases in the embryo itself. It may well be said that only few embrj^os which develop in butjTic acid and acetone solutions, while possessing blood vessels, have a circulation continuous with the extraembryonic area.
This accounts for the interesting fact that very often large lacunae filled with erythrocytes are seen in the bodies of some monstrous embryos as well as in the extraembryonic areas. These lacunae are very striking at the first glance at a living ovum on account of the bright red color of the erythrocytes. Another observation which was first recorded by Stockard ('15) and which I have also frequently made is that a large mesenchyme space, usualh^ m the head region or sometimes in other parts of the body, may frequently contain many leucocytes of apparently the polymorphonuclear varietj^ (fig. 24). The elements of the blood which come from different sources are thus seen to be isolated due to absence of a continuous system of circulation. The bearing of these data on the problems of vasculogenesis and haemogenesis is evident and I expect to discuss it elsewhere.
That these data may also have an important bearing on the genesis of some forms of hj'drocephalus has already been pointed out above. It is easy to imagine that even in man a local in
544
E. I. WERBER
Fig. 24 Photomicrograph of a transverse section through the e3'e region of a monstrum synophthahnicum bilenticum (Cf. fig. 2); I.e., leucocytes, o.c, optic chiasma. X 100.
hibition in the development of the blood and lymph vessels in the head region may result in accumulation of fluid in some parts of the head, even in the lateral ventricles of the brain due to lack of adequate drainage. Careful anatomical investigations of some well diagnosed cases of hydrocephalus may possibly bear out the validity of this assumption.
A ease described by Smith and Birmingham ('89) may be of interest in this connection. They re])ort a fetus aborted during the fifth month of j^regnancy with general oedema due to complete absence of lymphatic vessels. On microscopic examination of the skin and subcutaneous tissue they found in the sections large spaces, "in some parts clear and empty, in others filled
EXPERIMENTS ON MONSTROUS DEVELOPMENT 545
with a colloid material, evidently coagulated lymph." From their appearance and position they concluded that they were dealing with greatly distended lymph spaces, the overdistension being the result of the absence of a lymphatic drainage system.
6. Deformities of the appendages
One of the most frequently found malformations concerns the fins. Both the pectoral and caudal are very often rudimentary (figs. 7, 9 and 10) and diminutive in size. In some deformed embryos the pectoral fins or all fins may be absent altogether (figs. 11, 14, 19 and 20) and the tail in such embryos is club-shaped. I have also often recorded embryos in which only one pectoral fin was present (fig. 8) while only one embryo was found with three pectoral fins. One is reminded by these deformities of the fish's appendages of well known cases in the human being with which they seem homologous. The clubfoot in the human fetus and congenital absence of upper limbs in man have often been found. Likewise cases of supernumerary upper limbs have been recorded in man as well as in other mammals.
The fact that such defects can be produced in the fish by chemical action suggests that in mammals they may be due to like causes.
7. Duplicities
The tendency of butj'ric acid and acetone to produce twins seems to be only slight for I have observed only a few such cases. I have recorded only one case comparable to the 'Siamese twins' type of the human. In this egg two deformed embryos with a common heart had developed on opposite sides of the egg. Some other cases of twin formation found in these experiments belong to the type know^n as 'duplicitas anterior' (fig. 21 )which Speeman ('04) produced experimentally by tying a Hgature around the fissure between the first two blastomeres of the amphibian egg. Several cases of duplicity have also been recorded, the nature of which I expect to ascertain by microscopic exaiTiination.
54(3 E. I. WERBER
THE ^lORPHOLOGY OF TERATOPHTHALMIA
After this general survey of the recorded monstrosities I shall now briefly consider the morphology of some such ophthalmic defects as I have so far been able to examine in sections with a view of a more exhaustive treatment of the topic in the complete account of the results of these investigations. Before, however, presenting a few of the cases which I have studied, some terms may be defined, the use of which to me seems to be desirable.
The various degrees of the one-eyed condition which have been recorded b}^ previous writers (Speeman, Stockard, Lewis) as well as by myself, I shall henceforth classify in the following manner :
I ^lonophthalmia asjTnmetrica (s. lateralis)
(a) perfecta
II Cyclopia (s.iMonophthalmiamediana) i /, s i ^i i •
•^ ^ ^ 1^ (b) synophthalmica
TTT c 1+11- f (^) unilentica
111 bvnophthalmia < ,, s , .,
'^ (b) bilentico
. The first term is well known because it already has been used by Ahlfeld and adopted b}' Stockard and so it will need no further comment.
In the case of cyclopia the distinction is made between 'perfect' ('true') cyclopia and 'synophthalmic' cyclopia. The first is to indicate the presence of a single median eye which on microscopic examination is found to be single throughout and nowhere exhibiting evidence of its being composite in character. Under the latter one will be classified such cases where on macroscopic examination a single median eye (with one lens) is found present, while the microscopic examination of sections reveals its composite character. Under 'Synophthalmia' will be classified cases of 'fused' eyes where the composite character of the eye (or eyes) is plainly discernible iJi ioto. If such an optic apparatus should possess one lens only it will be designated as 'Synophthalmia unilentica,' while in the presence of two symmetrically placed lenses 'Synophthalmia bilentica' will be used as the descriptive term.
EXPERIMENTS OX MOXSTROUS DEVELOPMENT 547
A few illustrative examples will now be given, after which a discussion of the morphogenetic factors of teratophthalniia will be taken up.
Figure 5 represents a case of unilentic synophthalmia. The egg was subjected between the second and third cleavages, — i.e., about 2| hours after insemination^ — to the action of 10 cc. of tV gram molecular solution of butyric acid for twenty hours and then transferred to pure sea-water. Alread}' on superficial examination the composite structure of the eye is easily recognized. It is seen that the approximation of the two eye components is so close as almost to give the appearance of perfect cyclopia. Transverse sections (fig. 26, p. 549) show that the eye is composed of two incomplete optic cups facing each other and enclosing a single lens of about the usual size. The cornea and iris are normally developed and the retinal layer is well differentiated. Two optic nerves are seen to pass out of the eye in a fe\y loose bundles of fibers and, after having formed a chiasma, to enter the opposite sides of the brain.
The incompleteness of the fused optic cups is probably due to the circumstance that at a very earlj' stage of development a large part of the ophthalmoblastic material had been eliminated from development owing to the chemical alteration caused by the butyric acid. The injury sustained by the embryo must apparently have been the severest at the most anterior point of the main body axis, diminishing gradually posteriorwards. The following data seem to substantiate this interpretation. The forebrain is unpaired (fig. 25) and the rest of the brain is when followed in successive sections posteriorwards seen gradually to present more and more distinctly the condition of bilateral symmetry. The midbrain and hindbrain while being bilateral, exhibit, however, a certain other abnormality. The injury here was apparently mainly restricted to the blood and lymph vessels, the earliest anlagen of which seem to have been arrested in their development. This condition can be recognized by the great number of large, clear and empty spaces (fig. 26) in the tissues of the posterior parts of the brain, which in the living embryo have apparently been filled with fluid owing to
THE ANATOMICAL RECORD, VOL. 9, SO. 7
548 E. I. WERBER
the existing imperfection in the circulation. A condition of oedema has thus apparently resulted from lack of drainage. No other abnormalities of these parts of the brain or any other part of the embryo can be seen, which makes it appear very probable that the anterior part of the embryo body is the most sensitive one and thus subject to the highest degree of injury.
In figure 2 is seen an eighteen days old embryo exhibiting the condition of 'synophthalmia bilentica.' This egg had been subjected to the action of 10 cc. of a re gram molecular solution of butyric acid in 50 cc. of sea-water for twenty hours, after which time it was transferred to pure sea water. In the specimen in toto, the eyes, while being very close together, could not be considered as fused. On microscopic examination, however, it was found that the eyes were fused more posteriorly (fig, 24, p, 544) and that the fusion is the more intimate the more posterior is the section examined. If all sections be examined it can be clearly seen that the highest degree of the injury sustained by the egg due to the treatment is in the most anterior part of the embryo's body, i,e,, in the region of the eyes. The abnormalities found in this part of the body besides the fused eye involve also the olfactory pits, which are so closely approximated as to be partially fused, and the forebrain, which is unusually small in size and unpaired. The transverse sections of this embryo are somewhat oblique and the double eye being somewhat asymmetrically located in relation to the chief body axis, in the figure the lens of one eye can be seen to be sectioned about midway while of the other lens the most posterior part is cut. The two lenses, however, can be clearly made out in the figure. The choroid coats which are imperfectly developed and the well differentiated
Fig. 25 Camera lucida drawing of a transverse section anterior to the eye region of a monstruin synophthalmicum unilenticum (Cf. fig. 5), showing unpaired forebrain, f.b.. greatly enlarged cranial cavitv occupied by some fibrin, /. X 140.
Fig. 26 Fhotomicrograph of a transverse section through the eye region of the same embryo as in figure 25, showing the two components of the sjmophthalniic eye, facing one another, with one lens, /. Man}^ tissue spaces, t.s., are seen in the brain. The cranial cavity in the region of the optic lobes, o.L, is greatly distended, O.C., optic chiasma. X 100.
SmM
549
550 EXPERIMENTS OX MOXSTROUS DEVELOPMENT
retinal layers are seen to be perfectly coalesced ventrally, while dorsally they arch inwards to leave an opening for the optic nerves. The latter come as separate bundles of fibers from each one of the eye components and fuse into one trunk just at the point of emerging from the double eye. This common optic nerve trunk is in other sections seen to enter only one hemisphere of the brain. The cavity between the sclera and the brain has widened out and is on both sides near the head integument occupied by large leucocytes of apparently the polymorphonuclear variety. These leucocytes are seen in all sections of the embrv^o densely filling spaces between the tissues or they may also be found more scattered in the mesenchyme. The blood vessels of this embr^^'o as well as of its extraembryonic area are rather scarce, and the discontinuity of the latter ones being very striking, the suggestion is at hand that the circulation of the embryo was imperfect. This would account for the accumulation of white blood corpuscles m the mesenchyme as well as for the numerous apparently oedematous interstices which they often filled. Summarizing the abnormalities found in this embryo, it may well be said that the degree of injury sustained by it, being highest in the immediate region of the eyes, diminishes along the main body axis posteriorvvards, the midbrain and hindbrain, excepting the anomalies due to an inhibited circulation, being apparently normal in all other respects.
Of a much higher degree is the injury sustained by the case of perfect cyclopia with a supernumerary lens, cross sections of which are represented in figures 27 and 28. The egg had been subjected in the eight-cell stage to the action of 35 cc. of a gram molecular solution of acetone in 50 cc. of sea-water for fortyeight hours. The embryo when killed was tw^enty-seven days old. A single median eye is seen in a transverse section (fig. 27) which is smaller in size than a normal eye, the lens only being disproportionately large. All other structures of the eye, viz., cornea, iris, choroid coat and retina are present, the latter having differentiated in a defect i\e manner. At a somewhat more posterior level of the brain, the following view is presented (fig. 28). Laterally from the eye on one side is seen a large well
28
a tZltlselTaTy:7r'^^ ' '"^'^r"^ ^^^^^^^ ' ^^« -3'e region of
small medTar i±Lfe th r ff^'"t" 7 '°' ^^' "" ^^^^^^°S' «^ '--^ lens. The ?o;ebTin ^'^f ^^^^f ed. eye with a disproportionately large
From ace'one oludon ,3^' ^^'^"«™^^^^-,-' *"^"^ ^P^^^^' ^•^•' P^^^* *« -^eml fFiff 9Q p, . ^ ' ?• ^'^°' ™°'^^- sol- t« 50 cc. sea-water). X 160
^^"^ J^^^:^:\:^:T' ^^"^^^ °^ ^^^ san.eL?,ras in and optic anlagen on C X^^J^I^- ----- ^-: ^' ^ ^ side
551
552 E. I. WERBER
differentiated lens surrounded by an epithelial capsule and anteriorl}^ and laterally by a cornea, which is found to be in continuation with the cornea of the cyclopean eye. On the other side of the eye between its lateral border and the head integument is to be seen a large patch of cells with deeply stained nuclei and a little higher upwards on the same side (o. a.) and bordering the integument two more such patches of tissue can be seen in close approximation. On careful examination it was found that these three patches of cells represent fragmentary optic anlagen. The one of these optic anlagen which borders the eye has even differentiated into a small retinal layer which is in about the same stage of differentiation as the same structure of the cyclopean eye. There is only one optic nerve present which in more posterior sections can be seen to enter as a trunk one of the hemispheres. The brain is much deformed and more so anteriorly, where its symmetry is obscured, than posteriorly, where its bilateral symmetry can be recognized without difficulty. Many large clear tissue spaces can be seen in the brain in almost all sections which in the living apparently represented persistent early embryonic vascular anlagen and were filled with body fluid owing to lack of drainage caused by an imperfect circulation.
This embryo I regard as one instance of the perfect cyclopean condition, for only one complete eye has developed. Such indications as are present of the other optic anlage give us a clue to the genesis of the single-eyed condition in this case. Although the injury sustained by the embryo is not localized and is of a high degree, it is highest in the most anterior portion of the body, diminishing gradually posteriorwards along the main body axis. The anterior portion of the very early embryonic anlage has apparently undergone great destructive alterations of a blastolytic nature, owing to which the ophthalmoblastic material of one side was fragmented and dispersed while the corresponding material of the other side has been injured less severely and has given rise to an imperfectly developed median eye. The material which would in the normal embryo eventually be represented by the interocular area seems to have, owing
EXPERIMEXTS OX MOXSTROUS DEVELOPMENT
553
to some regulatory processes, moved lateralwards, where it has given rise to an independent lens. To the same regulatory process is it probably due that the ophthalmoblastic material of the less injured side has been shifted medianwards to give rise to the Cyclopean eye.
The case of 'asymmetric monophthalmia' which I shall now describe will, I believe, also point to unmistakable evidence that in teratophthahnic embryos the sustained injury is usually the severest at the anterior end of the body.
The egg had been subjected for forty-eight hours to treatment with acetone, 35 cc. of a gram molecular solution of which were added to 50 cc. of sea-water. The embryo was killed one day after hatching when it was sixteen days old. On examination in toto there was seen to be present only one eye in the usual lateral position while the mouth occupied exactly the position of the missing eye. In transverse sections (fig. 29) it is seen that the one eye present is well developed and apparenth" normal in structure. No indication of another eye or optic anlage can
Fig. 29 Photomicrograph of a transverse section through the eye region of a monstrum monophthalmicum asymmetricuin, from acetone solution (35 cc. gram molec. sol. to 50 cc. sea-water), sixteen days old, one day after hatching. The mouth, in., occupies the position of the missing eye. X 120.
554 E. I. WERBER
be found anj-^'here in the sections and onl\" one optic nerve is seen in the more posterior sections to pass out of the retina and terminate in the optic lobe of the opposite hemisphere. The brain is symmetric as far as its bilaterality is concerned, but is asymmetric in regard to the position occupied by the two hemispheres in relation to the main body axis. As can be seen in the figure, the hemisphere of the side where the eye has developed, is in its normal position while little can yet be seen of the other hemisphere, which has been shifted posteriorwards and comes to view in more posterior sections. The same posteriorward displacement has also affected the olfactorj^ pit and the semicircular canals of the side lacking the eye. There are no other abnormalities to be recorded for this embryo. The abnormalities fomid, however, justify the conclusion that the injury sustained by the early embryo was chiefly a unilateral one at the anterior end of the bodj^ Owing to this injury the ophthalmoblastic material of one side has suffered complete destruction. On the same side, owing to subsequent processes of regulation, the mouth has arisen in what was to be the position of the eye and a posteriorward displacement of the brain hemisphere of the injured side has taken place. The sustained injury, while being lateral from the median axis of the body, was severest at its most anterior end where it has elmiinated the material for an entire organ.
The cases described above and many more similar ones which have been studied led me to accept in the main the 'fusion theory' of cyclopia which has in recent years been advocated by Lewis ('09) and Speemann ('12) and to reject as untenable Huschke's ('32) view of the early single anlage of the eye which Stockard has quite recently ('13) adopted in his morphogenetic analysis of cyclopia.
Stockard, who was the first to produce experimental cyclopia in fish by chemical agents, has in his earlier work ('09) suggested that the developmental defect is due to the specific anaesthetic action of the chemicals (magnesium chloride, chloroform, ether, etc.) which he used. He thought that the giving off of the eye anlagen by the brain was inhi]:)ited often in an unequal manner
EXPERIMENTS OX MOXSTROUS DEVELOPMEXT 555
on both sides, thus giving rise by fusion of the inhibited anlagen to various transition stages between two normal eyes and cyclopia and anophthalmia. Recent investigations of McClendon ('12), however, who obtained similar results with a great variety of non-anesthetic substances, have caused Stockard ('13) to abandon his anesthetic theory of teratophthalmia. He now believes that the eye anlage in the medullary plate is single and median in position, that in normal development this single anlage eventually divides into two portions which move lateralwards, where they develop into the optic vesicles. If the embryo is subjected to the action of toxic chemicals this separation of the original single median optic anlage into two, may, he concludes, be inhibited to a greater or lesser degree and thus various degrees of the Cyclopean defect may result. He even submitted the theory of the single optic anlage to an experimental test which, he believes, answered the query in the affirmative. It is unfortunate that Stockard's statements are not substantiated by better evidence than that which he brings forth, since he has failed to support his statements by illustrations of specimens in toto and in sections of the material which he regarded as permitting of such important conclusions. Until this evidence is furnished, however, the theory of the single median optic anlage in the medullary plate would hardly seem to he acceptable. On these grounds, we cannot agree with Stockard's conclusion on the morphogenesis of Cyclopean defects.
The 'fusion theory' of cyclopia is, I believe, justified in the main, and has recently been ably supported by Speemann (I.e.), jNIall (1. c.) and W. H. Lewis (1. c). These authors assume a coalescence or fusion of two originally separate optic anlagen. Just how this fusion comes about may still be a matter of discussion. I think that Speemann's and Lewis' suggestions are very illuminating just on this point. Both these authors have produced all those conditions of one-eyedness which Stockard has obtained in his investigations, and which I have recorded in my experiments, a description of some of which is given above.
Both Speemann and Lewis have produced the teratophthalmic condition by mechanical injury of a small area at the ante
556 E. I. WERBER
rior end of the early embryo; and Lewis holds that the collapsing of the wound surfaces effected the approximation of the two optic anlagen. The degree of this approximation would depend upon the size of the fragment of interocular tissue which Speemann eliminated by constriction and Lewis by pricking. This, they conclude, would account for the various degrees of the Cyclopean condition. Even asymmetric monophthalmia can in this way be accounted for, as it is easy to imagine that in the experiment the injury inflicted to the embryo may sometimes be somewhat lateral from the embryos main body axis and that a complete eye anlage may thus be destroyed. The terat ophthalmic condition would then, as Speemann points out, have to be regarded as a 'defect' (evidently meaning a mechanical defect) rather than a developmental inhibition.
]My own conclusions are very similar to those arrived at by Speemann and by Lewis. I have, however, found it necessary to modify the fusion theory of cyclopia to conform to Stockard's and my own findings. The explanation offered by these authors is open to the criticism made by Stockard, namely, that cj^clopean eyes are rarely in size and extent equal to the sum of the two normal eyes combined." It is obvious that this objection is justified and that conclusions based on results obtained from mechanical experiments cannot be extended to cover the results of the chemical experiment or to account fully for the occurrence of the Cyclopean defect in nature under unfavorable environmental conditions.
The following modification of the coalescence theory of cyclopia meets the objection raised by Stockard, covers the results of the chemical experiment and thus, I believe, may be extended to the morphogenesis of various cyclopean defects in nature, where the causal factor is undoubtedly^ a chemical one.
"When the fertilized egg is subjected to the action of toxic substances it will sustain an injury, which — depending upon t'ne stage of development, the substance used and the strength of its solution — may be a general one, i.e., involving the whole or a large part of the embryo's body, or a locally restricted one. The results of Stockard's and ^IcClendon's work, as well as some
EXPERIMENTS OX MOXSTROUS DEVELOPMENT 557
results of my own experiments, point to a blastolj^tic injury of a restricted area at the anterior end of the early embryo's body in the case of teratophthalmia. Child's^ ('09, '12) important discovery of the 'axial gradients', according to which the anterior end of a flat-worm's body is the most sensitive one to the action of injurious substances w^ould seem well to justify this assumption. In the earh^ vertebrate embryo, before the organs have been differentiated, probably very similar physiological conditions obtain, and thus we may well assume that its anterior end is the point of least resistance. When the egg is acted upon by a toxic substance, a restricted area at the anterior end of the embryo's median body axis becomes so altered chemically as to be eliminated from further development or it may go on developing to a certain point beyond which it is chemically unable to proceed. This restricted area at the anterior end of the body axis is the region between the future optic anlagen or even the region of those anlagen. The size of the injured area at the anterior end is probably subject to considerable variation, and thus it ma}^ comprise the material which would normally correspond to the future interocular area and cause an approximation of the potential optic anlagen or it may extend even over the latter ones, thus eliminating parts of them, while the uninjured parts would coalesce after approximation and form any one of the various degrees of the synophthalmic condition. Or, finally, the injured area may comprise the whole of one potential optic anlage and little or no material of the future interocular area, thus causing the embrj^o to develop into a cj^lopean monster if the uninjured optic anlage is shifted medianwards, or into an asyimnetricalh' monophthalmic monster, if no such change in the position of the uninjured or less injured ophthalmoblastic material takes place.
^ Child offers an interesting interpretation for this high degree of sensitivity. According to him the rate of the metabolic reactions seems to be the highest at the anterior end of the planarian's body, decreasing gradually along the main body axis in the direction away from the head. This theoretical postulate of the metabolic 'axial gradient' was substantiated by the rate at which various points along the body axis have given off COo, the measurements having been made by Tashiro with his own biometer-method.
558 E. I. WERBER
Here too, the size of the chemically injured area may vary and thus in some instances a small fragment of the injured potential anlage may survive and develop into a diminitive whole, but somewhat rudimentary, eye, or into an eye fragment with a well differentiated retinal layer and choroid coat.
Howe^'er, it is necessary to assume that the injury which results in teratophthalmia is sustained at a very early stage of development. At that time the volume of the area which receives the injury is relatively small and such minute parts of the embryo as may represent the potential double anlagen of the eyes, are, since growth proceeds in a cubic proportion, relatively much nearer each other than they would be at a somewhat later stage, e.g., in the embryonic shield. This would account for the fact that the cyclopean eye may be very small in size, much smaller even than the normal eye. For, at this stage, the eliminated area may contain much more material than of one potential eye. The remaining ophthalmoblastic material may undergo some regulatory process to form eventually a single eye of corres])onding size or a double, fused eye if enough of the ophthalmoblastic material of both sides survives, or finally if the injury sustained by that material be so extensive as to comprise all of it, the condition of anophthalmia may be the result. This conception of the morphogenesis of teratophthalmia, while recognizing the coalescence theory of cyclopia as justified, also meets the objection which Stockard raised to Speemann's and Lewis' generalizations, namely, that the cyclopean eye is often much smaller than the sum of two normal eyes combined. At the same time, it makes it appear unnecessary to resort to a theory whose probability is so questionable as that of the single median optic anlago in the medullary plate" (Stockard).
The mechanism involved in the action of butyric acid and acetone in bringing about the great variety of recorded effects will have to be taken up as one of the further steps of this investigation. At the pres(>nt time, I can only state that there seems to occur in llic cgfis wIhmi subjected to the action of these solutions
EXPERIMENTS ON MONSTROUS DEVELOPMENT 559
an elimination of material of the blastomeres or of the germdisc and probabh' also of the yolk-sac. This elimination of material may be due either to the precipitating or solvent effect respectively of the chemicals which were used in these experiments. Since many eggs were found with living but extremely malformed embryos, the yolk-sacs of which were ruptured allowing some yolk to escape, the suggestion is at hand that another factor — some force like a temporarily increased internal osmotic pressure — may by its action have brought about the fragmentation of the germ-substance. On examining many eggs in which only amorphous tissue fragments can be found and the eggs with 'solitary eyes' (embryone absente) or eggs with 'isolated eyes' (embryone praesente) one gains the impression that either the blastomeres or the germ-disc had been blastolytically fragmented owing probably to both physical and chemical factors. These tissue fragments sometimes appear at such great distances from each other, or even from a malformed embryo, that there seems to be no doubt left that they have moved out of their original position along the axis of the embryo. Having examined many thousands of these pathological ova I believe that I possess enough evidence to justify the assumption of blastolytic processes of a physical or chemical nature, such as fragmentation due to a temporarily increased osmotic pressure or to chemical alteration of some parts due to the solvent or precipitating effect respectively of the chemicals used in these experiments. Mall speaks (1. c.) of a similar process which he assumes for some pathological human ova, and which he terms ' dissociation of cells. ' The effects of these apparently blastolytic processes vary enormously. For whatever parts survive them, may go on developing into a whole defective, or a meroplastic, embryo, or even into an isolated organ.
Just what brings about this great range of variation in the noted effects can at present hardly be more than conjectured. It would seem probable that the stage in which the egg is subjected to the action of the toxic solution ma}^ play an important part in that respect. For as Conklin ('12) has shown ". . . . stages during kinesis are more susceptible to modification than stages
560 E. I. WERBER
during interkinesis. Almost all persistent alterations of structure occur during cell division, few of those which occur during the resting period are permanent." However, this and similar other inquiries must be left to future investigation.
CONCLUSIONS AND SIBIMARY
1. Starting from the assumption that monstrous development is due to parental metabolic toxemia, experiments were performed in which Fundulus eggs were subjected to the action of some substances occurring in the blood or urine respectively of man during metabolic disturbances.
2. With two of these substances, i.e., butyric acid and acetone, positive results have been obtained, a great variety of monsters having been produced Vvhich are analogous or homologous respectively to human and other mammalian monsters. The monstrosities concern the eyes (cyclopia, synophthalmia, monophthalmia asjrmmetrica and anophthalmia), the ear vesicles, the olfactory pits, the mouth, the central nervous system, the heart and blood vessels, the fins (unpaired fin, absence of pectoral fins or all fins, club-tail, etc.) and body form.
3. A condition of hydrops was found in many embryos, due apparently to blood vascular abnormalities which might be considered as homologous to some forms of hydrocephalus in man.
4. In many eggs parts of the embryonic material have suffered destruction, while the remaining parts have developed into anterior hemiembryos or other meroplastic e]nbr3'os.
5. Eggs have been found in which one eye has developed from a small fragment of the medullar}' plate independently of an embryo (the 'solitary eye,' the 'isolated eye').
6. Regarding the mechanism of action of butyric acid antl acetone the conclusion has been reached and some evidence offered that the various monstrosities are brought about by a process of blastolytic fragmentation due to some factors not yet ascertained.
7. Regarding the formation of the various degrees of the 'cyclopean" dofoct it is concluded that the fusion theory of Speeman
EXPERIMENTS OX MOX'STROUS DEVELOPMEXT 561
and Lewis is justified in the main. An additional assumption is made, namely, that the blastolytic process which eliminates parts of the potential interocular or ophthalmoblastic material takes place at a very early stage of development (i.e., before the formation of the embryonic shield).
8. The results obtained tend to justify the assumption that monstrous development may be due to metabolic toxaemia.
The experimental part of this work has been carried out at the Marine Biological Laboratory at Woods Hole, ]\Iass., during the summer of 1914. It is a pleasure to me to acknowledge my indebtedness to the director, Dr. F. R. Lillie, for the facilities granted. I am also under great obligation to Professors ConkUn and ]McClure and to Dr. E.N. Harvey, whom I have often taken occasion to consult on some points of the work. For many courtesies I am indebted to Professor Dahlgren and ]\Ir. W. E. Hoy.
LITERATITtE CITED
Child, C. 'SI. 1909 The regulator}' change of shape in Planaria dorotocephala.
■, Biol. Bull. vol. 16.
1912 Studies on the dynamics of morphogenesis and inheritance in experimental reproduction IV. Certain dynamic factors in the regulatory morphogenesis of Planaria dorotocephala in relation to the axial gradient. Jour. Exp. Zool., vol. 13.
CoxKLiN. E. G. 1912 Experimental studies on nucle&r and cell division in the eggs of Crepidula. Journ. Acad. Natural Sciences of Philadelphia, vol. 15, Second Series.
Dareste, C. 1891 Recherches sur la production artificielle des monstrosities ou essais de teratologic experimentale. Second Edition.
FeriS, Ch. 1894 Etudes experimentales sur I'influence teratogene ou degenerative des alcools dits superieurs. Bulletin.? et Mem. dela Societe Medicale des Hopitaux des Paris.
Hertwig, O. 1896 Experimentelle Erzeugungen tierischer Missbildungen. Festschr. z. siebzigsten Geburtstage von Carl Gegenbaur. Bd. 2. 1910 Die Radiumbestrahlung in ihrer Wirkung auf die Entwicklung tierischer Eier. Sitzungsber. d. Konigl. Preuss. Akad. d. Wiss., Bd. 11.
HuscHKE. E. 18.32 t^ber die erste Entwicklung des Auges und die damit zusammenhangende Cyclopie. Arch. Anat. Physiol., vol. 5.
Lewis, W. H. 1909 The experimental production of cyclopia in the fish embryo (Fundulus heteroclitus). Anat. Rec, vol. 3.
562 E. I. WERBER
LoEB, J. 1893 t'ber die Entwicklung von Fischembryonen ohne Kreislauf.
Archiv. flir d. gesammte Physiologie, vol. 54. Mall, F. P. 1908 Astudy of the causes underlying the origin of human monsters.
Jour. Morph., vol. 19. McClexdox, J. F. 1912 The effects of alkaloids on the development of fish
(Fundulus) eggs. Am. Jour. Physiol., vol. 31. MoRG.\x, T. H. 1902 The relation between normal and abnormal development
of the embryo of the frog as determined by injury to the yolk portion
of the egg. I and II. Arch. f. Entw.-Mech., Bd. 15.
1904 The relation between normal and abnormal development of
the embr3^o of the frog (III) as determined by some abnormal forms
of development. Arch. f. Entw. .Mech., Bd. IS. Roux, \V. 1895 Gesammelte Abhandlungen zur Entwicklungsmechanik der
Organismen, Bd. 2. Speeman, H. 1912 Uber die Entwicklung umgedrehter Hirnteile bei Ainphi bienembryonen. Zool. .lahrb., Suppl. 15. (Festschrift flir J. \Y.
Spengel, Bd. 3). Stock ARD, C. R. 1907 The artificial production of a single median cyclopean
eye in the fish embryo by means of sea-water solutions of magnesium
chloride. Arch. f. Entw.-Mech., Bd. 23.
1909 The development of artificially produced cyclopean fish, "The magnesium embryo." Jour. Exp. Zool., vol. 6.
1910 a The influence of alcohol and other anesthetics on embryonic development. -Am. Jour. Anat., vol. 10.
1910 b The experimental production of various eye abnormalities and an analj'sis of the development of the primary parts of the eye. Arch. f. vergl. Ophthalmol., Bd. 1.
1913 An experimental study of the position of the optic anlage in Amblystoma punctatum, with a discussion of certain eye defects. Am. Jour. Anat., vol. 15.
1915 An experimental study of the origin of blood and vascular endotheliimi in the teleost embr\'o. Proc. Am. Assn. Anatomists, Thirty-first Session, Anat. Rec, vol. 9, no. 1, pp. 124-127. Smith, A. J. axd Birmixgham, A. 1889 .\bsent thoracic duct causing oedema ot a fetus. Jour. Anat. and Phvsiol., vol. 23.
THE DEVELOPMENT OF THE LYMPHATIC SYSTEM IN THE LIGHT OF THE MORE RECENT INVESTIGATIONS IN THE FIELD OF VASCULOGENESIS
CHARLES F. W. McCLURE
From the Department of Cotnparative Anatomy, Princeton University
Does the endothelium of the lymphatic system arise, at any time or place, in a discontinuous manner and independently of that of the veins? As we shall see, the determination of this question constitutes a solution of the lymphatic problem.
The view that lymphatic endothelium spreads continuously and uninterruptedly throughout the body of the embryo from the endothelium of the veins, is merely an extension, and application to the endothelium of the lymphatic system, of the wellknown view held by His, that the endothelium of the intraembryonic haemal vessels grows continuously and uninterruptedly into the embryo from the yolk-sac angioblast. Such a method of origin necessarily implies that all intra-embryonic endothelium arises only from a pre-existing endothelium which takes its origin in the yolk-sac, and that in the body of the embryo a discontinuity of origin never occurs.
The view opposed to the 'ingrowth' or 'angioblast' theory of His has been closely associated with the names of Ruckert and Mollier (1) . This view consists in the claim that the endothelium of the intra-embryonic haemal vessels develops in situ in the body of the embryo, and that it is not derived from the yolksac angioblast.
Since the Ij^mphatics merely represent a component part of a general vascular system, to which the haemal vessels also belong, the probability at least, is that, in the genesis of their endothelium, and in the establishment of a continuous system of vessels, the lymphatic and haemal vessels should follow a common genetic
563
THE ANATOMICAL RECORD, VOL. 9, NO. 7
564 CHARLES F. W. McCLURE
plan. Let us consider, in the light of the more recent investigations in the field of the vascular system, what this plan may be.
It is not the purpose of the present paper to give a review of the investigations of those who have consistently maintained a local origin for the endothelium of the intra-embryonic haemal vessels. It is only necessarj^ to refer to the more recent and excellent paper by Schulte in which such a review and critical analysis of their work is given.
The investigations of Schulte (2) on the mammalian embryo have shown in particular, that the yolk-sac angioblast cannot possibly aid in forming the endothelium of the umbilical veins; Schulte has also demonstrated in a most convincing manner, that the endothelium of other main intra-embryonic haemal vessels develops in situ from mesenchymal cells.
It should be clearly borne in mind that, until quite recently, the investigations which have dealt with the origin of intraembryonic endothelium have not been experimental in character, but have been based largely upon a study of fixed material in which, however, a local and discontinuous origin of bloodvascular anlagen has been observed.
Let us now see how the view that the endothelium of the intraembryonic blood-vascular system develops in situ, and does not grow into the embryo from the yolk-sac, has been borne out by experiment.
Two types of experiment have thus far been made to determine this question: (1) The partial separation of the embryo from the yolk-sac, or, the complete separation and isolation of a portion of the embryo from the rest of the embryo and from the ^olk-sac, at a time prior to the possible invasion of the embryonic axis by the yolk-sac angioblast; (2) by observing the effects produced on the developing blood-vascular system in embryos which have been allowed to develop under the influence of anesthetics or other chemical agents.
The experimental investigations of Hahn (3), and Miller and McWhorter (4) have shown, bj^ effecting a separation on one side between the body of the chick embryo and the yolk-sac, before vessels have appeared in the area pellucida, that blood
DEVELOPMENT OF THE LYMPHATIC SYSTEM 565
vessels make their appearance in the body of the embrj'o in a typical manner on the operated side. These vessels differ from those on the unoperated side only in size and rate of development, differences which may be correlated with their reduced drainage area and the consequent diminished quantity of circulatory fluid.
These experiments of Hahn, and ]\Iiller and McWhorter have conclusively shown that the yolk-sac angioblast cannot have grown into the embryo on the operated side. In order to eliminate the possibility, however, that the vessels on the operated side may not have been formed in situ, but by an invasion of angioblast from the normal unoperated side, Reagan (5) has recently completed a set of experiments in my laboratory which conclusively disprove this contention. Instead of separating only one side of the embryo from the yolk-sac, Reagan has been able to develop the heads of chick embryos, which had been completely separated from the rest of the embryo and from the yolk-sac, and in which endothelial lined vascular channels of mesenchymal origin invariably appear. As in the case of the experiments of Aliller and AlcWhorter, the operations were performed at a time before it would have been possible for the intra-embryonic tissue to have been invaded by yolk-sac angioblast.
Graper (6), under the direction of C. Rabl, performed a set of experiments on chick embryos, somewhat similar to those of Hahn, and Miller and McAATiorter, and, although he noted the presence of independent blood-islands in the body of the embryo, . he was unable to interpret them as having been formed in situ.
Jacques Loeb (7) was the first to observe the effects produced by certain chemicals (XaCN) on the developing blood vessels in fish embryos. He was able to produce a condition in which a beating heart and blood were present, but no circulation; a condition which, as stated by Schulte, can hardly be reconciled with the doctrine that the vessels of the embryo have a primitive continuity of lumen with those of the yolk-sac, for it is inconceivable that in such circumstances, a beating heart could fail to effect a circulation.
566 CHARLES F. "\V. McCLURE
The investigations of Stockard (8) supplement and coincide with those of Hahn, Miller and McWhorter, Reagan, and Loeb in a most decisive manner. Stockard has shown that, not only do anesthetics arrest the development of the intra-embryonic blood vessels in the embryos of Fundulus, at an earlj'- ontogenetic stage, but in such a manner that no doubt can now exist that, under normal conditions, these vessels are formed iJi situ by a concrescence of independent and discontinuous anlagen, and that their endothelium is derived directly from mesenchymal cells. It is interesting to note in this connection that Wenkebach (9) had already observed in the body and yolk-sac of the living fish embryo (Belone longirostris) , that mesenchymal cells play an important role in the formation of vessels and sprouts. In their general features the observations of Wenckebach have been confirmed by Raffaele (10).
It is thus seen that experimentation bears out the observations m^ade upon fixed and living material, that the intra-embryonic blood-vascular channels do not grow into the embryo from the yolk-sac, but are formed in situ by a concrescence of independent and discontinuous anlagen, whose endothelium is formed from intra-embryonic mesenchymal cells.
The vascular plexus formed in the extra-embryonic area of the vertebrate embryo, is as we know, at first represented by discontinuous, independent and circumscribed anlagen, the cells of which possess a local origin. Clefts or spaces, the future lumina of the plexus, soon make their appearance in a discontinuous manner amongst the cells of these anlagen, and it is by a concrescence of these vascular spaces that a continuous system of vascular lumina is finally formed. The cells which constitute the walls of these vascular spaces become transformed into the endothelium and, when blood-islands are present, the more centrally situated cells form the primary blood cells. It is interesting to note in this connection that McAVhorter and Whipple (11) have recently been able to demonstrate and record photographically the concrescence of separate vascular anlagen in the area pellucida of the chick's blastoderm in vitro.
DEVELOPMENT OF THE LYMPHATIC SYSTEM 567
If we compare the development of the intra-embiyonic bloodvascular channels, as determined b}' observation and experiment, with that of the plexus which arises on the j^olk-sac, we find, in the genesis of their endothelium from mesenchyme, and in their formation by a concresence of independent anlagen, that the intraand extra-embryonic blood-vascular channels follow exactly the same genetic plan.
If one attempted to follow the development of these intraor extra-embryonic blood-vascular channels by means of injections, it is evident that this method would reveal only the extent to w^hich a continuous system of injectible lumina had been established at the time the injections were made. It would fail completely to reveal the facts which have been definitely determined by experiment, that the injectible lumina had been previously formed by a concrescence of independent and uninjectible vascular'spaces, and that the endothelium which forms the walls of these lumina had been formed in situ, not from a pre-existing endothelium, but from mesenchymal cells.
Since we now know that the intra-embryonic blood vessels, like those in the yolk-sac, are formed by a concrescence of independent anlagen, and that their endothelium is formed in situ from mesenchymal cells, the question naturally confronts us as to the method by means of which these independent anlagen become connected with one another to form a system of vessels with continuous lumina, that extend throughout the body of the embryo.
There appear to be only three possible methods by means of which such connections could take place: (1) Either by means of a proliferation or migration of the cells of which the original independent anlagen are composed; (2) by a further local in situ differentiation into endothelium of the embryonic cells which intervene between the independent anlagen; or (3) by a combination of these two methods.
We all recognize the fact the endothehum, like other tissues of the body, is capable of growth after it has once been formed. In no other manner could we account for the increase in size
568 CHARLES F. W. McCLURE
which blood vessels undergo in the embryo after they have attained their adult structure and form. It is also possible for anastomoses to be formed between different blood vessels by means of a growth or sprouting of their endothelial walls, so that, in some cases, an increase in their extent, through growth, may actually take place. It is therefore quite probable that growth may play a considerable role in estabhshing a concrescence between the independent endothelial-hned anlagen of the bloodvascular system. From whatever standpoint it may be considered, however, the growth of an endothelium is a feature of secondary significance as regards the problem at hand, since the main question at issue does not concern the possibility that endothelium may or may not grow, but rather how the endothelium is formed that does the growing.
The distinction between the actual genesis of endothehum and the growth it may undergo after it has once been formed is naturally one that has been disregarded by those who maintain that intra-embryonic vascular endothelium is not directly a product of mesenchjixial cells. A special specificitj' has therefore been attributed by the supporters of the 'angioblast' theory to the endothelium of the intra-embryonic vascular system, on the ground that it takes its origin only from the yolk-sac angioblast. In accordance with this view, it is by means of one continuous and uninterrupted growth of a pre-existing endothehum (yolk-sac angioblast) throughout the bod}' of the embryo, that the endothelium of the blood-vascular and Ij^mphatic systems is formed.
Since the 'angioblast' theory of His no longer holds, the question of the specificity of tissues is involved in the vascular problem only to the same extent as is the case for any other tissue in the body. Whether the mesenchymal cells of the embryo are in an embryonic or undifferentiated state, and capable of further differentiation into cells which form muscle, connective tissue, endothehum, etc., is entirely beside the question; provided we know that the product of these intra-embryonic mesenchymal cells actually forms endothelium and that the latter is not derived from the 3'olk-sac angioblast. Also, the question con
DEVELOPMENT OF THE LYMPHATIC SYSTEM 569
cerning the origin of these mesenchymal cells, whether derived from entoderm, mesoderm or mesothelium, does not concern us here. The main point at issue is the establishment of the fact that the endothelium of the intra-embryonic haemal vessels is the product of a local in situ differentiation of certain cells in the embryo which have not been derived from the yolk-sac angioblast.
Let us now compare these conditions of the intra-embryonic blood-vascular system, as determined by sections and experiment, with those of those of the embryonic lymphatic system.
Our knowledge of the embryonic lymphatic system is gradually approaching a state where, in such forms as teleosts and am^ phibia, it may also be possible to determine by experiment how the IjTuphatic system is formed. A thorough knowledge of the lymphatic channels and the order of their appearance in the normal embryo would be quite essential, however, before experiment could be successfully applied. Since the anlagen of the lymphatics do not make their appearance in the embryo under normal conditions until after the veins have been established and have begun to function, it is quite possible, in cases of arrested development of the venous system, as demonstrated by Stockard in Fundulus, that development might never be successfully carried to the lymphatic stage. Be this as it may, until the problem has been tested by experiment, our knowledge and interpretation of lymphatic development must, for the present, be based upon the observation of fixed and of living material, and its comparison with the known developmental 'stages of the blood-vascular system, as observed in fixed and in living material, and as verified by experimental means. If it can be shown that the anlagen of the lymphatic system present exactly the same conditions in fixed and in living material, as those of the blood-vascular system, it is reasonable to infer that in their development the lymphatic and blood-vascular systems follow exactly the same genetic plan. // one were to observe that in certain cases intra-embryonic blood vessels were formed in the living embryo by a sprouting or growth of a pre-existing endothelium, would he now be justified in claiming that all of the remaining
570 CHARLES F. W. McCLURE
blood vessels of the embryo ivere formed in the same manner? In view of the fact that we now know that intra-embryonic blood vessels are not all formed in this manner, it would seem that a similar interpretation might also apply to the lymphatics.
Whatever else the case may be, in view of the above-mentioned experimental investigations of Hahn, Miller and McWhorter, Reagan, and Stockard, it can now be definitely stated that the endothelium of the lymphatic system is neither directly nor indirectly a product of the yolk-sac angioblast. Such being the case, it must either arise in situ, hke the endothelium of the intra-embryonic veins, from cells other than from a pre-existing endothelium; or, be a product entirely of the endothelium of the veins. If the former case be true, the endothelium of the lymphatic system should present exactly the same independent and discontinuous method of origin in the embryo as that of the extraand intra-embryonic haemal vessels; and, if the development of the lymphatics were followed by the injection method, the same restrictions as regards the injectibility of its independent anlagen should also necessarily apply. On the other hand, if the lymphatic system is entirely a product of the endothelium of the veins, its origin from mesenchyme should naturally never occur. As a viatter of fact, since intra-embryonic vascular endothelium has been shown by experiment to be a local product of mesenchyme, there now remains no valid reason or significance in the claim, as regards its specificity, that lymphatic endothelium is solely a product of that of the veins.
Let us "examine the evidence at hand and see whether the endothelium of the lymphatics, like that of the haemal vessels, develops in situ in the mesenchyme, or whether it forms an exception to that of the haemal vessels, and sprouts continuously and uninterruptedly throughout the body of the embryo from an endothelium already formed.
It is not the purpose of the present discussion to give an historical review of the literature bearing upon the development of the lymphatic system but merely, on the basis of comparison, to call attention to the evidence in favor of the view that the lymphatics, like the haemal vessels, are formed by a concres
DEVELOPMENT OF THE LYMPHATIC SYSTEM 571
cence of independent and discontinuous anlagen, and that their endotheUum arises in situ from intra-embryonic mesenchymal cells.
A principal contention of Huntington and McClure (12) regarding the development of the lymphatic system has been that its anlagen arise independently and discontinuously in the embryo, and that its endothelium does not spread continuously and uninterruptedly throughout the body from the endothehum of the veins. We have repeatedh^ shown that the lumina of the lymphatics are formed by a concrescence of discontinuous and independent lymph vesicles or lymph spaces, and that the cells which constitute the walls of these spaces are derived in situ from mesenchyme and not from the endothelium of the veins. In the early stages of our investigations we laid especial stress upon a plan of development for the lymphatic system of mammals which we described under the name of the 'extraintimaP theorj^ of Ijanphatic development, and which may be briefly described as follows: The development of the thoracic ducts (13) and mesenteric (14) l\Tiiphatics in the cat is correlated with the degeneration of certain venous channels, many of which are tributaries of the azygos division of the supracardinal veins (15). A series of independent lymph spaces arise discontinuously in the mesenchyme external to the intimal Hning of these degenerating vessels and, as these lymph spaces gradually become concrescent to form continuous channels, the latter, following a line of least resistance, utilize the static line vacated by these degenerating veins. In this manner certain of the main lymph channels of the mammalian embryo follow the course of and finally occupy completely the territory formerly occupied by veins. This principle of extraintimal replacement of abandoned venous channels by lymphatics accounts for the sinistral drainage plan finally assumed by the thoracic duct system in the embryo of the cat. The cranial or azygos division of the left thoracic duct of the embrj^^o persists as the main line of drainage in the adult, in correlation with a degeneration in the embryo of the left supracardinal (left azygos) and left postcardinal veins and the left duct of Cuvier.
572 CHARLES F. W. McCLURE
It is evident and appears clearly in our earlier publications, that the fundamental plan of development followed by these replacing lymph channels does not depart from that followed by other channels, either in mammals or in any other vertebrates where the development of the lymphatics is unaccompanied by the replacement of degenerating veins. Where, as in the case of the trout, lymph channels do not develop along the course of degenerating veins, an extraintimal replacement of a degenerating vein by a lymphatic necessarily does not occur. It is therefore plain that the extraintimal replacement, as described by us, possesses only a mechanical significance, and is merely an adaptation of a common plan of lymphatic genesis, through the concresence of independent anlagen, to the local conditions which prevail only in certain districts of the mamrnalian embryo.
The same general plan of development as outlined above by Huntington and McClure for the lymphatic system of the cat has also been found by Kampmeier (16) to occur in the embryo of the pig. His description of the independent and discontinuous anlagen of the thoracic ducts which he found in the injected pig embryo loaned him by Professor Sabin, needs no further comment.
F. T. Lewis (17) has described the presence of a chain of discontinuous 'lymphatic spaces' (endothelial-lined anlagen) in the rabbit embryo which lie along the azygos veins in the path of the future thoracic duct. He regards these anlagen, however, as having been detached from the veins. Concerning these multiple anlagen of Lewis, Sabin (18) has stated as follows:
Since these spaces are lined with a definite endothelium, they form a much more serious obstacle to the theory of growth of the lymphatics from the endothelium of the veins than the more indefinite spaces to be found in earlier embryos, and I cannot but think that if these multiple endothelial-lined isolated spaces do exist along the veins in later stages, they would form serious evidence against the theory of the origin of the lymphatics from the veins. Or at least if the lymphatics, in their growth, do pick up isolated endothehal-lined spaces, we shall again be left without a clue as to the origin of the lymphatic system.
It is significant to note that, although Sabin considers these isolated endothelial-lined anlagen of Lewis as having been
DEVELOPMENT OF THE LYMPHATIC SYSTEM 573
detached from the veins, she nevertheless now recognizes their existence in the pig embryo, and regards them as entering into the formation of the thoracic duct (19, 1911, p. 424).
The point I wish to emphasize in this connection is that Sabin now recognizes the fact that lymphatics may be formed by the concrescence of multiple and independent endothelial-lined anlagen and that she has thus far presented no valid evidence that these anlagen have been detached from the veins.
Sala (20) and more recently Miller (21) have shown that the thoracic ducts of the common fowl are formed by a concrescence of independent and discontinuous lymph spaces and that their endothelium is formed in situ from cells other than those which constitute the endothelium of the veins. ^Miller has further made the important discovery that groups of blood cells develop in the mesenchyme along the line of the thoracic ducts and that the latter subsequentlj^ convey these blood cells to the venous circulation. We therefore find that hematopoiesis may actually occur in connection with the development of certain lymph /channels, a condition which Huntington (22) has also recently verified for certain l}^iiphatics of mammals.
West (23), by a studj" of injected and uninjected embryos, has recently found that the posterior lymph heart of the conmion fowl develops in the mesenchyme and secondarily establishes a connection with the veins. He has also found that hematopoiesis occurs in the mesenchjmie in relation to the independent anlagen of the lymph heart, and states that the blood ceUs thus formed are not to be confounded with those which may later back into the lymph heart from the veins (see E. L. Clark, Anat. Rec, vol. 6).
Huntington (24), in a paper on the development of the lymphatic system in reptiles (chelonia,. lacertiUa), has shown that the systemic lymphatics develop in the mesenchyme independenth^ of the endothehum of the veins. A particular feature of his investigation is that he was able to demonstrate that the periaortic lymph channel in Chel^'dra serpentina arises in the mesenchyme in an area entirely free of veins.
574 CHARLES F. W. McCLURE
Stromsten's (25) investigations on the development of the lymphatic system in reptiles (chelonia) have led him to conclude that the lymphatics are formed by a concrescence of independent and discontinuous lymph spaces and that lymphatic endothelium is formed m situ from mesenchymal cells. His observations were based largely upon a study of injected embryos, and showed that the injecta did not reach the independent anlagen of the lymphatics, prior to their concrescence with one another to form a system of continuous lumina which had established a communication with the veins.
Fedorowicz (26) and more recently Kampmeier (27), have shown that the continuous lumina of certain lymphatics are formed in the amphibia (anura) by the concrescence of originally independent and discontinuous lumina. They conclude, however, that the endothelial walls of these lumina have been derived from the endothelium of the veins. E. R. Clark (28) regards the lumina of the developing lymphatics in amphibia (in tail of larval amphibia) as always continuous and capable of injection, while Fedorowicz and Kampmeier describe them as discontinuous at the start.
The writer (29) has demonstrated the presence of discontinuous and independent lymph vesicles in the trout embryo, which cannot be injected from other lymphatic anlagen or from the veins. Many of these vesicles arise in the mesenchyme remote from the veins, and no connection can be observed between their endothelium and that of the veins or that of any other lymphatic anlagen. One of these independent lymph vesicles, the subocular lymph sac, can actually be observed in the living trout embryo. On account of the relatively large size of this vesicle it has proved a most favorable object for experimentation and for study in sections, not only in proof of the fact that it arises independently of the veins, but also that its endothelium is of mesenchymal origin.
Allen f30) has investigated the development of the lymphatic system in Polistotrema (Bdellostoma) stouti and speaks of the lymphatics of fishes as veno-l^^mphatics. He states: I expect to show that the main factor in the' construction of the veno
DEVELOPMENT OF THE LYMPHATIC SYSTEM 575
lymphatic system is the same as was described for the caudal h'mph hearts, namely, the formation and union of certain mesenchymal spaces."
Allen, independently of ]Miller, has also observed that an active hematopoiesis occurs in the mesenchyme in relation to the developing caudal h^mph hearts of Polistotrema.
Except for differences of opinion regarding the origin of lymphatic endothelium, it may be observed that the above-mentioned investigators agree for the most part that the continuous lumina of the lymphatics, like those of the haemal vessels, are formed bj^ a concrescence of independent and discontinuous anlagen.
If we disregard entirely the personal equation which may have influenced any or all of the above-m.entioned investigators to interpret their findings in accordance \\ath one or the other view, the fact still remains that the anlagen of the lymphatic system, as observed in sections of injected and uninjected embryos, have been found to be identical with those of the intra-embryonic blood-vascular system, which has been shown by sections and experiment to be formed by a concrescence of independent anlagen, and its endothelium to be formed in situ from mesenchymal cells. Since we possess exactly the same kind of evidence in favor of the mesenchymal origin of lymphatic endothelium, as ive formerly did for that of the intra-embryonic haemal channels, before experiment urns applied, it therefore seems highly improbable that the endothelium of two sets of similarly appearing anlagen, belonging to the same general organ system and developing side by side, should differ in its mode of origin, rather thanfolloiv a common genetic plan .
■ We know that the lymphatics, both in the embryo and in the adult, establish a permanent communication with the veins at tj^pical points (31). The question therefore arises, what role, if any, may be played by the veins in establishing such communications with the independently formed h^mphatics? In the development of the general vascular sj^stem which includes the arteries, veins and lymphatics, the end result desired is the formation of a connected system of vessels which subserve definite functions in the economy of the general vascular system.
576 CHARLES F. W. McCLURE
If the general vascular system develops progressively in a uniform manner, by a concrescence of independently formed anlagen, and the lymphatics, as is the case, form the last link in completing the chain, it is evident that the same factors should account for the establishment of a connection between the veins and the independently formed lymphatics as between the independently formed anlagen of which the veins and Ij^mphatics are originally composed. Such a connection could alone be established, either by a growth or sprouting of the endothelium of the veins; by means of a further in situ differentiation into endothelium of the mesenchymal cells which intervene between veins and the independent anlagen of the IjTnphatics ; or, both of these factors might be involved. If a connection should be established by a sprouting of the endothelium of the veins, such sprouts would possess no significance beyond the fact which we all recognize, that all vascular endothelium is capable of growth after it has once been formed. The question, however, is not whether endothelium is capable of growth, but rather what are its limitations and how is the endotheUum formed that does the growing. It is therefore evident, if, at the points at which the lymphatics establish permanent communications with the veins, venous endotheUum should contribute to the formation of the lymphatics, it would play only a subsidiary role, and serve only as a means, in common with the endothelium of other independently formed vascular anlagen, of bringing two independently formed portions of the vascular system into communication with each other to form a continuous system of channels. Such veno-lymphatic connections have been hitherto described by Huntington and jMcClure (32) under the name of S^enolymphatics.' It. is evident, however, that these 'venolymphatics' would not differ, in any sense, from other connections, where a similar growth of endothelium is concerned, when made for the purpose of establishing a connection between any of the other independently formed anlagen of the general vascular system.
To sum up : The development of the general vascular system — haemal and lymphatic vessels — is a uniform process, which consists in a local origin (genesis) of endothelium from mesenchy
DEVELOPMENT OF THE LYMPHATIC SYSTEM 577
mal cells and a growth of endothelium after it has once been formed.
In view of what has been said above, it would therefore appear that the lyraphatic problem, in its broadest sense, should not be interpreted in terms either of a venous or non-venous origin, but rather in terms of the uniform phases of genesis and growth which may characterize the establishment of vascular channels in general.
LITERATURE CITED
(1) RtJCKERT, J., AND IMoLLiER, S. 1906 Jlandbuch d. vergl. und exper.
Entwickelungslehre d. Wirbeltiere, Bd. 1, Teil 1, zweite halfte.
(2) ScHULTE, H. vox W. 1914 Early stages of vasculogenesis in the cat (Felis
domestica) with especial reference to the mesenchjTnal origin of endothelium. Memoirs of The Wistar Institute of Anatomy and Biology, No. 3.
(3) Hahn, H. 1909 Experimentelle Studien iiber die Entstehung des Blutes
und der ersten Gefasse beim Hiinchen. I teil. Intraembryonale Gefasse. Arch. Entwickmech., Bd. .27.
(4) JMiLLER, A., AND INIcWhorter, J. 1914 Experiments on the development
of blood vessels in the area pellucida and embrj'onic body of the chick. Anat. Rec, vol. 8.
(5) Reagan, F. P. 1915 Vascularization phenomena in fragments of embry onic bodies completely isolated from yolk-sac blastoderm. Anat. Rec, vol. 9, no. 4.
(6) Graper, L. 1907 Untersuchung liber die Herzbildung der Vogel. Archiv.
Entwickmech., Bd. 24.
(7) Loeb, J. 1912 The mechanistic conception of life. Popular Science
JMonthly, vol. 80.
(8) Stockahd, C. R. 1915 An experimental studj' of the origin of blood and
vascular endothelium in the Teleost embryo. Proc. Amer. Ass. Anat., Anat. Rec, vol. 9, no. 1.
(9) Wenckebech, K. F. 1886 Beitrage zur Entwicklungsgeschichte der
Knochenfische. Arch. mikr. Anat., Bd. 28.
(10) Raffaele, F. 1892 Ricerche sullo sviluppo del sistema vascolare nei
Selacei. Mitt. Zool. Stat. Neapel., Bd. 10.
(11) McWhorter, J. E., AND Whipple, A. O. 1912 The development of the
blastoderm of the chick in vitro. Anat. Rec, vol. 6.
(12) Huntington, G. S., and McClure, C. F. W. 1907 The development of
the main lymph channels of the cat in their relation to the venous system. Anat. Rec, vol. 1.
(13) Huntington, G. S. 1911 The anatomy and development of the systemic
lymphatic vessels in the domestic cat. Memoirs of The Wistar Institute of Anatomy and Biology, No. 1. Also: Ueber die Histogenese des lymphatischen Systems beim Saugerembryo. Anat. Anz., Erganz. z. Bd. 37, Verb. 2, internat. Anat. Kongr., Briissel, 1910.
578 CHARLES F. W. McCLURE
(14) McCLrRE, C. F. W. 19i0 The extra-intimal theory and the development
of the mesenteric lymphatics in the domestic cat. Anat. Anz., Erganz. z. Bd. 37, Verh. 2, internat. Anat. Kongr., Briissel. Also: A few remarks relative to Mr. Kampmeier's paper on the value of the injection method in the study of lymphatic development. Anat. Rec, vol. 6, 1912.
(15) Huntington, G. S., and McClure, C. F. W. 1907 Development of
postcava and tributaries in the domestic cat. Am. Jour. Anat., vol. 6, Abstr. Anat. Rec, vol. 1.
(16) Kampmeier, O. F. 1912 The value of the injection method in the study of
Ij^mphatic development. Anat. Rec, vol. 6. Also: The development of the thoracic duct in the pig. Am. Jour. Anat., vol. 13, 1912.
(17) Lewis, F. T. 1905 The development of the lymphatic system in rabbits.
Am. Jour. Anat., vol. -5.
(18) Sabin, F. R. 1908 Further evidence on the origin of the lymphatic endo thelium from the endothelium of the blood vascular sj'stem. Anat. Rec, vol. 2.
(19) Sabin, F. R. 1911 A critical study of the evidence presented in scA'eral
recent articles on the development of the lymphatic system. Anat. Rec, vol. 5. Also: Der Ursprung und die Entwickelung des Lymphgefasssystems. Ergebnisseder AnatomieundEntwickelungsgeschichte, 1913. Also: The originand the development of the lymphatic system. The Johns Hopkins Hospital Reports Monographs, New Series, No. 5, 1913.
(20) Sala, L. 1900 Sullo sviluppo dei ctiori linfatici e dei dotti toracici nell
embriojie di polio. Ricerche fatta nel Laborat. di Anat. norm, della R. Univ. di Roma, vol. 7.
(21) Miller, A. M. 1913 Histogenesis and morphogenesis of the thoracic
duct in the chick; development of blood cells and their passage to the blood stream via the thoracic duct. Am. Jour. Anat., vol. 15.
(22) Huntington, G. S. 1914 The development of the mammalian jugular
lymph sac, of the tributary primitive ulnar lymphatic and of the thoracic ducts from the viewpoint of recent investigations of vertebrate lymphatic ontogeny, together with a consideration of the genetic relations of lymphatic and haemal vascular channels in the embrj'oe of amniotes. Am. Jour. Anat., vol. 16.
(23) West, R. 1914 The origin and early deA^elopment of the posterior lymph
heart in the chick. Proc Amer. Ass. Anat., Anat. Rec, vol. 8. Also: The origin and development of the posterior lymph heart in the chick. Am. Jour. Anat., vol. 17, no. 4, 1915.
(24) Huntington, G. S. 1911 The development of the lynii)hatic system in
reptiles. Anat. Rec, vol. 5.
(25) Stromsten, F. A. 1910 A contribution to the anatomy and development
of the posterior lymph hearts of turtles. Publication 132 of the Carnegie Institution of Washington. Also: On the relations between the mesenchymal spaces and the development of the posterior lymph hearts of turtles. Anat. Rec, vol. 5, 1911. Also: On the development of the prevertebral (thoracic) duct in turtles as indicated by a study of injected ;nid uninjected embryos. Anat. Rec, vol. G, 1912.
DEVELOPMENT OF THE LYMPHATIC SYSTEM 579
(26) Fedorowicz, S. 1913 Untersuchungen liber die Entwicklixng der Lymph gefasse bei Anurenlarven. Vorlaufige Mitteilung. Extrait dii Bulletin de I'Academie des Sciences de Cracovie, June.
(27) Kampmeier, O. F. 1915 On the origin of lymphatics in Bufo. Am. Jour.
Anat., vol. 17.
(28) Clark, E. R. 1911 An examination of the methods used in the study
of the development of the lymphatic system. Anat. Rec, vol. 5. Also: Further observations on living growing lymphatics; their relation to the mesenchyme cells. Am. Jour. Anat., vol. 13, 1912.
(29) McClure, C. F. W. 1913 The development of the lymphatic system in
fishes. Proc. 17th Internal. Cong, of Medicine, London. Also: The development of the lymphatic system in the trout. Anat. Rec, vol. S, 1914. Also: On the provisional arrangement of the embryonic h-mphatic system. (An arrangement by means of which a centripetal l>Tnph flow toward the venous circulation is conlrolled and regulated in an orderly manner, from the time lymph begins to collect in the intercellular spaces, until it is forwarded to the venous circulation). Anat. Rec, vol. 9, no. 4, 1915. Also: The development of the lymphatic system in fishes with especial reference to its development in the trout. ^lemoirs of The Wistar Institute of Anatomy and Biology, No. 4, 1915. (In press).
(30) Allen, W. F. 1913 Studies on the development of the veno-lymphatics
in the tail-region of Polistotrema (Bdellostoma) stouti. Quart. Jour. Micr. Sci., vol. 59.
(31) Silvester, C. F. 1912 On the presence of permanent communications
between the lymphatic and venous system at the level of the renal veins in adult South American monkeys. Am. Jour. Anat., vol. 12. Also: iMcClure and Silvester A comparative study of the Ivmphaticovenous commimications in adult mammals. Anat. Rec, vol. 3, 1909.
(32) Huntington, G. S., and McClure, C. F. W. 1910 The atiatomy and
development of the jugular lymph sacs in the domestic cat (Felis domestica). Am. Jour. Anat., vol. 10.
THE ANATOMICAL RECORD. VOL. 9, NO. 7
BOOKS RECEIVED
The receipt of publications that may be sent to any of the five biological journals published by The Wistar Institute will be acknowledged under this heading. Short reviews of books that are of special interest to a large number of biologists will be published in this journal from time to time.
THE INVESTIGATION OF MIND IN ANIMALS, E. i\I. Smith, Moral Sciences Tripos, Cambridge, 194 pages, 9 illustrations, bibliography and index. Cambridge, England, at the University Press, 1915.
Preface. There are few people who cannot relate some apparently striking instance of animal intelligence; the majority of such cases, however, will not stand critical examination. The science which has for its object the systematic investigation of the brute mind is Animal Psychology, and it would seem that the methods of this youthful discipline are still unknown to many, even among those who profess an interest in animal conduct. It is, then, with the purpose of presenting a brief account of the modes of procedure employed by Animal Psychology, its aims, trend, and the general nature of the results hitherto obtained, that this little book has been written. In a work of this character discussion and controversy would have been out of place, so the treatment has been confined as far as possible to description and illustration; at the same time attention has been drawn to some of the chief difficulties inherent in the inquiry. A complete and exhaustive presentation of facts was, of course, out of the question, and much that is of interest and importance has had, inevitably, to be omitted; but it is to be hoped that the interested reader of leisure will refer to some, at least, of the original articles mentioned in the bibliographj-, nearly all of which will be found to contain further references.
Under the headings Protozoan behavior; retentiveness: habit-formation, associative memory and sensorj- discrimination, instinct, homing, imitation, and the evidence for intelligence and for ideas, the author gives a most readable and informing account of the newer work on Animal Behavior. The book is written as a sketch and that plan of treatment is followed perfectly — with no lapses into detail and with a happj' exclusion of technicalities.
The author's purpose is to present a review of the investigations in the field of animal behavior and to point out the main views which are now current. Concrete illustrations of experimental results are given, conclusions are weighed and matters calling for further study indicated. The needs of both the layman and the behaviorist are met by this essay. H. H. D.
THE CLINICAL ANATOMY OF THE GASTRO-INTESTINAL TRACf , T. Wingate Todd. ]\I.B., Ch.B., F.R.C.S. (Eng.), Henry Willson Payne Professor of Anatomy in the Western Reserve University, Cleveland, Ohio; late Lecturer in Anatomy, University of Manchester, 264 pages,' 32 illustrations. Index of authors quoted and a subject Index, ^Lanchester at the University Press. Longmans, Green & Co., London and New York, 1915. $1.75.
580
THE SOUND-TRANSMITTING APPARATUS IN NECTURUS
H. D. REED
From the Zoological Laboratory, Cornell University
SIX FIGURES
The sound-transmitting apparatus of Necturus having been the subject of so much investigation, further comment would seem unnecessary. The results gained from the study of this apparatus in other groups of urodeles, however, and the uniform conclusions of Kingsbury ('05) and Norris ('11) and the significance of Herrick's ('14) and Brunner's ('14) observations bearing upon the rank of Necturus among tailed amphibia, demand a further investigation of the subject in order to determine the precise origin of every portion of the apparatus. Further studj^ seems especially important when coupling the evident neotitic status of this species with conclusions which accordingly might be drawn regarding the nature of the fenestral elements.
In Necturus the sound-transmitting apparatus is composed of a single plate that accurately fills the somewhat elliptical foramen vestibuli of the mature animal and is connected by a well-defined stilus with the suspensorium of the jaws. Any relation with other extraotic elements is wanting entirely.
In typical urodeles, as for example the Amblystomidae, this apparatus is a double structure (Kingsbury and Reed '08). It is composed of a cephalic piece, the columella, and a caudal portion, the operculum. Each element is distinct in its origin and they are, therefore, without morphologic relations. Of the two elements the columella is the first to appear and develops wholl}^ outside and independent of the ear capsule, only secondarily coming to lie against the membrane of the fenestra. It is essentially a flat plate connected with the suspensorium by a stilus. The columella becomes definitive in larval life and is to
581
582 H. D. REED
be considered as the more primitive of the fenestral elements. Upon the assumption of a terrestrial existence, the columella apparently becomes functionless and fuses wholly or in part with the ear capsule, while a new element, the operculum, becomes cut out from the walls of the capsule just caudad of the primary fenestra. The operculum bears a perilymphatic prominence from which the M. opercularis extends to the shoulder girdle. It is important to note that these elements are not only different in origin, but each possesses distinctive skeletal relations: the columella with the suspensorium, the operculum with the shoulder girdle.
As mentioned above, the sound-transmitting apparatus of Xecturus is composed of a single plate which possesses the anatomical relations of the columella of the amblystomid forms. In the Plethodontidae the sound-transmitting apparatus is in the form of a single plate filling the fenestra, but unlike that of Xecturus, the plate in this family possesses the anatomical relations of both columella and operculum; that is, the cephalic end is connected with the suspensorium of the jaws by a stilus, while the caudal end comes into relation with the shoulder girdle through the M. opercularis. Development shows that the morphology of the plate is in harmony with these conditions. It is a double structure; the columellar portion arises outside the ear capsule, while the opercular portion is formed from otic cells which unite during development with the columellar element, producing a single definitive plate, the components of which differ in their morphologic nature.
Conclusions appear to be uniformly in favor of a greater structural similarity between Xecturus and the larvae of the plethodontids than with those of any other group. In view of this and what has been said abo\'e respecting the characteristics of the fenestral elements in various urodeles, it seemed advisable to examine as complete a series of embryos and larvae as it was possible to obtain. The following were accordingly studied i^ embryos 11, 12, 15, 16, 17, 18, 19 and 20 mm. long, and larvae
1 An ini])ortant gap in the series was filled by the generous loan of preparations of the 44 mm. stage by Pnjf. II. II. Wilder.
SOUND-TEANSMITTING APPARATUS 583
21, 22, 23, 24, 25, 26, 35, 40, 44, 48 and 70 mm. in length, respectivel3\ The chief difficulty in arriving at conclusions respecting the morphology of the fenestral elements is the determination of the exact origin of parts. For the embryonic stages specimens 1, or at most 3 mm., apart in length proved satisfactory. Different specimens of a given length vary with regard to internal conditions of development. By employing several specimens of the various lengths it is possible to obtain every step in the development between two given stages and one is thereby enabled to trace the origin of elements cell by cell.
It is very evident that embryos from. 15 to 20 mm. in length are the important stages in determining the origin of the columella in this species. A careful study of these series substantiates in every respect the conclusions of Piatt ('97) and Kingsbury ('03) that in its origin the columella is independent of the ear capsule. It arises as a cord of cells extending between the squamosum and the fenestra vestibuli, but at all times in these early stages is clearly independent of both ear capsule and fenestral membrane. While all stages are necessary in tracing the development history of the fenestral structures, larvae from 35 to 70 mm. long are most important in determining the morphology of the plate itself.
The definitive plate in Necturus has a double origin, a suggestion made by Kingsbury and Reed ('09) as one of two possible interpretations of its nature.
Previous work renders it unnecessary to describe any given stage in detail. Only a brief sketch, therefore, will be given of the developmental changes taking place during the larval period. In larvae 23 mm. long the nature and relations of the columella are shown in figure 1. It consists in this stage of a well-defined chondrified rod extending from the level of the cephalic lips a third of the way across the fenestra. It is entirely free from all other skeletal structures, a relation which exists from its first appearance up to this stage. No stilus is present, the suspen•sorial connection being effected by the usual cord of cells. The structure and relations of the columella at this stage are very suggestive of similar stages in Spelerpes and as much in contrast
584
H. D. REED
R. j. VII.
Sq. Col
F. V,
Fig. 1 Drawing from a wax model of the ear capsule of a larval Necturus 23 mm. long. Col., columella resting against the fenestral membrane ; it is rodlike, without stilus, and is free from the skeletal parts of the ear capsule; Ec, ear capsule; F.V., fenestra vestibuli; Sq., squamosum.
Fig. 2 Drawing from a wax model of the ear capsule of a larval Xecturus 25 mm. long. Col., columella which has increased greatly in length and vertical diameter but is still without a stilus; Ec, ear capsule; F.V., foramen vestibuli; R.j.VTI, ramus jugularis of the nervus facilais; Sq., squamosal.
to the conditions which obtain in Cryptobranchus and Amblystoma, where the fenestral end of the columellar cord even before chondrification begins, spreads out over the membrane to form the plate portion of this element.
An examination of a specimen 25 mm. long ("fig. 2) shows that the columella has increased greatlj' in size as regards both diameter and length. Its increase in size is relatively greater than that of the fenestra. It is still free from the ear capsule and without a stilus. The growth of the columella combines the features of Amblystoma (Kingsbury and Reed '08) and Spelerpes (Reed '14) a condition which is not found in other urodeles, so far as they have been studied, unless the growth of this structure in Amphiuma is to be so interpreted. Increase in the size of the stilus is at first apparently uniform, so that thecylindrical shape in maintained. But beginning with the 25 mm. stage, growth is most active along the dorsal side and particularly
SOUXD-TRAXSMITTIXG APPARATUS 585
in the cephalic end. This method of growth causes the former rod to become more plate-like and while it lies wholly outside the fenestral membrane it rests against it. The nature of the whole plate is best shown in a series of sections of a specimen 40 mm. long. A section near the cephalic end (fig. 3) shows the columella filling the fenestra at that level. The shape of the plate indicates its rapid dorsal and slight ventral growth, and further, this stage (and others younger) shows clearly from the position of the columella upon the fenestral membrane and the method of growth of its peripheral cells that no otic tissue has taken part in its formation. Such is not the condition further caudad. In following this series in that direction one notes that the thickened central part gradually narrows with a corresponding increase in the width of the thin portion both above and below (fig. 4). Finally, behind the middle of the fenestra, the thicker portion comes to a point and disappears entirely. Thus the thick portion forms a triangular area, with the apex pointing caudad and the base filling the cephalic part of the fenestra. In models as weU as in sections the thick and thin regions are clearly marked (fig. 5). These two areas possess a deeper significance than that of mere topography. The thickened area represents columella resulting from the growth of the original extraotic rod, mentioned above and illustrated in figures 1 and 2. The thin portion of the plate arises as chondroblasts within the fenestral membrane independent of the columellar element. Figure 6 is from a section a little caudad of the middle of the plate. Here the thick columellar portion is in strong contrast with the thinner part of the plate found above and below. Both figures 4 and 6 show the mode of development of the thin part of the fenestral element . In the fenestral membrane near the edge of the previously formed plate, chondroblasts arise and through the subsequently secreted matrix come in contact with the edge of the plate itself. Thus new layers of tissue, the cells of which are derived from the fenestral membrane, are gradually added to the margin of the plate. This process, illustrated in figures 4 and 6, is quite different from that which obtains in the cephalic part of the plate, where it spreads out
I'ig. o Ti'uussection of tlic ear (•ai)sulo of a larval Nectunis 40 niiii. lon^^ through the cephalic part of the fcu(>stra. r.,arteria carotis interna; C.L,canalis lateralis; Col., columella spreading out over the fenestral membrane due to the proliferation of cells from its dorsal and ventral borders; Ec, ear capsule; F.M., fenestral membrane applied to the ontal side of the columella only; H., hyoid; O.C., oral cavity; O.c, oral epithelium; »SVy., squamosum; V.ii.l., vena petrosor lat oralis.
Fig. 4 Transseclion of the ear capsule of a lar\al Necturus 40 mm. long, atalcv(!lof the middle oi the fenestra. Ch., chondroblasts forming in the fenestral membrane independent of columellar tissue; Col., columella; C.I., canalis lateralis; Ec, ear capsule; F.M., fenestral membrane; O., otic i)ortion of fenestral plate (operculum) formed by chondroblasts in the fenestral membrane, which are later surrounded by bony tissue and thereby joined to the definitive plate.
586
SOrND-TRANSMITTING APPARATUS
587
Fig. 5 Drawing from a wax model of the ear capsule of a larval Neoturus 48 mm. long. Col., columella portion of fonestral plate derived from tissue outside the ear capsule; Ec, ear capsule; F.V., foramen vestibuli; 0., otic portion of fenestral plate derived from chrondroblasts in the fenestral membrane and comparable to the operculum of Amblj^stoma in origin and position; R.j.VII., ramus jugularis of the X. facialis; Sq., squamosal; St., stilus columellae.
Fig. 6 Transsection of the ear capsule of a larval Necturus 40 mm. long, just caudad of the middle of the fenestra. Ch., chondrobla&ts in the fenestral membrane; C.I., canalis lateralis; Ec, ear capsule; F.M., fenestral membrane; 0.. otic portion of the fenestral plate. At this level the columellar portion has almost disappeared; compare with figure 5.
over the fene.stral membrane through the growth of its own tissue, as is shown in figure 3.
Transsections during this stage of development show that the fenestral plate of Necturus possesses numerous areas of cartilage encircled by bone, so that it exhibits a decidedh^ ringed appearance. This is due to the mode of ossification in connection with the formation of cartilage about the periphery of the plate. As in Amblystoma, two plates of bone are formed,
588 H. D. REED
one upon the inner surface of the coknnellar portion and the other upon the outer. When growth of the cartilage ceases these plates meet above and below, thus forming a complete shell to this element. New chondroblasts in the membrane are consequently prevented from fusing with the previoush^ formed cartilage and in the end become joined only through the extension of bony tissue about them in groups, or individually, as may be seen by a glance at the figures, where various stages in the extension of bone are shown. In many cases everj^ stage in the formation and growth of cartilage and its inclusion by bony tissue ma}^ be seen in the same section. It is this method of independent origin of cells in the membrane and their fusion through ossification into a connected whole that accounts for the peculiar relations of the two elements shown in figures 4 and 6.
It appears evident from a study of the material available that the fenestral plate in Xecturus is of double origin. The anterior end and a portion of the center being formed from the columella proton represents that structure in Amblystoma. The caudal half being formed by the chondrification of the fenestral membrane belongs to the ear capsule and must be likened, therefore, to the operculum of Amblystoma, although the M. opercularis never appears. The plate, as a whole, in its morphological nature is like that of Spelerpes, the difference being found in degree only.
Otic connections of the fenestral plate. There appears in the literature (Wilder '03) some comment concerning a connection between the fenestral plate in Necturus and the otic capsule. Since this element is free from the ear capsule in 3"oung larvae and since there is no continuity between the two structures in the adult it appeared advisable to examine the various series carefully with this point in mind, for in the light of such connections in other forms, an early fusion and a later separation of the parts did not seem probable. As stated above, the position of the columellar proton is along the fenestra, outside the membrane. Chondrification begins in larvae about 22 mm. long and at this stage the columella is clearly distinct from the ear capsule. Very soon, however, it comes to lie close to the
SOUXD-TRAXSMITTIXG APPARATUS 589
fenestra! membrane, in which a decided impression is made. In somewhat older larvae (25 to 26 mm.) the cephalic and dorsal growth causes the columella to press closeh' against the lips of the fenestra, bringing about a relation which is maintained throughout life. In most instances the perichondrium, where the two structures meet, can be made out and in no place does there appear what might be considered a true fusion, but rather a firm articulation. An examination of the articulation in the adult strengthens the view that the lips of the ear capsule do not fuse with the fenestra! plate and contribute nothing to its formation. The close relations may result from physical causes alone, since the plate of the adult exactly fills the fenestra, or may be considered as reminiscent of such fusions as occur in Amblystoma and others.
The stilus. It is a well established fact that the stilus in Xecturus and Proteus is found below the jugularis branch of the facial nerve, a relation which is not the usual one among urodeles. It is believed, however, that this relation does not affect its homology with the stilus of other forms, a \iew which seems to be supported by the relations of columella and nerve which obtain in certain of the plethodontids. One observes that in the Amblystomidae the ramus jugularis ^^11 comes forth freely underneath the stilus columellae and follows a course caudad across the ventral portion of the fenestra and its plate. In the Plethodontidae — as Gyrinophilus, for example — it emerges close against the ventral border of the stilus, and turning somewhat abrupth', takes a dorsal course across the fenestra! plate. Whatever may have been the cause of the difference in the relative position of nerve and skeletal parts in these groups, it offers a possible explanation of the existing relations in Xecturus. The nerve not only has a more dorsal position, but is well defined before the mesenchyme becomes concentrated into the forerunner of the columella. The latter naturally develops underneath instead of above it. The apparent mingling of the R. jugularis VII and the cells of the columellar proton in some plethodontid embryos tends to strengthen such a belief.
590 . H. D. REED
Summary. The fenestra! plate in Necturus, while of the single type, is double in its origin. The columellar portion is extraotic, having no early developmental connection with the ear capsule. At about the beginning of larval life it spreads out over the fenestra! membrane and completely fills the cephalic portion of the fenestra, from which position it gradually narrows, coming to a point and disappearing near the center of the oval window. The remaining portion (by far the larger) of the fenestra is filled by tissue originating from chondroblasts in the fenestral membrane and therefore strictly otic. The columellar portion including the stilus is the homolog of the columella of Amblystoma. The otic part of the plate represents the operculum. The larval characteristics of the plate are shown not in its morphology but in the absence of the M. opercularis. The soundtransmitting apparatus of this species is a true morphologic intermediate between that of Amblystoma and the Plethodontidae.
LITERATURE CITED
Brunnkr, H. L. 1914 a The mechanism of pulmonary respiration in amphibians with gill clefts. Morphol. .Jahrb., Bd. 48, pp. 63-82. 1914 b Jacobson's organ and the respiratory mechanism of amphibians. Morphol. Jahrb.. Bd. 48, pp. 157-165.
Herrick, C. Judsox 1914 The cerebellum of Necturus and other urodele am])hibia. .lour. Comp. Neur., vol. 24, pp. 1-29.
KixGSBURY, B. F. 1903 Columella auris and nervus facialis in the Urodela. Jour. Comp. Neur., vol. 13, pp. 313-334.
1905 The rank of Necturus among tailed Batrachia. Biol. liulL, vol. 8, pp. 67-74.
Kingsbury, B. F., and Reed, H. D. 1908 The columella auris in ami)hibia (first paper). Anat. Rec, vol. 2, pp. 81-91.
1909 The columella auris in amphibia. Jour. Mor])h., aoI. 20, pp. 549-628.
XoRRis. H. W. 191! The rank of Necturus among the tailed amphibia as indicated by the distribution of its cranial nerves. Proc. Iowa Acad. Sci., vol. 18, pp. 137-143.
I'latt, Julia li. 1897 The development of the cartilaginous skull and of the branchial and hyoglossal musculature in Necturus. Morjihol. Jahrb , lid. 25, pp. 377-464.
Rkei), H. 1). 1914 Further obser\ations on the sound-transmitting apparatus inurodeles. (Proc. Am. Anat. Assn., Dec, 1913.) Anat. Rec, vol. 8, no. 2.
\Vii.ui;i{. II. H. 1903 The skeletal system of Necturus maculatus. Mem. Bost. Soc. Nat. Hist., vol. 5, pp. 387-439.
fm
ox THE COMPARATIVE OSTEOLOGY OF THE
LUMPKIN (ara:\ius vociferus) and its
PLACE IN THE SYSTEIVI
R. W. SHUFELDT
SIXTEEN FIGTTRES
In the world's avifauna there are two birds contained in the genus Aramus. They have the appearance of large rails, and, very much like these, they inhabit marshes and extensive swamps and bogs. One species — the Aramus scolopaceus of Gmelin — ranges through Brazil, Guiana, and Venezuela, while our United States species — A. vociferus — occurs in Florida, the Greater Antilles, Central America, northward to South Carolina, and, very rarelj^, westward into Texas.
A number of years ago I prepared an illustrated account of Aramus vociferus, comparing its skeleton with those of rails, cranes, and allied birds; but as no good opportunity offered at the time for its publication, the manuscript and figures were set aside with others I was at work upon then. Later on this material was taken up again, and I published a brief, illustrated synopsis of it, which, while useful in some ways, was quite inadequate for working avian osteologists. ^ In that contribution, however, there was presented the schemes of classification of the supersuborders Gruiformes and Ralliformes of a number of the most eminent avian taxonomists, of this and the last generation, as Merrem, Huxley, Garrod, Sclater, Newton, Reichenow^ Fiirbringer, Sharpe, Gadow, and others.
The crane-rail group at that time I considered to form the suborder Paludicolae, an opinion at variance with the one I now hold. ^ Shufeldt, R. W. on the osteology of certain cranes, rails, and their allies, with remarks upon their affinities. Jour. Anat. and Phys., Lond. October, 1894, vol. 29; X.S.. vol. 9, pt. 1, art. o, pp. 21-34; text figures.
2 Shufeldt, R. W. An arrangement of the families and the higher groups of birds. Amer. Xat., vol. 37, nos. 455-456, November-December, 1904, pp. 833857, figs. 1-6.
591
THE ANATOMICAL RECORD, VOL. 9, NO. 8 AUGUST, 1915
592 R. ^y. shufeldt
There are three little tables given in the above cited article in the Journal of Anatomy, arranged in parallel columns. These tablet" present, in a comparative w&y, the salient osteological characters of Rallus longirostris, Aramus vociferus, and 'Grus americanus, and are very useful. ^Moreover, the article is illustrated by figures of the skull of the limpkin, a rail, and a crane, which, while they show the characters pretty well, are by no means as good as similar figures made by me of more recent dates. It is not nw intention here to cite any of the numerous articles which have been published on the osteology of either the rails or the cranes and their allies, a number of which are from my own pen, contributed many years ago.
On the other hand, it is the specific object of this contribution to give a detailed account of the skeleton of Aramus vociferus, which is a chapter in avian osteology hitherto unpublished, and one which will prove to be highly useful to students of the subject in the future, particularly to paleontologists who may require such information at any time.
For many years the fact has been more or less generally recognized that Aramus is related to the cranes (Grus) on the one hand, and to the rails (Rallidae) on the other. This, however, is a question to be more thoroughly touched upon in the concluding remarks at the close of the present article. Further, it is fair to presume that many species of birds of past ages and eras have become extinct, which, were they in existence now, would not only fill in the above-mentioned gaps, but would render it at once clear exactly what the aforesaid relationships were among all these now most puzzling gruine and ratline genera and species of birds.
In so far as I am aware, none of the fossil cranes or rails discovered up to date have thrown much light upon this part of the subject, though there is no telling what future discoveries along such lines may have in store for us.
l^nfortunately there are, at this writing, but few skeletons of the Gruidae at my disposal for examination and comparison ; while among the Rallidae there are a larger number of species and genera of birds, both existing and extinct, which stand in
COMPARATIVE OSTEOLOGY OF THE LIMPKIN 593
need of osteological description and comparison with Aramiis, especially of the genera Rallus, Limnopardialis, Gymnocrex, Aramides, Aramidopsis, and other ralline groups, or birds supposed to be more or less nearly related to the typical forms we designate as rails.
Upon comparing the skeleton of such a species as Grus mexicana among the cranes or Gruidae with that of the limpkin, it would seem that the gap between these two genera and their affines upon either side is at least partially bridged over — that is, in so far as two representatives species can demonstrate it. Take the skull of Aramus for example. With equal truth one might say that it belonged to some bird that was either a rail-like crane or a crane-like rail; with such equality are the characters represented in it, that is, ralline and gruine ones. Especially is this true when we come to compare this skull with that part of the skeleton of Rallus longirostris crepitans on the one hand, and Grus mexicana on the other.
One of the most evident characters distinguishing it from Rallus is its ha^'ing a pretty well marked supraoccipital prominence, with an occipital foramen upon either side of it; this is a gruine character. The vacuity in the interorbital septum is smaller than it is in either Grus or Rallus. Aramus has a short pterygoid, much dilated behind, and not seen in either the crane or the rail, where the pterygoids, although short, do not show this dilation. Its palatines and its laterally compressed, sharppointed vomer, with its inferior edge cultrate, are exactly what we find in Grus, this latter bone being more spreading in the clapper rail.
Its maxillo-palatines, being plate-like, are decidedly more gruine than they are ralline, as is also the broader interorbital frontal region above. The merest paring is taken off the edge of the superior orbital margins, to indicate the presence of the nasal glands, while those depressions are better marked in Grus, and still better in Rallus. In type the quadrates and lacrj-mals agree in all three of these genera. The broad orbital process of the first-mentioned bones are squarely truncated, exacth' as we find them in Porzana. Aramus has its temporal fossae on the
^ 'b
•fi -i
1 .1
> ^
cc .2
G « 3 o
COMPARATIVE OSTEOLOGY OF THE LIMPKIN 595
lateral aspects of the skull concaved precisely as they are m Grus and Rallus, and all its supero-parietal region is beautifully rounded, as we likewise see it in the genera mentioned. Passing to the foramen magnum of Aramus, we find it to be quite circular in outline, while in Grus and Rallus it is cordate. Again, it differs from either in ha\'ing the anterior wall of its brain-case thoroughly completed in bone, a feature, by the way, not often met with in birds. In its fronto-premaxillary region above Aramus exhibits the most perfect type of schizorhinalism, the sutures among the several bones there being very distinct, and remaining persistent during the life of the individual. The descending limbs of the nasals are antero-posteriorly broader than they are in either Rallus or Grus, while the nasal processes of the premaxillary — which extend from those bones and the palatine forwards — are more or less rounded and rod-like, not showing the supero-longitudinal grooving seen in the latter genus.
Pars plana is small and very distinct on either side, and lacks that elegant scroll-like process seen above and in front of it in Grus, in which it connects upon either side the lateral ethmoidal wing with the nether side of the fronto-nasal roof above. In all the birds thus far considered, the basitemporal region of the skull is, with respect to its characters, in very close agreement, and in them, too, the osseous aural entrance is very open, lacking the bony protecting walls so generously supplied by the squamosal and its neighboring bones in not a few other birds.
The long, acutely V-shaped mandible of Aramus, with its symphysis extending back for at least one-fourth its length, more nearly approaches the bone as seen in Rallus than the mandible in Grus. Its articular ends are abruptly truncated behind, and a good-sized ramal vacuity exists at the usual site upon either side. Borders or edges of either ramus above and below are rounded, as is also the under side of the mandibular symphysis, the upper part of the latter being longitudinally grooved. The hyoidean apparatus, the sclerotal plates of the eyeballs, and the ossiculae auditus require no special description.
596 R. W. SHUFELDT
In all the North American cranes and rails, including Aramus, the thyro-hyal rods seem to be the only part of the skeleton of the tongue that ossify, and the ring of bony platelets in an eye are all small for the size of any particular species to which they may belong.
Of the remainder of the axial skeleton: Between the skull and pelvis, Aramus has 23 vertebrae. With the possible exception of the atlas they are all thoroughly pneumatic, as in Grus. The postero-external angles of the neural arch of the first cervical are produced as processes, extending backwards, while the superior periphery of its cup is roundly notched out. In the axis vertebra we find a very low, tuberous neural spine, and a well-developed hypapophysial one. But what is unusual is that it has a good pair of parapophysial processes directed backwards. Its neural canal is small for the size of the bird. This is also the case in the third vertebra, after which this canal gradually enlarges, to become small again as it passes through the dorsal series, where it is of markedly small caliber, as it is in the pelvis. In the third vertebra the neural spine is very inconspicuous, likewise in the fourth, to be entirely absent in the fifth to the eleventh inclusive; whereupon, in the twelfth, it gradually begins to make its appearance once more, until it assumes the well developed quadrate plate seen in the dorsal vertebrae. In the third vertebra we see interzygapophysial bars connecting the pre- and post-zygapophyses, while in the fourth these are long and reduced to the most hairlike dimensions. The lateral vertebral canals also begin in the third vertebrae, and persist as such to the 16th cervical inclusive. Commencing, as I have said, in the axis, the narial parapophysial spines are present down the chain to include the tenth, they being very long and slender from the fifth. In the eleventh they suddenly disappear altogether, which is an interesting fact. From the fifth to the thirteenth cervical a hypapophysial open channel is developed for the passage of the carotid arteries; a small spine takes its place in the fourteenth vertebra. But this never becomes very large thereafter, and disappears entirely as we pass to the last four dorsal vertebrae. This hypapophysial
COMPARATIVE OSTEOLOGY OF THE LIMPKIN 597
spine exhibits a tendency to trifurcate in the seventeenth segment of the column, but at the best it is but feebly accomplished. The 18th, 19th and 20th (dorsal) vertebrae are completely fused together, while the 21st, 22d and 23d are freely articulated. In these last three, long osseous metapophysial spines or ossified tendons of the spinal muscles ornament the neurapophyses and transverse processes, extending, as they do, both forwards and backwards. A slender pair of free ribs are suspended from the 17th vertebra; on the coossified 18th these connect with the sternum, as they do in the succeeding five dorsals. There is also a pair of pelvic ribs that meet the sternum by means of rather long hemapophyses. All these ribs are highly pneumatic, and the true thoracic ones have free unciform appendages upon their posterior borders. The second and third pairs of ribs are short and slightly stoutish, but they soon become, in the succeeding pairs, long, slender and narrow.
The pelvis of Aramus, in some important particulars, more closely resembles that bone in Grus than that in the Rallidae, though it is not lacking in the characters seen in the pelvis of the latter. Regarded upon superior view, we note that the faintly emarginated anterior ends of the ilia are squarely cut across, while the mesial margins of these bones rise up to fuse with the superior border of the sacrae crista, thus completely sealing in the 'neural canals' behind.
The surface of the preacetabular portion of either bone faces almost directly outwards. There are present no intervertebral foramina of the sacrum, such as we find in rails; and the postacetabular part of the pelvis, with the included sacrum, is bent dow^nwards, so as to make a considerable angle with the fore part of the bone. Posteriorly, the ischio-iliac extremities extend far back of the last sacral vertebra, the posterior margin of the latter lying in the curve formed by the mesial edges of the ilia.
Viewed laterally, it will be seen that the external iliac borders of the postacetabular region extend over the outer aspects of the ischia, but, proportionally, not to the extent they do in some of the Rallidae, though rather more than in Grus. The planes in which the ischia are found are nearly parallel to each other,
598
R. W. SHUFELDT
^
P
a
<H
m
0)
.
c3
a
•^^
j3
32
c3
^
a>
-c
s
'S
rf
(h
tS
"S
o
«
^
GO
'?
to
n
^
"oS
a
i!
h
_QJ
o
O
o >
-2
s
> -^
o
■r.
+^
^
Ti
g
-2
5=
c
03
o
<;
o3
■>
c
cc
o
O
"*
Ol
-Ti
03
p;
^
(h
3
<M
CO
3
CD
3
bC
S
3
s.
e
m
cC
'o
• i-t
fl
o
fl &
-*-3
o
tc
o
"o
g
-£3 03
3
tc
^
cc
C3
.2
a>
-<
03
'?
G
,.
fl
O
o
O
"rt
X!
^
!-i
OJ
QQ
03
-4J
s
?
a
O)
C3
"3
c3
>
CC
Ph
OJ
(M
CO
rj*
lO
b£
M
bio
bi)
^
Uh
h^
-In
COMPARATIVE OSTEOLOGY OF THE LIMPKIX 599
and the ischiadic foramen is relatively small. The flat, rather broadish pubic st^de is separated by an open line coming in contact with the lower edge of the ischium above, and its distal extremity is produced well beyond the latter bone, to run out as it slightly curves towards the fellow of the opposite side. The posterior ilio-ischiadic margin exhibits hardly the semblance of a 'notch,' the edge of the former being in a line nearly perpendicular to the plane of the postacetabular region, or to the superior edge of the pubic style that extends beyond it. Seen upon ventral aspect^ we find that the five leading sacral vertebrae throw out their lateral processes, to fuse with the ventral surface of the ilium upon either side; this agrees with Crex. In Grus, seven of the leading sacrals behave in this manner. In Aramus the next four vertebrae have their elevated diapophyses completely concealed upon this view by the much swollen sacral bod}'. Then follow five more vertebrae, which have stout transverse processes thrown out as braces against the inner margins of the ilia. The first pair of these are the longest, and thej^ rapidly graduate down to the ultimate sacral. In the case of the first pair, too, the outer ends of their transverse processes meet, upon either side, the superior periphery of a cotyloid ring, and posterior to this point they fuse with the outer ends of the diapophyses of the next vertebra behind. But these two only — the two true sacrals — are thus linked together. The renal fossae are deep and circumscribed, especially by the folding forwards of the postero-ventral parts of the ilia, as the}" do both in Grus and all typical Rallidae.
There appear to be six free caudal vertebrae, but they are small, and have all their outstanding processes considerably aborted. A sub-quadrilateral pygostyle of better proportions finishes off the distal extremity of the spinal column in this remarkable bird.
Passing to its shoulder-girdle, we find coracoids, scapulae, and OS furcula all highly endowed with pneumaticit3^ The first-mentioned are rather short bones; antero-posteriorly compressed, and lacking an epicoracoidal process, but developing an elegant, forward-curling scapular wing (from the mesial side
600 R. W. SHUFELDT
of the shaft), which presents, well up, that foraminal perforation seen in the coracoids of several other groups of birds. A generous articular facet occupies the mesial two-thirds of the lower border of the expanded sternal part of either one of these coracoids, and this facet is grooved for its entire extent in the transverse direction. This groove admits of perfect articulation with the rounded transverse eminence occupying nearly the entire length of either coracoidal facet of the sternum. Disagreeing with both Grus, and the Rallidae, the coracoids of Aramus slightly decussate in their sternal beds.
Os furcula is a broad U, without a hypocleidium, and with its transversely flattened rami, of no mean width, extending, without narrowing in the least, to include the symphysis, the anterior and posterior plane surfaces of which latter have graduall}' come to face the other way. The free, upper clavicular ends are bluntly truncated, and support each an articular face, for a coracoid and scapula of the corresponding side.
A scapula is a rather short bone, elegantly curved outwards, drawn gradually to a point behind, and its outer blade decidedly flattened in the vertical direction. Its head is large, and is occupied, at their usual sites, by extensive facets for the os furcula, the coracoid, and the demi-facet of the glenoid cavity. At the under or ventral side of this end of the bone, just within the coracoidal facet, we always find a pneumatic foramen, which is very large in Grus.
The sternum of Aramus is ver^- long and correspondingly narrow. Its carina is inclined to be shallow, and it extends the entire length of the sternal body, which latter is distinctly quadrilateral in outline, without any xiphoidal processes behind, but showing there a slight median emargination. The carinal angle is handsomely rounded off; the anterior carinal border above it is concaved and somewhat thickened; while the manubrial process is quite rudimentary. Either costal process is very much turned outwards, and is, comparatively, of no great height. The costal borders are transversely narrow; the haemapophysial facets upon them rather far apart, and their intervals occupied by deep little concavities. On the ventral aspect of the
COMPARATIVE OSTEOLOGY OF THE LIMPKIN 601
sternal body, on either side of the keel, a strong, rough, muscular line is seen, extending nearly its entire length. It is uniformly concaved outwards, the surface of the bone being roughened within it and smooth without. The convexity of this 'pectoral line' is well separated by an interval from the base of the keel.
On the thoracic aspect of this sternum we find it deeply concaved in front, gradually shallowing as we pass posteriorly. A median line of small pneumatic foramina are seen, and others are found, upon this view of the bone, in various localities. Between the costal borders, where the thoracic concavity is deepest, the two sides of the sternal body are flat, and face each other at an open angle. In front, the anterior wall, which bears the coracoidal grooves upon its outer side, is deep and nearly vertically disposed.
Of the appendicular sJceleton: Aside from the great difference in size, the humerus of Aramus is the very counterpart of that bone as we find it in Grus mexicana — even to the most trivial characters. In the former it has a length of 11 cm., just double that in the latter, and they differ from the humerus of the Rallidae in their being more completely pneumatic. The pneumatic foramen is a single hole, and the ulnar and radial crests are strongly developed; so is the thick humeral head, which is separated from the ulnar crest by a deep valley. The smooth shaft shows the double sigmoid curve and is subcylindrical on section. At the distal end we find the articular trochleae and of the usual form. There is a very rudimentary epicondylar process.
Having compared in detail the remaining bones of the pectoral limb of Aramus with the corresponding ones in Grus mexicana, I find them likewise, as in the case of the humerus, to agree, character for character, throughout, except in the matter of size. As a rule, the various bones have in Grus rather more than double the lengths of their counterparts in the antibrachium and pinion of Aramus. They are not pneumatic but inclined to be stoutish in proportions. An ulna is concavely bowed along its anconal border, while the palmar one shows a row of papillae for the quill-butts of the secondary wing feathers. The proximal end is, in proportion, larger than ordinary, as compared with the
602
R, W. SHUFELDT
distal extremity. This is so in order to support the articular cavities for the large trochleae of the humerus. An olecranon process is pretty well developed also. The radius is slightly bowed along its inferior border, and its carpal end is larger than its head. On the whole it exhibits the usual ornithic characters. This also applies to the two small bones of the wrist, and the various ones that make up the skeleton of the pinion. In the carpo-metacarpus the medius metacarpal is bowed and a trifle longer than the stouter and straight one of index. There is constantly present in both Grus and Aramus a minute hole on the anconal aspect of the head, between the tubercle that occurs there and the trochlear surface, which, inasmuch as it is not a pneumatic one, must be for the entrance of a nutrient vessel. The central portion of the expanded part of the proximal phalanx of index digit is very thin, and occasionally shows a single pin-hole perforation in the limpkin, but not in the crane, as a rule. Nothing peculiar is exhibited on the part of the terminal finger-joints, and pollex digit bears a claw.
None of the bones of the pelvic limb in either Aramus or (irus are pneumatic. In the majority of their essential characters, the corresponding ones in the two genera agree, except, of course, in the matter of size, thus being in Grus considerably longer and larger in every way. As to lengths, however, they differ in proportions, as will be seen from table 1, given in millimeters.
TABLE
1
FEMUB
TIBIOTARSUS
TARSO-METATARSUS
Grus
129 84
314 181
185
Aramus
136
Differences
45
133
49
In other words, twice the length of the femur in Aramus would l)e 168 mm. against 129 mm. -the length of the femur in (irus mexicana — showing the bone to be consideral)ly moi'e than twice the length in the former, as compared with the length of that in the latter. Whereas, in the case of the tibiotarsus,
COMPARATIVE OSTEOLOGY OF THE LIMPKIN 603
twice its length in Aramiis would be 362 mm. against 314 mm. of the bone in Grus, showing it to be nearly double the length — • as 39 mm. is to 48 mm., the differences as thus compared. The differences in the tarso-metatarsi are nearer what they are in the femora, as may be seen from the above table.
In Aramus the femur has a shaft that for its middle third is nearly cylindrical, the bone as a whole being slightly bowed in the anterior direction. Its great trochanter is very broad transversely, and its crest rises above the articular summit, being continued round on to the anterior aspect of the upper third of the shaft. A deep pit exists for the ligamentum teres on the semiglobular caput femoris. At the distal end the condylar portion is greatly developed; the rotular channel in front is very marked, as is its continuation below as the intercondyloid fossa. The inner projection of the external condyle is sharp and prominent, and the pits in the neighborhood for muscular and ligamentous attachment very distinct. The lowermost points of the condyles are nearly in the same horizontal plane, and the popliteal depression is not especially concaved. On the surface of the shaft we find the usual muscular lines well marked.
If Aramus and Grus possess patellae, they have been lost from all the material at present at my command, and at this writing I cannot speak with certainty upon that point.
The shaft of the tibio-tarsus is very straight, being flattened in front and rounded behind. Its pro- and ectocnemial crests are well developed, but they do not extend down on the shaft, the latter process being hooked with its plane at right angles to the former, which stands out about perpendicular to the surface of the anterior aspect of the shaft. The rotular crest is only very moderately developed. At the distal end of this bone we meet with the usual characters found there in all ordinary birds. For the confinement of some of the anterior tendons, we find the little osseous bridge spanning a deep groove, with the tubercles above it, one on either side, for the attachment of the ends of a ligament that fulfils a similar purpose. Of ordinary size, the condyles are separated by a well marked intercondjdoid groove, it being especially so in front. The internal condyle
604 R. ^y. shufeldt
is sharp behind, where it projects; is shallow in the vertical direction, and long in the antero-posterior — this being less the case in the external one. Both have the usual reniform outline.
Aramus has an incomplete fibula, its lower extremity fusing indistinguishabh^ with the outer side of the tibio-tarsal shaft at about its middle. The articular fibular ridge is high up on the latter, and has a length of about 2.5 cm. ; fusion does not take place at this point. he head of this bone is rather large, transversely flattened, and produced backwards. On the outer side of the upper part of its shaft is a small, deep, spiral groove for the tendon of the biceps flexor cruris muscle.
Tarso-metatarsus has its shaft quite as straight as that of the tibia. Its sides are flat, but before and behind it is longitudinally grooved for the passage of tendons. In either case, this grooving extends the whole length of the bone, but is deepest onthe anterior aspect for its upper third. On the summit of the bone the articular concavities are deep, and at the fore part between them the interarticular rounded prominence is conspicuous. The hypotarsus, though fairly well developed, is not especially so, and it is hardly at all extended down upon the back of the shaft. It has one main groove which is deep and nearl}^ closed over posteriorly. Within, this groove is partially subdivided into two by a more subordinate median longitudinal crest. Upon either side of the hypotarsus we find a small perforating foramen. The anterior exits of these are close together just above the tubercle for the tendon of the tibialis anticus muscle in front. At the distal end of the shaft the trochlear processes are rather large and distinctly separated from each other. The lateral ones curve towards the median
Fig. 6-16 Various bones of the liiiii)kiM (Animus vociferus); natural size. Drawn by the author in outline from spociinon Xo. 1179o, collection U. S. National Museum.
Pig. 6 Anterior aspect of right coracoid. Fig. 7 Coccyx, left surfac3.
Fig. 8 Front view of os furcula. Fig. 9 Anterior aspect, right tibio-tarsus and fibula. Fig. 10 Anconal aspect, left humerus.
Fig. 11 Right scapula, uijixt surface. Fig. 12 Left ulna.
Fig. 13 Left tarso-metatarsus, from in front. Fig. 14 Left carpo-metacarpus.
Fig. lo Left femur, anterior surface. Fig. 16 Proximal phalanx, index digit.
14
605
606 R. W. SHUFELDT
plane behind, particularly the inner one of the two, which is at the same time the highest on the shaft. The next in this latter respect is the external one, the lowest on the shaft being the mid-trochlear process. Between this and the outer one is found the usual perforating foramen for the anterior tibial artery.
The small free first metatarsal is but slightly twisted upon itself (about half a turn), and presents nothing beyond what we usually see in that bone in birds; it has a length of 9 mm.
Aramus has the normal t^'pe of foot for birds, that is, 2, 3, 4, 5 joints to 1 to 4 toes respectively. As a rule the phalanges are long and slender, agreeing in this and other respects better with this part of the skeleton in Rallus than -^ith Grus mexicana. Ungual joints are well developed and somewhat curved. All the sides of these latter are longitudinally marked by very distinct groovelets.
In both Aramus and Grus there is a great disposition for most of the tendons of the muscles of the pelvic limb below the thigh to ossify. In many cases they form stout, osseous rods, of very considerable length; others are beautifully dilated for half their lengths, like a small, partly opened fan, composed of the most delicate radii, being formed by that part of the tendon which is spread out in or on the muscle. In a disarticulated skeleton of Grus mexicana before me, there are between 60 and 70 of these ossified tendons, and about half that number in a similarly prepared skeleton from a specimen of Aramus.
From this description it will be seen that this rail-like bird belongs in a different family from the Rallidae — that is, in the family Aramidae, as pointed out by me in my classification of Aves, published in The American Naturalist in 1904 (vol. 38, nos. 455-456, Nov.-Dec, p. 852), which stands thus:
il)ersubor(ler .
Mli
KALLIFORMES
Suborder
XX
Fulicariae
Superfamily
I
Holiornithoidca
Family
I
Ih'liornithidae
Superfamily
II
Ralloidca
Family
I
Rallidae
II
Aramidae
THE PARAPHYSIS AND PINEAL REGION OF THE GARTER SNAKE
B. W. KUXKEL
Beloit College, Beloit, Wisconsin
FORTY-OXE FIGURES
The pineal region of the vertebrates has been studied by a large number of investigators whose interest has been directed especially to the parietal organ and epiphysis, the significance and functions of which are still so problematical. On account of the relatively high state of development of the parietal organ in Sphenodon and many lizards, the reptiles have received special attention and the papers of AYarren ('11) on the development of the pineal region in the lizard and turtle, Dendy ('99) on Sphenodon, and A^oeltzkow ('03) on the crocodilian furnish rather complete pictures of the development of the region in each one of these groups. The snakes have received ver\' little attention from this point of view and are only very imperfectly known. It was for the purpose of filling up certain serious gaps in our knowledge regarding this group that the present stud}^ was undertaken.
The literature dealing with the pineal region in snakes is very restricted. The earliest reference to this region is Hoffmann's ('85). In describing briefly the pineal region in Tropidonotus natrix, he called attention to a thickening of the roof of the brain at the boundary between the telencephalon and diencephalon. He offered no suggestion as to its significance but there can be little doubt that it was the anlage of the paraphj^sis which he saw. Hanitsch ('88) in a paper which I have not been able to see for myself, describes according to Leydig ('97) an organ which he supposed to be a well defined parietal eye in an embiyo Vipera berus. He described the eye as provided with a lens in
607
THE ANATOMICAL RECORD, VOL. 9, NO. 8
608 B.. W. KUNKEL
which were pigment masses. Ley dig thinks that Hanitsch was in error regarding this structure, and in his own investigations on Coronella he found nothing comparable to a parietal eye. In the stage described by Leydig, the paraphysis is characterized by the presence of many budUke evaginations and the epiphysis has become solid though it still retains its connection with the roof of the diencephalon by means of a hollow stalk penetrating the posterior commissure. Studnicka ('93) studied the epiphysis and paraphysis of an advanced embryo and a full-grown specimen of Tropidonotus and found the epiphysis to consist of a massive, ellipsoid body connected with the roof of the diencephalon by a thin stalk. The epiphysis itself was of a glandular structure. It was divided into lobes by connective tissue septa and exhibited a small cavity at the junction of the stalk and the body proper. In the adult form the same condition was found except that the cavity was wanting. Sorensen ('94) figured a sagittal section of the diencephalon of an embrj^o black snake in which the epiphysis was connected by an attenuated stalk with the roof of the diencephalon and was inclined caudad. He also figured a cross-section of the epiphysis of the garter snake which was a globular body. The failure of these investigators to find a parietal organ is probably because of the advanced age of the embryos studied and the temporary occurrence of that organ. The most complete account of the development of the pineal region in the snakes is that of Ssobolew ('97) on Tropidonotus and Vipera. In the former the epiphysis is described as a double evagination of the cerebral vesicle at the point of division between the di- and mesencephalon. The lumen of the epiphysis is still in wide open communication with the brain when the paraphysis first appears, and at its distal end exhibits eine kleine Vertiefung mit der Bildung von zwei rundlichen Kornern, — dem vorderen und hinteren." The former he interprets as the anlage of the parietal organ, the latter as that of the epiphysis. At this stage the wall separating the two does not quite reach the level of the wall of the diencephalic roof. Ssobolew's figures fail to show the parietal organ so that a clear picture of the relationship of this organ to the epiphysis is still
PARAPHYSIS AND PINEAL REGION OF SNAKE 609
lacking. In view of my own findings in Thainnophis and the irregularities which appear in the epiphysis, there may still be some question as to the exactness of his interpretation. In Vipera he described, without a figure, the anlage of the parietal organ as an anterior evagination from the epiphysis resting directly on the anterior epiphysial wall and accordingly elevated through the growth of the latter while not itself growing materially. All trace of the parietal organ disappeared later in both snakes. So far as I have been able to find, these are the only papers devoted to the development of the epiphysis and parietal organ in the snakes, although several have described the condition in the adult animal, for instance, Rabl-Riickhardt's ('94).
For a full account of the literature on this subject reference should be made to Studnicka ('05) and Gaupp ('98).
The observations upon which this stud\' is based were made upon a series of some 20 embryos of Thamnophis radix, the ordinary garter snake, ranging in length from about 10 to 100 mm. The lengths of the younger specimens could be obtained only approximately on account of their being coiled in a tight spiral. These embryos were projected and drawn by means of a camera at a magnification that was exactly determined in each case and was about 10 diameters; the length was then scaled off on the drawing by means of dividers. In the stages in which the cerebral flexures were marked, the most prominent point of the mesencephalon was regarded as the most anterior point. The older embryos were sufficiently flexible to permit a straightening of the coils and direct measuring. These embryos I have grouped into five stages for convenience of description. Stage I is characterized by its 2^ spirals and length of 11 mm. The choroidal fissure of the eye is invisible from the exterior; the mandibular, second, and third visceral arches are evident. The otocyst from the lateral aspect has a simple triangular form. In Stage II the olfactory pits have become constricted to form small nostrils. The body length is 18 mm. and there are 3| coils. Stage III has an approximate length of 30 mm. In Stage TV the length is about 42 mm. and the number of coils is 4| or 5. The phalli are well developed and everted at this time in the
610 B. W. KUNKEL
males. Stage V has a length of 80 to 100 mm. and is characterized by the disappearance of the mesencephalic flexure of the brain.
The roof of the forebrain included in the present study comprises the following parts enumerated from posterior to anterior :
(a) The posterior commissure, occupying the roof of the synencephalon (or pars intercalaris of Burckhardtj,
(b) The epiphysis or pineal body arising as an evagination immediately anterior to the posterior commissure,
(c) The superior commissure or commissura habenularis which connects the two ganglia habenulae, and which is situated immediately cephalad to the stalk of the epiphysis,
(d) The parietal organ or pineal eye situated in front of the epiphysis but separated later from it in the garter snake by the superior commissure,
(e) The postvelar arch (or dorsal sac of Goronowitsch or Zirbelpolster of Burckhardt),
(f) The choroid plexuses of the third and the lateral ventricles, arising from the roof of the anterior portion of the postvelar arch and sides of the telencephalon medium respectively,
(g) The velum transversum projecting into the encephalic cavity and separating the paraphysis from the postvelar arch,
(h) The paraphysis, a median diverticulum from the roof of the telencephalon medium.
Stage I
At this stage the roof of the diencephalon exhibits two minute, inconspicuous evaginations separated from each other by a short interval (fig. 1). The anterior one of the pair, I believe, represents the anlage of the parietal organ, the posterior one, the epiphj'sis. The posterior one of the two is broader at its base than the anterior one, but l)()th are of the same height and are inclined slightly so that the front wall of each makes somewhat less than a right angle with the roof of the diencephalon and the posterior wall rather more than a right angle. The anterior evagination is mucli compi-essed so that the himen is very small,
PARAPH YSIS AXD PINEAL REGION OF SNAKE
611
ABBREVIATIONS
ACC, process from the roof of the diencephalon marking the attachment of the accessory parietal organ.
ACP, accessory paraphysis
BV, blood vessel
CH, cerebral hemisphere
DCP, diencephalic choroid plexus
DI, diocoel
DM, di-mesencephalic groove
E, epiphysis
FM, foramen of Monro
IXF, infundibulum
LCP, lateral choroid plexus
P, paraphysis
PA, paraphysal arch
PC, posterior commissure
PO, parietal organ
SC, superior commissure
TM, telencephalon medium
TP, tuberculum posterius
TT, torus transversus
VT, velum transversum
Figs. 1-3 Sagittal sections of an embryo (C2) having a length of 11 mm. Stage I. Figure 1 is a combination of two adjacent sections through the brain, showing the relation of the epiphysis, parietal organ, velum, paraphysal arch, postvelar arch, and di-mesencephalic groove. X 20. ' Figures 2 and 3 are detailed drawings of the sections of the epiphj'sis and parietal organ from which figure 1 was made. X 230.
the posterior one opens by a wide mouth into the roof of the diencephalon. The distal ends of the two evaginations come in contact with the overlying external epithelium of the head. The velum transversum is present as a low ridge on the dorsal wall which continues laterally and ventrally and becomes more prominent at the sides than on the middle line and extends downward to the postoptic prominence. Externally ridgethis is manifest as a shallow groove.
612 B. W. KUXKEL
The paraphysis is not visible at this stage. The paraphysal arch is sunply a low, broad dome extending cephalad from the velum and passing over without interruption into the lamina terminalis as Tandler and Kantor have shown in their Stage II of Gecko. The position of the recessus neuroporicus is indicated by a space anterior to the paraphysal arch in which the encephalic wall is only about one-half as thick as it is immediately dorsal and ventral to it and where the basement membrane of the epithelium has disappeared and the space between it and the external epithelium of the embryo is filled up with a mass of polyhedral cells among which are many blood cells. The postvelar arch is very low. The diencephalon is separated sharply from the mesencephalon by a deep groove externally and a slight ridge internally.
The histological structure of the roof of the diencephalon exhibits several noteworthy features. In the postvelar arch there are two or three layers of nuclei situated toward the outer side of the brain. The preparations did not allow the outlines of the cells to be seen so that it is still a question how many layers of cells are represented by these nuclei. In the evaginations of the epiphysis and parietal organ the w^all is only about two-thirds as thick as the rest of the postvelar arch and apparently is only one layer of cells in thickness. In the interval between the two evaginations the wall is thin, as in the evaginations themselves. The thickness of the roof behind these evaginations becomes gradually and uniformly thicker toward the di-mesencephalic groove; but the roof of the mesencephalon becomes suddenly nearly twice as thick as the thickest part of the diencephalic roof. In the evaginations the nuclei lie rather toward the outer ends of the cells. The front wall of both evaginations appear somewhat thinner than the posterior one, as Tandler and Kantor have observed in Gecko.
The line of demarcation between the anterior parencephalic and the posterior synencephalic portions of the diencephalon is marked by the epiphysis which arises from the extreme posterior end of the former region. The anlage of the posterior commissure is characterized at this stage by a zone of cells m the roof
PARAPHYSIS AND PINEAL REGION OF SNAKE
613
of the synencephalon in which the cytoplasm is less dense than elsewhere. At this stage no traces of the other commissures of this region are visible.
The lateral and ventral prolongation of the velum transversum exhibits on its posterior aspect a very pronounced thickening having the form of a ridge that extends posteriorly and ventrally to the infundibulum. This ridge is very evidently anterior to the di-mesencephalic ridge which marks the division between the diencephalon and mesencephalon.
Stage II
This stage exhibits several slight advances over the preceding one in the region of the diencephalon. The epiphysis opens widely into the diocoel and is bowl-shaped, the cavity having a depth about equal to its antero-posterior diameter (fig. 5). The parietal organ has the form of a sohd outgrowth from the roof of the diencephalon. Its cells are elongated and placed perpendicular to the surface so that it has the appearance of having been originally an evagination, as in the precedmg stage, which has suffered a compression obliteratmg the lumen. In this
Figs. 4-5 Sagittal sections through the brain of an embryo (A2) Stage II, having a length of 18 mm., showing the parietal organ apparently constricted from the roof of the diencephalon so that its lumen is obliterated. The epiphysis is further removed from the parietal organ than in the embryo previously described. X 25. Figure 5 shows the epiphysis and parietal organ of the same embryo. X 230.
614 B. ^y. kuxkel
embryo the axis of the epiphysis is perpendicular to the roof of the dieneephalon but that of the parietal organ is inclined forward as before.
The histological structure of the epiphysis is very different from that of the surrounding portions of the encephalic wall. The cells composing it are arranged in a single layer with the nuclei toward the outer ends of the cells. The wall of the parietal organ is not as thick as that of the epiphysis, but like the latter organ, its walls are of a single layer of cells with their nuclei toward their bases. The cytoplasm of these cells differs slightly from that of the epiphyseal cells in being more dense.
The \'elum transversum is visible as a sUght ridge in the cavity of the brain and as a corresponding groove on the exterior. The paraphysis is just visible as a shght outpocketing of the apex of the paraphysal arch. Immediateh' dorsal to it is a large blood vessel. The cells composing it have a less dense cytoplasm than those forming the brain generally. The posterior commissure is quite clearly differentiated and the very slight constriction between the parencephalon and synencephalon is clear.
In another embryo of essentially the same size as the one just described, the paraphysis is apparently more highly developed. It has the form of a fingerlike evagination which is directed posteriorly. Its length is somewhat less than that of the epiphysis. Its posterior wall exhibits several wrinkles, as Ssobolew described in his ^'ipera embryo number 4. In this embryo I cannot be sure of the presence of a parietal organ. The plane of the sections (parallel to the roof of the synencephalon) was not favorable for the display of such an outgrowth, but a careful study of the series revealed nothing that could be interpreted as a parietal organ. The epiphysis has become somewhat larger than in the previous embryo described in this stage and slants distinctly backwards.
The telencepluiloii exhibits the two hemispheres and the telencephalon medium which is posterior and dorsal to the foramina of Monro. The paraphysis projects from the roof of this as a fingerlike evagination or, as in another embryo, as two very small evaginations. The telencephalon medium is greatly
PARAPHYSIS AND PINEAL REGION OF SNAKE 615
compressed laterally and its sides are parallel so that the velum transversum is very narrow. It projects very slightly into the ventricle.
The posterior commissm^e in this stage seems very distinctly to lie in the second diencephalic segment of the brain and not in the mesencephalon as it has been usually described. It is situated a short distance caudad to the epiphysis. Thamnophis agrees in this regard completely with Lacerta and Chrysemj^s as described by Warren. In its earhest manifestation in the reptiles, the posterior commissure is wholly diencephalic. As will be shown later, however, it apparently encroaches upon the mesencephalon.
Stage III
The changes that have taken place in the pineal region at this stage are shght, but nevertheless perfectly evident. The epiphysis has increased in length so that it is now about twice as long as wide, and its axis inclines dorsally and posteriorly. Its cells are arranged in several layers and the posterior wall continues to be thicker than the anterior one and is sUghtly wrinkled. In one section the posterior wall bears a small budlike projection (fig. 11) the proximal half of the slender lumen is larger than the distal half although at the apex itself it is again slightly larger so that in sagittal section it looks almost as if it were divided mto a cephaHc and caudal branch. The apex of the epiphysis lies in contact with the external epithelium of the head and is surrounded by numerous large blood sinuses in the mesenchyme. In another embryo having a length of 33 mm. the epiphysis communicates with the ventricle of the diencephalon by a constricted neck (figs. 12 and 17). The parietal organ shows an appreciable advance in development over that of the earliest stage observed where it consists of a budlike outgrowth in wide open communication with the diencephalon on the middle line a short distance in front of the epiphysis. In this stage, as in the one immediately preceding, it has become solid, apparently by a pinching together in an antero-posterior direction. It has increased somewhat in
616
B. W. KUNKEL
ft y- ,' rt\ -S I V'} * "f ?■ f Jf ^-^
PARAPH YSIS AND PINEAL REGION OF SNAKE 617
Figs 6-11 Sagittal sections of an embryo (J) having a length of 29 mm. Stage III. Figure 6 is a section through the entire head to show the general topography of the roof of the diencephalon. The section does not pass exactly through the paraphysis. X 20. Figure 7 is a detailed drawing to show the histology of the paraphysis which in this specimen was very slightly developed. X 230. Figure 8 shows the accessory parietal organ situated to the left of the mesial plane. X 230. Figure 9, the roof of the diencephalon on the mesial line at the level of the accessory parietal organ. X 230. Figures 10-11, adjacent sections through the epiphysis and parietal organ. The epiphysis is directed posteriorly; the parietal organ is suffering a lateral constriction from the roof of the diencephalon but is still in continuity with it ; the epiphyseal wall has become much thicker than in Stage II. X 230.
Figs. 12-16 Transverse sections through an embryo (G) having a length of 33 mm. Stagelll. X 25. Figure 12 is a section passing through the epiphysis and showing it slightly constricted off from the roof of the diencephalon. Figure 13 is a section 105 m in front of the preceding and passing through the parietal organ. Figure 14 passes through the posterior portion of the accessory paraphysis derived from the roof of the telencephalon medium. Figure 15 shows the paraphysis cut through its attachment to the telencephalon. Figure 16 passes through the middle portion of the paraphysis.
length and inclines cephalad so that its anterior face is for the most part in contact with the roof of the diencephalon in front of it. The portion of the roof of the diencephalon between the two evaginations continues to be made up of a single layer of
618 B. W. KUXKEL
Figs 17-18 Portions of figures 12 and 13 respectively, to show the histological structure of the epiphysis and parietal organ. X 230.
cells. The ])aiietal organ is made up of a single layer of cells in which the nuclei lie toward the basement membrane. In one embryo of this stage I could find no trace of a parietal organ.
At some distance in front of the parietal organ — about midway between the epiphysis and the \'eluin transversum — the roof of the diencephalon exhibits a curious structure ffig. 9) whose significance may still be somewhat doul^tful. Partl}^ by an increase in the number of cell layers and partly by the increased length of the innermost layer whose clear cytoplasmic portion is relatively taller than elsewhere, the roof thickens to form a conical projection of small size. Separated by a short distance from the thickening mentioned there is a tiny nodule of cells lying in the mesenchyme similar in staining qualities to the roof of the brain (fig. 8). It lies 0.06 mm. to the left of the apex of the thickening but at such a level that its ventral limb is on a level with it. There is very distinctly a complete separation between the two structures but at the same time their form and position leave little room for doubt as to their earlier continuity. This evidently is an unusual structure for in my whole series of embryos it appeared only once. For the present it may perhaps be looked upon as an accessor}' parietal organ.
The velum transversum is in the form of a broad fold of the roof and sides of the brain projecting into the ventricles. The anterior and posterior limbs of the fold make an obtuse angle with each other. The anterior limb bears two very slight budlike evaginations whicli may l)ecome iinolved in the anlage of the paraphysis. In Lacerta niuralis eml)ryos 23. and 4.5 mm. long.
Figs. 19-21 Sagittal sections through the brain of an embrvo (D2) 35 mm in length. Stage III. X 20. Figure 19 shows the relation oi the paraphvsis and velum transversum and the earliest trace of the diencephalic choroid pleius l-igure 20 passes 45 m to the right of figure 19. to show the connection of the paraphvsis and telencephalon medium. Figure 21 shows especiallv the prominence tornied by the ganglion habenulae and the development of the choroid plexus ot the lateral ventricle.
619
620
B. AV. KUXKEL
Warren has sho\\ni three similar evaginations which he regards as the anlage of the paraphysis.
The paraphj'sis in an embryo having a length of 33 mm. consists of a triangular pocket which is very much compressed from side to side and which opens into the ventricle of the telencephalon medium by an openmg greatly extended anterior!}^ and posteriorly. In another embrj^o of slightly greater length, 33 mm., the paraphysis extended as a fingerlike pouch somewhat compressed laterally and extending forward between the two cerebral hemispheres ( fig. 16) . The roof of the telencephalon medium
'M-^
Figs 22-23 Sections of the same embryo as figures 19 to 21. X 230. Figure 22 shows the parietal organ in continuity with the roof of the diencephalon in front of the superior commissure. Figure 23 shows the anlage of the diencephalic choroid plexus.
slants laterally from the middle line at a low angle which increases to a right angle immediately behind the origin of the paraphysis.
In one of the specimens of this stage there is a small outgrowth behind the paraphysis, but in front of the velum transversum which resembles closely the organ described in another embryo of this stage as an accessory parietal organ (fig. 8). It has the form of a solid bud only 30 ^l in diameter which is slightly constricted from the roof by a groove on its posterior side. Figure 14 shows a section immediately posterior to its connection with the roof of the brain. It may become involved in the paraphysis later, but the ])araphysis is so well advanced in this stage
PAIL\PHYSIS AND PINEAL REGION OF SNAKE 621
that it hardly seems Hkely. On the other hand it can scarcely be regarded as an accessory parietal organ since it originates from a segment of the brain anterior to the diencephalon.
The posterior commissm'e at this stage is clearly outlined. It does not lie immediately behind the epiphysis as at first, but is separated by a short portion of the roof, in this respect resembling Lacerta and Chrysemys as described by Warren. Later this portion of the roof of the synencephalon between the commissure and epiphysis becomes invaded by transverse fibers. In one embr\'o of this stage the transverse fibers seem to extend very slightly behind the groove separating diencephalon and mesencephalon.
Stage IV
At this stage the development of the pineal region exhibits considerable variation. Although the various specimens appeared externally to be of the same age, the relative development of the parts of this region varied greatly.
The posterior commissure has extended both caudad and cephalad, so that it reaches the stalk of the epiphysis in front and passes beyond the di-mesencephalic groove behind and occupies the entire roof of the synencephalon. Its fibers are very clearly differentiated by this time.
The epiphysis has still a simple form, being cylindrical or clubshaped, showing a differentiation into body and stalk. In two cases there was a wide open connection between the lumen of the epiphysis and the ventricle of the diencephalon; but in two other instances the pinching off seemed to be complete. The wall of the stalk is thinner than that of the body and in several instances the anterior wall is slightly thinner than the posterior one. Generally the body and stalk make an obtuse angle with each other, but in one case the stalk itself is bent. In one embryo in which the cavity of the epiphysis was completely severed from that of the diencephalon, a large blood vessel presses obhquely into the anterior wall (fig. 25).
The superior commissure can be distinguished as a small band of transverse fibers immediately cephalad of the stalk of the epi
622 B. W. KUNKEL
physis in the three embryos of this stage in which the epiphysi. is still in communication with the diocoel, but in the other specimens which are more advanced in regard to the epiphysis, no trace of a superior commissure could be seen (fig. 28).
As might be expected, the parietal organ exhibits the greatest diversity of development. In one instance it was simply a solid budlike evagination as has been noted already in several of the earlier stages (fig. 22). In a second instance it was in the form of a solid pyriform mass of cells with the smaller end reaching almost to the roof of the diencephalon but not coming in contact with it (fig. 24) . The axis of the organ in this embryo is inclined upward and forward at an angle of about forty-five degrees with the roof. The stem is a single column of cul^oidal cells, the body itself exhibits a single layer of spherical nuclei around the margin with only a few situated internally. It has no trace of a lumen. It is notable in this embryo that the roof of the diencephalon is made up of only a single layer of tall columnar cells in the region from which the parietal organ apparently has separated whereas in the other specimens of this stage there are distinctly several layers. In a third embryo, the parietal organ has separated completely from the roof of the diencephalon and is in the form of a solid ovoid mass of cells with its long axis extending forward and upward. There was no trace of a lumen or stalk in this case (fig. 26). The roof of the diencephalon, however, was characterized by a slight irregularity in which the nuclei were rather crowded and squeezed toward the outer side of the roof,
Figs. 24-28 Sagittal sections of the roof of the diencephalon of three different embryos of 42 mm. length (Embryos B, R, and S). Stage IV. Figure 24 shows the parietal organ as a pyriform outgrowth separated from the roof of the diencephalon. X 230. Figure 25 is from the same embryo as the preceding and shows the lumen of the epiphysis closed off from the diocoel and a large blood vessel entering the anterior wall of the epiphysis. X 230. Figure 26 shows the parietal organ completely separated from the roof of the diencephalon in the form of a solid ovoid mass of cells and the epiphysis in open communication with the diocoel. X 230. Figure 27 shows the general relation of the parts of the brain of the same embryo as the preceding. X 20. Figure 28 shows the lumen of the epiphysis completely cut off from ilic diocdcl and limited to the body of that organ, and the parietal organ as a liollow sphere of cells completely separated from the brain. X 230.
PARAPHYSIS AND PINEAL EEGION OF SNAKE
623
THE ANATOMICAL RECORD, VOL. 9, NO. S
624 B. W. KUNKEL
indicating probably a very recent separation. Several blood vessels were in close relation to the organ. Still a fourth condition of the parietal organ was met with in another embryo in which it was in the form of a hollow spherical vesicle (fig. 28) . It was in this instance made up of a single layer of cells and was completely removed from the roof of the diencephalon. The postvelar arch is large and domelike in this stage and is characterized especially by the numerous small wrinkles especially in its anterior portion (figs. 19 and 20). The anterior and posterior limbs make approximately a right angle with each other but the posterior limb is much longer than the anterior one. The velum transversum is well differentiated, with its two faces making an angle of about thirty degrees with each other. The epithelium of the front wall in the middle line is conspicuously thicker than that of the posterior wall.
The paraphysis has the form of an irregular conical or oval evagination in open connection with the cavity of the telencephalon medium. Usually it is slightly constricted at the base. It is considerably more voluminous than the epiphysis and its walls are more irregular. At this stage the blood vessels surrounding the paraphysis are very conspicuous.
The choroid plexuses exhibit at this stage marked advance in development over that of previous stages. The diencephalic plexus apparently appears later than the telencephalic, for at this time it is evident only as several bud-like thickenings of the post velar arch projecting into the ventricle of the brain, and separated from each other by fairly deep clefts (fig. 23). The telencephalic plexus consists of an outgrowth extending anteriorly into the lateral ventricle from the wall immediately dorsal and lateral to the foramen of Monro (fig. 21). It is rounded at its free anterior margin and contains a large blood vessel. The histological differentiation of the telencephalic plexus is marked. The cells which are arranged in a single layer are more slender and taller than in the paraphysis and the cytoplasm is denser, and their nuclei are more elongated.
PARAPHYSIS AND PIXEAL REGION OF SNAKE 625
Stage V
The paraphysis at this stage has become much longer and has a horizontal position, its blind end pointing caudad (fig. 39), the choroid plexus of the lateral and the third ventricles are much folded and the epiphysis has become greatly elongated and its stalk has become solid.
The paraphysis extends from the roof of the median telencephalic ventricle at first dorsally and then turns at a right angle so that its distal half is horizontal. Its lumen is of somewhat irregular form. Proximally it is T-shaped, distally it becomes flattened dorso-ventrally. It lies in contact with the postvelar arch causing a distinct depression along the median line. The opening of the paraphysis into the ventricle of the telencephalon medium is exactly at the level of the opening of the foramina of Monro so that the ventricle widens quite suddenly immediately ventral to the opening of the paraphysis.
The velum transversum is difficult to delimit accurately. Its cephalic wall passes directly into the paraphj'sis and its caudal wall into the choroid plexus of the diencephalon. It is rather narrow from side to side, and at its sides it is very short. j\Iedially it exhibits a longitudinal groove with parallel sides so that in cross-section it appears like a bilobed tongue depending from the dorsal wall of the brain. In general the velum has a vertical position. Its caudal surface bears several large oval or irregular prolongations which represent the choroid plexus of the diencephalon. These processes are not to be distinguished from those which hang down from the post velar arch. They are in fact continuous with them.
The choroid plexuses of the telencephalon and diencephalon at this stage are quite complex in form. The latter has already been mentioned as made up of a number of irregular masses or folds from the caudal wall of the velum and roof of the diencephalon. It extends caudad as far as the level of the parietal organ and almost completely fills up the dorsal portion of the third ventricle which is nearly circular in cross-section. A study of the transverse series of sections of this region showed so
Figs -^9-34 Transverse sections through the brain of two embryos (D and C) having a length of 80 and 90 mm. respectively. Stage V. Figure 29 represents a section through the paraphysis of embryo D. X 230. Figure 30 passes thi-ough the epiphysis and the superior and posterior commissures of embryo L. X -UFigure 3l\epresents a portion of the previous figure to show the histx)logical structure of the epiphysis. X 230. Figure 32 is a section through the same embrvo as the preceding but further anterior, passing through the parietal organ, ganglia habenulae, and superior commissure. X 20. Figure 33 shows the histological structure of the parietal organ of the same embryo. X 1-30. I'lgure 34 shows the histological structure of the epiphysis of embryo C, showing several
clefts within. X 230.
626
PARAPHYSIS AND PINEAL REGION OF SNAKE 627
great irregularity on the two sides that there can be no definite arrangement of parts. The telencephaUc plexuses are likewise complicated. That of the lateral ventricle has the form of a plate of nearly horizontal position with its distal, lateral margin much thickened and turned up dorsally so that it has a concave upper surface. This plate extends anteriorly from the posterior side of the foramen of Monro so that the latter is partially obliterated by it. Warren describes the choroid plexus lateralis as springing from the paraphysal arch immediately in front of and lateral to the mouth of the paraphysis, invaginating the dorsomesial wall of the hemispheres. The plexus extends medially from its connection with the margin of the foramen of Monro so that it projects freely into the ventricle of the telencephalon medium and unites with the lateral wall of this ventricle by a slender stalk, ventral and lateral to the origin of the paraphysis (fig. 35). This median prolongation of the lateral plexus represents probably the telencephalic choroid plexus of Warren or the choroid plexus inferioris. It is wanting, according to Warren, in Lacerta but in Chrysemys there are described two paired masses growing back from the origin of the lateral plexus into the diencephalon. In Thamnophis they are confined to the unpaired ventricle of the telencephalon and do not extend posterior to the velum. The telencephalic plexus of the Amphibia, in which group it is most highly developed, arises from the paraphysal arch in front of the paraphysis. Warren thinks that these paired masses in Chrysemys may be the homolog of the amphibian telencephalic plexus.
The parietal organ is present in an embryo having a length of 90 mm. as a hollow, ovoid bodj^ entirely separated from the roof of the third ventricle (figs. 32-33). Its long axis extends antero-posteriorly. It is situated about 50ai dorsal to the diencephalon, slightly nearer the stalk of the epiphysis than the distal end of the paraphysis (fig. 39). It is composed of a single layer of cuboidal cells with large spherical nuclei. A large blood vessel lies immediately dorsal to it.
The stalk of the epiphysis is very slender and has attained a considerable length. It passes into the roof of the diencephalon
628
B. W. KUXKEL
P'igs. 35-38 Transverse sections through the dorsal portion of an embrj'o (D), having a length of 80 mm., arranged from anterior to posterior, showing the paraphysis, lateral, median, and diencephalic choroid plexuses, ganglia habenulae, superior commissure, and a pocket of the roof of the diencephalon cut off bj^ the superior commissure. X 20.
between the superior and posterior commissures and is directed almost exactly dorso-ventrally. The body of the epiphysis is large and pear-shaped. In one specimen it exhibited a small lumen in the anterior and ventral portion near its connection with the stalk, as has already been described in the adult serpent by Studnicka ('93). Its histological structure has changed markedly for now it is made up of an irregular mass of cells packed closely together with occasional connective tissue fibers. In another specimen there were several irregular clefts instead of a single lumen (fig. 34),
PAEAPHYSIS AND PINEAL EEGION OF SNAKE
629
The posterior commissure has increased much in size and occupies the entire roof of the third ventricle from the very attenuated stalk of the epiphysis posteriorly to the anterior portion of the mesencephalon. The superior commissure is strongly developed, forming a distinct projection from the roof of the diencephalon (fig. 32). The. fibers of the superior commissure are much longer than those of the posterior and curve forward toward their ends, presenting thereby a concavity in front.
The diencephalon is very much compressed laterall5^ Anteriorly in the immediate region of the velum its sides are parallel, but further caudad it becomes widened in its dorsal portion around the choroid plexus so that it becomes nearly circular in cross section. Immediately ventral to the widened portion of
630
B. W. KUNKEL
Fig. 39 Reconstruction of the pineal region of an embryo (C) having a length of 90 mm. Stage V. The right half of the model is viewed from the middle line. X 36.
the ventricle it becomes greatly narrowed by the development of the optic thalami until the space between the two walls is obliterated. The ependyma at the line of fusion becomes flat and indistinguishable except as a layer of cells which appear like fibers in cross-section. Dorsal to this line of fusion and extending caudad to it, in the region caudad to the superior commissure, the ependymal cells become very different from elsewhere, increasing greatly in height. These taller cells pass over directly into the cuboidal cells of the stalk of the epiphysis (iig. 40).
PARAPHYSIS AND PINEAL REGION OF SNAKE
631
Fig. 40 Represents a transverse section through the stalk of the epiphysis of an embryo having a length of 90 mm. The section was accidentally broken across the stalk. X 50.
Fig. 41 Represents a sagittal section through the pineal region of an embryo having a length of 100 mm., showing the large blood vessel dorsal to the epiphysis. X 50.
Each ganglion habenulae forms a dorsal projection on the diencephalon in the region of the posterior end of the paraphysis. These two ganglia are separated dorsally by the much folded roof of the diencephalon making up the choroid plexus of the third ventricle. In one specimen the superior commissure apparently cuts off a portion of the roof of the diencephalon between the two ganglia so that there is a short pocket formed extending posteriori}' above the commissure (fig. 38).
DISCUSSION
The present study on the development of the parietal region in the garter snake has thrown light upon several matters which have been the subject of much debate.
First of all, there can be little doubt left regarding the presence of a parietal organ in certain stages of the ophidians. Either
632 B. W. KUNKEL
the organ described in this paper is the parietal organ or it is an organ entirely different from anything else described in the vertebrates. In favor of the interpretation here placed upon it, there may be mentioned; first, the method of origin a& an evagination from the roof of the diencephalon anterior to the epiphysis and posterior to the postvelar arch; second, its form when at its maximum development; namely, a hollow spheroid consisting of a single layer of cells. Against this interpretation may be mentioned; first, the appearance of the parietal organ later than the epiphysis; second, its wide separation from the epiphysis with the superior commissure separating the two ; and third, its failure to migrate dorsally to lie in close proximity to the dorsal side of the head. The first and third objections are not significant since in those groups in which there is no doubt whatever regarding the identity of the parietal organ there is much difference in the relative time at which the parietal organ and epiphysis appear, although the parietal organ in the lizards always seems to be in advance of the epiphysis. The failure of the parietal organ to pass far dorsally may be explained by its distance in front of the epiphysis which does not lie ventral to it and which otherwise by its growth might push the parietal organ dorsally as it does for example in Lacerta. Besides this, the extreme development of the paraphysis caudad in such formxS as Lacerta may also tend to displace the parietal organ toward the dorsal surface of the head. This last factor is wanting in Thamnophis because of the shorter paraphysis which does not insinuate itself beneath the parietal organ. The wide separation of the epiphysis and parietal organ constitutes a valid objection to the notion here put forth, for a similar relation has not been noted in any other type. There are two facts, however, which seem to make this objection less significant. The parietal organ is known in various forms to originate in slightl}^ different places, as will be shown later; and it has also been noted in this study that the segment of the roof of the diencephalon lying between the epiphysis and parietal organ in the earlier stages of the garter snake exhibits a different histological structure from the rest and resembles more nearly that of the epiphysis itself. It
PAR.\PHYSIS AND PINEAL REGION OF SNAKE 633
seems reasonable, on the whole, to regard the present organ as a parietal organ rather than one sui generis.
Another matter upon which evidence has been produced by the present study is the independence of the parietal organ and epiphysis.
The earlier view regarding the relationship between the two was that the parietal organ was pinched off from the distal end of the epiphysis (Strahl '84). Thus it would appear that the two organs were in the closest possible relation. A few years later Baldwin Spencer ('86), as a result of an extensive study of the parietal organ of many lizards, confirmed Strahl's conclusion, for in all the species studied the parietal organ and epiphysis seemed to be connected by means of the 'parietal stalk.' Beraneck ('87) studying the development of the parietal organ and epiphysis in Lacerta agilis was the first investigator to regard the two as independent, arising from two separate anlagen, one of which lay close in front of the other. Francotte ('96) recognized that the parietal organ might arise in two different waj^s. According to the first type of origin the anlagen of the parietal organ and epiphysis appear as independent buds from the roof of the diencephalon. The anterior one, the parietal organ, appears first and is larger than the posterior one, which develops into the epiphysis. In the beginning the parietal organ is larger than the epiphysis. According to the second type of origin, the posterior bud does not arise from the roof of the diencephalon but from the postero-dorsal border of the anterior one. Francotte regarded this latter type to be derived from the former which he thought was the more primitive. This apparent origin from the same outgrowth Francotte explained by assuming that the parietal organ elongates more rapidly than the epiphysis at first and drags the latter along with it so that the lip between the two at an early stage really represents the segment of the roof of the diencephalon which lay between them in the first place. This is also confirmed in a measure by Klinckowstrom ('93) on Iguana. In this lizard the roof of the diencephalon at one time exhibits a single evagination directed anteriorly, which is secondarily separated into two parts by a
634 B. W. KUXKEL
ringlike furrow. The parietal organ is in this way cut off from the epiphysis. The subsequent growth of the epiphysis is not, however, in the direction at right angles to the furrow just mentioned, but dorsally so that the scar marking the earlier connection of the parietal organ and epiphysis comes to lie on the anterior wall of the epiphysis. This means, as can readily be seen, that the material from which the greater part of the definitive epiphysis is derived, lay originally posterior to that of the parietal organ just as is the case where there are two separate anlagen for these organs. The fundamental difference between the condition in Iguana and Francotte's first type of origin of these organs lies in the greatly retarded appearance of the epiphysis in Iguana, until after the parietal organ's anlage has grown to considerable size.
Further evidence of the independence of these two structures is afforded by a study of the innervation of the epiphysis and parietal organ in certain vertebrates. The parietal organ may be supplied by the parietal nerve whose fibers enter the brain by wslj of the superior commissure, in front of the epiphysis, and from there may be traced to the right ganglion habenulae (Strahl and Martin '88); the epiphysis may be supplied by a nerve, the pineal nerve, which enters the posterior surface of the epiphysis and arises from the roof of the brain from the posterior commissure (Klinckowstrom '93, on Iguana).
The development of the parietal organ and epiphysis in the garter snake leaves no rooni for doubt as to the complete independence of these organs in this form; since there is a considerable interval between them on the roof of the diencephalon from their earliest appearance, in which space the superior commissure later appears.
PARAPHYSIS AND PINEAL REGION OF SNAKE 635
BIBLIOGRAPHY
No attempt at completeness in this bibliography has been made because of the availability of Gaupp's ('98) and Studnicka's ('05) work.
Beraxeck, E. 1S93 L' individualite de 1' oeil parietal. Anat. Anz., Bd. 8, pp. 669-677.
BuRCKHARDT, R. 1894 Die Homologien des Zwischenhirndaches bei Reptilien und Vogeln. Anat. Anz., Bd. 9, pp. 320-324.
Dbndy, a. 1899 The pineal eye in Sphenodon punctatus. Q. J. M. S., vol. 42, pp. 111-153.
1910 On the structure, development, and morphological interpretation of the pineal organs and adjacent parts of the brain in Tuatara. Phil. Trans. Roy. Soc. London, ser. B, vol. 201.
Dexter, F. 1902 The development of the paraphysis in the common fowl. Am. Jour. Anat., vol. 2, pp. 13-24.
Fraxcotte, p. 1888 Sur le developpement de 1' cpiphyse. Arch. d. Biol., T. 8, pp. 757-821.
1894 Sur I'oeil parietal, 1' epiphyse, la paraphyse. Bull, de 1' Acad, roy. de Belg., Ser. 3, T. 27, pp. 84-112.
Gaupp, E. 1898 Zirbel, Parietalorgan, und Paraphysis. ^lerkel u. Bonnet's Ergebn. d. Anat. u. Entwickl., Bd. 7, pp. 208-285.
Haxitsch, R. 1888 On the pineal eye of the j'oung and adult of Anguis fragilis. Proc. Liverpool Biol. Soc, vol. 3.
Herrick, C. L. 1892 Embryonic notes on the brain of the snake. Jour. Comp. Xeur., vol. 2, pp. 160-176.
Hoffmann, C. K. 1886 Weitere Untersuchungen zur Entwicklungsgeschichte der Reptilien. Morph. Jahrb., Bd. 11, pp. 176-219.
Humphrey, O. D. 1894 On the brain of the snapping turtle. Jour. Comp. Xeur., vol. 4, pp. 73-116.
DE Klinckowstrom, A. 1893 Le premier developpement de 1' oeil parietal, 1' epiphj'se et le nerf parietal chez Iguana tuberculata. Anat. Anz., Bd. 8, pp. 289-299.
VON KtJPFFER, C. 1903 Die Morphogenie des Xervensj'stems. Hertwig's Handb. d. vergl. u. experimen. Entwickl.
Leydig, Fr. 1897 Die Zirbel und Jacobson'sche Organe einiger Reptilien. Arch. f. mikr. Anat., Bd. 50, pp. 385-402.
LocY, W. A. 1894 The derivation of the pineal eye. Anat. Anz., Bd. 9, pp. 169-180.
MiNOT, C. S. 1901 On the morphology of the pineal region based upon its development in Acanthias. Am. Jour. Anat., vol. 1, p. 81.
636 B. W. KUNKEL
SoREXSEX, A. D. 1894 Study of the epiphysis and roof of the diencephalon. Jour. Comp. Neur., vol. 4, pp. 153-170.
SsoBOLEW, L. W. 1907 Zur Lehre von Paraphysis und Epiphysis bei Schlangen. Arch. f. mikr. Anat., Bd. 70, pp. 318-329.
Studxicka, F. K. 1905 Die Parietalorgane. Teil V. Oppel's Lehrb. d. vergleich. mikroskop. Anat.
Taxdler, J., u. Kaxtor, H. 1907 Die Entvvickelungsgeschichte des Geckogehirns. Anat. Hefte, Bd. 33, pp. 553-665.
VoELTZKOW, A. 1903 Epiphysis und Paraphysis bei Krokodilien und Schild" kroten. Abhandl. d. Senckenberg. naturf. Gesellsch., Bd. 27.
Warrex, J. 1905 Development of the paraphysis and pineal region in Necturus maculatus. Am. Jour. Anat., vol. 5.
1911 The development of the paraphysis and pineal region in Reptilia. Am. Jour. Anat., vol. 11, pp. 313-392.
THE GASTRIC VASA BREVIA
H. M. HELM
From the Anatomical Laboratory of the University of Wisconsin
THIRTY-SEVEN FIGURES
The gastric vasa brevia are those branches of the splenic artery and veins and their terminal divisions which pass by way of the gastro-splenic omentum to the fundus of the stomach.
In this consideration of the gastric vasa brevia the aim has been: first, to determine the most usual arrangement of these vessels in the adult, as regards number, size, origin, and distribution ; and, second, to ascertain the time and manner of their development in the embryo. Conclusions regarding the adult arrangement are based on a series of twenty-five drawings of the spleens and splenic vessels of as many dissecting-room subjects. The first eight drawings were made by Dr. Bunting, and it was at his suggestion that the study was carried farther.
Number. A glance at the outstretched arteries (figs. 1-25), shows that there were three in two cases, four in six, five in six, six in eight, and seven in three. Thus there are usually more than three and seldom as many as seven ; the most usual number is six, but four and five are hardly less frequent. The average number in the series was a little over five.
Size. The vasa brevia are always small. They were measured in fifteen of the subjects examined and in no case were they more than 4 mm. in diameter. Not infrequently they were mere threads, less than 0.5 mm. in diameter; the most common size was about 2 mm.
Origin. The splenic artery ordinarily divides into two main divisions, a superior for the supply of the upper half of the spleen, and an inferior for the supply of the lower half of the spleen, a part of the great omentum and a part of the greater curvature
637
638
H. M, HELM
Figs 1-25 Diagrammatic sketches of twenty-five specimens showing the origin of the vasa brevia arteries of the stomach from the splenic artery and .is branches. The vasa brevia are shown dark. The lighter branches not otherwise labeled are splenic branches. 0, gastro-epiploio ; P, pancreatic branch S (hg. D , branch to accessory spleen.
GASTRIC VASA BREVIA
639
of the stomach. The left gastro-epiploic may be a branch of the inferior di\dsion, as was true in thirteen cases, or it may be a branch of the main splenic trunk, occurring before the division into superior and inferior di^dsions. In any case the gastroepiploic trunk usually gives off vasa brevia (e.g., this was true in thirteen of the seventeen gastro-epiploic vessels examined). It was true in every one of the eight instances in which the gastroepiploic arose from the splenic artery proper.
THE ANATOMICAL RECORD, VOL. 9, Xo.
640
H. M. HELM
There may be anastomosis between the superior and inferior divisions, as in figures 16 and 17, and there may be accessory superior branches from the splenic trunk, as in figures 2 and 5, or accessory inferior branches as in figure 9.
As we have seen, the vasa may arise from (1) the splenic artery itself; (2) the accessory splenic branches springing from the main splenic trunks midway between the coeliac axis and the spleen, and (3) the superior and inferior divisions (the latter including the gastro-epiploic) and their secondary branches.
GASTRIC VASA BREVIA 641
Vasa brevia arising from the splenic artery itself are the exception; figures 9 and 16 show them. That shown in figure 9 was about 1 mm. in diameter and it passed to the dorsum of the stomach, low down on the fundus (VI, fig. 26). That shown in figure 16 was larger (3 or 4 mm.) as were all the vasa brevia in this specimen. It passed to the dorsum of the stomach toward the cardiac orifice. Both these vessels arose close to the spleen. More frequently the splenic artery gives off a gastric branch near its origin. This branch may be a typical vas breve, but as often it is a coronary or accessory coronary branch destined for the supply of the lesser curvature. Such a branch was present in four of the twenty-five specimens, as shown in figures 11, 14, 18 and 24. In figure 11 the vessel was 1.5 mm. in diameter and about 5 cm. long. It arose 3 cm. from the origin of the splenic artery and 9 cm. from the hilus of the spleen, and passed to the dorsum of the fundus about 3.5 cm. below and 2 cm. to the left of the cardiac orifice. It did not anastomose with any other vessel; it was a true vas breve.
The vessel shown in figure 18 was similar in size, origin, and distribution; it was also a true vas breve. This specimen also showed a vas breve originating nearer to the spleen and similar to those described as occurring in figures 9 and 16.
In case of those splenic arteries which give off accessory splenic branches before reaching the spleen, the accessory branches almost always give rise to vasa brevia. Such branches were present in five cases, and all but one (fig. 2) gave off vasa brevia. But since these branches occur in but twenty per cent of the cases, we come to the rather self-evident conclusion, that typically, the vasa brevia arise from the superior and inferior splenic divisions and their branches.
We have said the usual number of vasa brevia is five or six. Table 1 shows the number of branches arising from the superior and inferior divisions .
The vasa brevia arising from the superior division tend to be smaller than those arising from the inferior. This was true in ten of the fifteen cases in which comparative measurements were made. In another case the smallest vessel arose from the supe
642
H. M. HELM
Figs. 26-37 Diagrammatic sketches of twelve stomachs, showing the distribution of the vasa brevia of twelve of the specimens illustra^ted in figures 9 to 25. The distribution of the vasa brevia of the specimen shown in figure 9 is illustrated in figure 26; that of figure 10 in figure 27; figure 11 in figure 28; figure 12, in figure 29; figure 13, in figure 30; figure 14, in figure 31; figure 15, in figure 32; figure 17, in figure 33; figure 18, in figure 34; figure 20, in figure 35; figure 24, in figure 36; figure 25 in figure 37. In figure 36 the area of distribution of branch V was not determined accuratelj- and is not shown. This is also true of branch VI in figure 31. In figures 29, 32, 33 and 34 the general area of distribution is shown instead of the approximate area of each branch. In figures 27 and 33 the stomach is turned so as to show the line of omental attachment. Figures 28 and 31 show the ventral, the other figures the dorsal surface of the stomach. In figures 28 and 31 the area of distribution of branch I is really on the dorsal surface so that the stomach is represented as transparent over this area. The other branches are distributed near the line of omental attachment.
GASTRIC VASA BREVIA 643
TABLE 1
Figs
1
2
3
4
5
6
7
8
9
10
11
12
13
Sup
Inf
2 2
2 3
3 3
2 4
3 2
1 3
2 2
2 4
3
4
3
2
2 1
4 2
3 2
Figs
14
15
16
17
18
19
20
21
22
23
24
25
Sup
Inf
3 2
4 3
2
o
2 3
2
2 2
3
2
2 1
3
1
1
2
2 2
4 3
rior division, but the other superior branches were as large as the inferior. In the other four instances all the vessels were of about the same size. In many cases the superior branches, notably the first two, are mere threads, less than 0.5 mm. in diameter, whereas the inferior branches are apt to be 1.5 to 3 mm. in diameter.
A glance at the sketches shows that the point of origin of most of the vasa brevia is very uncertain so soon as one attempts to localize it to a secondary branch of the superior or inferior division. This is because the secondary splenic branches themselves are so variable. However, the first vas breve is relatively constant. It is small, as we have said, and usualh' arises from the highest splenic branch of the superior division — the terminal branch, virtually — close to where it sinks into the spleen. This was the case in sixteen of the twenty-five spleens examined; in another it arose slightly lower, from the superior division itself; in another it arose from the second instead of the first splenic branch ; and in two others it arose from a superior accessory splenic branch. Its distribution is likewise relatively constant. Typicallj^, it passes to the highest point on the fundus. It is not always the highest branch, however; thus in figure 15, vessel IV had the highest position on the stomach.
The second vas breve usually arises from the superior division or from one of its uppermost branches close to the first, runs parallel with the first, and has the next lower position on the fundus; like the first it is usuallv small. The third also usually
644 H. M. HELM
rises from the superior division in case there are five or six vessels in all. The fourth arises close to the bifurcation of the splenic artery, sometimes from one main division, sometimes from the other. The fifth and sixth arise from the inferior division, frequently from the gastro-epiploic trunk. The vessels tend to run parallel and to reach the stomach in the order of their origin.
Distribution. A consideration of the distribution of the vasa brevia gives a somewhat more satisfactory result. Since the vessels run in the gastro-splenic omentum, they reach the greater curvature of the stomach in the region of the fundus. Some small twigs may pass to the fundus just ventral to the line of omental attachment, but in every case virtually the whole area of vasa brevia supply was dorsal to the line of omental attachment: i.e., on the dorsal or original right side of the fundus. The uppermost vasa brevia tended to remain practically in the line of omental attachment; the lower branches, on the other hand, usually passed well onto the dorsum. In no case did a vessel pass much distance onto the ventral surface of the fundus.
In no case did a vas breve form anastomotic loops with other vessels. Thus they differ from all other gastric vessels, i.e., the coronary and gastro-epiploic branches; they are end arteries.
A glance at the sketches, figures 26 to 37, will indicate the location and extent of the area of distribution of the vasa brevia. In figure 31 the vessels are confined to the region of omental attachment, as many twigs passing anteriorly as posteriorly. No anterior twigs pass far, however, while the first vessel is distinctly dorsal in position. In figures 27 and 28, likewise, the vessels remain close to the line of omental attachment, though the tendency is toward dorsal distribution. In all other cases the area is distinctly dorsal to the line of attachment of the omentum. Furthermore, it is to the left of a line dropped from the esophagus to the greater curvature. That is to say, the vasa brevia are confined to the fundus. Figure 26 illustrates the fact previously mentioned that vessels arising from the inferior division do not necessarily take the lower positions on the stomach.
Development. The vasa brevia develop very early in the embryonic life as primary branches of the splenic artery. They
GASTRIC VASA BREVIA 645
later become tributary to the splenic arteries as these are differentiated and in adult life are always small. Their number and origin is variable, but their distribution is constant: they pass practically wholly to the dorsum of the fundus. They are end arteries and in no case anastomose with other vessels. In a pair of duplicate twin fetuses there were in one body five vessels, in the other six and the origins of the vessels differed in the two specimens.
To summarize: In the adult there are usually five or six vasa brevia, but there may be fewer or more. The vessels usually arise from the superior and inferior divisions of the splenic artery, but they may arise from the main trunk of the splenic artery or from accessory splenic branches; the branches of the superior division tend to be smaller, more numerous and to take a higher position on the fundus than th,e inferior branches, but the reverse may be true. The vasa brevia are never very large — at least under normal conditions; they are terminal or end arteries; they pass to the dorsum of the fundus of the stomach.
MEMOIRS
OF
THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY
Xo. 5
THE DEVELOPMENT OF THE ALBINO RAT, MUS NORVEGICUS ALBINUS
G. CARL HUBER
From the Department of Anatomy, University of Michigan, and the Department of Embryology, the Wistar Institute of Anatomy and Biology
I. FRO.M THE PRONUCLEAR STAGE TO THE STAGE OF :MES0DERM AXLAGE; EXD OF THE FIRST TO THE EXD OF THE XIXTH DAY
THIRTV-TWO FIGURES
COXTEXTS
Introduction 3
Material and methods 5
Ovulation, maturation, and fertilization 9
Pronuclear stage 13
Segmentation stages 21
2-cell stage 21
4-cell stage .^ 29
8-cell stage 31
12 to 16-cell stage 35
Summary of segmentation stages, rate, and volume changes 36
Completion of segmentation and blastodermic vesicle formation 42
Blastodermic vesicle, blastocyst, or germ vesicle 56
Late stages of blastodermic vesicle, beginning of entypy of germ layers. ... 63
Development and differentiation of the egg-C3'linder 73
Late stages in egg-cylinder differentiation, and the anlage of the mesoderm 92
Conclusions 108
Literature cited 112
II. ABXORMAL OVA; EXD OF THE FIRST TO THE EXD OF THE XIXTH DAY
TEN FIGURES
CONTENTS
Introduction 115
Half embryos in ^lammalia 117
Degeneration of ova at the end of segmentation 120
Incomplete or retarded segmentation 121
Abnormal segmentation cavity formation 126
Degeneration of ova as a result of pathologic mucosa 129
Imperfect development of ectodermal vesicle 132
Two egg-cylinders in one decidual crypt 1?8
Conclusions 140
Literature cited 142
Price, post paid to any country, $2..i0
PHILADELPHIA, PA.
Reprinted from the Journal of Morphology, Volume 26, Xo. 2, June, 191.5
646
ON THE INFLUENCE OF EXERCISE ON THE GROWTH OF ORGANS IN THE ALBINO RAT
SHINKISHI HATAI
The Wislar Institute of Anatomy and Biology, Philadelphia
For the past three years experiments have been carried on to determine the effect of long-continued exercise on the growth of organs in the albino rat. The main object of the present investigation was twofold: (1) to repeat the observations of Donaldson ('11) who found a shght increase (2.6 per cent) in the brain weight in albino rats which had been subjected to exercise for the period of six months, and (2) to extend the observations to other organs besides the central nervous system. The present paper includes the data obtained by tjie previous study, just mentioned, as well as the results of my own investigations.
TECHNIQUE
The opportunity for exercise was given by placing the test rats in a form of cage which was used by Slonaker ('08) in his studies on the daily activity of the albino rat. The revohdng cage consists of a large cyhndrical drum 58.5 inches in circumference, made of i^-inch wire mesh, which revolves on a stationary axle. On the axle was fastened the nest box and the food and water pans. The number of revolutions of the cage is registered by a cyclometer.
Albino rats one month old were placed in these cages for a period of three to six months, which is equivalent to seven to fourteen years of human hfe. Each cage contained a single rat. The rats used for the control were litter brothers and sisters of those in the revolving cages and were placed in the ordinarj^ laboratory cages (one foot high, one foot wide and six feet long) . The total number of rats examined was 36 controls
647
648 SHINKISHI HATAI
and 42 test animals. For the material of the 1914 series the present writer is under great obligation to Miss Caroline Holt, a graduate student at The Wistar Institute, who permitted the use of her exercised rats with their controls, and I wish to thank her for her courtesy in this matter.
METHOD OF COMPUTING THE AMOUNT OF ALTERATION
in determining the deviations of the organs in the exercised rats from those in the controls, the following method was used: As a first step, the weights of the various organs corresponding to the observed body length of the rats were computed by means of formulas (for various formulas see Hatai '13 and '14). This computation was made for both the controls and the exercised rats. The differences between the observed and computed values of both control and exercised are now transformed into percentages by taking the computed values as 100 per cent. Thus we obtain two sets of percentage values, one expressing the difference between the computed and observed values in the control rats, and the other expressing the difference between computed and observed values in the exercised rats. If exercise has not altered the organs of the animals at all, then these two sets of percentages should be alike within the limits of the normal fluctuations. If, on the other hand, exercise has altered the organs, these two sets of percentages should differ more or less according to the nature of the response to exercise! If in the case of any organ we now take the difference between the percentage value obtained for the controls and that for the exercised rats, this difference represents the amount by which the exercised rats, as compared with controls, depart from the value obtained by the formula. By the use of this procedure, successive series are thus referred to the formula values in each instance and since the deviations are measured by this means always from the same standard, they may be directly compared with one another.
In the case of the thymus gland, the age of the rats was taken as the basis for the computation, since the weight of the thymus is much more highly correlated with the age than with either the
INFLrENCE OF EXERCISE ON GROWTH IN RAT
649
weight or the length of the body (Hatai '14). The example taken from the 1911 male series (the second series in table 3), may serve to illustrate the method of comparison described above (table 1). Expressing in words the results as given in the last line of this table, these show that the exercised rats are 13.4 per cent heavier than the controls and have a tail 1.62 per cent longer.
AMOUNT OF EXERCISE TAKEN
The rats in the revolving cages often take a large amount of voluntarj^ exercise. To illustrate this I have chosen an example from the 1912 series and given the average distance run during every 24 hours for the entire period of 93 days.
As will be seen from table 2, the distance run is almost incredible in some instances, the greatest average run for 93 days being 10 miles per 24 hours; a record made by one female. Since the rat is most active during the night (Slonaker '12) this daily run
BODY LENGTH
TAIL LENGTH
BODY WEIGHT
(MM.)
(MM.)
(GM8.)
Controls: observed values
212
176
238.3
Controls: values calculated from
formulas (body
length taken as
basis of computa
tion)
180
238.1
Percentage deviation of ob
served from calculated val
ues A
-2.22%
0.08%
Exercised: observed values
195
164
202.0
Exercised: values calcvxlated
from formulas
(body length
taken as basis of
computation)
165
178.1
Percentage deviation of ob
served from calculated val
ues B
-0.60%
13.48%
Amounts by which the percent
ages for the exercised
animals differ from those for
the controls; i.e., B-A
1.62%
13.40%
650
SHINKISHI HATAI
TABLE 2
Showing the distance run by the exercised rats in a period of 93 days
EXERCISED RAT, XO.
NO. OF MILES PER 24 HOURS
SEX
EXERCISED RAT, NO.
NO. OF MILES PER 24 HOURS
SEX
8 BlO
0.4
M.
7 As
7.6
M.
7 A,,
0.5
M.
7 A,
7.6
M.
8B9
0.6
\i.
7B2i,
6.2
F.
7 B"i6
0.6
M.
7 B20
7.4
F.
7Bi9
4.8
M.
GB^go
7.7
F.
7B,8
6.0
M.
6B=83
8.3
F.
6 B^g.,
6.7
M.
SB„
10.2
F.
was accomplished for the most part within a twelve-hour period. With the exception of the first four sluggish male rats, the average run was about 6| miles for males and 8 miles for females. Slonaker found that female rats were more active than the males and he further noticed considerable individual variation in the activities. The present results agree nicely with, the observations of Slonaker. It is interesting to note also that the maximmn average run by Slonaker's rats w^as 11 miles (average for one month) while that of mine was 10 miles per day for 93 days, thus showing a close agreement even in this respect. We do not know how active the wild Norwaj^ rats may be, but a considerable curtailment of the normal activity in the albino rats under domestication seems highly probable.
The effects of long-continued exercise on the body of the Albino as a whole, and on the organs, was the object of the present investigation.
INFLUENCE OF EXERCISE FOR 90 TO 180 DAYS OX THE EXTERXAL MEASUREIMEXTS AXD OX THE ORGAX ^YEIGHTS
1. Alterations of the external measurements
The external measurements of the exercised albino rats contrasted with those of the controls are given in table 3.
To prevent misunderstanding or misinterpretation, a word of explanation touching the tables is in order. Using the series for 1911 males, the second one in table 3 (the same which has been used as an illustration for procedure on page 649), attention
INFLUENCE OF EXERCISE ON GROWTH IN RAT
651
is called to the following points: The entries for the control are the observed values— the entries for the exercised are the observed values. The percentages deviations of both controls and exer TABLE 3 Shoiviiig the external measurements in both control and exercised rats
LENGTH (mm.) OF
Body
Tail
WEIGHT
OF body:
GH.«IS
DTrR-\TIOX
exercise:
DAYS
.\ge:
D.\YS
NO. OF R.4.TS USED
Stock albino rat (Donaldson, '11)
.1/. and F
Control
199
165.00
199.60
180
242
31
Exercised
195
163.00
192.80
242
24
Per cent: Exerc.
-cont.
0.42
2.90
Stock albino rat,
1911 series M.
Control
212
176.00
238.30
180
215
21
Exercised
195
164.00
202.00
195
11
Per cent: Exerc.
-cont
1.62
13.40
Inbred albino rat, 1912 series M.
Control
212
184.00
217.30
180
212
9
Exercised
214
181.00
245.30
213
10
Per cent: Exerc. -cont
-2.77
8.42
Inbred albino rat, 1912 series F.
Control
191
174.00
155.40
180
215
6
Exercised
198
173.00
179.80
213
7
Per cent: Exerc .-cont
-4.77
2.52
Stock albino rat, 1914 series M.
Stock albino rat, 1914 series F.
Control
176(F)
167.00
145.60
90
1.35
3
Exercised
198
178.00
220.90
135
4
Per cent: Exerc. -cont.
-2.56
11.35
Control
Exercised
176 182
167.00 171.00 -1.60
145.60 159.40 -1.87
90
1.35 135
3 4
Per cent: Exerc.-cont
Percentage by which exercised rats differ from controls. Same weight given to each series
-2.02
6.76
652 SHIXKISHI HATAI
cised from the formula values have been determined but are not entered in the table. The difference between these two percentage deviations is alone given in the table after '%: exerccontr. ' To test the correctness of these last figures it is always necessary to carry out the operations which have been described (p. 648). This same explanation apphes to the other series in this table and also to the other tables 3, 4 and 5.
Body length. The average absolute length of the body is practically identical in both the exercised and control rats. The difference amounts to 0.5 per cent in favor of the controls. The data given bj^ Donaldson ('11) for a single series show also a small difference of 2 per cent in favor of the control. We may conclude therefore that exercise does not alter the growth of body in length to any noticeable extent.
Tail length. The tail length with respect to body length tends to be slightly shorter in the exercised than in the non-exercised rats. The average difference is 2.02 per cent in favor of the control. The data given by Donaldson show practical identity in his series. However, the difference is noted in four out of my five series and I conclude therefore that exercise tends to retard the growth of the tail in length.
Body weight. A slight increase of 6.76 per cent is given in the average body weight for the exercised rats when compared with that of the non-exercised. This difference appears in four series out of five. Data given by Donaldson shows also an increase of 2.90 per cent in favor of the exercised. This relative gain in body weight of the exercised rats may be due in part to the absence of lung disease in the exercised rats. As will be seen later, most of the control rats were suffering from a lung infection, while most of the exercised rats were free from this. It is therefore possible that the increased body weights shown in table 3 are in some measure due to the sUghtly emaciated condition of the control rats, thus raising the relative value in favor of the exercised rats. I am therefore incUned to believe that in the case of the albino rat the body weight is not much affected by the form of exercise here given — though it may be slightly increased.
INFLUENCE OF EXERCISE ON GROWTH IN RAT
653
Considering these three external characters together, we may say that long-continued exercise does not modify significantly any of the characters here mentioned with the possible exception of the tail length, which shows a slight tendency to a deficit with respect to the body weight.
2. Alterations of the viscera
TABLE 4
Showing the weights of viscera of the exercised rats compared icith those of the controls. Arrangement of data explained on p. 650.
BODY
LENGTH
(mm.)
HEART
WEIGHTS (GM.S.) OF
ALIM. TR.VCT
Kidneys
Liver
Spleen Lungs
Stock albino rat, 1911 series, M.
Control
212
0.886
1.768
10.21
0.611
1.796
21
Exercised
195
0.904
1.867
10.35
0.589
1.588
11
Per cent: Exerc.
-cent.
26.980
32.170
23.15
25.980
17.250
Inbred albino rat, 1912 series M.
Control
Exercised
212 214
0.784 0.874 7.380
1.627
2.032
17.300
8.50 11.36 22.46
0.445
0.415
-6.790
2.325
1.228
-82.990
7.86 9.15 9.49
9 10
Per cent: Exerc. -cont.
Inbred albino rat, 1912 series F .
Control
191
0.587
1.246
6.48
0.395
2.205
6.77
6
Exercised
198
0.787
1.530
8.85
0.363
0.999
8.12
7
Per cent: Exerc.
-cont.
17.430
8.310
17.24
-14.900
-89.970
7.52
Stock albino rat, 1914 series M.
Control
(F)176 198
0.595
1.044
37.680
1.054
1.790
25.280
6.19
9.06
13.12
0.651
0.444
-61.870
1.000
1.128
-15.160
8.27
8.82
-14.80
3
Exercised
4
Per cent: Exerc.
-cont.
Stock albino rat, 1914 series F.
Control
176 182
0.595
0.822
27 . 130
1.054
1.317
11.930
6.19
7.74
12.72
0.651
0.457
-63.210
1.000 1.105 1.390
8.27
8.09
-9.76
3
Exercised
4
Per cent: Exerc. -cont.
Average percentage by which exercised albino rats differ from controls. Same weight given to each series . .
23.320
19.000
17.74
-24.160
-33.090
-1.89
654 SHINKISHI HATAI
From table 4 we note the following modifications.
Heart. In all the series the weight of the heart is considerably heavier in the exercised rats than in the controls. The average difference is 23.32 per cent in favor of the exercised rat.
Kidneys. The kidneys of the exercised rats are also heavier than those of the controls in all the series. We note the average difference of 19 per cent in favor of the exercised.
Liver. The weight of the liver is also heavier in the exercised rat in all the series. The difference is 17.74 per cent in favor of the exercised rat.
Spleen. On the other hand, the weight of the spleen is greater in the control than in the exercised. The difference amounts to as much as 24.16 per cent in favor of the controls. This difference does not appear to be due to a greater variability of the spleen, since it occurs in four series out of five, and furthermore, the same phenomena occurs in another series, which will be considered later (table 7). The reason why exercise retards the normal growth of the spleen with respect to body length is not clear.
Lungs. We notice the average difference of minus 33.90 per cent in the exercised lungs compared with the controls. In the case of the lungs, however, those heavier than normal for the body weight are associated with a lung infection. As a matter of fact, most of the control rats were infected, while the exercised rats were free from infection. It has already been noted that the infection of the lungs in the control rats was probably responsible in part for the relatively small body weight of the controls. It is highly interesting to see that exercise prevents, or at least delays, an onset of a very prevalent pulmonary infection in the albino rat.
Alimentary tract. The ahmentary tract as here designated includes not only the digestive tract proper, such as the stomach, intestine, etc., but also all attached structures, as the pancreps, omentum, as well as fat deposited in them. As is shown in table 4, the alimentary tract is not evidently modified. We note an average difference of 1.89 per cent in favor of the controls. This small difference, associated as it is with the bal
INFLUENCE OF EXERCISE ON GROWTH IN RAT 655
anced distribution of plus and minus variations, justifies our conclusion that exercise has not affected this part of the visceral system.
3. Alterations of ductless glands
Testes. The testes of the exercised rats show an average increase of 12.33 per cent when contrasted with those of the controls. This increase occurs in both the male series, thus showing the significance of the reaction.
Ovaries. The ovaries in the exercised rats show an increase of 84.33 per cent when contrasted with those of the controls. The difference between the exercised and controls in these organs is evident even at a glance. It is interesting to note that the sex glands of the Xorw^ay rats are normally heavier than those of the Albinos. The increase here shown as the result of exercise may possibly indicate a return to the wild form, not in size alone but also in fertihty.
Hypophysis. The hypophysis responds to exercise differently according to sex. We note an increase of 10.25 per cent in the case of the male and a deficit of 22.23 per cent in the case of the female. The approach of the weights in the two sexes and the larger loss in the female bring about relations which I have observed in the wild Xorw'ay (Hatai '14 b).
Suprarenal glands. As in the case of the hj^pophysis, the suprarenal glands show also dissimilar reaction to exercise according to the sex. Thus we note practically no alteration in the male (0.84 per cent), while there is an increase of 47.76 per cent in the female — again relations such as the wild Norway shows (Hatai '14 b).
Thyroid. We note a difference of nearly 13.44 per cent in favor of the control rats.
Thymus gland. The thymus of the exercised rat shows a slight relative increase of 4.80 per cent when contrasted with that of the control. It occurs, however, in only two cases out of four, and furthermore the greater variability in the alteration suggests that the difference here noted may not be significant at all. We must await the results of future experiments to make any positive statement.
THE AXATOMICAL RECORD, VOL. 9, XO. 8
656
SHINKISHI HATAI
TABLE 5
Showing the weights of the ductless glands in the exercised rats compared with those in the controls. Arrangement of data explained on page 650.
BODY LENGTH
(mm.)
iSEX GL.\NDS
WEIGHTS (GMS.) OF
Hypophysis Suprarenals Thj-roid
Thymus
NO. OF RATS USED
Stock albino rats, 1911 series M.
Control
212
2.2280
21
Exercised
195
2.3380
11
Per cent: Exerc
cont.
19.7800
Inbred albino rat, 1912 series M.
Control
212 214
2.3430 2.5030
0.0085 0.0096
0.0328 0.03.39
0.0344 0.0.3.54
0.0984 . 1085
9
Exercised.
10
Per cent :
Exerc
cont.
4.8800
10.2500
0.8400
0.2700
7 . 1800
Inbred albino rat, 1912 series F.
Control . . .
191
Exercised.
198
Per cent :
Exerc- |
cont.
0.0368 0.0602
46.4200
0.0113 0.0127
-2.0200
0.0425 0.0622
27.3200
0.0273 0.0282
-5.4600
1051
1698
41
7000
Stock albino rat, 1914 series M.
Control . . .
176 (F.)
0.0282
0.1931
3
Exercised.
198
0.0290
0.1355
4
Per cent :
Exerc
cont.
-25.8100
-25.4900
Stock albino rat, 1914 series F
Control
Exercised... . Per cent: E
176
182
xerc
0.0557 0.1145
0.0076 0.0058
0.0417 0.0656
0,0282 0.0247
0.1931 0.1836
3 4
cont.
122.2500
-42.4300
47.7600
-22.7400
-4.2000
Average percentage by which exercised albino rats differ from control. Same
12.33(M.)
10.25(M.)
0.84(M.)
-13.4400
4.8000
weight given to each series
84.33(F.)
-22.23(F.)
47.76(F.)
INFLUENCE OF EXERCISE ON GROWTH IN RAT 657
From the above it is clear that most of the ductless glands are subject to a considerable alteration as the result of exercise. Beyond pointing out that the exercised rats show, in the case of the testes, ovaries, hypophysis and suprarenals, relations similar to those found in the wild Noi-way, I am unable to give an interpretation of the changes observed. The data are therefore presented without further comment except to repeat that the alterations here noted are constant and are not the result of a great inherent variability of these organs.
4. Alterations of central nervous system and of eyeballs
Brain weight. On the average the brain weight of the exercised rat is 4.02 per cent heavier than that of the controls. This relatively greater weight of the exercised rat is true not only on the average, but also for all the series given in table 6. A variation of 4 per cent is not usually regarded as a large figure, nevertheless it is certainly significant for this particular organ. Indeed, this gain of 4 per cent is the largest plus alteration so far obtained from our experiments. It seems safe to conclude from its constancj', as well as from relative uniformity of the value, that exercise increases the brain weight with respect to the bodj^ length. It should' be noted also that the present results agree with the finding of Donaldson ('11) in this respect. It is not clear, however, whether this gain was due to a uniform enlargement of an entire mass, or to the increase of special divisions or structural components of the encephalon. A detailed analysis may settle this question in the future.
Spinal cord weight. The weight of the spinal cord evidently is not significantly altered. This is shown not only by its slight average modification of 0.88 per cent but also by the fact that the variation is not uniform in all the series. Donaldson found a difference of 0.65 per cent in favor of the control rat. All we can say about this part of the central nervous system is that exercise produces no significant alterations.
Amount of water in the brain and spinal cord. As shown in table 6, the percentage of water in the brain and spinal cord is
658
SHINKISHI HATAI
TABLE 6
d and of eyeballs of the exercised rats Showing the weights of brain and spinal cof^i:^^^ ^j ^^^^ explained on p. 650.
compared with those of the controls. Arrangeme.
KRCENTAGE OF
BODY LENGTH
(mm.)
WEIGHT.S (GMS.)
Brain Sp. cord
Brain
ISp. cord
WEIGHT OF NO. OF
EYEB.'^LLS B.\TS
(GMS.) USED
Stock albino rat (Doncddson '11) M. and F.
Control
Exercised
Per cent: Exerc.-cont.
199
1.920
0.597
78.41
71.45
195
1.951
0.580
78.12
71.37
2.570
-0.650
-0.29
-0.08
31
24
Stock albino rat, 1911 series M.
'A 21 ! 11
Control
Exercised
Per cent: Exerc.-cont.
212
1.872
0.614
78.09
70.39
195
1.866
0.564
78.08
71.05
3.670
3.000
-0.01
0.66
Inbred albino rat, 1912 series M.
Control
212 214
1.784 1.868 3.940
0.610
0.603
-2.420
78.57 78.40 -0.17
71.40
71.20
-0.20
0.294 0.299
96
Exercised
Per cent: Exerc.-cont
10 7
Inbred albino rat, 1912 series F .
Control
Exercised
Per cent: Exerc.-cont.
191
1.715
0.547
78.24
70.89
198
1.826
0.558
78.57
71.65
4.430
-3.220
0.33
0.76
0.288 0.279 -9.920
Stock
albino
rat, 1914 series M.
Control
Exercised
Per cent: Exerc.-cont
176(F) 198
1.648 1.836 5.990
0.288
0.288
-16.950
3
4
Stock albino rat, 1914 series F.
Control
176
1.648
0.288
3
Exercised
182
1.701
0.276
4
Per cent: Exerc.-cont
2.060
-10.910
Percentage by which exercised
albino rats differ from controls.
Same weight given to each series.
4.020
-0.88
-0.02
0.41
-9.45
INFLUENCE OF ENERGISE ON GROWTH IN RAT
659
not modified. This result agrees with the findings of Donaldson ('11).
Eyeballs. As a sample of the sense organs, the eyeballs were examined. As will be seen from table 6, the average weight of the eyeballs of the exercised rats is 9.45 per cent less than that of the non-exercised rat. The meaning of the smaller eyeballs in the exercised rats is not at all clear. I may mention, however, that from the data so far accumulated, the wild Norway rat has somewhat smaller eyeballs than the Albinos of the same body length.
THE EFFECT OF EXERCISE TAKEN FOR A PERIOD OF 30 DAYS
We have found that exercise taken for a period of 90 days produces changes in the organs to the same extent as exercise given for the period of 180 days (tables 3 to 6). It was thought interesting to determine the minimum period necessary to produce all the typical alterations. For this purpose the rats one month old were kept in the revolving cages for 30 days. The number of rats used was 6 controls and 6 experimented.
Without discussing the individual characters separately, we may make the following general statement.
(1) In no case are the alterations as large as those shown bj' the rats which had been kept in the revolving cages for the period of 90 or 180 days.
TABLE 7
Showing percentage values by which the several characters of the rats exercised for 30 days differ from those of the non-exercised. Males = 3 controls and 4 exercised. Females = 3 controls and 2 exercised. In this table only the final percentage values {i.e., for "% exerc.-contr.") are given.
PER CENT
PER CENT
PER CENT
Tail length
0.4
2.4
0.2 4.2
— 7.7
4.7
Heart Kidnej's
Liver
Spleen
Testes
Ovaries
8.0 8.2
-9.5
-52.9
9.2
51.4
Thyroid Thymus
Hypophysis < ^
, /M. Suprarenals < ^
2.0
Body weight
Brain
-1.1 25.7
Eyeballs
10.6
Alimentary tract
Lungs
- 8.0 30.1
660
SHINKISHI HATAI
(2) The amount of time necessary to produce the typical alterations varies according to different organs.
(3) While the heart and kidneys show a typical change, the liver shows a contrary modification. This may be due to a rapid utilization of reserve materials following a rapidgrowth, as well as the greater activity of the animal at this younger period.
(4) The early response of the sex glands to exercise may be the result of two combined factors; a greater supply of nutrition following the rapid circulation, and a strong tendency for growth of sex glands at this period of 50 to 60 days of age.
(5) Changes shown by other organs, particularly by the hypophysis, suprarenals and spleen, are very large. It is to be noted, however, that the weights of the hypophysis in the two sexes, despite the fact that both show a gain, tend to come together as in the 90 and 180 day series, while in the case of the suprarenals, the increase in the female is much the greater, a result again agreeing with the earlier series.
We may conclude from this short experiment that the effect of exercise is clearly shown in the rats which have been kept in the revolving cages for one month only, though the amount of modification varies considerably according to different organs.
TABLE 8 Showing the relatioti between heart weight and amount of exercise taken
DESIGNATION OF RAT
BODY LENGTH
(mm.)
heart weight
(cms.)
NO. OF MILES PER 24 HOURS
SEX
8 Bio
222
0.725
0.4
M.
7 A47
212
0.771
0.5
M.
8 B9
215
0.799
0.6
M.
TB^ie
218
0.853
0.6
M.
7Bi9
201
0.720
4.8
M.
7B48
220
0.888
6.5
M.
6B282
217
1 . 149
6.7
M.
7 A5
205
0.907
7.6
M.
7 A4
212
0.915
7.6
M.
7B2i7
198
0.745
6.2
F.
7 B20
191
0.724
7.4
F.
6 B%2
205
0.838
7.7
F.
6 6^83
201
0.891
8.3
F.
8B„
197
0.762
10.2
F.
INFLUENCE OF EXERCISE OX GROWTH IN RAT 661
WEIGHT OF HEART IN RELATION TO THE .\:M0UNT OF EXERCISE TAKEN
The variability^ in the activities of rats in the revolving cage suggested that there might exist a definite relation between the heart weight and the amount of exercise taken. To test this point table 8 was prepared. The data were taken from the 1912 series.
Among normal rats kept in the ordinary cages the correlation between heart weight and body length is very high (Hatai '13). Table 8 shows, however, that the correlation between the heart weight and body length in the exercised rats is almost zero, while, on the other hand, the correlation between the heart weight and amount of exercise taken is very high. To illustrate these points, I have divided the male records into three groups and the females into two groups according to the following plan:
Male Group 1 Average for the first four sluggish rats
Group 2 Average for the rats which ran 4.8 to 6.7 miles
Group 3 Average for the rats which ran 7.6 miles
Females Group 1 Average for the rats which ran 6.2 to 7.7 miles
Group 2 Average for the rats which ran 8.3 to 10.2 miles
These average values are tabulated below.
Males.
Females.
Body length (mm.)
Heart weight fgms.)
Mean distance ruu (miles)
Group 1 Group 2 Group 3
217 213 209
0.787
0.919 0.911
0.5 6.0
7.6
Group 1 Group 2
198 199
0.769 0.827
7.1 9.3
From the above we notice that despite a greater body length in Group 1 (males) the corresponding heart weight is less in accordance with the least distance run. On the other hand, Group 3 (males) which has the least body length (209 mm.) gives almost as large a heart weight as Group 2, whose body length is 213 mm. In coincidence with this non-correlation
662 SHINKISHI HATAI
between the body length and heart weight, we notice a harmonious relation between the heart weight and amount of exercise taken.
In the female series we notice that despite the practical identity in their body length, Group 2, which had run the greater distance, surpasses in heart weight Group 1, which had run the lesser distance. These facts indicate clearly that the heart increased in weight in relation to the amount of exercise taken, as was anticipated.
GENERAL REMARKS
It is clear from the foregoing that long-continued exercise (equivalent to a period of 7 to 14 years in man) in the albino rat produces many striking alterations in the organs. The modifications here given are found to be true not only for all the series, but in most cases even for the contrasted pairs of rats, and thus the results are not dependent on the variability of these organs. I am confident that the alterations are the result of the longcontinued exercise. It maj' not be out of place to mention here that a careful analysis of data has been made to see whether or not the infected lungs are in any way responsible for the changes observed in the organs. The results were negative.
The medical literature abounds with writings on the subject of 'exercise.' We find a universal recognition of hypertrophy of the heart following severe and long-continued exercise. I am not aware, however, that there are any similar statements concerning the modifications of other organs. Although from the results of physiological investigations on metabohsm during or after severe physical exercise in man and in mammals, the occurrence of modifications in organs (such as the kidneys, liver, lungs, etc., besides the heart) are quite conceivable, yet we still lack the anatomical data for man.
From a purely biological standpoint, long-continued exercise has a special interest in the case of the domesticated albino rat. Former investigation has estabhshed the fact that the albino rat is a strain of the Norway rat (Hatai '07) and further, that these
IXFLUEXCE OF EXERCISE OX GROWTH IX RAT 663
two forms of the rat possess central nervous systems of a dissimilar weight, the Norway having an absolutely heavier brain and spinal cord for a given boby weight (Donaldson and Hatai '11). Again, recently, I have shown that the relative weights of some of the ductless glands differ also in the Norway and albino rats (Hatai '14 b).
In rabbits, Darwin ('83) noted several physical differences, particularly in the cranial capacity between the wild and domesticated forms. The wild rabbits possessed a noticeably greater cranial capacity than the domesticated variety. More recently Lapicque and Girard ('07) have accumulated extensive data on the brain weights in wild and domesticated races. These two authors conclude also that the wild forms surpass the domesticated in their relative brain weights. Domestication, however, involves numerous interrelated factors which can be only slowly isolated by systematic study, and this experiment was devised to test the value of exercise, which seems to be one of the important factors forming the complex of domestication. The present investigation shows that at least in such organs as the brain, eyeballs, sex glands, hypophysis and suprarenals, the exercised Albinos show an approach to the Norway rat.
In conclusion, I may repeat that from the anatomical side the question of exercise is usually taken rather lightly in the case of man, nevertheless when we consider its striking effect on some of the organs of the rat, a further careful investigation of the subject, not only in the rat but in man also, seems certainh' worth while, both from the general biological standpoint and for its bearing on hygiene.
CONCLUSIONS
1. The following determinations were made on exercissd as compared with non-exercised rats: (1) External measurements; body and tail length and body weight. (2) Visceral organs; heart, kidneys, liver, lungs, spleen and alimentary tract. (3) Ductless glands; testes, ovaries, hypophysis, suprarenals, thyroid
664 SHINKISHI HATAI
and thymus. (4) Nervous system and sense organs; brain and spinal cord, and ej-eballs.
2. The albino rats allowed to exercise in the revolving cages for 90 or 180 days show modifications in most of the organs. Among them, the following ma}' be mentioned:
(a) The heart,, kidneys and liver show an average excess of about 20 per cent, while the spleen shows a similar amount of deficiency.
(b) The brain weight shows an average excess of 4 per cent, while no change is noticed in the case of the spinal cord. (This result agrees with the observation of Donaldson '11).
(c) The ovaries give an excess of 84 per cent, while the testes give an excess of 12 per cent.
(d) The hypophysis, as well as the suprarenals, respond differently to exercise according to sex. Furthermore, these two organs show, as the result of exercise, an approach to the relations characteristic for the Norway rat.
(e) The exercised rats were either entirely free from lung infection or but slightly affected. The control rats, on the other hand, had badly infected lungs and in some series several of them were lost, presumably from the lung disease. Analysis of the data shows that the lung infection is not responsible for the changes observed in the organs.
3. Exercise for the period of 30 days showed in most organs modifications similar to those observed in rats exercised for 90 to 180 days.
4. In the exercised rats the heart weight and amount of exercise taken are highly correlated.
LITERATURE CITED
Darwin, C. 1883 Variations of animals and plants under domestication. 2nd
Ed. D. Applcton & Co., vol. 1, p. 135. Donaldson, H. H. 1911 On the influence of exercise on the weight of the
central nervous system of the albino rat. Jour. Comp. Neur., vol. 21.
pp. 129-137. Donaldson, H. H., and Hatai, 8. 1911 A comparison of the Norway rat with
the albino rat in respect to body length, brain weight, spinal cord weight
and the percentage of water in both the brain and the spinal cord.
Jour. Comp. Neur., vol. 21, pp. 417-458.
INFLUENCE OF EXERCISE ON GROWTH IN RAT 665
Hatai, S. 1907 On the zoological position of the albino rat. Biol. Bull., vol. 12, pp. 266-273.
1913 On the weights of the abdominal and the thoracic viscera, the sex glands, ductless glands and the eyeballs of the albino rat (Mus norvegicus albinus) according to body weight. Am. Jour. Anat., vol. 15, pp. 87-119.
1914 a On the weight of the thymus gland of the albino rat (iSIus norvegicus albinus) according to age. Am. Jour. Anat., vol. 16, pp. 251-257.
1914 b On the weight of some of the ductless glands of the Norway
and of the albino rat according to sex and variety. Anat. Rec, vol. 8,
pp. 511-523. Lapicque, L., and Girard, p. 1907 Sur le poids de I'encephale chez les animaux
domestiques. Compt. rend. Soc. de Biol., vol. 62, p. 1015. Slon.\ker, J. R. 1908 Description of an apparatus for recording the activity
of small mammals. Anat. Rec, vol. 2, pp. 116-121.
1912 The normal activity of the albino rat from birth to natural
death, its rate of growth and the duration of life. Jour. Animal
Behavior, vol. 2, pp. 20-42.
. MEMOIRS
OF
THE WISTAR- INSTITUTE OF AXATO^^IY AND BIOLOGY
Xo. 6
THE RAT
COMPILED AND EDITED BY
HENRY H. DOXALDSON
REFEREXCE TABLES AXD DATA FOR THE ALBIXO RAT (MUS
XORVEGICUS ALBIXUS) AXD THE XORWAY RAT
CMUS XORVEGICUS)
To be published in October
Cloth bound. Price, post paid to any country, $3.09
PREFACE
For a number of studies on the growth of the mammaUan nervous S3'stem made by ni}' colleagues and myself we have used the albino rat. In the course of the work we frequently felt the need of referring to other physical characters of the rat to which the nervous system might be related. This led us to collect such data as were already in the literature and also led us to make further investigations. The facts gathered in this way have proved useful to us and are here presented in the hopes that they will be useful to others also.
CONTENTS
Preface. Introduction. Classification. Earlj- records and migrations of the common rats.
Part 1. Albino rat — Mus norvegicus albinus. Chapter 1 — Biology. Chapter 2 — Heredit}'. Chapter 3 — Anatomy. Chapter 4 — Phj^siology. Chapter 5 — Growth in total bodj' weight according to age. Chapter 6 — Growth of parts or systems of the body in weight. Chapter 7 — Growth of parts and organs in relation to body length and weight in relation to age. Chapter 8 — Growth in terms of water and solids. Chapter 9 — Growth of chemical constituents. Chapter 10 — Pathology.
Part 2. Mus norvegicus. Chapter 11 — Life historj'. Chapter 12 — Growth in weight of parts and systems of the body. Chapter 13 — Length of tail and weights of body, brain and spinal cord in relation to body length. Chapter 14 — Growth in terms of water and solids. Chapter 15 — References to the literature. Index
666
THE GROWTH OF THE FETUS OF THE ALBINO RAT
FROM THE THIRTEENTH TO THE T^\"ENTY SECOND DAY OF GESTATION
J. M. STOTSEXBURG
The Wistar Institute of Anatomy and Biology
TWO FIGURES
For the prenatal growth in weight of the human body Jackson ('09) has presented data gathered by himself and compared these with such data as had already been pubhshed. Similarlj^, Lowrey ('11) has studied the prenatal growth of the pig. In both cases the authors have found it necessary to depend in part on preserved material. WTiile preservation may not alter materiall}' the relative weights of parts of the bodj', it undoubtedly does alter the total body weight and the records for that character must be interpreted, therefore, with this fact in mind.
The following study on the growth of the fetus of the albino rat forms another series of observations on the prenatal growth of the mammal and has the special virtue of being made throughout on fresh specimens.
The fetuses were all from second litters, the female having been allowed to breed once and to raise her litter under observation, so that we might be assured of her normal beha^dor as a breeding animal. The data have been gathered from the colony at The Wistar Institute between 1907 and 1913.
\IETHODS The female was mated for the second time, under observation, and after the lapse of the desired interval, was killed with ether. The fetuses were removed, cleared of membranes, and then each placed in a previously weighed stoppered vial and the weight determined to a tenth of a milHgram. The operation was aloe?
668 J. M. STOTSENBURG
ways conducted within a protecting chamber to prevent the loss of moisture by evaporation. The fetus of 13 days was found to be the youngest which would stand manipulation without damage and the observations begin with that age. The litters of 38 females have been thus studied. The total number of fetuses removed was 336, gi\ing an average of 8.8 per litter, with a range of from 3 to 16 fetuses per litter. For 330 of these the exact weights have been obtained.
The observations furnish records for the weights of the fetus from the 13th to the 22d day inclusive — the latter being about the time of birth — under usual conditions, and within these Umits they give the weights of the fetus at approximately twentyfour-hour intervals. The observed weights for this series are entered in table 1.
Allien the data of table 1 are combined and the means taken, we obtain the mean fetal weights given in table 2. The values given in table 2 are the means of the averages for the several litters of like age. Thus the average value for each litter was given the same weight irrespective of the number of fetuses in the litter. "WTien the data of table 2 are plotted they furnish the graph in chart 1. The form of this graph illustrates the rate of growth as given in table 2 and this agrees with the general observation that in the growing fetus the rate tends to diminish with advancing age.
It has been pointed out by Donaldson ('06) that we may assume the span of life in the albino rat to be three years — ;between birth and natural death — and that this span in the rat is equivalent to ninety years in man. On this assumption the rat grows thirty times as rapidly as man. If we apply this ratio to the gestation period it follows that one-thirtieth of the human gestation period is about 9 days, but the rat has a gestation period of some 22 days. The explanation of this discrepancy between the rate of prenatal and that of postnatal growth is still wanting but the recent observations of Huber ('15) on the early stages of development in the Albino show that the first phases go very slowl3\ Taking 22 days for the gestation period of the Albino and 271 days Qlall. in Keibel and Mall '10) for
GROWTH OF FETUS OF THE RAT
669
TABLE 1
Giving the observed iveights of the fetuses at different ages from the 13th to the 22d day of gestation. Where the horn of the uterus from which the fetus came and its relative position in the horn were not noted, the fetus weights are given in ascending values under the heading 'Horn and position not noted.' When these facts were noted the weights of the fetuses are given under the respective horns and in serial order, no. 1 being the fetus nearest to the ovary. In a few cases the horn was noted but the order of the fetuses not determined. All the weighings were made to the tenth of a milligram, but in the table only three digits are entered. The diet of the mother, which appears to influence the number of fetuses, is also given
.SEHI-\L
.\GE OF LITTER
DIET
HORX OF UTERU.S .\XD POSITION IN HORN
HORN AND
Days
Hours
Left
Right
NOT NOTED
39:...
13
Scrap
0.036 1 0.038 2 0.044 3
1
0.045 2 0.038 3 0.029 4 0.031 5
21....
13
2
Bread and
milk
0. 1
0. 2
0.041 3
0.010 1 0.055 2 0.028 3 0.068 4 0.037 5 0.042 6 0.047 7
22....
13
2
Bread and
milk
0.032 0.034 0.034 0.036 0.038 0.041 0.—
0.033 0.043 0.050
23....
13
2
Bread and
milk
0.046 1 0.046 2 0.036 3 0.049 4 0.041 5 0.046 6
0.058 1 0.049 2 0.046 3 0.038 4
670
J. M. STOTSENBURG
TABLE 1 (Continued)
SERI.\L
AGE OF LITTER
DIET
HORN OF UTERrS AND POSITION IX HORN
HORN AND
Days
Hours
Left
Right
NOT NOTED
42....
14
Scrap
0.092 0.088 0.093 0.107 0.092
1 2 3 4 5
0.081 0.108 0.107 0.085 0.101 0.091 0.097 0.104 0.059
1 2
3
4 5 6
7 8 9
17....
14
2
Bread and milk
0.122 0.145 0.098 0.127 0.116 0.136 0.101
1 2
3
4 5 6
7
20....
14
2
Bread and milk
0.117
0.
0.135
1 2 3
0.085 0.099 0.131
1 2
3
24....
14
2
Bread and milk
. 103 0.144 0.115 0.120 0.100
1 2
3
4 5
0.108 0.101 0.091 0.096 0.102 0.104
14....
14
2
Bread and milk
0.080 0.109 0.118 0.119 0.121 0.124 0.126
43....
15
Scrap
0.104 . 109 0.114 0.094 132 0.118
1 2 3 4 5 6
0.119 0.098 0.111 0.088 0.109 0.097
1
2
3 4 5
GROWTH OF FETUS OF THE RAT
671
T.\BLE 1 (Continued)
SEKI.\L
.\GE OF LITTER
DIET
HORN OF irTERUS .\ND POSITION IN HORN
HORN .\ND
Days
Hours
Left
Right
NOT NOTED
38....
15
Scrap
0.223 0.202 0.205 0.217 0.206
1 2
3 4 5
0.228 1 0.217 2 0.244 3
16....
15
Bread and
milk
0.148 0.167 0.182 0.183 0.197 0.226
0.158 0.170 0.176 0.189 0.190
7
15
Bread and
milk
0.119 0.143 0.176 0.186 0.193 0.196
41....
16
Scrap
0.319 0.306 0.336 0.329 0.315 0.256
1 2
3 4 5
6
0.342 1 0.360 2 0.336 3 0.327 4 0.322 5
15....
16
Bread and
milk
0.348 0.310 0.322 0.347 0.300
1 2
3 4
5
0.258 1 0.306 2 0.306 3 0.353 4
0.288 5
12....
16
Bread and
milk
0.257 0.291 0.336
25....
16
Bread and
milk
0.320 0.326 0.328 0.332 0.351 0.352 0.356 0.373 0.390
THE AN.\TO.\UC.^L REroRD, VOL. 9, NO. 8
672
J. M. STOTSENBURG TABLE 1 (Continued)
SERIAL NO.
AGE OF LITTER
Days Hours
16
Bread and milk
18...
40.
17
17
Bread and milk
Scrap
17
Bread and milk
13....
18
Bread and milk
HORN OF UTERUS AND POSITION IN HORN
Lett
0.536 1
0.474 2
0.617 3
0.543 4
Right
0.419 1
0.529 2 0.608 3
HORN AND
POSITION NOT NOTED
0.220 0.233 0.237 0.252 0.263 0.269 0.274 0.276 0.298 0.304 0.311
0.482 0.491 0.493 0.508 0.530 0.531 0.543 0.625
0.518 0.536 0.580 0.595 0.649 0.650
0.898 0.934 0.955 0.101 0.105 0.105 0.106 0.108
GROWTH OF FETUS OF THE RAT
673
TABLE 1 (Continued)
AGE OF LITTER
SERIAL NO.
Hours
HORN OF UTERUS AND POSITION IN HORN
Left
Right
HORN AND
POSITION NOT NOTED
30....
18
Scrap
18
Bread and milk
36.
18
37.
19
Scrap
Scrap
1.480 1
1.550 2
0.530 3
1.450 4
1.310 1
1.690 2
1.540 3
1.340 4
0.825 0.938 0.941 0.944 0.958 0.961 0.962 0.973 0.978 0.980 0.983 0.986 1 .000 1.010 1.020 1090
0.930 0.950 1.030 1.020 1.090 1.130 1.170 1.230 1.250
0.819 0.859 0.866 0.939 0.943 0.954 0.961 0.967 1.010 1.150
674
J. M. STOTSENBUEG TABLE 1 (Continued;
AGE OP LITTEB
SERIAL SO.
Days
Hours
19.
19
Bread and milk
31,.
19
19
Bread and milk
Scrap
35....
20
Scrap
33.
20
Scrap
26....
20
Scrap
HORN OF UTERUS AND POSITION IN HORN
Left
1.930 1
1.020 1.440 1.440 l.ooO
Right
1..560 2.020 1.740 1.910 1.630 1.860 1.900
2.280
1
2.780
2.310
2
2.480
2.700
3
2.520
2.520
4
2.690
2.390
5
2.510
6
HORN AND
POSITION
NOT NOTED
.510 .•590
1.670 1.670 1.690 1.700 1.710
1.730
730 740
2.130
2.280
2.470
2.480
2.500
2.550
2.600
2.770
2.830
2.870
2.900
GROWTH OF FETUS OF THE RAT
TABLE 1 (Continued)
675
SERIAL
.\GE OF LITTER
DIET
HORN- OF UTERUS ."VXD POSITION IX HORN
HORN .^ND
Days
Hours
Left
Right
NOT NOTED
26....
20
Scrap
2.940 3.000 3.060 3.200
34....
21
Scrap
3.750 3.950 4.000 4.020 4.160 4.420
3.980 4.050 4.280
35....
21
Scrap
3.920 3.990 4.030 4.110 4.150 4.220
32....
21
Scrap
3.580 3.720 3.790 3.810 3.930 4.080 4.090 4.260 4.370
28....
21
Scrap
3.030 3.400 3.470 3.470 3.550 3.580 3.590 3.720 3.760 3.950
27....
21
Scrap
4.000
\
676
J. M. STOTSENBURG
TABLE 1 (Continued)
SERIAL
AGE OF LITTER
DIET
HORN OF UTERUS AND POSITION IN HORN
HORN AND
PO.SITION
NO.
Days
Hours
Left
Right
NOT NOTED
27.... 44. . . .
21
22
Scrap Scrap
4,460 1
4.710 2
3.910 1 4.710 2 4.950 3 4.700 4 4.540 5 4.820 6 4.750 7 4.710 8
4.020 4.120 4.150 4.220 4.330 4.380 4.460
TABLE 2
Derived from the data in table 1, and showing the mean weights of the fetuses at ten
ages during gestation
.\GE IN DAYS
ND.MBER OF FETUSES
AVERAGE WEIGHT OF FETUS IN GRAMS
RATE OF INCREASE IN WEIGHT
13
34 44 37 44 21 43 30 25 42 10
'0.040 0.112 0.168 0.310 0.548 1.000 1.580 2.630 3.980 4.630
per cent
14
179
15
50
16
83
17
77
18
83
19
58
20
65
21
51
22
16
GROWTH OF FETUS OF THE RAT 677
Fetus of albino rat Weight in granns
13 14 15 16 17 18 19 20 21 22 Days
Chart 1 Showing the mean weights in grams of the fetuses of the albino rat at 24-hour intervals, from the 13th to the 22nd- day of gestation, inclusive, based on the values given in table 2. The weights here given would be slightly increased if the mothers had all been fed on 'bread and milk,' and slightly diminished had the mothers all been fed on 'scrap.'
that of man we find the actual time ratio to be .1 to 13. Applying this ratio to the human records, the 13th daj^ of gestation in the rat would correspond to the 169th day in man. If the weights in the two species — man and the rat — correspond during the gestation period, then at birth, 4.6 grams for the rat would represent 3250 grams for man. At the 13th day the rat fetus weighs 0.04 grams, so by proportion we would obtain a weight for the human fetus of 283 grams at 169 days. Speaking broadly, this seems to be too small a fetal weight for man (Jackson '09).
678
J. M. STOTSENBURG
TABLE 3 Crown-rump length of fetus in millimeters; scrap diet only
SERIAL XO.
AGE IX DATS
XrXIBER IX LITTER
.WERAGE WEIGHT OF
FETCS IN GRAMS
AVERAGE CROWN-RCMP LENGTH IN MM.
rAnge of length
IN MM.
42
14
8
0.093
9.5
9 -10
43
15
12
0.107
9.4
9 -10
38
15
8
0.218
12.1
12 -12.5
41
16 17 18 19
11
7 9 8
0.322 0.525 0.947 1.490
13.0 16.3 19.1
22.7
12.5-13
40
16 -17
36
18 -21
37
20.5-24
35
20
10
2.510
27.7
24 -32
34
21
9
4.070
36.7
35 -39
44
22
10
4.630
39.2
36 -41
It might be interpreted, however, as evidence for a still greater slowness of growth in the rat during the earlier period of gestation, but any attempt to follow the matter further must await better data on the weight of the human fetus at different ages.
The data contained in table 1 are sufficient to justify some further discussion of the characters and relations of the fetus during the period covered by the observations.
It is often desirable to have the data for fetal weight correlated with fetal length. Table 3 gives in a number of cases the crown-rump measurements of the fresh fetus after the membranes had been cleared away. The litters from scrap-fed mothers only have been used for this purpose.
A word of explanation touching the diets is here in place. In the course of these observations the general diet of the colony was changed. The earlier litters were from females fed on a diet in which bread and milk were the chief features — this is designated 'bread and milk,' while the later records were from females fed on a 'scrap' diet — i.e., table scraps from which materials known to be injurious to the rats had been excluded — ^this is designated as 'scrap.'
Two differences which are apparently related to the diet, appear, as can be seen from table 4, in which the data are arranged according to the diet of the mother.
GROWTH OF FETUS OF THE RAT
679
In seven out of eight comparisons the litter number for the scrap diet is greater, while the average weight of the fetus is less for the scrap diet litters. Also in seven out of eight comparisons; the exceptional records are in parentheses. The lower average weights of the fetuses from the scrap-fed rats are about what we should expect to follow from the increase in the number in the litter (King '15) but the appearance of the larger number per litter in the scrap-fed series was an unexpected result. The distribution of the litter size (number of individuals) is a fairly symmetrical one and is shown in chart 2.
The mean value for the litter size is 8.8. Our laboratorjrecords show a mean litter size of about 7.0 for the general
TABLE 4
Effect of diet on the number oj fetuses in the litter and on the mean fetus weight
AGE, DAYS
DIET
NO. OF
UTTERS
LITTER DESIGX,\TION
.AVERAGE NO.
IN LITTER
AVE. WT. OF FETUS, GRAMS
13
14
15
16
17 18 '.
Bread and milk Scrap
Bread and milk Scrap
Bread and milk Scrap
Bread and milk Scrap
Bread and milk Scrap
Bread and milk Scrap
Bread and milk Scrap
Bread and milk Scrap
3
1
4 1
2
2
4 1
2
1
2 2
2 2
1
2
(21) (22) (23) (39)
(20 (17) (14) (24) (42)
(7) (16) (38) (43)
(12) (25) (15) (5) (41)
(6) (18) (40)
(13) (4)
(36) (30)
(2) (19)
(37) (31)
(26)
(33) (35)
9
(7)
8 14
8.5 9.5
8 11
6.5
8
8.5 13
6 9
8 8.5
0.041
0.037
0.117 0.093
0.174 0.162
0,305 0.322
0.560 0.525
1.05
19
20
0.95
1.59 1.58
2.95
2.47
General average: Bread and milk. Scrap
7.8 10.0
0.848 0.781
680 J. M. STOTSENBURG
population of the colony (King and Stotsenburg '15). In the present case, however, it is to be remembered first, that we are dealing only with second litters, which tend to be large (King and Stotsenburg '15) and second, that there may be some tendency also for the fetuses to be more numerous than are the young actually born.
Litter size— albino rat Frequency
/\
\
7 8 9 10 11 12 14 16
Number of fetuses per litter
Chart 2 Showing the frequency, as indicated on the ordinate, of the litters containing from 3 to 16 fetuses, as indicated on the abscissa. The mean value is 8.8 fetuses per litter.
DISTRIBUTION OF THE FETUSES BETWEEN THE TWO HORNS
OF THE UTERUS
This distribution was noted in the case of 20 litters and the details are given in table 5. In /o/o there were 90 fetuses in the right horn, 94 in the left. In two cases the right horn was sterile, and in four cases there was the same number of fetuses in each horn. If the comparison is made between the average weights of the fetuses in the two horns it is seen that in nine out of the fourteen possible comparisons the average weight of the fetus is greater in the horn containing the smaller number.
GROWTH OF FETUS OF THE RAT
681
THE WEIGHT OF THE FETUS ACCORDING TO POSITION IN HORN
An examination of the fetal weights according to the position of the fetus in the horn has not revealed any correlation. At the same time, inspection of table 1 shows that marked variations in the weights of the fetuses in the same htter and even within the same horn may occur.
For the growth of the Albino from the beginning to the end of gestation, we already have the observations of Huber ('15) giving weight data for the first 3 days and 17 hours, so that there still remains to be filled the interval of about 10 days between the end of Huber 's weight records, and the 13th day, which marks the beginning of the records here presented.
TABLE 5
Showing the number of fetuses in each horn of the uterus and their average iceight
.\GE OF
LITTER
NO. IN LITTER
LEFT
HORN
RIGHT HORN
Days
Hours
Xo.
Weight
No.
Weight
21
13
13
13
13
14
14
14
14
15
15
15 ■
16
16
17
19
19
19
20
21
22
2 2 2
2 2 2
10
10
10
8
7
6
11
14
11
8
12
10
11
7
4
8
8
10
9
10
3 7 6 3
7 3 5 5 6 5 6 5 6 4 4 1 4 6 6 2
0.041
0.043
0.044
0.039
0.121
0.126
0.116
0.094
0.184
0.210
0.112
0.325
0.310
0.542
1.36
1.93
1.50
2.45
4.05
4.58
7 3 4 5
3 6 9 5 3 6 5 5 3
7 4
0.041
22
0.042
23
0.048
39
17
0.036
20
0.105
24
42
0.100 0.092
16
38
43
0.176 0.229 0.104
15
41
18
9
0.302 0.337 0.519
19
1.80
37
35
1.47 2.62
34
4.08
44
4.63
Total
94
9i
T
682 J. M. STOTSENBURG
LITERATURE CITED
DoxALDSox, H. H. 1906 A comparison of the white rat w ith man in respect to the growth of the entire body. Boas Anniversary Volume, pp. 5-26. G. E. Stechert & Co., New York.
HuBER, G. Carl 1915 The development of the albino rat (I\Ius norvegicus albinus). Part I. From the pronuclear stage to the stage of mesoderm anlage; end of the first to the end of the ninth day. Jour. ^Morph., vol. 26, pp. 247-358.
Jacksox, C. M. 1909 On the prenatal growth of the human body and the relative growth of the various organs and parts. Am. Jour. Anat., vol. 9, pp. 119-161.
Keibel, Fraxz, axd Mall, Fraxklix P. 1910, 1912 r^Ianual of human embryology. 2 vols. J. B. Lippincott Co., Philadelphia.
KiXG, Helex D. 1915 On the weight of the albino rat at birth and the factors that influence it. Anat. Rec, vol. 9, pp. 213-231.
KiXG, Helen D., and Stotsexburg, J. M. 1915 On the normal sex ratio and the size of the litter in the albino rat (Mus norvegicus albinus). Anat. Rec, vol. 9, pp. 403-420.
Lowrey, Lawsox, G. 1911 Prenatal growth of the pig. Am. Jour Anat., vol. 12, pp. 107-138.
a^
OBSERVATIONS ON THE DIFFERENTIATION OF
THE GRANULES IN THE EOSINOPHILIC
LEUCOCYTES OF THE BONE-ALARROW
OF THE ADULT RABBIT
PEELIMINARY NOTE A. R. RINGOEN
From the Histological Laboratory of the Department of Animal Biology, University of Minnesota, Minneapolis
In a recent paper^ the writer has described the mast myelocytes and mast leucocytes in the bone-marrow of the rabbit. No evidence could be found in support of the theory that the mast cell of the rabbit represents a young or ' unripe ' eosinophil or special cell,- or for the view expressed by Proscher that the mast granules are products of a mucoid degeneration of the spongioplasm of a lymphocyte. The preparations show that mast leucocytes are true granular cells, equivalent in all respects to the other granular cells, wdth both the myelocyte and fully differentiated forms represented in the marrow.
The supposed relationship of mast leucocytes to eosinophil and special leucocytes necessitated a detailed study of their development also. The same material which was used for the study of the mast leucocytes was found to be excellent for the investigation of the other granulocytes, and of these the eosinophil leucocytes were of special interest on account of the many theories regarding the origin of their granules.
The exact origin of the eosinophil leucocytes of mammals has been the subject of considerable investigation, and up to the present day there is no unanimit}^ of opinion among investigators as to the source and nature of the granules of these cells. The literature relative to the subject is enormous; it shows that
1 Anat. Rec, vol. 9, no. 3, 1915.
-As claimed by Pappenheim's students: Benacchio, Kardos, and Szecsi.
683
THE AXATOMICAI. RECORD, VOL. 9, NO. 9, SEPTEMBER, 1915
684 A. R. RINGOEN
the most divergent theories and explanations are held with reference to the histogenesis of these cells.
Various investigators of the eosinophil problem have come to recognize that the supply of eosinophilic leucocytes in the adult animal is not necessarily limited to mitosis of pre-existing eosinophilic myelocytes, but that a heteroplastic means of regeneration of these cells must also be taken into consideration. The investigations of Tettenhamar ('93), Sacharoff ('95), Brown ('98), Weidenreich ('01), Howard and Perkins ('02), AscoH ('04), Maximow ('09), Pappenheim ('09), Badertscher ('13), Downey ('13, '14), and Barbano ('14), have shown the importance of the heteroplastic form of development. Downey confined his studies to the differentiation of the eosinophilic leucocytes of the bone-marrow of the guinea-pig and found that heteroplastic development of these cells from non-granular cells is by no means an exceptional process.
Haematologists who believe in a heteroplastic form of differentiation of eosinophils, however, do not agree on the derivation of the granules, and consequently a number of views have been advanced which have sought to account for their source and nature. The older view, that eosinophils are formed from polymorphonuclear neutrophils by a direct transformation of their granules, has within recent years lost support. Brown ('98) has described such a direct transformation of neutrophil granules into eosinophil granules in human muscle infected with trichinae. This theory has been advanced a number of times by different investigators, but it appears that the conclusion for such a direct transformation of one type of granule into another type is not based on sufficient evidence. The mere fact that Brown found a marked increase in the number of eosinophils, with a corresponding decrease in the number of polymorphonuclear neutrophils, does not necessarily indicate that a direct transformation process was going on.
The eosinophils are so wideh^ distributed throughout the tissues, and in certain pathological conditions become so numerous, that it seems quite reasonable to believe that they multiply in these situations by homoplastic means also, since the cor
ORIGIN OF EOSINOPHIL GRANULES 685
responding myelocytes are often present (Gulland and Goodall, Herzog).
At the present day the Uterature relative to the origin of the eosinophil granules during heteroplastic differentiation of eosinophilic leucocytes may be said, in the main, to be rather sharply centered about the belief that the granules are of an exogenous origin. Weidenreich is the chief exponent of this theory; he believes that the granules of all eosinophils are hemoglobincontaining products of degenerated erythrocytes,^ i.e., he believes that the granules are not the products of protoplasmic activities of the cells which contain them.^ Weidenreich states explicitly that this is the only source of the eosinophil granules, and furthermore, that there are no observations on record which prove a gradual differentiation of eosinophil granules in the protoplasm of non-granular cells, ^
The theory that the eosinophil granules are of an exogenous origin, and that they are not related to hemoglobin, has been given additional support by the recent observations of Badertscher ('13), who believes that the granules of eosinophil leucocytes seen in the neighborhood of degenerating muscle fibres and erythrocytes in Salamandra atra during metamorphosis are products of the degenerating fibers and red cells, and that they are, therefore, related to hemoglobin or its dissociation products.
Weidenreich "s hemoglobin theory has met with many staunch supporters. Downey ('13, '14), however, has recently taken exception to the theory in so far as the eosinophils of the bonemarrow^ are concerned. He states that his preparations showed nothing which would indicate hemoglobin was concerned in the elaboration of these granules; he believes that the eosinophil
^Anat. Rec, 1910, vol. 4, p. 327, "Die eosinophilen Granula der Saugetiere sind als exogene Plasmaeinlagerung zu bezeichnen und zwar als hamoglobinhaltige Telle, grosstenteils von Erj-throcj^ten herrlihrend, die durch hamolytische Yorgange zerstort, oder in toto phagocytiert wurden."
Anat. Anz., 1901-1902, vol. 20, p. 197, "Die eosinophilen Leucocyten sind also nichts anderes als sog. Lj-mphocyten, welche die durch den Zerfall roter Blutkorperchen entstehenden feinen Trummer in ihren Plasmaleib aufnehmen, wobei ihr Kern in die pohonorphe Form iibergeht."
5 Die Leucocyten und Verwandte Zellformen, p. 250.
686 A. R. RINGOEN
granules are real intracellular formations (endogenous differentiations), which are the products of specific activities of the protoplasm. Downey's view is, therefore, in strict opposition to that of Weidenreich and others, who believe in an exogenous origin for all eosinophil granules.
Barbano ('14) has also recently favored the view that the granules are endogenous formations; he believes that they are secretory granules which may be extruded from the cells. His conclusions are based entirely on a studj^ of local eosinophilia in various pathologic conditions.
Under normal conditions the eosinophil leucocytes have been reported bj^ various investigators as being widely distributed throughout the tissues, appearing in great numbers in the gastrointestinal tract, in the walls of the trachea, in the connective tissue surrounding the bronchi, in lymph glands, the thj-mus, and hemolymph glands. Under certain conditions the number of these cells may be materially increased, so that great numbers of them may appear throughout the section. It now seems certain that many of them are the products of local development, while others have emigrated from the vessels.
The local development of eosinophils, however, is still denied by many authors. Barbano believes that, in local eosinophilia, he can exclude the emigration of myelocytes from the vessels, and that the cells which are found in these local accumulations are, therefore, new differentiations from non-granular cells. The latter are typical small and large lymphocytes. That they may differentiate into granulocytes is shown by the fact that Barbano finds many mononuclear eosinophils whose nuclei are indentical with those of the lymphocytes. That lymphocytes, especially small lymphocytes, are concerned in the production of acidophil granules was shown also by Downey and Weidenreich,^ Howard and Perkins,^ and others.
^ Downey, H., andWeidenreifh, Fr. 1912, t'bcr die Bildung der Lj^mphozyten in Lymphdriisen und Milz. Arch. f. mikr. Anat., Bd. §0.
" Howard, W. T., and Perkins, R. G. 1902, Observations on the origin and occurrence of cells with eosinophile granulations in normal and pathological tissue. The Johns Hopkins Hospital Reports, vol. 10.
ORIGIN OF EOSINOPHIL GRANULES 687
The theory that the eosinophil granules are derived from phagocytosed material (Weidenreich, Badertscher, Brown, and a great many others), is based largely upon the presence of free eosin-staining granules among erythrocytes and muscle tissue which are undergoing degeneration. These free granules are believed to be ingested by lymphocytes which are then converted into eosinophils, many of which are distinctly mononuclear, ^
Benacchio ('09), in considering the bone-marrow of the rabbit, was primarily interested in the origin of the mast leucocytes of this animal; consequently, he did not make detailed investigations as to the histogenesis of eosinophils and special cells. He concluded, however, that the myelocytes with basophilic granules which he found in great numbers in his preparations were not real mast myelocytes, but simply 'unripe' stages of young eosinophil and special cells. ^ Pappenheim, Kardos, and Szecsi also regard all the basophilic granulocytes in the marrow of the rabbit as immature eosinophils and special cells. Many haematologists, in fact, believe that the early myelocyte stages of eosinophil and special cells have a granulation, which has a predominant basophilic element when first differentiated. These granules, however, do not retain their basophilic element, as they should if they were real mast granules, but they undergo a gradual transformation or 'ripening process,' during which thej^ change their staining reactions and are finallj^ transformed into eosinophil granules. That such transformation actually takes place has been reported by Ehrlich ('78, '79), Schwarze ('80), Hirschfeld ('98), Benacchio ('09), Kardos ('09), Pappenheim and Szecsi ('09), Maximow ('10, '13), and Downey ('13, '14).
The early basophilic granules of eosinophils and special cells, however, are in no way related nor similar to the basophilic
Sternberg claims (Ueber die Entstehung der eosinophilen Zellen. Beitr. z. path. Anat. und allg. Pathol., Bd. 57) that he can distinguish real eosinophil granules from erythrocyte fragments.
5 Mast myelocytes or fully differentiated mast leucocytes could not be found when Benacchio's methods were used. For the detection of these cells in the marrow of the rabbit, methods of technique m.ust be used in w^hich water is absolutely avoided. After lucidol-acetone fixation, however, the granules are more resistant to water and are able to withstand its action while being stained in watery staining combinations.
A. R. RINGOEN
granules of mast leucocytes. The mast granules are endowed with certain specific and diagnostic characters at their first appearance within the cell body, and they are readily distinguished from the granules of eosinophils and special myelocytes. This is contrary to the statements of Weidenreich, who believes that if all of the granules of the eosinophil and special leucocytes were basophilic when first formed, it would be impossible to distinguish their myelocytes from the basophilic myelocytes of mast leucocytes. I have found, however, that mast granules can always be distinguished from the granules of other basophilic myelocytes, provided that the proper methods of fixation have been applied to the marrow.
Mention has already been made of the fact that the methods of technique emploj^ed by Benacchio failed completely to demonstrate mast leucocytes, although his preparations did show great numbers of eosinophil and special cells. These results would indicate that the chemical composition of mast granules in the rabbit, at least, is quite different from that of the ordinary basophilic granules of young eosinophils and special cells. Mast granules are more soluble in water than are the basophilic granules of either eosinophil or special myelocytes. As far as resistance to water is concerned, it also appears that the granules of the mast leucocytes of the circulating blood are quite different from the granules of the mast myelocytes and the more fully difTerentiated mast cells of the marrow. The granules of the mast leucocytes of the blood are less soluble in water than are the granules of the mast myelocytes and the corresponding leucocytes of the marrow. This difference in the constitution of the granules is shown by the fact that the most ordinary methods will preserve the granules of the blood mast cells, while the strict elimination of water is necessary for the preservation of the granules of both the mast myelocytes and the fully difTerentiated mast leucocytes found in the bone-marrow. In the older mast leucocytes, or those found in the blood stream, there may be some chemical change within the cell body — initiated by the blood plasma — as soon as the leucocytes are thrown out into the circulation, whiclj in turn acts upon the mast granules
ORIGIN OF EOSINOPHIL GRANULES 689
changing their composition to a greater or less extent and thus rendering them more resistant to the action of water. The mononuclear mast leucocytes are never found in the blood of the normal adult rabbit, but are confirmed under ordinary conditions to the marrow. In these cells the basophilic granules are very sensitive to the action of water. As the number of granules increase the nucleus gradually becomes polymorphous, while in the fully differentiated mast leucocyte of the circulating blood the nucleus is very polymorphous and the granules are comparatively resistant to the action of water.
The presence of basophilic granules in eosinophil myelocytes is no longer doubted nor questioned by haematologists, their occurrence having been reported by a number of investigators, including Arnold, Hirschfeld, Hesse, Benacchio, Kardos, and others. Maximo w and Pappenheim have called particular attention to the very decided basophila of young eosinophil and special granules in the eosinophils and special cells of the rabbit.
An early 'primitive' granulation which is also basophilic has been reported by several investigators. Pappenheim regards it as an early or 'prodromal' granulation that is not related to the final eosinophil or special granulation which appears later as a new difi'erentiation. According to Pappenheim the 'prodromal' granulation is derived from the nucleus of the cell and is basophilic; it disappears when the specific granulation appears later. He believes that the eosinophil granules are a new development and that they too are basophilic when first differentiated. Weidenreich admits the presence of basophilic granules in some of the eosinophils, but claims that they are either fragments of the nucleus or endogenous differentiations which are in no way related to the eosinophil granules. Alaximow described a primitive azurophil granulation in eosinophil myelocytes, and he also claims that it is not related to the specific granulation which is developed later. Hertz and Pappenheim have also described an azurophil granulation in the leukoblasts and myelocytes of myelogenous leukemia.
690 A. R. RINGOEN
As early as 1895, Arnold, in studying the morphological features of the cells of the marrow of the rabbit, observed that many granulocytes contained basophilic granules. He also noticed that many myelocytes of the same type contained both basophilic and acidophilic granules within the same cell body. Arnold thought it probable that the basophilic granules were transformed into acidophil granules, since so many of the former showed considerable variation in their staining reactions even within the same cell body. Arnold, in fact, seems to have made the correct interpretation of his preparations; however, he gave no detailed descriptions of the gradual changes in the morphology of the granules during the transformation process; neither did he work out in a detailed manner the gradual changes in the staining reactions of the early basophilic granules.
Other investigators of the bone-marrow have also noted changes in the staining reactions of the basophilic granules of eosinophilic myelocytes and have, in fact, reported a transformation of the basophilic into the acidophilic type, but they have not made a particular study of the transformation process itself and of the other phenomena which are seen to accompany it. The life-history of the eosinophil granule has not been worked out with sufficient detail to warrant the statement that all basophilic granules in eosinophilic myelocytes represent unripe eosinophil granules. No attempt has been made to give minute descriptions of the gradual changes in staining reactions which the basophilic granules pass through in becoming transformed into acidophil granules.
Maximow, in his earlier investigations of the bone-marrow, was interested primarily in the role which lymphocytes played in the elaboration of granules in their protoplasm, and in the regeneration of the granular leucocytes, consequently, he also failed to make a detailed study of the histogenesis of the eosinophil series. In 1913, however, his figures show that the very youngest granules of eosinophil myelocytes are basophilic when first differentiated, and that they are gradually transformed into acidophil granules, although Maximow does not emphasize this point in particular. He used cover-glass preparations (Helly
ORIGIN OF EOSINOPHIL GRANULES 691
fixation followed by staining in alcoholic thionin) and found that the granules of the eosinophils showed considerable variation in their staining reactions depending on the stage of differentiation that they had attained. The older or more fully differentiated granules were stained green in the alcoholic thionin, while the myelocytes contained granules which were stained blue, in addition to other granules which were of a green color.
Downey ('13, '14) has made a special study of the life-history of the eosinophil granules based on the variations in staining reactions and form of these granules. He also finds that the eosinophil granules are basophilic when first differentiated. Gradually, however, these early basophilic granules change their staining reactions, taking on the eosin of the stain when subjected to the action of the indulin-aurantia-eosin staining combination, instead of staining in the indulin, as Downey found to be the case when the granules were first differentiated. Furthermore, he found that in the later stages of differentiation the granules lost their avidity for the eosin of the staining mixture and stained with the aurantia of the same staining combination. In regard to the staining reactions, Downey states that
Som3 of the granules change their staining reactions while they are still small and basophilic, while others remain basophilic until they have reached a size even greater than that of the fully differentiated granule before such change takes place. That these larger granules do not disappear, and that they are transformed directly into the eosinophil granules is shown by the fact that many of the largest ones are stained in the acid component of the staining mixture, while others are of a mixed tone.
These gradual progressive changes in the staining reactions of basophilic granules in eosinophilic myelocytes together with changes in their shape and size have led Downey to conclude that as far as the bone-marrow is concerned, eosinophil granules are true endogenous formations resulting from special activities of the protoplasm, and that they are not related to hemoglobin or its dissociation products, as maintained by Weidenreich and others.
Barbano, who also regards the eosinophil granules as real endogenous differentiations, has also noted differences in the
692 A. R. RINGOEN
staining reactions of eosinophil granules, although he does not give detailed descriptions of these changes. In using the hemalum and eosin staining combination, he found that there were great individual differences in the avidity with which the granules stained in the eosin. In an epithelioma of the uterus, also in other similar cases, Barbano found that many of the granules in cells of the lymphocyte type stained only slighth^ in the eosin, while others scarcely stained at all with this dye. The latter appeared clear and refractive and were colored by only the slightest tinge of eosin. Bacbano, how^ever, does not interpret these differences in staining reaction as indicating a gradual 'ripening' process of the eosinophil granules. He believes that h'mphocj^tes under certain conditions differentiate granules which are typical eosinophil granules from the very beginning, although in some instances the early formed granules may not exhibit a remarkable afhnit}' for the acid component of the staining mixture.
Downey, however, has shown that the first granules of the eosinophil myelocytes are not the typical granules of the fully differentiated cells. He believes that the fully differentiated granule is the end product of a series of gradual, complex changes in chemical constitution, as well as in form and size, of the small 'unripe' granule which first appeared in the protoplasm of the myelocyte.
In view of the varied opinions concerning the origin and nature of the eosinophil granules, and since the majority of those authors who believe in the hemoglobin nature of the granules have based their studies on local eosinophilia, it is of importance that new studies of the bone-marrow be undertaken, with their results in mind, in order to determine whether the eosinophils of the marrow develop under conditions which might indicate that their granules are also related to hemoglobin products. Fortunately I have had the great pleasure of carrying on an investigation of this kind under the direction of Professor Downey, to whom I am greatly indebted. As far as the marrow of the rabbit is concerned, my observations on the life-history of eosinophil granules do not differ essentially from those of Professor
ORIGIN OF EOSINOPHIL GRANULES 693
Downey. In strict corroboration with his findings, my preparations showed nothing which would indicate that hemoglobin was a contributing factor in the formation of these granules.
Many of the myelocytes, in smears prepared according to Pappenheim's method, i° contain basophilic granules only. These cells might easily- be taken for mast myelocytes if it were not for the fact that other similar cells contain oxyphilic granules also. The oxyphilic granules may be very numerous or there may be only a few of them in any one cell, and, in general, the presence of a greater number of oxyphilic granules in a cell seems to be conditioned on a corresponding diminution in the number of basophilic granules. This, together with the fact that many of the granules are intermediate in staining reaction, shows that there is a gradual change in the staining reaction of the granules from basophilic to oxyphilic. The intermediate stages in this gradual process of 'ripening' are so numerous that there is no question but what all of the basophilic granules seen in preparations prepared according to this method are eventually transformed into granules whose chemical constitution becomes such that they finally stain only in the acid component of the staining combination. Such granules are surely not mast granules, for the latter have never been known to change their staining reactions in this way. They remain basophilic throughout their existence.
No other type of basophilic granules, besides those which eventually become oxyphilic, could be found in the preparations prepared by the above mentioned method. We must, therefore, give Benacchio credit for a correct interpretation of the nature of these granules when he stated that the cells which contain them are young 'unripe' eosinophil and special leucocj^tes. Benacchio, however, did not go into the details of the gradual differentiation and transformation of these granules, and such detailed study would have been impossible with his methods, because with them the vast majority of the cells are distorted, their granules are swollen, and the outlines of the nucleus are usually indistinct.
1° Folia Haem., Archiv, Bd. 13.
694 A. R. RINGOEN
The results obtained by Kardos ('09), in working with sections of bone-marrow of the rabbit fixed in 100 per cent alcohol and in Helly's mixture, are difficult to understand. He found neither mast cells, nor cells of any kind which contained basophilic granules. Contrary to the findings of Kardos, I find that in sections of bone-marrow (material fixed in 100 per cent alcohol and stained in alcoholic thionin) myelocytes with basophilic granules are very numerous, and furthermore, that in these same preparations it is also possible to demonstrate mast myelocytes and fully differentiated mast leucocytes. Sections stained in ]May-Giemsa not only show many granulocj^tes which contain basophilic granules, but also other cells in which both basophilic and acidophilic granules are intermixed, and still others in which all of the granules are of the acidophil type. The preparations also show the various other types of granulocytes which are characteristic of the marrow of the rabbit. The basophilic granules of the eosinophil and special myelocytes are also seen in sections of material fixed in Helly's fluid. True mast granules, however, are not preserved by this method, and with their granules dissolved it is difficult to identify the mast cells.
From the above it is seen that it is not a difficult matter to demonstrate basophilic granules in the marrow of the rabbit even with the most ordinary methods. These granules, however, are not the mast granules. If the latter are also desired it is necessary to avoid the use of fixing fluids which contain water. P'or material fixed in bulk absolute alcohol proved most satisfactory, and for smears the lucidol-acetone method of Szecsi.
Of the various methods tried for working out the life-history of the eosinophil granule, none gave sharper and more decisive results than did Benacchio's method of staining bone-marrow smears in a mixture of indulin-aurantia-eosin. These smears were fixed in Helly's fixative for fifteen minutes and then washed in running water from three to four hours. The preparations were dehydrated and finally stained in the indulin-aurantia-eosin mixture, in a thermostat at a temperature of 38°C. This gave excellent results for the study of both the eosinophil and special myelocytes. The chief advantage of this method is that it
ORIGIN OF EOSINOPHIL GRANULES 695
practicallj' eliminates the special myelocytes from our consideration, at least in the later myelocyte stages, since the granules of the special cells at no time in their evolution show any great affinity for the eosin of the staining mixture. The vast majority of the special cells have dark-grayish-black granules, but in the youngest mj^elocj'tes these granules also have a slight affinity for the eosin of the mixture, a condition which often makes it difficult to distinguish the earliest eosinophil myelocytes from those of the special cells. The special granules, however, very soon develop their strong affinity for the indulin to the complete exclusion of the eosin, while many of the granules of the eosinophil myelocytes become strongly oxyphilic, causing them to stain intensely with the eosin.
In the earlier stages of the eosinophil myelocytes in which there are only a few acidophil granules, there are a great many small basophilic granules, with a few medium-sized and large basophilic granules scattered among them. In the later nwelocy te stages, however, most of the granules are large and many of them are acidophilic. However, the later stages, including those in which most of the granules are acidophilic, contain a few small basophilic granules as well as a few larger ones. It is very probable that these smaller granules are the youngest ones; all of them probabh^ increase in size before developing an affinity for the eosin of the stain. The presence of the small indulinophilic granules in the later myelocyte stages in which the eosinophilic granules are very numerous could not be accounted for by Hirschfeld. Downey, however, believes that these smaller granules represent recent differentiations, since they are small and still basophilic.
In addition to the changes in the staining reactions of young eosinophil granules which at first are basophilic, there is further evidence in favor of the view that these young basophilic granules are the precursors of eosinophil granules. T]^e life-history of the eosinophil granule, in the rabbit at least, is not completed with the change in staining reaction, since the granules in the fully differentiated eosinophil leucocytes of the blood of this animal are typicalh^ spindle shaped, while the early granules
696 A. R. RINGOEN
are spherical. Downey has already shown (as far as the eosinophil leucocytes of the guinea-pig are concerned, and the histogenesis of these cells seems to be very similar in the bone-marrow of the rabbit) that in addition to the changes in the staining reactions there are further changes in the morphological features of these granules which prove conclusively that the basophilic granules are in reality the 3'ounger eosinophil granules.
The granules of both the eosinophil and special leucocytes are very small when they are first formed, and both types of granules are stained dark with the indulin, with possibly a slight tinting with the eosin of the mixture. This uniformity in size and staining reaction frequently make it impossible to distinguish the earliest special myelocytes from those of the eosinophil leucocytes. ]Maximow encountered the same difficulty in rabbit embryos, but claims that he could always distinguish the two types of cells in the marrow of post-natal animals. In the latter he finds that the first eosinophil granules are from the very beginning brighter and coarser than are those of the special cells; at first the youngest granules possess a clear basohpilic quota and stain a bluish tinge, but are not metachromatic as are the granules in the special cells.. In spite of these slight differences there are times when it is almost impossible to classify basophilic myelocytes with any degree of certainty. ]\Iaximow admits that in alcoholic thionin preparations it is very difficult to distinguish between eosinophil and special myelocytes. He states, however, that the granules of the eosinophils can often be distinguished from the granules of the special cells by the fact that some of the eosinophil granules enlarge very rapidly "und dass man infolgedessen schon in den noch ganz granulaarmen Zellen typische, grobe, gliinzende azidophile Granula neben nur sehr sparlichen feineren erblicken kann. "
Downey also states that when the granules are few in number and of small dimensions there may be considerable difficulty in distinguishing between eosinophil and special myelocytes. He also found that the granules of eosinophil myelocytes enlarged shortly after they were differentiated and that some of them changed their staining reactions, while others remained basophilic
ORIGIN OF EOSINOPHIL GRANULES 697
for a longer period of time. The basophilic granules in young special myelocytes, on the other hand, remained basophilic for a longer period of time which was followed by a rapid change in staining reaction involving all of the granules.
The same difficulty in determining the exact position of the very earliest myelocytes, i.e., those with a few graimles only, was encountered in the present investigation.
In the indulin-aurantia-eosin preparations there is no difficulty in finding basophilic myelocytes which contain a few decidedly oxyphilic granules. Myelocytes containing the latter can be diagnosed as eosinophil myelocytes, since the granules of the special cells are never stained intensely with the eosin of the mixture. It must be admitted, however, that there were always a few cells which it was difficult to classify. With more intensive study it might be possible to properly place these cells also. However, it is not the object of the writer to describe the morphological features of the very earliest myelocyte stages of eosinophil and special myelocytes. The indulin-aurantia-eosin preparations show beyond all doubt that, in so far as the adult animal is concerned, the bone-marrow contains two distinct types of myelocytes, the precursors of eosinophils and special cells respectively. The inability to diagnose the specific type of basophilic myelocj^te in every case — before some of the granules stain in the eosin — does not invalidate the conclusions in regard to the life-history of the eosinophil granules.
The results of the study outlined above show that these granules are differentiated gradually from the basophilic protoplasm of non-granular cells, and that they pass through a gradual progressive development which is expressed in changes in staining reaction, as well as in shape and size. When first formed they are indulinophilic (with Ehrlich's triglycerine mixture) or basophilic with heterogeneous mixtures containing a basic dye. At first there are only a few granules in the cell, but their number is gradually increased, the youngest granules always being basophilic (or indulinophilic), while the older ones have become distinctly acidophilic. The number of granules which are intermediate in staining reaction, i.e., which have an affinity for
698 A. R. RINGOEN
both the acid and the basic component of the heterogeneous mixtures (or for induHn and eosin of the triglycerine mixture), shows that there- is a gradual transformation from one type of granule to the other, and not a replacement of one kind by another. There is no evidence for Weidenreich's view that the basophilic granules, which he admits are present in the myelocytes of eosinophil leucocytes, are endogenous formations which are not related to the eosinophil granules which are supposed to be developed later from products of hemoglobin dissociation.
The changes in the character of eosinophil granules are similar to those which are seen during the development of the special granules. For the latter these changes are universally conceded to indicate a process of gradual differentiation, progressive development and 'ripening.' It is difficult to see why the same conclusion should not apply to the eosinophil granules, especially when there is nothing to indicate that in the normal bone-marrow fragmenting erythrocytes or other hemoglobin products are in any way concerned in their development.
The development of eosinophil leucocytes in the tissues is a different question. For them there is much evidence to show that their granules are closely related to hemoglobin products. Their granules apparently do not pass through the same series of changes during their development as was outlined above for the eosinophil granules of the bone-marrow, at least no such changes have ever been described. This may be due to the fact that local eosinophilia has never been investigated from this same standpoint. Barbano's recent study of the subject indicates merely that the granules of the tissue eosinophils are quite variable in their staining capacity with eosin. These variations, however, do not seem to be bound with any particular stage in the development of these leucocytes, and there was nothing to show that those granules which had the least affinity for the eosin were the youngest ones. Barbano, however, has peculiar ideas about eosinophil leucocytes and his results do not seem very trustworthy.
Giitig beheves that there are two types of eosinophils, those of the tissues and those of the marrow and blood. The granules of the hematogenous eosinophils are true endogenous differenti
ORIGIN OF EOSINOPHIL GRANULES 699
ations, while those of the tissues are of an exogenous origin. Downey agrees with this conclusion, provided that the observations of Weidenreich and others on the development of eosinophil granules in local eosinophilia are correct. With these questions in mind a renewed study of local eosinophilia would be of great importance.
soniARY
The bone-marrow of the normal adult rabbit shows no evidence for the support of the theory that the eosinophil granules are exogenous formations which are derived from hemoglobin or its dissociation products (Weidenreich and others).
Application of the proper methods of technique, however, show that these granules are real manifestations of protoplasmic activities, and that they are gradualh^ differentiated in the cytoplasm of mononuclear cells (Downey).
The indulin-aurantia-eosin preparations show that the youngest granules of the myelocytes are indulinophilic. They do not remain indulinophilic for any length of time, but pass through a series of progressive evolutionary processes during which they change their shape and staining reactions. Finally they are transformed into tj'pical eosinophilic granules. In the fully differentiated eosinophil leucocyte of the marrow all the basophilic (with Giemsa, etc.) granules have been converted into acidophilic granules and no new basophilic (indulinophilic with trigh'cerin) granules are formed.
These gradual changes in the staining reactions and morphology of basophilic granules in the myelocytes of eosinophil leucocytes must be interpreted as indicative of progressive evolution on the part of the eosinophil granules.
LITERATURE CITED
Arnold, J. 1895 Zur ^Morphologie unci Biologie der Zellen des Knochenmarks.
Virch. Archiv, Bd. 140.
1896 Ueber die feinere Struktur der hamoglobinlosen und hamoglobin haltigen Knochenmarkszellen. Virch. Archiv, Bd. 144. AscoLi, M. 1904 Ueber die Ent.stehung der eosinophilen Leukocyten. Folia
Haem., Bd. 1.
THE AXATOMICAL RECORD, VOL. 9, XO. 9
700 A. R. RINGOEN
Badertscher, J. A. 1913 Muscle degeneration and its relation to the origin
of eosinophile leucocytes in amphibia (Salamandra atra). Am. Jour.
Anat., vol. 15. Barbaxo, C. 1914 Die lokale Eosinophilie. Virch. Archiv, Bd. 217, Heft 3. Bexacchio, G. 1911 Gibt es bei Meerschw-einchen und Kaninchen ISIastrnj-e locyten und stammen die basophilgekornten Blutmastzellen aus dem
Knoehenmark? Folia Haem., Archiv, Bd. 11. Brown, T. R. 1898 Studies on trichinosis, with special reference to the increase
of the eosinophilic cells in the blood and muscle, the origin of these
cells and their diagnostic importance. Jour. Exp. Med., vol. 3. Browxixg, C. H. 1905 Observations on the development of the granular
leucocytes in the human foetus. Jr. Path, and Bacter., vol. 10. DowxEV, H. 1914 a Heteroplastic development of eosinophil leucocytes and of
haematogenous mast ceils in bone marrow of guinea-pig. Anat. Rec,
vol. 8, no. 2.
1914 b The origin and development of eosinophil leucocytes and of
haematogenous mast cells in the bone-marrow of adult guinea-pig.
Folia Haem., Archiv, Bd. 19. Ehrlich. p. 1891 Farbenanalj^tische I'ntersuchungen zur Histologic und
Klinik des Blutes. Berlin, August Hirschwald. GuTiG, K. 1907 Ein Eeitrag zur Morphologic des Schweineblutes. Arch. f.
mikr. Anat., Bd. 70. Herzog, G. 1914 tjber adventitielle Zellen und liber die Entstehung von
granulierten Elementen. Verh. d. deutsch. Pathol. Gesellsch., Bd. 17. He.s.se, F. 1902 Zur Kenntniss der Granula der Zellen des Knochenmarks,
bezw. der Leukocyten. Virch. Archiv, Bd. 167. HiRSCHFELD, H. 1898 Zur Kenntniss der Histogenese der granulirten Knochen markzellen. Virch. Archiv, Bd. 153. Kardcs, E. 1911 t^er die Entstehung der Blutmastzellen aus dem Knoehenmark. Folia Haem., Archiv, Bd. 11. INIaximow, A. 1906 Uber die Zellformen des lockeren Bindegewebes. Archiv.
f. mikr. Anat., Bd. 67.
1907 E.xperimentelle Untersiichungon zur pos-ttotalen Histogenese
des mj'eloiden Gewebes. Eeitr. z. path. Anat. und allg. Pathol.,
Bd. 41.
1910 Die embryonale Histogenese des Knochenmarks der Saugetiere.
Archiv. f. mikr. Anat., Bd. 76.
1913 Untersuchungen uber Blut und Bindegewebe. VI. Uber Blutmastzellen. Archiv. f. mikr. Anat., Bd. 83, Abt. 1. Opie, E. L. 1904 a The occurrence of cells with eosinophile granulation and
their relation to nutrition. Am. Jour. Med. Science, vol. 127.
1904 b An exi)erimental study of the relation of cells with eosinophile
granulation to infection with an animal parasite (Trichina spiralis;.
Am. Jour. Med. Science, vol. 127. Pappexheim, a. 1905 Zur Frage der Entstehung eosinophiler Leukozyten.
Folia Haem., Bd. 2.
ORIGIN OF EOSINOPHIL GRANULES 701
Pappexheim, a. 1909 t'ber die Deutung and Bedeutung einkerniger Leukozytenformen in entziindlichen Zellanhaufungen mit besonderer Ri'icksicht auf die lokale Eosinophilie. Folia Haem., Bd. 8.
Pappexheim, A., und Szecsi, St. 1912 Hamozj'tologische Beobachtungen bei experimenteller Saponinvergiftung der Kaninchen. Folia Haem., Bd. 13.
Pro.scher, Fr. 1909 t'ber Experimentelle ba.sophile Leukozj'tose beim Kaninchen. P'olia Haem., Bd. 7.
RiXGOEN', A. R. 1915 Ob.servations on the origin of the mast leucocytes of the adult rabbit. Anat. Rec, vol. 9, no. 3.
Sacharoff, X. 189o Uber die Entstehung der eosinophilen Granulationendes Blutes. Archiv. f. mikr. Anat., Bd. 45.
ScHWARZE, G. 1880 Ueber eosinophile Zellen. Inaug. Diss., Farbenanalytische Untersuchungen, Berlin.
SzECSi, St. 1913 Lucidol. ein neues Fi.xiermittel. Deutsche Medizinische Wochenschrift, no. .33.
Tettexhamar, E. 1893 Ueber die Entstehung der acidophilen Leukozytengranula aus degenerirender Kernsubstantz. Anat. Anz., Bd. 8.
Weidexreich, Fr. 1901 Uber Blutlymphdriisen. Die Bedeutung der eosinophilen Leukoc\-ten, iiber Phagocytose und die Entstehung von Riesenzellen. Anat. Anz., Bd. 20.
1905 Uber die Entstehung der weissen Blutkorperchen im postfetalen Leben. Verh. d. Anat. Ges., Bd. 19.
1910 Die ^lorphologie der Blutzellen und ihre Beziehung zu Einander. Anat. Rec, vol. 4.
1911 Die Leucocyten and verwandte Zellformen. \Yeisbaden, J. F. Bergmann.
MEMOIRS
OF
THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY
Xo. 6
THE RAT
COMPILED AND EDITED BY
HENRY H. DONALDSON ^
REFERENCE TABLES AND DATA FOR THE ALBINO RAT (MUS
NORVEGICUS ALBINUS) AND THE NORWAY RAT
(MUS NORVEGICUS)
To be published in October
Cloth bound. Price, post paid to any country, $3.00
PREFACE
For a number of studies on the growth of the mammahan nervous system made by my colleagues and myself we have used the albino rat. In the course of the work we frequently felt the need of referring to other physical characters of the rat to which the nervous system might be related. This led us to collect such data as were already in the literature and also led us to make further investigations. The facts gathered in this way have proved useful to us and are here presented in the hopes that the}' will be useful to others also.
CONTENTS
Preface. Introduction. Classification. Early records and migrations of the common rats.
Part 1. Albino rat — ^lus norvcgicus albinus. Chapter 1 — Biology. Chapter 2 — Heredity. Chapter 3 — Anatomy. Chapter 4 — Physiolog3^ Chapter 5 — Growth in total bodj^ weight according to age. Chapter 6 — Growth of parts or systems of the bodj^ in weight. Chapter 7 — Growth of parts and organs in relation to body length and weight according to age. Chapter 8 — Growth in terms of water and solids. Chapter 9 — Growth of chemical constituents. Chapter 10 — Pathology.
Part 2. The Norway Rat — Mus norvegicus. Chapter 11 — Life history. Chapter 12 — Growth in weight of parts and systems of the body. Chapter 13 — Length of tail and weights of body, brain and spinal cord in relation to body length. Chapter 14 — Growth in terms of water and solids. Chapter 15 — References to the literature. Index.
702
AX ABNORMAL FROG'S HE.ART AYITH PERSISTING DORSAL MESOCARDIUM
"^ JAMES CRAWFORD WATT
Department of A)wto»)y, University of Toronto
SIX FIGURES
The frog forming the subject of description in this paper belongs to the species Rana pipiens, bred in tanks in the Medical Building of the University of Toronto. After pithing, the sternum was cut away to expose the heart, by a student working in the Pharmacological Laboratory and the abnormal heart was so striking in appearance that the frog was sent to Prof. J. Playfair McMurrich who kindly turned it over to me for investigation.
After observing the beating of the heart, the frog was placed in strong formalin solution, and the heart was injected with warm wax. The wax on hardening enabled the necessary dissection to be easily performed, and was easily removed from the heart later, to admit of examination of the cavities. The heart and great vessels were also dissected in a couple of normal frogs for purposes of comparison with this abnormal specimen.
.\fter opening the pericardium the heart is seen (fig. 1) as a long tubular organ extending forward under the floor of the mouth, with its anterior portion displaced to the left of the median line by the hyoid bone and mass of the tongue.
The sinus venosus is situated in its usual location and receives the postcaval and the two precaval veins. It extends forward to open into the right atrium which lies directly cephalad of it. The atria are only incompletely divided from each other on the surface, and they are marked out from the sinus venosus at one end, and the ventricle at the other, b}^ constrictions in the wall. They are of large capacity and appear as long as the whole heart
703
704 JAMES CRAWFORD WATT
of a normal frog of equal body size, and form one-half the total length of this heart. They have an oblique position, being nearer the median line posteriorly, as they are forced out to the left side under the floor of the mouth as they proceed forward. The groove on the ventral surface, marking off the left atrium from the right, starts just cephalad of the sinus venosus on the left side, and runs obliquely forward toward the origin of the conus arteriosus, but is lost before reaching this point, so that there is no visible division externally of the atria in the part nearest the ventricle.
The ventricle continues forward in the line of the atria and is of normal size. Its apex is situated almost in contact with the inner side of the mandible. Xo conus arteriosus and no aorta can be seen with the heart undisturbed, but by lifting the atria and ventricle slightly, and pulling them out to the left, the conus arteriosus is seen leaving the ventricle on the right side, and running medially and dorsally (fig. 2). Immediately back of the hyoid bone it reaches the left side of the esophagus and bifurcates, sending the left ventral aorta directly dorsally alongside the esophagus, where it divides, and the right one horizontally' across the esophagus, after which it divides, giving off the
ABBREVIATI0X8
Au., aortac P.CA., i)ulinocutaneous arteiy
,1/., atrium Post.Cav., postcaval vein
B.C., bulbus cordis or conus arteriosus P.Pc. parietal pericardium
C.C.A., common carotid artery R., right half of pericardial cavity
D.A., dorsal aorta R.Ao., right aorta
D.M., dorsal mesocardium R.Al., right atrium
L., left half of pericardial cavity R.Pr.Cav., right precaval vein
L.An., left aorta R.P.V., right pulmonary vein
L.At., left atrium S.V., sinus A'enosus
Li v., liver V., ventricle
L.Pr.Cav., left precaval vein
t
Fig. 1 Dissection of frog to show abnormal heart. Sternum and ventral muscles have been cut away, also the parietal pericardium; the heart lies undisturbed.
Fig. 2 Same as figure 1 except that heart has been drawn over to left by hooks, to show persistent dorsal mesocardium attached to its dorsal surface. Cut edge of j)ariotal pericardium is also shown.
C.C A
705
706 JAMES CRAWFORD WATT
common carotid trunk, the rest turning dorsally to form the puhno-cutaneous artery and dorsal aorta. No abnormalities exist beyond this point, all the arterial trunks appearing normal.
When the heart is lifted a structure which is probably the mechanical cause of the abnormality comes into view. This is a double fold of pericardium (fig. 2) reflected from the wall of the sac to the dorsal surface of the heart. It extends all the way from the sinus venosus to the apex of the ventricle, and at this latter point exhibits a free edge. Running off at right angles to this fold as it lies over the ventricle, is a process of the fold passing to the right over the conus arteriosus and ventral aortae, right out to where these latter structures leave the pericardial cavity, so that this smaller fold has no free border.
I interpret this membrane as a completely persistent dorsal mesocardium. Its presence may be the cause of the abnormal shape of the heart, for the membrane normally disappears as flexion begins in the heart tube, and its continued existence would I think, be an obstruction to this flexion, especially to the bend which would project the heart toward the ventral surface (see arrow, fig. 3). It would thus hold the heart in the extended tubular condition. There has been a certain amount of flexion, however, in this heart, shown by the conus arteriosus coming off the ventricle dorsally and running caudally and medially. The free anterior border of the dorsal mesocardium is thus accounted for, as that part of the membrane lying in the long axis of the heart represents the mesocardium only up to the point of flexion which forms the apex of the ventricle. The mesocardium (fig. (5) originally anterior to this point has been carried back with the conus arteriosus, over which it is still found, and now appears as a process running off from the right side of the rest of the membrane.
The presence of this flexion seems to be an argument against regarding the mesocardium as a force resisting the folding up of the heart tuVx;, but on analysis of the nature of the bend exhibited, this objection is found to be more apparent than real. The bend which has occurred, it will be remembered, is one which brings
ABNORMAL FROg's HEART
707
D.M
R.?r Cav.
Fig. 3 Diagram of normal unflexed heart tube seen from the right, lying in pericardial cavity, showing dorsal mesocardium. Movement of heart tube in direction of arrow would be resisted. (See under figures 1 and 2 for list of abbreviations).
Fig. 4 Diagram of abnormal frog's heart seen from right, lying in pericardial cavity and showing persistent dorsal mesocardium.
Fig. 5 Diagram of normal unflexed heart seen from above to show line of attachment of dorsal mesocardium.
Fig. 6 Diagram of abnormal frog's heart seen from above, to show line of attachment of persistent dorsal mesocardium. Arrow indicates direction in which flexion has occurred.
708 JAMES CRAWFORD WATT
the part of the tube displaced, slightly above and to one side (figs. 4 and 6) of the part still retaining its original position. This means that no stretching of the suspending membrane of the tube — the mesocardium — is necessary, but only a folding of this layer on itself to one side. There is no increase in the distance between any part of the heart and the roof of the pericardium. There would of necessity be a great increase in this distance over a part of the heart if the fold had occurred which projects the ventricle (see arrow, fig. 3) into a position ventral to the atria, and this would of course bring great strain and stress to bear on the mesocardium. in order to stretch the part attached to the regions of the heart involved, sufficiently to permit of the flexion. As the mesocardium forms a continuous double-layered membrane suspending the whole heart tube, its strength would be much greater than any adventitious bands or adhesions. Its thickness, also is considerable, relatively to the size of the heart in early stages of development, and so I think it is probable that it was possessed of sufficient strength to resist stretching, and so to prevent the normal atrioventricular bend. The fact that the conus bend occurred was because it could take place without stretching the attached membrane, merely folding it on its side.
The left pulmonary vein comes directly into the left atrium through the dorsal mesocardium. The right \ein runs across the roof of the pericardial cavity (figs. 1 and 2) in which it forms a transverse fold lying caudal and parallel to the right ventral aorta, and in this fold it runs to the mesocardium which it enters to reach the atrium.
The flexion of the anterior portion of the heart can be accounted for by the recession of the branchial region during development, carrying back the aortic arches, and necessitating a bending back of the arterial end of the forwardly directed tubular heart. This bend is in the normal direction for that always found between the ventricle and conus arteriosus, but appears on the dorsal surface instead of the ventral because of the absence of the atrioventricular bend, which would have projected all the heart cephalad of it to the \'entral surface.
ABNORMAL FROg's HEART 709
From the foregoing description it will be seen that there is nothing in the condition of this heart that cannot be explained on embrj^ological grounds. The primary cause of the deformity appears to be the failure of the dorsal mesocardium to disappear after the formation of the simple tubular heart. Development in the heart tube has proceeded in a normal way throughout, and the final form of the heart has been influenced almost solely by the mechanical action of the dorsal mesocardium, accompanied by the recession of the branchial arches.
Of course there is another explanation for the condition found here. It is that there was a primary failure of the atrioventricular region to undergo any flexion, and that the presence of the dorsal mesocardium is a second and distinct anomaly, not related in an}^ way to the failure of flexion to occur. Both of these abnormalities are exceedingly rare. As far as I have been able to ascertain neither of these conditions has been previously described, and from the previous discussion I do not see why they cannot be regarded as cause and effect.
I have tried to find accounts of any similar conditions recorded in the literature on the heart and pericardium. Anomalies of the pericardium mentioned are fringes, apertures communicating with the pleural cavity, also complete absence of the pericardium. I have seen no description of a persistent mesocardium or any membranous septum or band which could be interpreted as such. Hundreds of cases of cardiac abnormalities are listed, but none of them were due to incomplete flexion of the heart tube. I have not read all the papers entitled simply 'Cardiac anomaly' as there are scores of them, and experience showed that what were investigated were common occurrences, such as failures of septa, and the like; thus I may possibh^ have missed a case similar to the present.
There is no doubt, however, that both the conditions here present are extremelj^ rare. Both the disappearance of the mesocardium and the flexures of the heart tube are quite early occurrences in embr3'onic life. Usually, of course, the earlier a process occurs, the more fundamental it is, the more likely it is not to present abnormalities, being more firmly impressed on
710 JAMES CRAWFORD WATT
the organism, and the graver are the consequences of departures from the normal. If the supposition is correct that the dorsal niesoeardium would offer mechanical opposition to flexion in the heart, then its persistence is fraught with much graver consequences than at first thought would be supposed, and will lead to most extreme cardiac deformity. The deformity in the frog was not incompatible with full activity and functional efficiency, but it is conceivable that in higher animals it might be incompatible with existence, at least beyond fetal life. An homologous deformity in man would postulate a heart tube partly at least located in the _ neck, subject to compression and to stretching from the movements of this part of the body, and lacking all protection from the outside, such as is afforded by the ribs.
If the dorsal mesocardium should persist, and yet permit of flexion going on normally in the heart, it should be found in the adult as a membrane forming a complete cephalocaudal partition across the sinus transversus of the pericardial cavity, and I have seen no mention whatever of any membrane in this region. Text-books of embryology and of pathology pay scant attention to the mesocardium, unite in describing its very early and complete disappearance, and recount no case of its persistence in whole or in part. It is evidently worthy of consideration in view of the fact that it can be retained and may have far-reaching effects on the heart.
In conclusion I wish very heartily to thank Professor Henderson and Professor ^Ic^NIurrich. through whose kindness this interesting specimen has come into my possession: and f or friendlj' criticism modifying the opinions expressed in this paper I am also indebted to Professor McMurrich.
April 21, 1915
A MECHANICAL DEMCE TO SHIPLIFY DRAWING WITH THE MICROSCOPE
RAPHAEL ISAACS
From the Anatomical Laboratory, University of Cincinnati
THREE FICrRES
It is often desirable in making rapid drawings of objects under the microscope, to have some inexpensive mechanical aid, especially when the outline is complex. For some purposes the camera lucida is either inconvenient or undesirable, being too expensive for use in large classes, besides gi\ing a limited field. A simple accessory to aid in making fairly accurate outlines of any desired magnification, from note-book to chart size regardless of the size of the field, should thus find a place in the laboratory equipment, especially if at the same time it can be furnished at a low cost.
The device here described makes use of the fact that the mapping out of the outlines of an object is easier if some fixed points of reference on the specimen can be established and related to corresponding points on the drawing paper. These requirements will be fulfilled if a piece of glass, ruled in squares (fig. 2) is placed in the eyepiece of the microscope so that the squares appear superimposed on the object. The enlarged outlines of the object can then he made on a similarly ruled sheet of drawing paper, by noting the position at which any point of the object appears in the squares in the microscope and placing a mark on the corresponding square and position on the paper. The squares on the paper are lightly ruled in pencil and therefore easily erased. The magnification of the drawings can be regulated at will by making squares of appropriate size on the paper. To rule the squares accurately on a small piece of glass, a mounted 'writing diamond' (a small diamond mounted at the end of a piece of wood or metal) is necessary, although no doubt a carborundum pencil or other substitute may be used. The diamond, in its holder, is fastened at the side of the body tube of a microscope with two wide rubber bands (fig. 1) so that although firmly held, the diamond will have some freedom when pressure is used. The microscope is used merely as a convenient way of holding the diamond so that it can be raised or lowered without changing its position, or throwing it out of adjustment. The lines are ruled l)y moving the glass beneath the point of the diamond with a mechanical stage. This makes it possible to secure lines at right angles to each other and squares correct to the tenth of a millimeter. The following procedure will l)e found advisaljle:
711
712
RAPHAEL ISAACS
A smooth, preferably' thick piece of glass is selected and fastened to a shde with a drop of thick xylol-balsam which has been evaporated and melted on the slide. On pressing the glass down so as to insui'e a flat upper surface, the balsam soon cools and sets firmly. If the upper surface slants or is uneven, the ruled lines will not be of uniform thickness. A thick round cover-glass or a square cut from a slide can be used for this purpose. The slide, carrx-ing the glass to be ruled is now fitted tightly into the mechanical stage so that there is very little free movement. Bv using the vernier to mark off the distances, lines
Figs. 1-3 ■Metliod of using the mechanical stage for ruling an eyepiece scale to be used in making drawings with the microscope. Fig. 1 Microscope, with diamond holder in position. Fig. 2 Ruled glass. Fig. .3 Method of mapping out in drawing.
can be ruled one millimeter apart (a convenient distance for most purposes) accurate to the tenth of a millimeter. To rule the lines, the diamond is lowered with the coarse adjustment and brought into contact with the slide using v(n\y light pressure with the fine adjustment (fig. 1). The reading of the fine adjustment, when the conditions are ideal, will help to give the same pressure for every hne. The pressure is tested by starting the line outside the field to be ruled. The elasticity of the rubber bands will give the diamond enough freedom to overcome any slight uneveness of the glass. The slide is moved using the mechanical stage with a steady, sweeping motion. The diamond is then lifted with the fine adjustment, and the slide ])rought back to the starting position, the slide being moved up and the new position adjusted until lh(> reading on the vernier shows it to be correct. The horizontal lines should l)e made In- moving the slide towards the side
DEVICE TO SIMPLIFY MICROSCOPICAL DRAWIXG 713
having the fixed support, thus giving it the backing of the sohd shoulder of the mechanical stage. For the same reason, vertical scratches should be made by pushing the slide forward. To make the lines at right angles to the ones first drawn, the pressure of the diamond must be as light as possible, as too great pressure in crossing scratches is not considered good for the diamond. It is unwise to go over a line once drawn, as this will often start cracking of the surface. Care should be used in remo\'ingthe glass, as it is liable to break. The balsam is warmed and when soft the glass is slid of^'. It should then be mounted with the ruled surface exposed, on another piece of glass of a size to fit neatly into the ej'epiece. using melted balsam. The ruled lines may be made more distinct by rubbing them lightly with a lead pencil, vSO that some of the black graphite is deposited in the grooves. In case it is desirable to cut the ruled glass into smaller pieces, special lines slightly deeper should be ruled along the places to be broken, as it is unsafe to try to break the glass along the light hues already ruled.
The whole operation of ruling takes but a short time, and a good slide can be finished in less than five minutes. With a little practice, especially if the diamond is good, the lines will be fairly even. In a properly adjusted eyepiece, the lines will be in focus, when the ruled glass is placed on the diaphragm inside, the ruled surface being placed facing downward. The glass can be used eciually well in a compound microscope or in a binocular. In the latter, where a single picture of a solid object is desired, it is well to put the ruled glass on the side of the eye usually used with the ordinary microscope. Of course the two tubes will give different pictures.
To use the glass in drawing, sciuares of any desired size are lightly ruled on the drawing paper. It is usually simpler if the scjuares on the paper appear the same size as those in the microscope. Students can get good results if the}- u.se the corner of a slide as a right angle and measure the distance on the lines with a ruler. The approximate position of the object is noted on the squares in the micrscope and the corresponding position found and mapped ofT on the squares on the paper (fig. 3). For low-power work, the lines appear better if the light is cut down. Further details of the drawing can be filled in later and the squares erased. As these squares are only a help in drawing, and as the process may tend to become mechanical, the instructor should be somewhat reserved in placing it freely in the hands of elementary students, where the efforts to draw a microscopic object are often an aid in analysing the specimen. Under high power, however, the squares are a decided help in analysing a complex field and the real educational value of the specimen comes in the synthesis of the parts in the finished drawing.
MEMOIRS
OF
THE WISJAR INSTITUTE OF ANATOMY AND BIOLOGY
No. 5
THE DEVELOPMENT OF THE ALBINO RAT, MUS NORVEGICUS ALBINUS
G. CARL HUBER
From the Department of Anatomy, University of Michigan, and the Department of Embryology, the Wistar Institute of Anatomy and Biology
I. FROM THE PRONUCLEAR STAGE TO THE STAGE OF MESODERM ANLAGE; END OF THE FIRST TO THE END OF THE NINTH DAY
THIRTY-TWO FIGURES
CONTENTS
Introduction , 3
Material and methods 5
Ovulation, maturation, and fertilization 9
Pronuclear stage 13
Segmentation stages 21
2-cell stage 21
4-cell stage 29
8-cell stage 31
12 to 16-cell stage 35
Summary of segmentation stages, rate, and volume changes 36
Completion of segmentation and blastodermic vesicle formation 42
Blastodermic vesicle, blastocyst, or germ vesicle 56
Late stages of blastodermic vesicle, beginning of entypy of germ laj'ers. ... 63
Development and differentiation of the egg-cylinder 73
Late stages in egg-cylinder differentiation, and the anlage of the mesoderm 92
Conclusions 108
Literature cited 112
II. ABNORMAL OVA; END OF THE FIRST TO THE END OF THE NINTH DAY
TEN FIGURES
CONTENTS
Introduction • 115
Half embryos in Mammalia 117
Degeneration of o\ a at the end of segmentation 120
Incomplete or retarded segmentation 121
Abnormal segmentation cavity formation 126
Degeneration of ova as a result of pathologic mucosa 129
Imperfect development of ectodermal vesicle 132
Two egg-cylinders in one decidual crj^pt 138
Conclusions 140
Literature cited 142
Frice, post paid to any country, $2.50
Hepriiitwl from the JorRN,\i. of MoHpnoi.oGy, Voluii e 26, No. 2, June, 1915
714
A SIMPLE FORM OF DRAWING APPARATUS^
ALEXANDER S. BEGG From the Harvard Medical School, Boston
ONE FIGURE
This apparatus was designed for use by students in the Harvard Embryological Laboratory, where it has served with success during the courses just closed. It has effected a saving in time spent on outhnes, and a gain in accuracy. Bj^ the addition of an inexpensive lens to our ordinary set of microscope objectives, it has also been utihzed in neurology. The apparatus has the advantages of simphcity, cheapness, and freedom from the care and dirt of carbon arc lamps. No special room or electrical connections are needed, since it is attached to an ordinary socket in the open workroom, the back of the box being towards the windows. It must be understood, however, that it is not designed to be used in place of an Edinger, or similar apparatus for higher powers.
The apparatus (fig. 1) consists of a box, blackened on the inside, measuring 32 inches in height by 18 inches in width and depth. One side is left open, and in the center of the upper end is a hole for the reception of the main condenser system. This system consists of two plano-convex lenses, 3 inches in diameter, mounted in a cell around the upper rim of which is a flange. The cell hangs by the flange in the aperture mentioned above. The focal length of the combination is 2 inches. Above the condenser is held a Mazda projection bulb of 100 watts, mounted in an Edison keyless socket at the end of a horizontal tube, through which the conducting cord pasSes. This tube is held bj^ means of a right angled sliding body on a perpendicular rod, which is mounted on top of the box; this gives an adjustment of the light source in all directions. It is found better to turn light off and on at wall fixture, and to avoid disturbing lamp when once adjusted in optical axis. A cheap tin shield around the lamp prevents light from escaping into the room.
Inside of the box are two cleats, one on either side, to support a wooden rack or frame upon which rests the stage and microscope. These cleats (b) are so placed that the upper surface of the microscope stage is 5 inches from the lower surface of the condenser. The stage is 6 inches square, with a center opening of 2j inches, which may be
^ This apparatus was designed during the fall of 1914 while working under a grant from the Carnegie Institution of Washington, D. C.
715
THE ANATOMICAL RECORD, VOL. 9, XO. fl
716
ALEXANDER S. BEGG
reduced by a washer to f inch. Beneath the stage is fixed, in inverted position, the arm and body tube of a microscope. The one pictured is from a discarded instrument of the type found lying idle in most laboratories. It has a coarse adjustment by means of the slip tube and sleeve, and a fine adjustment by micrometer screw. When once the optical axis is found, the rack is prevented from being displaced backward by means of small screws inserted behind the ends of the rack and into the upper surfaces of the cleats. This merely acts as a check to
Fig. 1
backward movement, and permits removal of rack from the box in front. When working with large slides, as of brain stem, the larger stage opening is used, and in order to cover the slide, the rack and stage are moved to the upper cleats (a) about 2 inches closer to the condenser.
The lenses used are microscope objectives of 48, 32, 25, 4, 16 and 6 mm. focus, without eyepieces, and a special achromatic lens of 4f-inch focus, obtained locally from Pinkham and Smith, for neurological work. The mounting of this lens is threaded to fit the eyepiece end of the body tube. A 3- or 4-inch diaphragm of cardboard may be held in place by objectives to cut out rays from around object and tube. No
SIMPLE FORM OF DRAWING APPARATUS 717
curtains have been used, since the body of the worker blocks the hght from without fairly well.
An accessor}' condenser is mounted above the stage. This is a single lens of 2-inch focus which is adjustable up and down a pillar by means of a sleeve. This, however, is not needed in the ordinary routine and may be omitted. The image is received upon paper placed either upon the lower end of the box, or upon a blackened compo-board shelf placed on either of the lower cleats, depending upon the magnification desired.
The cost of each apparatus, based upon figures for the lot of eight, in use in this laboratory, is as follows:
Boxes, painting, racks, etc 83.00
'Slain condensers, in flanged cell 2 .50
Accessory condenser in ring 1 .00
Brass stage and washer, with pillars for accessory condgnser, etc. . 4.50
100-watt Mazda projection bulb 1 .50
Lamphouse, tubing, socket, cord and plug 2 .85
Special 4i-inch lens in threaded mount 2 .50
Total 817 .85
After the above article was in the hands of the printer, we had an opportunity to trj- the 250- Watt and 400- Watt gas-filled bulbs as a source of light. These lamps are highly satisfactory, the higher power light being almost equal to a 4-ampere arc in brilHancy, and of course much superior in the matter of convenience.
BOOKS RECEIVED
The receipt of publications that may be sent to any of the five biological journals published by The Wistar Institute will be acknowledged under this heading. Short reviews of books that are of special interest to a large number of biologists will be published in this journal from time to time.
THE DEVELOPMENT OF THE HOIAX BODY, A manual of human embryology, J. Playfair McMurrich, A.M., Ph.D., LL.D., Professor of Anatomy in the University of Toronto; formerly Professor of Anatomy in the University of Michigan. Fifth edition, revised and enlarged, with two hundred and eightyseven illustrations, several of which are printed in colors. Philadelphia, P. Blakiston's Son & Ca, 1012 Walnut Street, 1915.
From preface to the fifth edition : The increasing interest in human and mammalian embryology which has characterized the last few years has resulted in many additions to our knowledge of these branches of science, and has necessitated not a few corrections of ideas formerly held. In this fifth edition of this book the attempt has been made to incorporate the results of all important recent contributions upon the topics discussed, and, at the same time, to avoid any considerable increase in the bulk of the volume. Several chapters have, therefore, been largely recast, and the subject matter has been thoroughly revised throughout, so that it is hoped that the book forms an accurate statement of our present knowledge of the development of the human body.
718
THE USE OF GUIDE PLANES AND PLASTER OF
PARIS FOR RECONSTRUCTIONS FROM
SERIAL SECTIONS: SOME POINTS
ON RECONSTRUCTION
WARREN H. LEWIS
From the Anatomical Laboratory, Johns Hopkins University
FIVE FIGURES
Owing to the generositj' of the Carnegie Institution of Washington, funds were made available for reconstructing the head of a 21 mm. human embryo, No. 460, ]\Iall collection. During the process of this reconstruction, some rather helpful methods were developed, which others who are engaged in similar work may find useful, and with this in view I have been urged to publish the methods employed.
In regard to the preservation and staining of the embryos, as well as the cutting and mounting, it is assumed that these operations have been carried out in such a manner that the series is practical^ perfect, and that no unavoidable shrinkage has occurred during the dehydration and embedding; and that in cutting the orientation is such as to give as nearl}^ perfect horizontal, sagittal or frontal sections as is possible. Great care must be taken in mounting the sections on the slides in order to avoid distortions. Reconstruction is undoubtedly greatly facilitated by the use of guide lines in the sections, such as are produced by the ordinary camphor black method. Unfortunately, such guide marks were not present on any of the series of sections used and it was necessary to depend for the form on photographs or camera drawing of the whole embryo, made before cutting. These should be as nearly as possible from lateral, frontal at horizontal ^dews.
PHOTOGRAPHS OF SECTIONS
I have been able to abandon the laborious and time-consuming method of drawing or tracing each section projected in the usual manner onto paper, and have substituted the more expensive but far better method of photographing on large plates, and using the line bromide or azo G hard (matte) prints. While this method is more expensive as regards the immediate outlaj" of money, it is much cheaper in the end than the old method of tracing, when account is made of the time involved. The photographs are far superior to any drawing that can possibl}^ be made and greatly facilitate the work, both on account of
719
720 WARREN H. LEWIS
the greater accuracy and the greater wealth of detail. The various structures to be reconstructed may first be colored on the photograph, thus enhancing the clearness and sharpness of the picture. The ordinary Sussner creta poly color pencils were used. The photographs of the sections were made in a dark-room with the ordinary Zeiss projection apparatus and for ordinary diameters, 40 or 50, the Zeiss 5 cm. planar lens was used. This is an ideal lens for such work since there is no measurable distortion of the image. In the place of an ordinary arc lamp, we substituted a 250-watt mazda stereopticon bulb, a round bulb with filaments grouped together in a small ball at the center. To avoid unequal illumination from the spiral filaments a ground glass plate was interposed directly in front of the light. The advantages of the mazda light over the arc are (1) it remains constant, and (2) it can easily be turned on or 'off for time exposures.
The arrangement of the Zeiss optical bench is shown by the diagram, figure 1. I am aware that this may not be the most perfect arrangement, still we were able to obtain excellent results. The fnost important point consists in focusing each section by moving the slide carrier back and forth, the lens remaining fixed, instead of the usual method of moving the lens back and forth. When the magnification is once adjusted to the required diameter, the lens " (7) is securely fixed in position and the plate-holder (14) likewise. Thereafter, the object or slide is brought into focus by means of a fine adjustment connected with the focusing-rod (10). Magnification is thus not altered from section to section or from slide to slide, since the sections are thus brought in the focus of the lens. This insures equal magnification of every section, the most important condition for accurate reconstruction. This method of focusing was introduced into the laboratory by Doctor Essick.
The dark-room by this method becomes the camera, in which a perpendicular board (14) with slots for the plates takes the place of the plate-holder. This board is pivoted in the center and can be freely turned at any angle in a perpendicular plane and clamped there by means of a thumb-screw at the back.
The plate-holder-board is attached to a movable stand, which can be moved to or from the lens in a straight line only, and can be securely clamped in position when in the proper place. At 50 diameters' magnification with the 50 mm. planar lens the plate-holder should be about 6 ft. and 8 in. from the lens. A series of holes can be so placed on the plate-board,, as at a, b, c, figure 1, into which the clamp bolt can be changed and the board given a new center for different sized plates. An old plate having the same thickness as those about to be exposed was used for focusing, after having had a fine piece of whiie paper tightly pasted over it. An undeveloped plate is still better.
Standard orthonon and Stanley commercial plates were used. The latter are apparently just as good and much cheaper than the former. An exposure of about 5 seconds gave the best results; diaphragm of the planar lens at 4. Flat negatives with very little contrast give the l:)est
SOME POINTS ON RECONSTRUCTION
721
1
— •
^
• -v
-n
rt
c>
r/5
_c
'^^
o
ci
iC
'35
in
^
<l)
hr
o
C
^
_o
J2
vT
-ti
c3
3
cc
o
s
1-:
n
^
o
O
tr
^
c
O
o
7.
^
o
TS
3
rf
02
^
o
Tl
'-4^
T3
N!
lo
•-H
C
rt
s
c3
I- o
722 WARREN H. LEWIS
prints, while 'contrasty' plates, which appear very beautiful, give very poor prints since the shadows and high hghts are in ioo strong contrast. When the shadows and high hghts are in strong contrast in the sections — as was the case in the series used, since the central nervous system was deeply stained with alum cochineal and the delicate connective tissue but faintly stained — it is not easy to get negatives which will give prints showing detail in both regions. With full development, a strong hght and short exposure give flatter, softer negatives than a dim light and long exposure. The diaphragm of the lens should be as wide open as is consistent with a sharp image, to increase light and reduce time of exposure. It is important to use a developer that will help to give softness to the negative and the following formula is recommended. There is a decrease in the amount of sodium carbonate usually employed for the purpose of increasing the softness of the negative :
Formulae Jor pyro developer
Solution No. 1 Solution No. 2 Solution No. 3.
Water 470 cc. Water 470 cc. Water 470 cc.
Oxalic acid... .700 mg. Sod. sulph 60 gms. Sod. carb 60gms.
Pyrogal. acid. 30 gms.
For tank development use 30 cc. each of solutions 1, 2, 3, to 1000 cc. water; develop 20 minutes at 70° F. For tray use 60 cc. each of solutions 1, 2, 3, to 1000 cc. water; develop 5 minutes at 70° F.
Add 1 drop of a 10% solution of Potassium Bromide to every 30 cc.
Fixing bath formula
Hypo 480 gms.
Water 1920 cc.
With acid hardener
Water 150 cc. Acetic acid (28%) 90 cc.
Sulphite soda 30 gms. Powdered alum 30 gms.
The azo G hard (matte) paper should be given short exposure with bright light and developed with formula recommended for azo portrait prints. If the amount of elon is doubled and the hydrochiiion decreased one-half, a still softer print with more detail is obtained. For contrast plates a still softer developer may be used for the prints with the following formula:
Water 300 cc. Hydrochinon 4 gms.
Elon 4 gms. Carbonate of soda 22 gms.
Sulphite of soda 30 gms. Potassium bromide 1 gm.
SOME POINTS ON RECONSTRUCTION 723
MODEL OF THE EXTERNAL FORM
With the photographs or drawings of the sections complete, our next step is to make a model of the external form of the embryo or of a large enough part of it to insure that we have as accurate a reproduction as possible. For this the ordinary Born wax plate method is used. Wax plates of the proper thickness are cut out for the external form and piled either by the orientier guide lines or according to the photographs of the external form. It is extremely important that this external form shall be as perfect as possible, for on it all subsequent reconstructions are based, as will be seen later.
The wax plates should be piled without fusing so that they can be easily unpiled later. In making this external form of a whole embryo, it is usually best to begin piling with the larger plates from the middle of the trunk towards the top of the head, and another pile from the same region towards the tail, as when horizontal sections are used. In fact, two or three piles may be used, provided the guiding curves coincide. Or one may pile the head with the trunk, then hft off the head and turn the trunk piece upside down and pile the tail encl^on it. In this case the guide curves will overlap.
It is a great help to use a guide curve from a negative outline in cardboard of the external form of the embryo, made by magnifjdng the photograph of the external form to the proper diameter, as shown in figure 2. In horizontal series this curve should be made from a direct sagittal view of the embryo, and will of course be in the same plane as the median sagittal plane of the model. It is also important to establish the relation of the plane of the sections to this enlarged outline in order to give the proper angle to the guide curve.
The sections of different embryos, for example, may be cut at somewhat different angles to the frontal plane of the embryo, as in figure 3. So if one were piling the head end of the embryo from the middle of the trunk, it would be necessary to determine the plane of the sections in relation to the whole embryo, whether in the direction of a or 6 or c, etc., in order that the base-line of the cardboard guide curve may coincide with the proper plane a or h or c, etc., when it rests on the baseboard. It may also happen that the sections in a horizontal series are cut obliquely to the median sagittal plane, as is often the case. In piling the plates for the external form from such a series the median sagittal plane of the plates must be made to start at a corresponding angle to the base-board, leaning either in the right or left, as the case may be. In this way the plates are piled up with their flat surfaces parallel to a horizontal base-board.
In a similar manner, the proper angle should be used in piling plates from sections cut more or less obliquely to the other planes of the embryo. It is important that the piling be done on a board with a true surface. Overhanging parts which are likely to sag should be supported from the base-board.
THE ANATOIIICAL RECOBD, VOL. 9, NO. 9
724
WARREN H. LEAVIS
SOME POINTS ON RECONSTRUCTION
/:^o
Fig. 3 Outline of embryo to show different arrangements of cardboard guide a, b, c, for cross-sections, cut at different angles to the frontal plane of the embryo.
EST.^BLISHIXG GUIDE PLANES
Since in most series there are no guide marks, it was necessary in some manner to establish guide hnes on our photographs that could be used for all subsequent reconstructions. I first thought of drilling two perpendicular holes through the entire series of plates as they stood together in the piled-up form, and then by placing each plate on its own photograph the position of the hole could be traced onto the photograph and we would thus have two orientier marks on each photograph (or drawing) that could be utilized in piling plates for future models.
A somewhat different and better method was finally adopted, which has proved verj^ successful. After the piling of the external form was complete and satisfactory and the cardboard outline removed from' the heacl end (for example, from a series of horizontal sections) two perpendicular posts were erected from the base-board in such a manner that a line drawn between them passed along the median plane of each plate or parallel to it (fig. 4). With a straight-edge rule a fine was then drawn with a needle across the top wax plate, the straight edge resting against the edges of the two perpendicular posts, and a second line was drawn at right angles to this at a given distance from one of the perpendicular posts. The top plate was then carefully lifted off and sunilar lines drawn on the next and each succeeding plate in turn.
726
WARREN H. LEWIS
3
Fig. 4 Method of making guide lines: 1, baseboard; 2, perpendicular posts; 3, straight edge; 4, piece at right angles to it.
Care must be taken, of course, not to disturb the position of the plates until the lines are drawn. Thus there are established on each wax plate two lines, at right angles to each other, which coincide with two planes through the model or embryo that are perpendicular to the plane of the sections. One of the yrinci'pal planes either coincides with the median plane or is parallel to it, while the other is. at right angles to this and at a given distance from one of the perpendicular posts. With these two principal planes established, it is possible to repile the external form or any other part of the embryo with mathematical precision, since these two planes are likewise perpendicular to the plane of the sections.
After the guide lines have been drawn on each wax plate, they must next be transferred to the photographs or drawings by super-imposing each wax plate on its photograph or drawing and marking at the ends of the guide lines. The wax plate is then lifted off and the points on the photograph connected by lines similar to those on the plates. When the two principal guide lines are established we have found it convenient to establish secondary guide lines by drawing other lines, 5 cms. apart, parallel to those over the entire surface of the photographs.
We have introduced lately a still better method : namely the printing of lines in red ink over the photographs from a lithographic stone. These lines form squares of 1 cm. with slightly heavier lines every 5 cms. The lines are printed to correspond with the two principal guide planes. Such lines greatly facilitate not only the plastic reconstruction work but are of especial value for graphic reconstructions. These hnes will of course coincide with various planes through the embryo. The advantage in having such a number of planes will become apparent when one wishes to reconstruct small structures that are limited to a parti(;ular part of the embryo. The whole section or any part of it will thus ])e included in rectangles or squares of various sizes depending upon the extent of the part.
SOME POINTS ON RECONSTRUCTION
727
WAX MOLD FOR PLASTER OF P,\RIS CAST
With the development of these guide planes we have abandoned the usual Born wax plate method of making models, and use instead wax plate negatives into which plaster of Paris is poured. The method is as follows: Structures to be modeled are outhned on wax plates in the usual manner, and at the same time, while the plate is still in proper position under the photograph, points are pricked through into the
Fig. 5 Method of piling wax mold: 1, baseboard; 2, perpendicular right angle corner; 3, glass plate; 4, wax mold; 5, vent; 6, galvanized iron wire bridge; 7, gate for plaster between parts of mold.
wax with a fine needle at the corners of the rectangle in which the structure or structures outhned are included. The outline is transferred from the photograph onto the wax plate by the use of carbon paper; tracing on the photograph with a smooth glass point. Each plate is then carefully trimmed to the rectangular shape corresponding to that outlined by the four needle points The outhned structures are then cut out, leaving holes in the plates. The plates are then piled into a perpendicular rectangular corner (fig. o). Bridges of galvanized iron wire (ordinarj^ iron wire will rust and discolor the cast) can be.placed
728 WARREN H. LEWIS
in position as the piling progresses to hold the various parts of the cast together. Wire, string, or cloth may be inserted into the holes of the finer structures to give strength. Gates and vents to carry plaster from one part of the model to the other and to allow air to escape were cut through suitable places as the piling progressed. It is better not to smooth the inside of the mold. Owing to the fact that the plates were trimmed into rectangles, having the four sides in the same perpendicular planes, the structures which are represented by the holes in the plates must necessarily come into the proper relation with each other. If some of these structures come to the edge of the plate at any place, they would necessarily be cut off by one of these planes. After the piling is completed the outside edges of the plates are fused together to prevent the plaster from leaking out. A wax plate is also fused over the side of the block if any holes come to the edge. In piling the wax plates, it is necessary to measure the height of the block of plates after the addition of each new plate in order to be sure that the plates are piled properly as regards height. It is usually necessary to scrape off a slight burr which comes along the edge of the cut.
We have often found it advisable to build up models in rather small blocks, usually about 50 mm. in thickness, and where the models are large these blocks can be hmited in other directions as well and the casts later fused together or merely fitted together as a dissectable model. Such combinations can be varied to suit special conditions.
THE PLASTER OF PARIS CAST
After the piling is completed and the edges of the plates are fused, the bottom of the pile which rests on the glass plate is made fast. Plaster of Paris is then poured into mold until it rises above the top and before it has completely hardened the excess on top is usually scraped off level with the top plate.
We used a grade of plaster known as potter's plaster. The mass consists of about equal weights of water and plaster. The latter is sifted into the water until it just begins to show drj^ on the top. It is then stirred a little and poured into the mold. After setting for an hour or so, the wax is melted off in boiling water. The cast is taken out and washed in very hot water and dried in the air. Plaster of Paris is wonderful material to work with and requires but little experience to handle it with considerable facility. The plaster cast of course shows the lines of the plates just as wax models do unless considerable l)olishing is done. It is easier to smooth off a plaster cast than a wax model. The plaster is easily trimmed with a knife or sandpaper and the angles remaining between the edge of the plates can easily be filled in with fresh i:)laster painted on with a l)rush to any desired thickness. Corrections and additions can also be made on such plaster models by cutting off and building up with fresh i)laster, using wire, string, cloth, etc., if necessary. The different structures are easily tinted with water colors (suspensions in water or ordinary commercial house paint pig
SOME POINTS OX RECONSTRUCTION 729
raents; such as, ultramarine blue, j'ellow ochre, burnt sienna, chrome yellow, Enghsh vermiUion, etc.). A better method if one wishes to polish the model, is to mix up fresh plaster by using water colored with the pigment and to do the final smoothing with this mixture. If models are made in sections there is no very great difficulty in putting these together in the proper relation to each other. Finalh- the entire model can be toughened by soaking in hot paraffin until all the air is driven out by the paraffin which penetrates through the plaster. It is best to have the paraffin bath somewhere between 95 and 100°C. when the model is lifted out, in order that the surface will not be heavily coated with })araffiii.
\Y-\X PLATES
The wax plates were made according to the following formula:
Bees' wax 6 parts
Paraffin 4 parts
White lump rosin 2 parts
The ordinary o6°C. Standard Oil paraffin was used; lump rosin is much better than powdered rosin. 2000 grams poured on very hot water — surface 3 by 4 feet — gives plate 2 mm. in thickness."^ The hotter the wax and water the better. The air bubbles which form in the wax are driven off before the plates cool by playing a Bunsen flame over the surface.
The points which I wish to emphasize are: first, the use of photographs; second, the use of the series of guide lines which coincide with planes that are at right angles to each other and perpendicular to the plane of the sections; and third, the use of plaster of Paris. Although the preliminary steps are somewhat more complicated than those usually employed, the}' are nevertheless essential and in the end save both time and expense. The advantages of a plaster model are obvious to those who have worked with wax and realize the dangers of distortion and the difficulties involved in strengthening the wax and in model ing fine structures.
M E M O I E S
1^
The publication of this series of Anatomical Monographs has been undertaken with the purpose of presenting the results of original investigation in anatomy which are too extensive for incorporation in the already over-crowded current periodicals.
No. 1. The Anatomy and Development of the Systemic Lymphatic Vessels in the Domestic Cat, by George S. Huntington, Professor of Anatomy, Columbia University, New York City. This monograph states in a few pages the various theories held in regard to lymphatic development in general and then proceeds to present the result of six years' careful investigation on mammalian lymphatic development. Part I deals with the development of the sj'-stemic h-mphatic vessels in their relation to the blood vascular system. Part II deals with the development of the preazygos and azj'gos segments of the thoracic duct. This memoir contains 175 pages of text, 8 text figures (two in color), 254 photomicrographs and 21 colored plates. Sent post paid to any country for S4.00.
No. 2. Contribution to the Study of the Hypophysis Cerebri with Especial Reference to its Comparative Histology, by Frederick Tilney, Associate in Anatomy, Columbia University, New York City. Part I contains a historical review of the literature. Part II deals with the comparative histology of the pituitar}^ gland and gives a report of six hypophysectomies performed upon cats. This memoir contains 72 pages of text, 2 text figures, 60 photomicrographs and plates. Sent post paid to an}^ country for SI. 50.
No. 3. Early Stages of Vasculogenesis in the Cat (Felis Domestica) with Especial Reference to the Mesenchymal Origin of Endothelium,
by H. Von W. Schulte, Department of Anatomy, Columbia University, New York City. This memoir contains 90 pages of text and 33 figures of which 14 are in colors. Sent post paid to any country for SI. 50.
No. 4. The Development of the Lymphatic System in Fishes, with Especial Reference to its Development in the Trout, by C. F. W. McClure, Department of Comparative Anatomy, Princeton University. In press, to be published in October.
No. 5. The Development of the x\lbino Rat, Mus Norvegicus Albinus: L From the pronuclear stage to the stage of mesoderm anlage; end of the first to the end of the ninth day: H. Abnormal ova; end of the first to the end of the ninth day; by (1. Carl Huber, Department of Anatom3% T'niversity of ^Michigan, and the Division of Embrj'ology, Wistar Institute of Anatomy and Biology, Philadelphia. This memoir contains 142 pages of text and 42 figures from drawings by the author. Sent post paid to any country for S2.50.
Orders should be addressed and checks made payable to
THE WISTAR INSTITUTE OF ANATOMY AND BIOLO&Y
Philadelphia, Pa.
7.30
7)i
co:mparatia E osteology of certain rails and
CRANES, AND THE SYSTEMATIC POSITIONS OF THE SUPERSUBORDERS GRUIFORMES AND RALLIFORMES
R. W. SHUFELDT
NINE FIGURES
Owing to the existence of such peculiar birds as the Kagu TRhinochetus) ; the finfeet (Podica); the Ortygometra ; the sun bittern (Euryp}'ga) ; the trumpeter (Psophia) ; the jMadagascaran genus Mesites, and the Seriema (Dicholophus) , the exact Hmits of the rail and crane groups of birds still existing in nature have long been a mooted question with avian taxonomers ; and far more light from anatomical sources is required before the true relationships of many of the groups and species just enumerated can be definitely determined.
Such classifications of the Class Aves as have appeared within comparatively recent times are much at variance with respect to opinions upon the groups to be discussed in the present article.
In order to demonstrate this, it is but necessar}^ to present in brief several of the classifications of birds, which have been offered by writers of authority on the subject during the last centur\\ For example, in the year 1813 the Academy of Sciences of Berlin, in its Abhandlungen (pp. 237-250), published a scheme of bird classification proposed by Blasius Merrem, entitled Tentamen Systematis Naturalis Avium." That gifted writer distinguished his fourth group of birds as 'Aves palustres,' and thus divided it:
4. Aves palustres :
A Rusticolae: (a) Phalarides — Rallus, Fulica, Parra; (b) Limosugae — Xumenius, Scolopax, Tringa, Charadrius, Recurvirostra
B Grallae: (a) Erodii — Ardeae ungue intermedis serrato, Cancroma; (b) Pelargi — Ciconia, Alycteria, Tantali quidam, Scopus, Platalea; (c) Gerani — Ardeae cristatae, Grues, Psophia
C Otis
731
THE ANATOMICAL RECORD, VOL. 9, NO. 10, OCTOBER, 1915
732 R. W. SHUFELDT
Fo]- the time it was given, this classification is by no means lacking in merit, and there are taxonomers of the present day who may, with profit, still contemplate it.
Fourteen years later I'Herminier, in his '^Recherches sur I'appareil sternal des oiseaux," published in the ^A.ctes' of the Linnean Society of Paris (T. 6, pp. 3-93), showed upon osteological grounds that the Rallidae and the true cranes (Grus) were affined ; that neither family was especially related to the herons (Ardeidae), and so should be separated from them.
Passing to the classifications of many of the other early writers — always more or less conflicting in their views — we come to the famous scheme of Professor Huxley, so frequently quoted in previous papers of mine (P. Z. S., 1867). In his Order III (CARINAT.AJE, Merrem), Group 2 (Geranomorphae), Huxley arraj^s two families thus:
Family 1 : Gruidae
Intevmediate forms: Psophia, Rhinochetus Famil}^ 2 : Rallidae
Intermediate forms: Otis, Cariama
He remarks thereon that he considered the cranes and the rails as constituting the t3'pical forms of the group. The herons he places in a distinct suborder, the Desmognathae, in a group Pelargomorphae, containing the Ardeidae, the Ciconiidae, and the Tantalidae.
In the curious and unique classification of Garrod (P. Z. S., 1874, p. 116), we find his subclass Homalogonatae divided into numerous orders, the first of which is the Galliformes. This latter he divides into 'cohorts,' x, B, y, etc., — ^Cohort B being the Gallinaceae, and it is thus divided:
Faniih' 1 : Palamedeidae 2: Gallinae 3: Rallidae 4: Otidae
Sul)family 1: Otidinae
2: Phoenicopterinae Family o: Musophagidae (i: Cuculidac
Subfamily 1: ("cntropodinae 2: Guculinae
COMPARATIVE OSTEOLOGY, RAILS AND CRANES 733
It is not surprising, in a classification of this nature, to find the herons in another Order, placed between the Cathartidae and Steganopodes, while the Gruidae are consigned to still a difierent order, from the rails, for example, and placed between the plovers and gulls (among the Limicolae), the only other cohort in the same Order being the Cohmibae.
Order XXI of Doctor Sclater's scheme is the Fulicariae, which is created to contain (a) the Rallidae, and (b) the Heliornithidae. In the XVIIth he places the (a) Aramidae; (b) Eurypygidae; (c) Gruidae; (d) Psophiidae; (e) Cariamidae, and (f) Otidae. The Parridae he places in Order XVHI — the Limicolae.
Professor Xewton, ever cautious and far-seeing in everything he did in ornithology — although the present writer does not always coincide with him in all his views, as in the present instance — places in the Herodiones the Ardeae, the Ciconiae, and the Plataleae. Fulicariae and Grues — the latter made to include the Gruidae, Psophiidae, Aramidae, Eurypyga, and Phinochetus — make up the Grallae.
Doctor Reichenow, who in 1882 gave us "Die Yogel der zoologischen Garden," divides the Class Aves into 'Series.' Series III is the Grallatores, split up into orders, suborders, families, etc., and I extract the following from it:
Suborder B: Arvicolae
Family 19: Otididae 20: Gruidae Suborder C : Calamocolae
Family 21 : Rallidae
Subfamily A: Rallinae
B: Gallinulinae C: Parrinae Familj- 22: Eurj'pygidae Suborder D : Deserticoeae
Family 23: Thinocoridae 24: Turnicidae 25: Pteroclidae Order VII: gressores
Family 26: Ibidae
27: Ciconidae
28: Phoenicopteridae
29: Scopidae
30: Baloenicipidae
31 : Ardeidae
734
R. W. SHUFELDT
In 1884 appeared the second edition of Doctor Coues's 'Key' to North American birds, and in it we meet with the following scheme :
Order
ALECTORIDES
Suborders
Gruiformes Ralliformes
Families
J Gruidae ) Aramidae / Parridae \ Rallidae
Subfamilies
fRallinae < Gallinulinae [Fulicinae
This is one of the most natural arrangements I have met with, and in most particulars closely agrees wdth what I will probably have to suggest in the sequel, excepting the position of the Parridae.
Dr. Leonhard Stejneger in 1885 proposed a classification in that now^ wellknown work "The Standard Natural History." In it he places the Jacanidae in his superfamily Scolopacoidae of the order Grallae, and the Gruidae, Aramidae, and Rallidae in another superfamily, the Gruioideae of the same order. The storks and herons are removed to another order, the Herodii.
In the first edition of the 'Check-List' of the American Ornithologists Union, the following scheme is given:
Order
Suborders
Families
Subfamilies
Genera
Grues
Gruidae
Grus (3 sp.)
Aramidae
Aramus (1 sp.) Rallus (6 sp.) Porzana (5 sp.)
Ralli
Rallidae
Crex (1 sp.)
Gallinulinae
lonornis (1 sp.) Gallinula (1 sp.
Fulicinae
Fulica (2 sp.?)
Here the Jacanidae arc placed as the last family of the Limicolae.
Ftirbringer places the Parridae in a gens Parrae of a suborder Charadi-iiformes, which belongs to his order Charadriornithes. Between this order and the order Alectorornithes we find inserted two intermediate suborders, thus:
COMPARATIVE OSTEOLOGY, RAILS AND CRANES
735
Suborder
Gens
Family
Eurypygidae
Eurypygae
Rhinochetidae Aptornithidae
Gruiformes
Gruidae
Grues
Psophiidae Cariamidae
Fulicariae
Heliornithidae Rallidae
Hemipodii
Mesitidae Hemipodiidae
The late Doctor Sharpe, in his exceedingly useful "Review of recent attempts to classify birds," also places the Parrae in an order Charadriiformes (Order XVIII) , while the Grues and Arami are in an order Gruiformes, along with the Rhinochetides, Mesitides, Euryp3^gae, Psophiae, and Dicholophi (Order XIX).
Dr. Hans Gadow, who has accomplished so much in the morphology of birds, has given us at least two schemes of classification for the class. One of these appears in Brown's 'Thierreich' (Aves), and the earlier one in the P. Z. S. ('92, p. 229), a paper "On the classification of birds." In this latter he places the Parridae among the Limicolae, and divides his Gruiformes into the Eurypygae, Ralli, Grues, Dicholophi, and Otides.
There are a number of others we might quote, but enough has been presented for my purpose: to show that a great variance of opinion still exists among the best authorities on the subject, but that the tendency seems to be to keep the Gruidae, the Rallidae, and the Aramidae more or less closely together, and well removed from the herons and storks and ibises, while the Parridae are placed among the limicoline forms, more or less near the plovers.
We will next proceed to examine the osteology of several forms representing such genera as Grus, Rallus, Crex, lonornis, Gallinula, and FuHca. As long ago as July, 1888, I printed an account of the osteology of the sora rail; and as that contribution is now probably well known to comparative anatomists, I will not reproduce a.ny part of it here beyond making references to
736 R. W. SHUFELDT
it in the course of the comparisons to be made with Crex, Rallus, etc., further on in the present paper. ^
My ability to compare the skeleton of our sora rail (Porzana Carolina) with the corn crake of Europe (Crex crex) is entirely due to the kindness of Dr. F. E. Beddard, F.R.S., Prosector of the Zoological Society of London, who, wdth great generosity, presented me with a fine skeleton of the latter ralline form. Such a comparison need not detain us long, for the osteological characters that distinguish Crex and Porzana are but a kind of distinguishing two genera osteologically, while, as a matter of fact, in many of their skeletal characters, these two short-billed rails are very much alike. Indeed, to such an extent is this the case, and so gradually do the skeletal characters of Crex shade through those of Porzana to typical Rallus, that the distinction between 'land-' and 'water-rails' possesses, in such premises, no foundation in fact, the opinion of some authorities to the contrary.
Essentially, the corresponding characters of the skull and associated parts of the same are identical in Crex and Porzana, the former simply being of greater size, as the corn crake is about one-third larger than the Carolina rail. The same observation applies to all the elements of the shoulder-girdle. The scapulae in Porzana are, however, relatively somewhat longer and more curved. With regard to the ribs and spinal column, they are the same in these two genera, while a few differences are to be found in the pelvis. These are of interest. The most important one is the rising up of the mesial borders of the ilia anteriorly, to to meet the superior edge of the sacral crista in Crex and not in the Carolina rail.
Passing to the sternum, we find but two good distinctive characters worthy of mention. Upon its dorsal aspect in Crex there is a median osseous bar extending from its anterior border abruptly downwards, and but slightly backwards, to fuse at its narrower end with the sternal body; this is absent in Porzana.
R. W. Shufeldt. Osteology of Porzana Carolina, Jour. Comp. Med. and Surg., New York, July, 188S, vol. 9, no. 3, art . 17, i)p. 231-248; numerous cuts in text.
COMPARATIVE OSTEOLOGY, RAILS AND CRANES 737
In Crex, too, the bod}^ of the sternum is both actually and relatively larger than it is in the Carolina rail, while its lateral xiphoidal processes are drawn more closely towards the free external margins of the sternal body behind. This last character is to be well noted. Aside from the matter of size, the characters seen in the skeleton of the pectoral limb are essentially the same in the two genera, while a few distinctive ones are found in the pelvic Iwih. For example, the cnemial crests of the tibio-tarsus are more conspicuously developed in Porzana than in Crex, while the several joints of pes in Porzana are comparatively slenderer and actually longer than they are in the corn crake.
In comparing the lengths of the corresponding long bones of the upper extremity in these two genera, the differences are very well marked. In the case of the corresponding bones in the lower extremity, the difference, with respect to lengths, is but slightly in favor of Crex in any particular case. The femora form an exception to this last statement. It can be shown thus:
Humerus Femur Tibio-tarsus Metatarsus
Crex 45 mm. 49 mm. 65 mm. 41 mm.
Porzana 36 mm. 37 mm. 60 mm. 38 mm.
Differences in length 9 mm. 12 mm. 5 mm. 3 mm.
The patellae are absent both in Porzana and in Crex.
Fulica, lonornis, and Gallinula are all genera that have characters in their skulls and associated skeletal parts which essentially agree with the corresponding ones in Porzana. So slight are the differences that, were we to be guided by these parts of the osseous system alone, it would be impossible to find even characters for generic distinctions to separate the forms in question. They will each and all constitute examples to show that the skull in birds is not always an all-sufficient guide to correct taxonomy, much less a never-failing index of remote or near affinities in this class of vertebrates. The skull in Fulica might be attached to a skeleton of a Porzana, of a size corresponding to that of the former; and there is not an ornithotomist, living or dead, who would for an instant believe it was the slightest shade out of the way. It would simply be regarded as having come
738 R. W. SHUFELDT
from a very large-sized Porzana, and assigned to that genus as a matter of course.
Coots (Fulici) and Gallinules (Gallinula) have, too, the same character in their vertebral column, ribs, shoulder-girdle, and sternum that we found in Crex and Porzana. In the sternum of Fulica, the lateral ziphoidal processes are long, and slightly inclined to flare outwards more than they do in the Carolina rail; and a manubrial process is also better developed, although it is still very small.
Coming to the pelvis, we find Fulica has the preacetabular parts of the ilia as we find them in Porzana ; but the pelvis is actually as well as relatively longer and narrower in the firstnamed genus. The lateral postacetabular projections are not so well marked, while the pubic styles are quite different from what we have described for those parts in Porzana. They each become broader as they proceed backwards, until they are met by a^conspicuous out-turned process, given off by either ischium at its postero-inferior angle. At this point, a pubic style turns abruptly downwards, and is continued for some little distance to its truncate free posterior extremity. This downbent portion is broader and flatter than the style is at its commencement beneath the cotyloid cavity. On the dorsal postacetabular surface of the bone, the parial and scattered inter-diapophysial foramina are usually entirely absent in Fulica, and always so in the skeletons of adult Gallinules. Otherwise the pelvis in Gallinula closely coincides with that part of the skeleton in Porzana, with the exception of the ilia meeting the sacral crista; in that particular it is more like the pelvis of Crex.
The appendicular skeleton in the coots and Gallinules is distinctly ralline in character. They lack patellae in the pelvic limbs, the several bones in Gallinula being more like the corresponding ones in Crex than in Porzana ; the reverse of this is the case for Fulica. An example of this is well seen in the development of the cnemial crests of the tibio-tarsus , they being much suppressed in Gallinula, as they are in the corn crake, while in a coot they are conspicuously developed, as we find them in the Carolina rail.
COMPARATIVE OSTEOLOGY, RAILS AND CRANES 739
From these short thick-billed ralline birds, the passage to the long and comparatively slender-billed representatives of the genus Rallus is easily made. To illustrate this latter genus I have before me skeletons of both Rallus longirostris crepitans and R. 1. obsoletus.- A glance at either of them is sufficient to satisfy us that they are closely related to the forms we have already passed in review above.
In Rallus 1. crepitans all the characters of the skull and associated skeletal parts are typically ralline, agreeing with what we found in Porzana. The chief difference to be noted is an elongation of the bones of the frontal portion of the skull, but more particularly the bones of the face and the mandible. This latter gives the bird its long beak and the elongate external narial apertures. Another point to be noticed is the well-marked, though narrow supra-orbital glandular depressions. These are confined to the entire superior margin of either orbit, and are very faint in Porzana but slightly better marked in Fulica.
As in Fulica, Rallus has much better defined temporal fossae; they are entirely restricted to the lateral aspects of the skull. In this species of Rallus, the maxillo-palatines may be in contact with each other in the median line; the}^ are always very close to each other there in all true rails. The number and characters of the vertebrae between the skull and the pelvis, as well as the number and general arrangement of the ribs, agrees with the Carolina rail and with Crex.
The shoulder-girdle, the sternum, and the -pelvis of this species of rail also agree exactly with what we find in Crex, and this is an interesting fact. The limb bones are more like those in the corn crake also than those of either Porzana or the coots and Gallinules. In other words, Crex stands between Rallus and Porzana, rather than between Porzana and the Gallinulinae, to which latter place it has been incorrectly^ assigned by some authorities.
2 For the first-named species I am indebted to Mr. Philip Laurent, of Philadelphia, and for the last-named to Dr. T. S. Palmer, of the U. S. Dept. of Agriculture, who collected the material for me at Berkeley, California, for use in the present connection, and my thanks are here tendered to him and to Mr. Laurent for their timely assistance.
740 E. W. SHUFELDT
With respect to the osteology of Aramus vociferus, I have already written a complete and fully illustrated account of the skeleton of that puzzling species (The Anatomical Record, vol. 9, no. 8).
OSTEOLOGY OF GRUS AAIERICAXUS AND OTHER CRANES OF THE GEXU8 GRUS
Of the genus Grus I have before me, at this tune, a complete, disarticulated skeleton of G. americanus, a complete skeleton of G. mexicana, and two extra skulls of G. canadensis. For this
Fig. 1 Left lateral view of the skull and mandible of the little brown crane (Grus mexicana); photographed by the author, natural size; reduced in reproduction to three-fifths; Specimen No. 820, Coll. U. S. Nat. Museum.
material I am indebted to the U. S. National Museum, of Washington, D. C., and from its examination I am enabled to demonstrate the following facts:
The skull. Except in point of size, the skull of Grus canadensis presents the same characters as those found in G. americanus, the latter being about one-third larger, and lacks the sculpturing along the superior orbital margins for the nasal glands.
There is but little else to be said here about the skull of this crane, as I incidentally described it when writing out the abovementioned account of the skull in Aranuis. Special attention is invited, however, to the additional perforation above the central
COMPARATIVE OSTEOLOGY, RAILS AND CRANES 741
Fig. 2 Left lateral view of the sternum and os furcula of Grus mexicana. Fig. 3 Left coracoid, anterior aspect. Fig. 4 Left carpo-metacarpus, anconal aspect. Fig. 5 Left tibio-tarus, distal portion on anterior aspect. Figures 3-5 are all from the same skeleton that furnished the sternum in figure 2 (Xo. 820, Coll. L'. S. Nat. ]Mus.). Drawn by the author, natural size, and here reduced to two-fifths.
one in the interorbital septum, and to the large, leaf-Hke maxillopalatines; these are entire, though very thin. Another point is the pecuUar manner in which the descending part of the lacrymal stands out at right angles to the rest of that bone, not approaching the zygoma as it does in so many other birds where it is found, and as it does to a great degree in Aramus. This is likewise the case in Grus mexicana (fig. 1).
In the mandihle of this crane I find the ramal vacuities nearly filled in by the splenial elements, and its symphysis is wider than it is in the Imipkin, which gives rise to a wider longitudinal groove upon its upper side. Hardly any evidence of a coracoid process exists in the lower jaw of either of these birds, the merest tubercle being present at the site where it is coixunonly found on the superior ramal border. In both Aramus and Grus, this part of the skull is only partially pneumatic, and the pneumatic foramina at the supero-mesial aspects of the inturned processes of the articular ends are always single and small.
Of the remainder of the axial skeleton. 'WTien Garrod gave us his account of the anatomy of the limpkin, in a paper in the Proceedings of the Zoological Society of London (76, pp. 275
742 R. ^y. shufeldt
277), which he entitled On the anatomy of Aramus scolopaceus," he remarked of the bird: The sternum is completely Gruine, as are the other parts of its skeleton" (p. 275), by which he meant, I presume, in a general way; for if he meant anything else by the word 'completely,' what he said will by no means strictly apply to Grus americanus, as may be seen from what follows.
Now I have shown in my former paper that Aramus has 23 vertebrae between skull and pelvis; Grus americanus has one more than this — 24. They are all highly pneumatic, and each one is about double the size of its corresponding representative in the spinal column of Aramus, so far as the latter can be satisfactorily ascertained or determined.
The first eleven cervical vertebrae in G. americanus have exactly the same characters as the first eleven in the column of Aramus; but from that point on, as we pass from vertebra to vertebra, we come to appreciate the fact that a change is slowly taking place. The 12th cervical is relatively shorter in Aramus, and its neural spine, or rather eminence, is longitudinally divided; not so in Grus. These differences still obtain in the 13th, while in the 14th the neural spine in Aramus is represented by a remarkable saddle-shaped enlargement, and the hypapophysial canal is replaced by a spine. This canal in the 14th vertebra of the crane is still continued, and the neural spine is a low, median eminence near the center of the centrum. Again, these differences are carried on to the 15th vertebra, wherein the peculiarly enlarged neural eminence of this one in Aramus lends to the bone a most extraordinary appearance. Strange to relate, the 16th cervical in these two birds have the same identical characters. To some extent, this also applies to the 17th in each; but the 17th in Aramus bears a pair of free cervical ribs of no mean length, while the pleurapophyses in this vertebra in Grus are short and fused with the bone. In Grus the 18th and 19th vertebrae are still free, and the 18th supports the first pair of free ribs; they do not connect with the sternum as they do in the case of the 19th. In Aramus the 18th vertebra is fused with two others, and, by means of haomapophyses, its ribs do con
COMPARATIVE OSTEOLOGY, RAILS AND CRANES 743
nect with the sternum. From this point on, to include the pelvis, the sternum, and other parts of the axial skeleton, the skeleton in Grus differs so much from what we find in Aramus, that a more exact description is required for the former. In Grus americanus, the 20th, 21st and 22d vertebrae of the spinal column are coossified so as to form a single bone. In it the neural canal is markedly small in caliber, while a large pneumatic canal, of fully three times its size, occurs above it. It is almost entirely unobstructed by any slender osseous trabeculae. Nothing whatever of this nature occurs in Aramus. This canal is open at both ends, and the entire bone otherwise is more or less riddled with pneumatic openings. Very extensive ossification of the tendons attached to it upon its neural aspect has taken place, and they lead forwards and backwards as lengthy interlacements.
The 23d and 24th vertebrae in the column of Grus americanus are again free, and are remarkable bones in many particulars. The lofty neural spines are quadrilateral in form, and great, lengthy, osseous rodlets, with fringed ends, project directly forwards and backwards from their superior borders. Similar coossified tendons, directed in the same manner, are found on the supero-external aspects of the diapophyses. The neural canal is particularly small, when we come to take the size of the bird into consideration. The pneumatic canal above it, which is completed by the neural arches, is, like the neural canal, a cylindrical passage, but has certainly five or six times the caliber, and is filled with a very open, spongy, osseous tissue. Of especial interest are the forms assumed by the post- and prezygapophyses. Antero-posteriorly, they are exceedingly narrow, while at the same time they are very long. Both in front and behind the mesial ends of the pair meet, and make an angle with each other of about ninety degrees. Lying in the middle line, the apex of this angle is found in the peripheiy of the entrance to the neural canal, at its highest dorsal point. The aperture of the angle embraces the immediate entrance to the 'pneumatic canal' described above — the whole being in a plane perpendicular to the longitudinal axis of the centrum. The mesial ends of the prezygapophyses of the 23d vertebra do not quite meet with each
744 R. W. SHUFELDT
other; the articular surfaces are upon their upper aspects. This is their character also in the next succeeding vertebra; but in both of these the mesial ends of the postzygapophyses do meet, and the articular surfaces of them have the appearance of being continuous.
The general form of the pelvis in Grus americanus is the same as we see in the pelvis of Aramus, but there are to be found a few differences in details. In front, the antero-mesial margins of the ilia are so rounded off that they fail to meet the superior border of the fore part of the sacral crista to more than the extent of the neural spine of the first vertebra composing it. In the postacetabular region small perforating foramina are seen among the diapophyses of the vertebrae. Some of these have a parial arrangement., and some are scattered. Upon lateral aspect of the bone we are to note how, upon either side, the ilium is produced as a conspicuous ledge overhanging the antitrochanter and the ischiadic foramen. A much smaller ledge of this kind is found just before the ilium terminates posteriorly; the two are separated by a considerable interval. This formation is the very reverse of what we find in the Rallidae. Turning to the under side of the pelvis of this crane, we find that it is seven of the leading vertebrae that throw out their lateral processes, to fuse by their other ends with the ventral surface of the ilium upon either side, instead of only five as in Aramus. Then follow three more, which have their apophyses thrown directly upwards against the pelvic roof, being thus concealed entirely upon direct ventral view. The 35th vertebra of the spinal column in this specimen, on its left side only, sends out a weak parapophysis, to reach over to the pelvic wall just above the cotyloid ring. But in the last six (five in Aramus) 'sacrals,' or true sacral vertebrae plus 'urosacrals,' these lateral apophysial braces are big and strong, and have their outer ends, and the adjacent ventral surfaces of the pelvic basin, all indistinguishably fused into one mass at their several points of meeting. Other characters of the renal fossae in Grus agree, in the main, with what we find in Aramus.
Both the pygostyle and the last two or three caudal vertebrae,
COMPARATIVE OSTEOLOGY, RAILS AND CRANES 745
in this museum specimen of Grus, have been lost (or cut off by the taxidermist), and I have but the leading four vertebrae of the skeleton of the tail before me. Judging from these, however, one is enabled to say that they are more highly pneumatic than they are in Aramus, but that there is the same stunting of the outstanding processes and spines as in that genus. And all these vertebrae are very small compared with those in the prepelvic part of the vertebral chain.
It has already been said above that the 18th vertebrae supports a pair of free ribs. They are pneumatic, nearly 3 cm. long, and without unciform processes and haemapophyses. Both these latter characters pertain to the ribs of the 19th vertebrae, and the uncifomi appendages may anchjdose to the borders of the ribs, though this is not the case with those in the middle of the dorsal series. Vertebrae 20 to 23 also have fully developed, highly pneumatic ribs, all connecting with the sternum by costal ones.
Then there are three pairs of pelvic ribs ; but the haemapophyses of the ultimate pair fail to have their costal ribs quite reach the sternum. Nothing peculiar marks any of these free pleurapophyses of the dorsal region; if we select one from the middle of the series, w^e find it considerably curved so as to conform to the outline of the thorax. It is constricted opposite the facet for the large, flake-like epipleural process; while from this point, both in the dorsal and ventral direction, it very gradually widens again. But at the best, these are narrow, though at their lower ends they become sufficiently broad to support a pretty goodsized facet for the costal rib.
The last pelvic pair of all are vers^ long and slender, and at their vertebral ends they are more or less rudimentarily developed — ^the tubercles being entirely aborted.
Grus americana, as well as Grus mexicana (fig. 2), has its os furcula completely fused with the sternum at the carinal angle, the point of fusion being broad and firm.
The 'body' of the sternum is long and narrow as in Aramus; its xiphoidal margin is incHned to be a little jagged, and presents no definite pattern of notching. The thoracic aspect of the
746
R. W. SHUFELDT
sternal body is more generally and roundly concaved than it is in Aramus, and the pneumatic foramina far more open. Small air-holes of this kind absolutely riddle the bone upon this surface in front. They are found upon either side of a prominent median mound which exists in this locality, it being the super D-external e\ddence of the coiling of the windpipe within the carina, to which reference will be made presently. This crane has seven haemapophysial facettes upon either costal border, to the six found there in Aramus. Its costal processes are very low, curl outward, and each is squarely truncated. The anterior border of the of the body of the bone is thickened, and upon either side of the median mound spoken of abo\'e, it overhangs the fossae it contributes to form in those localities; and it is within these fossae or recesses that the very numerous pneumatic perforations are seen, to which reference has just been made.
The 'coracoidal grooves' are very extensive, and are the same in general character as they occur in the limpkin, except that in Grus they do not decussate.
Below, the carina extends the entire length of the sternal body; transversely it is very thick, but this part is entirely hollowed out for two very different purposes. These are: the admission of air to the posterior moiety, while anteriorly a chamber is created in which the trachea is once or twice looped and coiled.
Fig. 6 Anconal aspect of the right humerus of Grus mexicana. Fig. 7 High'" scapula, dorsal surface. Fig. 8 Left femur, posterior aspect. Fig. 9 Distal extremity of the right tarso-metatarsus, anterior view. These bones are all from the same skeleton that furnished those seen in figures 2-5 (antea; No. 820, Coll. U. S. Xat. Mus.). Drawn by the author, natural size, and here reduced to two-fifths.
COMPARATIVE OSTEOLOGY, RAILS AND CRANES 747
The latter room, however, is not sufficient for the purpose just mentioned, so the cavity has ossified by extension far out in front, appearing as a median, rounded, transversely-compressed, closed protuberance between the costal grooves and the point below where the trachea enters the carina. The tracheal loop which is lodged in this part is the continuation of the same that passes round through the median mound on the thoracic aspect of the body of the sternum, to which I have already made reference above.
The sternum I have just described is, as stated above, from a specimen of Grus americanus, but the arrangement of the tracheal coils within the carina exist very much as they are seen in the figure showing them for Grus mexicana (fig. 2). Age and sex may have something to do with this; and a great many more sterna of these birds, at all ages, must be examined before the complete and full account of this very interesting feature can be written out. No such material is at hand at this writing.
Os furcula, as I have already stated, is, in Grus americanus, fused below with the angle of the carina of the sternum. It is quite a differently characterized bone as compared with the os furcula of Aramus. In the first place, it is more on the V-order of pattern than on the U; its rami are far more cylindrical; indeed, in the limpkin the rami are very decidedly flattened, and its free clavicular ends are very differently fashioned from those parts as found in the latter bird. Here, in Grus, they are pointed at their extremities. The articular facets situated below these points are comparatively very small; while below them, again, about a centimeter down either ramus, we meet with a moderate, more or less abrupt swell in the bone, which gradually dies away about half way down towards the symphysis. These enlargements occur upon the antero-external border of the clavicular ramus of either side — a border the more easily defined here by the very distinct muscular ridge-like lines that pass longitudinally down over the ramal surface. This clavicular fourchette of Grus differs quite as much from that bone as seen in Aramus, as the one in the latter differs from the bone in Rallus.
Apart from the great differences in size, the coracoids and
THE ANATOMICAL RECORD, VOL. 9, NO. 10.
748 R. W. SHUFELDT
scapulae in Grus americanus simply repeat, in all their essential characters, the corresponding elements of the shoulder-girdle as they are found in Aramus. WTien articulated in situ, however, the end of the clavicle on either side does not come in contact with the scapula.
A scapula in Grus americanus has just twice the length of that bone in the limpkin, and it is not as much curved; while the pneumatic foramen on the under side of its head is verv^ large and deep. There is a large pnemnatic foramen, too, on the posterior aspect of the expanded sternal portion of either coracoid in G. americanus, and these bones, in this species of crane, are relatively much shorter and stouter than are the coracoidsof Aramus; otherwise they have the same form and character.
So far as the axial skeleton is concerned, then, in the genera here compared, we can hardly say with Garrod that the "sternum in Aramus is completely Gruine, as are the other parts of its skeleton," for the differential characters are too marked, too important, and too numerous, when w^e come to regard them carefully in such representatives as Aramus vociferus and Grus americanus.
The appendicular skeleton. Little requires to be said here upon this subject, as the various bones were practically described during the comparisons made, on a former page, with those of the pectoral and pelvic limbs of Aramus. All the essential characters of the skeleton of the arm as seen in A. vociferus are repeated in the corresponding bones of that limb in Grus.
In the lower extremity we find, in the femur of Grus, a very deep 'rotular channel' between the condyles in front, and posteriorly there is constantly seen above the fibular groove of the external condyle a small, concave excrescence, to which is attached in life the femoral head of the flexor perforans digitorum profundus muscle.^ Only a slight roughness occurs upon the femur at the same site in Aramus.
Tibio-tarsus and fibula have identically the same morphological characters in the two genera. In the tarso-metatarsus, the groove seen in the hypotarsus of that bone in Aramus is com ' Myology of the raven, p. 1G7, fig. 46, a. Macmillan and Co., London, 1890.
COMPARATIVE OSTEOLOGY, RAILS AND CRANES 749
pletely sealed over in Grus, converting it into a cylindrical perforation passing vertically through the center of the projection. In addition, the back of the hypotarsus is shallowly grooved in two or three places for the passage of tendons. At the posterior aspect of the shaft the outer edge of the tendinal groove is, for its middle third, elevated as a conspicuous crest in this crane, and it serves well to keep the ossified tendons in the channel where they belong. For the rest, these bones in the two genera under consideration have similar characters.
Joints of the phalanges of pes are relatively much stouter and shorter in Grus than they are in Aramus ; in other words, the skeleton of the foot in Aramus, in so far as the toes are concerned, is markedly ralhne in type, notwithstanding the fact that all the other bones of its pelvic lunbs, with the exceptions mentioned, are so typically gruine.
In my former brief abstract of the osteologj^ of the birds which have been compared in the present paper, I published a 'synoptical table,' in which the skeletal characters of Rallus longirostris, Aramus vociferus, and Grus americana were critically compared in parallel columns. As stated above, this paper appeared a great many years ago in the Edinburgh Journal of Anatomy; and as that publication is to be found in the majority of the larger scientific hbraries in America, and is therefore accessible to students of avian osteology in this country, the aforesaid table has been omitted in the present contribution.
COXCLUSIOXS
In so far as the rails, cranes, and their allies, are concerned, as they are represented in North America and other regions where they occur, they are allied or related to each other as I have, in the main, provisionally pointed out in another contribution,^ thus:
R. W. Shufeldt. An arrangement of the families and higher groups of birds. Amer. Nat., vol. 38, nos. 455-456, Nov.-Dec, 1904, pp. 833-857. The part of the classification referred to is to be found on pages 851-852. As this classification of birds first appeared (that is, in the paper here cited), the family Aramidae was arrayed among the Ralliformes, while here it is placed among the Grues, where it really belongs.
750
R. T\'.
SHUFELDT
Supersuborder
XII
GRUIFORMES
Supersuborder ;
XIII
RALLIFORMES
Suborder XIX
Grues
Suborder
XX
Fulicariae
Superfamily
I
Gruioidea
Superfamily
I
Heliornithoidea
Family
I
Gruidae
Family
I
Heliornithidae
II
Aramidae
Superfamily
II
Ralliodea
III
Psophiidae
Family
I
Rallidae
Superfamily
IV
Cariamoidea
Family
I
Cariamidae
Superfamil}^
III
Eurypygoidea
Family
I
II III
IV
Eurypygidae Rhinochetidae ^Mesitidae Aptornithidae
These groups are arrayed between the supersuborders Stereornithif ormes and the Apterygif ormes, where they naturally belong.
My examination of the skeleton of the courlans demonstrates the fact that, in that part of their anatomy at least, those birds have the gruine characters predominating. But these gruine characters are not always typical, and the departures seen are frequently of a class that distinguish families among birds rather than genera. This settles the position of the courlans in the system as a family — the Aramidae of the crane-group (Grues).
In the A. O.U. "Check-List of North American Birds, published in 1895 (second edition), it will be noted that the Aramidae was placed as a family among the Ralli or rail-group; and that after my paper appeared in the American Naturalist ('04) it was removed, as a family, to the crane-group (Grues). where up to this writing, it has been very properh^ retained. There are not a few similar errors yet to be rectified in that surely not-up-to-date volume. However, its last edition has the Rallidae properly and naturally arranged^ — a fact upon which we may congratulate ourselves, perhaps all the more for the reason that the Grues have also been relegated to their true position in the system in that work ('10).
As elsewhere pointed out, Fiirb ringer is of the opinion that the Apteryges are far more closely related to the Rallif ormes than has heretofore been realized; and if this proves to be true, another linking line for the cranes and rails leading to the generalized struthious types is in evidence, with all the gallinaceous birds more or less related.
THE GROWTH AND VARIABILITY IN THE BODY \^^IGHT OF THE ALBINO RAT
HELEX DEAX KIXG
From the Wistar Institute of Anatomy and Biology
FIVE FIGURES
For several years investigations have been in progress in the animal colony of The Wistar Institute for the purpose of ascertaining the environmental and nutritive conditions most favorable for the development of the rat. The very flourishing state of the colony at the present time seems to indicate that these investigations have solved the problem of the care and breeding of this animal, and that in future it will be possible to supply a 'standardized' type of rat for laboratory use.
Growth records for the body weight of the albino rat have already been given by Donaldson ('06) and by Jackson ('13), but it seems worth while to publish additional data obtained from a study of a series of rats bred in The Wistar Institute colony in order to show the rapid and continuous growth of this animal in response to a particularly favorable set of environmental conditions. It is hoped that these records may also serve as standards with which the body weights of various strains of rats, raised under similar conditions, may be compared.
The thirteen litters used in this study were taken from the general colony of albino rats kept as 'stock' supply. In choosing the litters care was taken to select only those in which the individuals were of good size at birth and appeared strong and vigorous. The animals used, therefore, represent the best stock in the colony. A random selection of any stock Utters available for the purpose in mind was not feasible, as experience has shown that rats that are small and weak at birth do not, as a rule, grow at a normal rate and that they usually die at an early
751
752 HELEN DEAN KING
age: records from litters of this kind would have seriously affected the results.
Since the removal of young rats from the nest at or soon after birth often results in their being destroyed by the mother when returned, the first w^eighings were made when the Utters were thirteen days old. As at this age differences in the weights of the members of the litter are usually very slight, individuals of the same sex were weighed together and the average weight of the group recorded. The rats were weighed in a similar way when they were thirty days old, but thereafter the individuals of the litters were weighed separately at intervals of one month. Records were taken over a period of sixteen months, by which time so many of the rats had died that the investigation was brought to an end.
Members of the same litter were kept together and allowed to breed. This, of course, introduced the possibility of an error in the records for the body weights of the adult females. In the rat pregnancy can usually be detected by the twelfth or thirteenth day, at which time, as Stotsenburg's ('15) records show, the average weight of each fetus is only about 0.04 grams. At this stage the increase in the weight of the female as a result of pregnancy is comparatively slight, and as females known to be pregnant were never weighed, the records have probably not been affected to any great extent by this factor. Onlj^ a very few of the litters cast were reared, and the weights of nursing mothers were not recorded if they were below those of the last weighing before pregnancy was noted.
In any series of weighings of live animals there is always an unavoidable error due to the presence of a greater or less quantity of undigested matter in the alimentary tract. To minimize this source of error in these records the rats were always weighed in the morning before they had been fed.
An illness of any kind has a marked effect on the body weight of the albino rat, and cases are not uncommon in which animals have lost 100 grams in weight in the course of two or three weeks. Except in very old rats, a steady loss in weight for a period longer than one month is an almost certain indication of a chronic disease
GROWTH AND VARIABILITY IN WEIGHT OF RAT 753
that will eventually kill the animal. The body weights of animals obviously ill were not used in making up the final records, and as far as known the body weights given in the accompanying tables are those of animals in good physical condition.
All the rats were reared under similar environmental conditions. The food given them was a 'scrap' diet consisting of carefully sorted table refuse which was fed once each day, while corn on the cob was always available as an extra ration. Each cage had an abundant supply of water, which was renewed daily to prevent contamination.
GROWTH IN BODY WEIGHT OF THE ALBINO RAT
In order to obtain a series of records that would represent the normal increase in the body weight with age it was considered advisable to take a sample of the stock at two different periods rather than to rear a large number of litters at one time. This plan has extended the work of collecting records over the greater part of three years, but it has amply justified itself since it has shown that, when environmental conditions are uniform, there is comparatively little variation in the rate and in the extent of growth of rats born in different years. Records for the growth in body weight obtained from one set of stock albino rats can therefore be used as 'standards' for the colony, as long as the animals are reared under similar external conditions.
The first series of rats used in this investigation comprised seven litters born between December 26, 1912 and March 10, 1913. These litters contained a total of 46 individuals, 23 males and 23 females. The average weight of the indi\'iduals of each sex, together with the extreme body weights for each age at which records were taken are given in table 1. Individual data for this and also for the second series of rats studied are filed at The Wistar Institute.
In a series of investigations recorded in a previous paper (King '15), it was found that the average body weight of stock albino males at birth is 4.6 grams, and that the average birth weight of the females is 4.5 grams. According to these records the male
754
HELEN DEAN KING
rat is heavier than the female at birth, and, as shown in table 1, the male also exceeds the female in body weight at all ages at which data were collected, the difference in the weight of the two sexes becoming more marked as the animals grow older.
The growth of the albino rat is very rapid dm'ing the first 120 days of postnatal life, as Donaldson has already shown. After this age growth in body weight is relatively slow, as is
TABLE 1
Data on seven litters of stock albino rats, showing the increase in the weight of the
body with age {Series 1)
MALES
FEM.4.LES
AGE IN DAYS
Bod
\- weight in grams
No.
indi-'
viduals
Body weight in grams
No. indi
Average
Lowest
Highest
Average
Lowest
Highest
viduals
13
18.2
16
21
23
16.0
14
20
23
30
49.7
43
60
23
47.4
40
57
23
60......
131.0
87
170
23
113.2
77
153
23
90
188.5
125
238
23
148.9
99
178
16
120
230.3
146
284
23
173.1
125
197
17
151
252.1
196
307
23
181.0
152
215
18
182
268.0
195
343
23
197.5
147
245
21
212
276.0
221
366
22
192.3
136
245
18
243
281.7
194
355
21
200.2
141
256
19
273......
270.6
ao2
344
20
203
158
248
16
30J
281.4
198
366
IS
201.1
65
230
18
334
291.1
238
368
16
212.4
178
239
17
365
301.7
248
370
12
214.2
176
247
15
395
310.0
254
381
9
213.7
178
247
15
425
316.3
255
368
8
207.9
169
238
12
455
316.0
249
349
6
215.2
196
242
7
485
313.5
255
374
3
219.5
210
241
4
indicated by the data in table 1 . Increase in body weight does not entirely cease when the rats have reached maturity, and it usually continues as long as the animals are in a healthy condition. The later weight increase, however, is not true growth, but chiefly the accumulation of adipose tissue, as is the case in many other forms.
The second series of rats used in this study consisted of six litters, born during the first week of October, 1913. These lit
GROWTH AND VARIABILITY IX WEIGHT OF RAT
/OO
ters contained a total of 54 individuals, equally divided as to sex. Growth data for this series of rats are given in table 2. The growth records obtained for this series of rats are so nearly like those for the rats of the first series that the rate and extent of body growth in the individuals of the two series can best be compared through the growth graphs constructed from the average body weight of the animals as given in table 1 and in table 2.
TABLE 2
Data on six litters of stock albino rats, showing the increase in the weight of the body icith age (Series 2)
MALES
FEM.VLES
AGE IX
daVs
Bod
• weight in grams
No. individuals
Bod
• weight in grams
No. indi
Average
Lowest
Highest
Average
Lowest
Highest
viduals
13
16.4
13
18
27
15.3
13
18
27
30
47.4
41
54
27
44.2
39
50
27
60
116.0
64
149
27
101.8
66
124
27
90
181.6
103
226
27
147.7
95
177
23
120
217.1
151
274
27
173.7
132
212
25
151
238.6
169
294
27
189.8
147
225
27
182
250.1
172
303
27
195.1
149
232
21
212
261.3
176
324
26
201.4
167
238
24
243
277.8
206
334
23
216.5
171
256
24
273
290.7
218
349
21
216.6
170
254
22
304
310.8
220
379
18
235.2
177
262
20
334
309.8
240
385
17
231.8
170
276
18
365
310.6
245
377
16
231.6
177
276
16
395
317.4
246
376
15
226.9
175
284
16
425
310.0
243
397
15
221.0
178
271
18
455
329.2
289
414
9
223.4
198
264
11
485
330.3
276
437
9
241.4
197
324
9
Chart 1 shows the growth graph for the 23 males belongmg to the first series and also the growth graph for the 27 males of the second series of animals studied. In this, as m other charts where the graphs would properly run ver^^ close together at the beginning, the distance between the graphs has been shghtly exaggerated in order to keep the hues distinct.
756
HELEN DEAN KING
8 8 $ S
GROWTH AND VARIABILITY i:S WEIGHT OF RAT 757
In this chart the growth graph for the males belonging to the second series runs somewhat below that for the males of the first series until the rats have reached 240 days of age. The two graphs cross at this point and the graph for the second series then runs above the other until the end, except for a slight dip at the 425 day period. The pronounced drop in the graph for the first series that occurs at 280 days is not to be considered as normal. It is due to the fact that, when the majority of the rats in the series were about eight months old, the animals were removed to new quarters and the change, which unfortunately took place during a spell of cold, damp weather, so affected the animals that many of them, particularly the males, did not show a normal gain in body weight for about two months.
Graphs for the females of the two series, constructed from the average body weights as given in table 1 and in table 2 are shown in chart 2.
Growth graphs for the females belonging to the two series bear about the same relation to each other as do those for the males. The graph for the second series runs slightly beneath that for the first series in the beginning, it meets the other graph at the 120 day period, and subsequently runs higher for the remainder of its course. There is no drop in the graph for the females of the first series comparable to that found in the graph for the males at the 280 day period. The change of quarters which so adversely affected the growth of the males seemed to have had so little effect on the females that the growth graph has not been lowered at any point.
In both chart 1 and in chart 2 the growth graphs for the two series of rats born nearly a year apart run verj^ close together throughout their entire length. This indicates that, when environmental conditions are uniform, the growth of stock albino rats takes a very definite course and tends to produce animals having a like weight at any given age.
To summarize the results the growth data for all the individuals in the two series are combined in table 3. From the average body weights given in this table the graphs in chart 3 have been constructed.
758
HELEN DEAN KING
|,.,,|||||ll[l|||M||||||||g
LJU_L4^
!
" " ^ jT Ti"! ' ---_-
5 tn
- - ' 3
'-----^jh
•
" " '11 -I" "
. - |-i
i-- -3
] J :.: ::::. : :
■ " " ti -- -
""ij: - -i -
- _ - - - a
- - -- f i
..?
J:± --- - -i
• = - = = ---'— :::::::: + ::
|
J
- -- 5-5
-3 »
-c
-. -_ -1 S
< ^ - »
, + \
---^
- x::::::::::
I Jt
df \
%
% \ :: + :::::
•>
2 r -
.g — ' _ _ ,
-o ^^
D
< " V : _
N
~_ _ .1
1 S"5 :::i:;
"-::::::. .:.: :5
_§ ' :: .:: : :
_. . . - -3
-° 1
- ±± : :::::::.
: : ;:::""::::: : :::?
£ - T _. _
1 - J -
5 '" " :
c--- 1
; :
\
I
' " ""'::::::::
- ire .
"" " $ ::::. ,
- ar:..:___
3J --
_. _. __ \
"'!,- 5
" |: " :::_
Sjs - *
_ - _ --?S.
!"" "
::
--:!*.
!!j, 3
" 8" " "
__ _. _«h....^
a " . -
. _ , !!l._
% % % % % % %
s § i k i ^ "^
o
•M
T3
'hu
c3
a;
^
m
>>
.o
o
cS
Xi
c!
G
'-'
o3
a
c
GROWTH AND VARIABILITY IN WEIGHT OF RAT
759
A comparison of the graphs shown in this chart brings out very clearly the great difference between the growth of the male and that of the female rat. The graphs run very close together until the 60-day period. At this point they begin to diverge rapidly, the graph for the males soon appearing far above that for the females. At 200 days of age the average weight of the males is
TABLE 3
Data on thirteen litters of stock albino rats, showing the increase in the weight of the body with age {sumynary of the data in table 1 and in table 2)
MALES
FEMALES
AGE IN DAYS
Body weight in grams
No. individuals
Body weight in grams
No. indi
Average
Lowest
Highest
Average
Lowest
Highest
viduals
13
30
60
90
120
151
182
212
243
273
304
334
365
395
425
455
485
17.2 48.5 122.9 184.8 223.2 244.8 258.4 268.0 279.7 280.9 296.1 300.8 306.1 314.1 312.2 323.9 326.0
13 41 64 103 146 169 172 176 194 202 198 238 245 246 243 249 255
21
60 170 238 284 307 343 366 355 349 379 385 377 381 397 414 437
50 50 50 50 50 50 50 48 44 41 36 33 28 24 23 15 12
15.7 45,7 107.1 148.0 173.4 186.3 196.5 197.3 209.6 210.8 219.1 222.4 223.1 220.5 215.8 220.2 234.7
13 39 66 95 125 147 147 136 141 158 165 170 176 175 169 196 197
20 57 153 178 212 225 245 245 256 254 262 276 276 284 271 264 324
67 50 50 39 42 45 42 42 43 38 38 35 31 31 30 18 13
about 50 grams more than that of the females (table 3) , and this difference, as the growth graphs in chart 3 show, tends to become still greater as the animals grow older.
The growth of the albino rat in bod}^ weight has been studied by Donaldson, by Ferry ('13), and by Jackson. Growth graphs constructed from data obtained by the first two of these investigators and graphs from my own series of animals are given in chart 4 and in chart 5.
760
HELEN DEAN KING
GROWTH AND VARIABILITY IX WEIGHT OF RAT
761
The first data recorded for the growth of the albino rat in bodyweight were obtained by Donaldson from a series of animals reared in the laboratory of the University of Chicago. Bread and milk formed the staple food of the rats used in this investigation: this diet being varied occasionally by the addition of vegetables, meat, corn and sunflower seeds.
~20 55 50 yb
Chart 4 Graphs showing the growth in body weight of three series of male albino rats reared under different environmental conditions. A, graph constructed from data obtained by Donaldson; B, graph constructed from data furnished by Ferry; C, graph constructed from data recorded by King.
Donaldson's growth graph for the male albino rat (A, chart 4) is based on data for 19 animals weighed at varying intervals for one year. This graph has practically the same form as that for the male rats reared in The \yistar Institute colony (C, chart 4), but it runs considerably lower throughout its entire length. The space between these graphs represents, for the adult rats,
762 HELEN DEAN KING
a difference of about 20 grams in the average body weight of the two series of animals.
Graph B in chart 4 was constructed from growth data for about 50 male albino rats bred in the laboratory of Yale University. These data were kindly furnished by Miss Ferry, who used them in her study of "The rate of growth of the albino rat." As regards the food given the rats Miss Ferry states in a letter: The diet of the rats consisted of Austin's dog biscuit, sunflower seeds with fresh vegetables (chiefly carrots or corn and string beans) two or three times a week, and a small amount of cooked meat twice a week. A little salt was alwaj^s kept in the cages."
The growth graph for the albino males based on Ferry's data closely follows Donaldson's graph, running a little above it in the beginning but falUng below after the 60-day period. The drop in graph B at the end is due to the fact that Ferry included in her records the body weights of some adult animals that were probably ill.
The body weight of the albino rat at any given age depends to a considerable extent on the character of the food, as has been shown experimentally by Hatai ('07) and by Osborne and Mendel ('11). The fact that the male rats from which the data for Graph C were obtained grew^ more rapidly and attained a greater absolute weight than the rats reared by Donaldson and bj'Ferry can doubtless be ascribed in great part to the difference in the food that the animals received. The 'scrap' food given the rats in The Wistar Institute colony contains a relatively larger amount of meat and of fresh vegetables than the diet supplied to the rats used in the other two series of investigations noted. Considerable quantities of these substances are evidently needed by the rat to supply the materials needed to stimulate vigorous and continued growth.
Growth graphs for the body weights of the female albino rats used in these three series of investigations are given in chart 5.
The growth graphs for the three series of femalp rats bear very different relations to each other from those shown by the graphs for the males of the series (chart 4). The graph constructed from the data furnished by Miss Ferry (B, chart 5) is consider
GROWTH AND VARIABILITY IN WEIGHT OF RAT
763
ably lower than either of the other two graphs. This indicates either that exceptionally small rats were used for her investigations, or that the animals were not in good physical condition and so did not gain in weight at a normal rate. Whether these females were allowed to breed is not stated.
As investigations made by Watson ('05) had shown that female rats that are allowed to breed are heavier than mimated females of the same age, the growth graph for female albino rats as given by Donaldson, and reproduced as graph A in chart 5,
i I M I I I : I 1 i H I i I I I I H Growth in body weight Albino Rat ! ' ' ' I j i ' [^
j_LLm^— 1 I ! I ■ ; : .XI
z6 40 60 6U 100 120
Chart 5 Graphs showing the growth in bod}^ weight of three series of female albino rats reared under different environmental conditions; lettering as in chart 4.
was constructed from the actual weight data of unmated females up to the 90-day period, and beyond this point the data used were the body weights of unmated females corrected to accord with the weights of breeding females as calculated by Watson's formula.
The female rats belonging to Donaldson's series grew very slowly at first, as is shown by the position of graph A in chart 5. At 120 days of age the average body weight of the females used in my investigations was 173 grams. This is practically the es
THE ANATOMICAL, RECORD, VOL. 9, NO. 10
764 HELEN DEAN KING
timated weight for breeding females of this age as calculated by Donaldson. In chart 5, graph A meets graph C at this point and subsequently runs parallel with but slightly above it. It is, of course, impossible to obtain, at stated intervals, the true body weights of albino females that are allowed to breed. Pregnancy produces a temporary increase in weight that may amount to 50 grams or more, while the nursing of a large litter usually causes the female to lose considerable weight. To weigh females only after they have recovered from nursing a litter makes the intervals between the weighings too long to give weight records of much value. Positive errors in the weighings of females during early pregnancy undoubtedly occur in my records, but, as previously stated, such errors would be very slight, and they have probably been balanced by negative errors due to the weighing of nursing females or of females that were ill. Graph C is probably, therefore, fairly representative of the average body growth of albino females that are allowed to breed. It is of interest to note that the graph constructed from actual weight records for breeding females runs close to the theoretical graph.
In his paper on the postnatal growth of the rat, Jackson gives some data for the body weight of male and of female albino rats of various ages, but these data are not extensive enough to make it possible to construct from them growth graphs similar to those in chart 4 and in chart 5. From the records given it would appear that the rats used by Jackson were relatively small, since the males had an average weight of 213.0 grams when they were one year old, while the average weight of the females at this age was 163.7 grams: corresponding figures for the rats in my series give 306.1 grams as the average weight of the males, and 223.1 grams as the average weight of the females at one year of age.
An interesting comparison between the rate of growth of the male and of the female albino rat during the early stages of development has been made by Donaldson. His records show that as early as the seventh day after birth the female rat grows more actively than the male and is, as a rule, a relatively heavier animal up to about 55 days of age. At this point the male
GROWTH AND VARIABILITY IN WT:IGHT OF RAT 765
shows a very rapid gain in its rate of growth and he is much heavier than the female at all subsequent ages.
Jackson found that the excess of average weight was invariably in favor of the male at birth, and also in the majority of cases at all succeeding ages." In general, however, his data seem to confirm Donaldson's conclusion that the growth of the female is more -vigorous than that of the male during the first two months of postnatal life.
In the thirteen litters of rats used in my investigations there was only one litter in which the average weight of the females at thirteen days of age exceeded that of the males. In this case the average weight of the females was 18 grams while that of the males was 17 grams. In one htter the average weight of the females at 30 days of age was exactly the same as that of the males, but in all other htters the males averaged from one to eight grams heavier than the females. At 60 days of age a few of the females weighed slightly more than the smallest males in the same Utter, but in no case did any female have a weight equal to that of the largest male. In this series of animals the male tends to exceed the female in body weight in early as well as in late stages of development, but up to about 60 days of age the difference is ver>^ slight. Growth graphs for the two sexes, shown in chart 3, run veiy close together at first and begin to diverge only after the animals have reached 60 days of age. Female rats attain their maximum weight sooner than do the males, thus indicating that their rate of growth exceeds that of the males, although their absolute body weights may be less than those of the males at any given age. Evidence that the female rat tends to develop somewhat more rapidly than the male i's also shown by the fact that, as a usual thing, the female rats in a litter open their eyes several hom-s sooner than do the males.
VARIABILITY IX THE BODY ^\^EIGHT OF THE ALBINO RAT
Even with environmental and nutritive conditions as nearly uniform as possible, rats of the same sex belonging to the same Utter show marked differences in body weight that must be attributed to factors inherent in the 'germplasm' from which the individuals were derived. By studying the magnitude of the
766 HELEN DEAN KING
variation in these body weights we can obtain some idea as to the range in the action of these unknown intrinsic factors of growth even if we get no hint as to the nature of the factors themselves.
The relative extent of variability in body weight, as well as in other characters, is at the present time best determined through the coefficients of variation obtained for the body weights at each age at which weighings are taken. Since the frequency curves for both male and female rats when plotted from the data given in table 3 are fairly symmetrical, it w^as possible to obtain the coefficients of variation for the body weights according to Pearson's formula as given by Davenport ('14).
The index of variation (o-) was first obtained by the following method :
<^tf. ^^'■'^^'^ _ ^/sum of [(deviation of class from mean) ^ x frequency of class]
a =
/
number of varieties
htf
The coefficient of variation (C) was then found by the following formula in which the number of the varieties is indicated by the letter N.
(T
V I.- ;•, I C = — X 100 per cent.
The coefficients of variation for this series of body weights are relatively small, being in many cases less than 10 per cent. It was possible, therefore, to calculate the probable error in these coefficients (EC) by the formula:
c
EC = 0.6745
V2N
Table 4 gives the coefficients of \^ariation with their probable errors for the body weights of the male and of the female rats used in the two series of investigations described above. For the 13- and for the 30-day periods grouped data were used in making the calculations, as onl}^ the average body weight of the individuals of each sex was recorded in the weighings of the various litters at these ages: for all other ages the individual data were used.
GROWTH AND VARIABILITY IN ^^IGHT OF RAT 767
The range of variability in the body weights of the male rats is practically the same for the two series of htters weighed, as the difference between corresponding coefficients of variation is not more than three in any instance (table 4). The coefficients given in table 4 for the body weights at 13 and also at 30 days are doubtless lower than would be the case had the coefficients been calculated from indi\ddual and not from average body weights. The high coefficients for the males at 60 and at 90
TABLE 4
7^^
ERRATA
The Anatomical Record, volume 9, number 10, October, 915, in the middle of page 766, substitute for the lines and ormulas there printed these corrected formulas and lines.
_ ^|sum of [(deviation of class from mean)^ X frequency of class]
V^
number of variates
The coefficient of variation (C) was then found by the followig formula in which the mean of the variates is indicated by le letter A.
C = - X 100 per cent. A
days indicate that the maximum variabihty for this sex comes at this period. For the adult males the range of variabihty in body weight is practically constant for all the ages studied up to one year. From this point it tends to diminish shghtly, as Minot ('91) found to be the case in guinea-pigs.
The high coefficients of variation for the body weights of very old males, as shown in table 4, can have little significance although they occur in both series, owing to the small number of individuals
768 HELEN DEAN KING
weighed at this time and to the fact that the probable errors in the coefficients are relatively large.
Corresponding coefficients for the body weights of the female rats in the two series do not accord as well as do those for the males. There is, in fact, no agreement whatever between the coefficients for the first three ages at which the animals were weighed. In the first series the highest coefficient (16.1) is that for the females at 60 days of age ; in the second series the highest coefficient (14.9) is that for the 13-day period. Coefficients for the body weights of the females of the first series at 90 and at 120 days of age are practically the same as the corresponding coefficients for the second series, and the coefficients for later stages show no significant differences. The range of variability in the body weights of the females that lived to 485 days of age is curiously unlike in the two series. Females of the first series that lived to this age varied little in their body weights, as the coefficient of variation is only 5.6; while the females of the second series exhibited a very wide range of variability in their body weight, as is indicated by a coefficient of 14.1. The relation of body size to longevity in the rat is a point that will be considered in detail when a larger series of records is available for analysis.
The coefficients of variation for body weights at different ages, calculated from the data for all of the individuals in the two series, are given in the last two columns of table 4. From these coefficients it is possible to compare the range of variability in the body weight of 50 albino males with that of 50 females. At 13 and at 30 days of age the females seem to be quite as variable in body weight as the males, as the coefficients for body weight of the sexes at these ages are much the same. Males and females show their maximum range of variability in body weight at 60 days of age, but the coefficient of variation for the body weight of the males is 17.0 against 15.7 for that of the females. The males are more variable in body weight than the females from this time on, as at every subsequent age the coefficients for the male weights are higher than those for the female weights, although when the probable errors in these co
GROWTH AND VARIABILITY IN WEIGHT OF RAT 769
efficients are taken into consideration the difference in favor of the males is very' shght in many cases.
Other studies on the rat also seem to show a greater variability in the males than in the females. Hatai ('08) found that in skull measurements the males show a greater tendency to variability than do the females. Jackson's data show that males are more variable than females in body weight except at 20 days of age, and he concludes that "variability in body weight is lowest at birth (the coefficient being about 12) and is not much higher at seven days (16). It appears highest at three weeks (28), and at later periods varies from 19 to 21. The average coefficient, taking all ages together, is 19." In my investigations I find that the maximimi coefficient of variation for the body weight of the rat comes at 60 days in both sexes, and that the average coefficient is much smaller than that calculated by Jackson, being 13.6 for the males and 12.1 for the females. The fact that' Jackson's calculations were based on data obtained from rats that represented "for the most part a random sample of the general population at each age," while mine were based on the weighings of the same series of individuals throughout the entire period of observation, undoubtedly accounts for the differences in our results.
It has been held by many investigators, among whom are Darwin ('71) and Brooks ('83), that throughout the organic world the male tends to be more variable than the female. The known facts regarding variability in man have been collected by Ellis ('11) who sums up his discussion of the subject with the following statement: "In man, as in males generally, there is an organic variational tendency to diverge from the average; in woman, as in females generally, an organic tendency * * * to stability and conservatism involving a diminished indi\ddualism and variabihty." Pearson ('97) attempts to refute this theory, which he states is based chiefly on the fact that pathological variations seem to occur more frequently in man than in woman. After a critical examination of a large mass of growth statistics for various races Pearson concludes that, although men are more variable than wom.en at certain ages and in certain characters,
770 HELEN DEAN KING
yet there is not a pronounced difference between the sexes as regards variabihty, the weight of evidence indicating shghtly greater female than male variabihty. Data regarding variabihty in the rat as collected by Hatai, by Jackson and by myself do not support Pearson's contention, since these data show that, in the characters tested, there is decidedly greater variabihty in the male than in the female rat.
Boas ('97) and Porter ('05), among others, have shown that in man variabihty in body weight is correlated with rapidity of growth. A similar correlation in the rat is not shown by Jackson's data, although it seems to be indicated by my results judging from the relative size of the coefficients for body weight at various ages as given in the last two colunms of table 4. Since in this table the coefficients given for the 13- and for the 30-day periods were calculated from the average body weights of groups of rats and not from individual data they cannot justly be used to give evidence on this point. At 60 days of age, when the rats are still growing verj- rapidly as is indicated by the growth graphs in chart 3, the coefficients of variation are higher for both males and females than at any subsequent period. The rate of growth is beginning to slacken at 90 days of age, and the coefficients indicate a corresponding lessening in the range of variability of the body weights. At 120 days of postnatal life the period of rapid growth is at an end, and it is significant that at this point the coefficients for both sexes drop to the level that is maintained, with no important change, up to one year of age, when growth has practically stopped. Correlation between the rate of growth and variability in body weight exists in the rat in a late period of adolescence according to these records, but further investigations will be necessary^ in order to determine whether this correlation also exists at birth and during the early stages of development.
Two of the litters of rats used for this study contained an unusually large number of individuals of the same sex: one litter was composed of nine males and three females; the other had seven females and two males. For the purpose of comparing the variability within the litter with that of the general popula
GROWTH AND VARIABILITY EN WEIGHT OF RAT
771
tion as determined from the data for all of the Utters weighed, growth records for these two litters are given in table 5 and in table 6. Data for the males, together with their coefficients of variation, are presented in table 5.
On comparing the data in table 5 with the corresponding data in table 3, it is fomid that the average body weight of these nine males greatly exceeds that for the males of the general poptable 5
Showing the increase in body weight with age and the coefficients of variation for nine males belonging to the same litter of stock albino rats
AGE IX DATS
BODY
WEIGHT IX
J RAMS
COEFFICIEXTS
OF
VARIATIOX
XO. IXDI
Average
Lowest
Highest
VIDU.VLS
13
15.0 50.0
115.1 177.7 220.5 246.6 262.2 273.0 296.0 311.5 338.1 341.5 334.0 336.3 331.2 345.6 341.4
98 162 199 231 231 234 241 252 269 278 270 273 267 289 297
135 206 242 269 287 298 329 349 379 385 377 376 397 '414 437
9.6±1.52
7.9±1.25
6.5*1.03
5.1±0.80
7.1±1.12
6.8±1.14
10.1±1.69
9.5±1.59
10.0±1.68
8.6±1.44
10.0=^1.68
9.8±1.65'
13.3±2.23
13.7±2.66
15.0±3.19
9
30
9
60
9
90
9
120
9
151
9
182
9
212
8
243
8
273
8
304
8
334
8
365
8
395
8
425
8
455
6
485
5
ulation after the animals reach the age of 120 days. This result may be due, possibly, to the fact that the litter to which the males belonged happened to be by far the best of the 13 litters weighed, judging from the size and vigor of the individuals and from their longevity.
Growth data for the seven females belonging to the same litter with fhe coefficients of variation for their body weights at various ages, are given in table 6.
772
HELEN DEAN KING
After the age of 60 days the average weight of these seven females exceeded that of the females representing the general population, as may be seen by comparing the corresponding data in table 6 and in table 3. The largest females in the series studied were not contained in this litter, but were members of the litter containing the nine males whose growth data are given in table 5. \Miether there is any correlation between the number of individuals of the same sex in a litter and the size of the indi TABLE 6
Showing the increase in body weight with age and the coefficients of variation for seven females belonging to the same litter of stock albino rats
AGE IN DATS
BODY
WEIGHT IN
GBAMS
COEFFICIENT
OF VARIATION
NO. INDI
Average
Lowest
Highest
13
12.5 41.0 105.5 164.0 176.5 204.2 205.4 215.6 225.0 231.4 242.8 240.0 239.0 221.6 224.4 220.1 234.8
99 143 160 163 190 197 204 193 212 215 220 198 203 198 226
121 177 195 220 223 228 256 254 265 259 255 243 250 239 250
6.4*1.15 9.2*1.78 6.5*1.16 8.9*1.59 6.6*1.40 4.9*1.04 7.7*1.38 8.5*1.53 7.4*1.32 6.3*1.22 5.3*1.12 6.5*1.68 5.4*1.15 6.0*1.16 4.3*0.91
7
30
7
60
7
90
6
120
7
151
7
182
5
212
6
243
7
273
7
304
7
334
6
365
5
395
5
425
7
455
6
485
5
viduals, as the records from these two litters seem to indicate, can only be determined from the data of a much larger series of litters.
The coefficients of variation for the males of one litter, as given in table 5, are considerably lower than those for the males of the general population (table 4) up to the time that the animals were 425 days old: the small number of individuals weighed at an advanced age make it impossible to draw any conclusions from
GROWTH AND VARIABILITY IN WEIGHT OF RAT 773
the data given. Coefficients of variation for the body weight of the seven females from the one htter are hkewise considerably below those for the females of the general population (table 4). From these facts it follows that variability within the litter unit is less than that of the general population, as Jackson has already shown.
The variability in the body weights of the male members of a litter is greater than that in the female members of a litter, as is shown by comparing the coefficients of variation given in table 5 with those in table 6. The difference, as might be expected, is about the same as that between the males and females of the general population.
From an examination of the coefficients of variation for the body weights of the individuals belonging to several litters, Jackson concludes that "in general the variation in body weight within a given litter of albino rats is probably less than half that of the general population of the same age under similar environment." Taking all ages together I find that the average coefficient of variation for the body weights of the nine males from the one litter is 9.5, while that for the males of the entire series is 13.6 (table 4); for the seven females belonging to the same litter the average coefficient of variation is 6.7, that for the females of the general population being 12.1. In this instance the range of variability within the litter is about 70 per cent that of the general population in the case of the males, while for the females it is about 55 per cent. These figures seem to indicate that the relation between fraternal variability and racial variability for the body weight of the rat is much the same as that for human stature which, according to Galton ('94), is approximately 62 per cent. Definite conclusions regarding this point cannot be drawn from a consideration of the records from such a small number of litters, but must be deferred until a larger series of data is available for analysis.
774 HELEN DEAN KING
SUMMARY
1. The present paper gives the data for the increase in body weight with age for two series of stock albino rats reared under similar environmental conditions. Altogether thirteen litters containing 100 individuals, 50 males and 50 females, were used in this study.
2. Growth graphs for each sex, constructed from the average body weights at various ages, are practically the same for the two series of rats used (charts 1 and 2). From this fact it follows that, when environmental conditions are uniform, the growth of albino rats within a given colony tends to follow the same course and to produce individuals having a like weight at anystated age.
3. As a rule, the male rat is heavier than the female at birth and also at all subsequent ages at which records were taken. During the first 60 days of postnatal life the body weight of the female tends to approach that of the male, but after this age the male grows more rapidly than the female and soon greatly exceeds her in body weight. At 200 days of age the male rat weighs, on the average, about 70 grams more than the female of the same age (chart 3).
4. The female rat tends to increase in body weight at a much more rapid rate than does the male during the early stages of development, and she reaches her maximum weight much earlier than does the male.
5. The environmental and nutritive conditions under which rats are reared have a marked influence on their body weights, as is indicated by the relation of the growth graphs constructed from data obtained from three different series of rats reared under different conditions (charts 4 and 5).
6. Variability in the body weight of the albino rat, as measured by the coefficients of variation, is greatest when the animals are about GO days of age. It decreases slightly at 90 days, and
GROWTH AND VARIABILITY IN WEIGHT OF RAT 775
after 120 days remains practically constant until the animals are about one j^ear old.
7. Very young female rats seem to show as great a range of variability in body weight as do the males, but the males are more variable than the females at all later stages of growth.
8. The average coefficient of variation for the body weights of the 50 male rats used in this study is 13.6; that for the females is 12.1.
9. In the rat there is apparently a direct correlation between the rapidity of growth and the variability in body weight after the animals have reached 60 days of age. The records collected are not in a form to give evidence regarding the correlation that exists at earlier stages of growth.
10. Fraternal variability in the rat is less than racial variability. For the male rat the fraternal variability is about 70 per cent that of the general population; for the female it is about 55 per cent.
LITERATURE CITED
Boas, F. 1897 The growth of Toronto school children. Report U. S. Commissioner of Education, Washington.
Brooks, W. K. 1883 Heredity. Baltimore, Md.
Darwin, C. 1871 The descent of man. London.
Davexport, C. B. 1914 Statistical methods with special reference to biological variation. Third edition; New York.
Donaldson, H. H. 1906 A comparison of the white rat with man in respect to the growth of the entire body. Boas Anniversary Volume, New York.
Ellis, H. 1911 Man and woman. Fourth edition; Scribners', New York.
Ferry, E. L. 1913 The rate of growth of the albino rat. Anat. Rec, vol. 7.
Galton, F. 1894 Natural inheritance. New York.
Hatai, S. 1907 The effects of partial starvation, followed by a return to normal diet, on the growth of the body and central nervous system of albino rats. Am. Jour. Phys., vol. 18.
1908 Studies on the variation and correlation of skull measurements in both sexes of mature albino rats (Mus norvegicus var. alb.) Am. Jour. Anat., vol. 7.
Jackson, C. :\I. 1913 Postnatal growth and variability of the body and of the various organs in the albino rat. Am. Jour. Anat., vol. 15.
776 ' HELEN DEAN KING
King, Helen Dean 1915 On the weight of the albino rat at birth and the factors that influence it. Anat. Rec, vol. 9.
MiNOT, C. S. 1891 Senescence and rejuvenation. I. On the weight of guineapigs. Jour. Phys., vol. 12.
Osborne, T. B., and Mendel, L. B. 1911 Feeding experiments with isolated food substances. Carnegie Inst. Washington.
Pearson, K. 1897 The chances of death. London.
Porter, W. T. 1905 Growth of St. Louis children. Trans. St. Louis Acad, of Science, vol. 6.
Stotsenburg, J. M. 1915 On the growth of the fetus of the albino rat from the thirteenth to the twenty-second day of gestation. Anat. Rec, vol. 9.
Watson, T. B. 1905 The effects of bearing young upon the body weight and the weight of the central nervous system of the female albino rat. Jour. Comp. Neur., vol. 15.
TAILLESSNESS IN THE RAT
SARA B. CONROW
From the Wistar Institute of Anatomy and Biology
THREE FIGURES
CONTENTS
Introduction 777
Historical survey 777
Material and methods 779
Description of the rats examined 780
Summary 783
Literature cited .^ 784
INTRODUCTION
There seems to be a general opinion that when a rat appears without a tail it means the loss of the tail by accident early in the animal's life, and it is usually suggested that it was bitten off by another rat at the time of birth or soon after. The objects of this paper are to describe skeletal conditions in the region posterior to the thoracic vertebrae of several tailless rats, and to correct the existing impression that a tailless rat occurs through the accidental loss after birth of a once existing tail. As to the word tailless; by tailless we mean here "with no caudal vertebrae." There are cases, of course, where, from disease or from some accident after birth, the tail has become simply a stub, but in these cases some caudal vestebrae remain.
HISTORICAL SURVEY
Little data seems to have been recorded concerning the tailless condition of the higher animals. Short tails have been noted among cats, fowls, and dogs, while the number of caudal vertebrae 'has been found to vary in some other animals, but only
778 SARA B. CONROW
among dogs have cases been recorded where the caudal vertebrae of a mammal were completely absent.
Hind ('89), Anthony ('99), Kennel ('01), and Davenport ('05) all give accomits of mating ^lanx cats or short-tailed cats with the long-tailed variety and having the short tail appear in many of the offspring. The ordmarj^ cat, according to Flower ('85), has twenty-two caudal vertebrae, and accordng to Jayne ('98), eighteen to twenty-six. As to the number of caudal vertebrae of the short tailed cats. Flower ('85) gives three for the Manx cat, Anthony ('99) describes six, and Kennel ('01) speaks of six 'post-sacral' vertebrae in a so-called tailless cat.
Concerning fowls, Godron ('65) in a footnote says that complete lack of coccyx has been observed in a large number of fowls and that the character is ver>" readily transmitted. He does not give any details of vertebral conditions in these cases, but, from descriptions of other 'runipless' fowds, we conclude that the cocc>^ here probably was not completely lacking. In the ordinary fowl Davenport ('06) gives the number of free, caudal vertebrae as five, followed by a fused portion, the uropygial bone. Davenport ('06) describes a rmnpless game female as having two unsymmetrically formed and intimately fused caudal vertebrae, followed by a knob of bone about 1 mm. in diameter, Darwdn ('83) speaks of the caudal vertebrae in three rumpless fowls as being few in number and anchylosed together into a misformed mass. He also reports the inheritance of rumplessness in fowls, as does Davenport ('06).
Some other animals have been recorded as showing variation in the number of their caudal vertebrae. Bateson ('94) gives the following:
Man: (Normal number of caudal vertebrae, according to Flower, '85, three to four). A male with sacral and caudal vertebrae anchylosed together and of uncertain number; a female with the coccyx of three pieces anchjdosed together. Anihrojyoid apes:
Chimpanzee: (Nonnal number of caudal vertebrae, according to Flower, '85, five). One animal had six caudal vertebrae and others had two to four.
TAILLESSNESS IN THE EAT 779
Orang-utan: (Xonnal number of caudal vertebrae, according to Flower, '85, four). One animal had three caudal vertebrae, three each had two caudal vertebrae, a fifth had the caudal vertebrae anchylosed with the sacral, and a sixth had only one caudal vertebrae.
Sloths: Among the sloths there was considerable variation from the normal number of caudal vertebrae.
None of the above refer to an absolutely tailless condition. We have, however, recorded by Godron ('65), a complete absence of caudal vertebrae in the dog. He examined by palpation the sacral region of a tailless female \vater spaniel and found at the end of the vertebral column a rounded, bony surface w^hich he took to be the last sacral vertebra, and concluded that the coccyx was gone completely. (According to Flower, '85, the caudal vertebrae of the dog number from fifteen to twenty-three.) In his account of this female he states that, w^hen she w^as mated with a tailless brother, six of the litter of seven were absolutely tailless. "WTien she w^as mated wdth a long-tailed male water spaniel, the litter of four were all tailless like the mother. The grandfather of these dogs had a rudmientary tail 3 cm. long. Godron ('65) speaks also in a footnote of a tailless species of dog, the Dalmatian hound or brach-hound of Bourbonnais.
In all of these accounts of short-tailed and tailless animals the conditions referred to are congenital and not the result of accident.
MATERIAL AND METHODS
The material for this studj" was obtained from The Wistar Institute rat colony and consisted of rats of the species J\Ius norvegicus albinus and ]Mus norvegicus (pied). Specimens for the work were not abundant, since, during the past nine years, only five tailless rats have appeared in the colony, although forty thousand rats have been observed. Of these five, one was eaten by an older rat soon after birth, one male is at present mated in the colony, and the remaining three rats were killed and examined.
The method of examining these rats was as follow^s: The animal was chloroformed, its body weight and body length taken,
THE AXATOMICAL RECORD, VOL. 9, NO. 10
780
SARA B. CONROW
and its skin and viscera removed. The vertebral column with pehdc girdle attached was partly cleaned of its masses of muscle, covered with a boiling hot, 2 per cent solution of 'Gold Dust' washing powder (a 1 per cent solution was used for the youngest rat), and kept at about 95°C. until the flesh was softened so that it could be removed. This was done in tap water and, in addition to the usual instruments, a bone-scraper and a toothbrush were used. Care was taken not to separate the vertebrae in the region of and caudal to the pelvic girdle attachment. The bones now were dried and placed in corked vials with tar camphor balls to protect them from Anthrenus.
DESCEIPTION OF THE RATS EX.AJMINED
Of the three tailless rats whose vertebrae were examined, one was a female and two were males. The data concerning them are presented in table 1. The normal rat of this species (Mus norvegicus) has six lumbar vertebrae, four sacral, and twentynine to thirty-one caudal.
Rat No. 1, the female, was of the species Mus norvegicus (pied). Its parents were among some pied, pet rats of the colony. It was two A^ears old when killed, its body weight was 171.3
TABLE 1 Data on tailless rats
RAT NO.
SPECIES
SEX
AGE DAYS
BODY WEIGHT
BODY LENGTH IN MM.
PARENTS
DATE OF KILLING
grams
1
Mus norvegicus (pied)
9
730
171.3
193
among pied pet rats
12-14-' 14
2
Mus norvegicus albinus; half inbred; 11th generation
cf
388
190.9 (month earlier 210)
194
9X6" 10th stock gen. inbred
o-19-'14
3
Mus norvegicus albinus; stock strain
d"
30
21.7
89
2- 8-' 15
TAILLESSNESS IN THE RAT
781
Fig. 1 Rat No. 1, ventral aspect; 1, pelvic girdle; 2, sixth lumbar vertebra; 3, three modified sacral vertebrae.
Fig. 2 Rat No. 2, ventral aspect; 1, pelvic girdle; 2, fifth hmibar vertebra; 3, sixth lumbar vertebra; 4, two modified sacral vertebrae.
Fig. 3 Rat No. 3, right lateral aspect; 1, thirteenth thoracic vertebra; 2, two first lumbar vertebrae; 3, two or three modified lumbar vertebrae; 4, pelvic girdle.
grams, and its body length was 193 mm. The arrangement of the most posterior vertebrae of this rat and their relation to the pelvic girdle may be seen in figure 1. Supposing all the lumbar vertebrae to be present, we have here, posterior to the lumbar vertebrae, only three modified sacral vertebrae. Thus one sacral vertebra and all of the caudal vertebrae are missing.
782 SARA B. CONROW
The striking feature here then is that the vertebral column ends posteriorly about midway of the long axis of the pelvic girdle, far in from the posterior end of the body.
Rat No. 2, a male, was a half-inbred albino (Mus norvegicus albinus) of the eleventh generation. Its mother was a strict inbred of the tenth generation and its father a stock male. When killed it was 388 days old, its body weight was 190.9 grams (one month earlier it had weighed 210 grams), and its body length was 194 mm. The arrangement of the most posterior vertebrae and their relation to the pelvic girdle may be seen in figure 2. Supposing all the lumbar vertebrae to be present, we have here, posterior to the lumbar vertebrae, only two modified sacral vertebrae. Thus two sacral vertebrae and all the caudal vertebrae are missing. Here again the vertebral column ends posteriorly far up the long axis of the pelvic girdle, even anterior to the middle of its axis.
Rat No. S, a male, was an albino rat (Mus norvegicus albinus) of unknown parentage. It was thirty days old when killed, its body weight was 21.7 grams, and its body length was 89 mm. This rat was small and in rather poor condition. The arrangement of the most posterior vertebrae and their relation to the pelvic girdle may be seen in figure 3. We have here no sacral vertebrae, but two good lumbar vertebrae and two or three modified lumbar vertebrae. The last (sixth and perhaps fifth also) lumbar vertebra, all of the sacral vertebrae, and all of the caudal vertebrae are missing. In this case then the vertebral column is more modified than in the other two rats, for here the vertebrae reach only to the anterior part of the pelvic girdle, and the girdle is attached to the column merely by a small surface near its anterior end. This mode of attachment allows the posterior end of the girdle to hang very low down, almost at right angles to the column. In the living rat the sagging of the girdle was very noticeable, as it allowed the head of the femur to drop far down and thus gave an odd appearance to the posterior part of the rat's body.
This completes the description of the three tailless rats whose bones have been examined. As to the tailless male albino rat
TAILLESSNESS IN THE RAT 783
which has been mentioned as at present mated in the colony, we are confident from inspection that in this animal also the vertebral column ends far forward along the long axis of the pelvic girdle, for this condition may be felt distinctly by pressing the finger on the rat's back in the pelvic girdle region.
The vertebral structure of all the tailless rats which we have examined seems to show that the deformity is not due to an accident after birth, since in each case the column ends in the pelvic region far from the posterior end of the body. We conclude, therefore, that the rats were born tailless, and even more than tailless, since they lack more than the caudal vertebrae.
SUMMARY
An examination of the vertebrae of three tailless rats showed that all of them lacked all of the caudal vertebrae; and besides, the first lacked one sacral vertebra; the second, two sacral vertebrae; and the third lacked one (and perhaps two) lumbar vertebra and all four of the sacral vertebrae.
In each case the vertebral column terminated in the pelvic region far anterior to the posterior end of the body, showing that the tailless condition was due not to accident after birth but to a congenital deformity of the vertebral column.
784 SARA B. CONROW
LITERATURE CITED
Anthony, R. 1899 Heredity in Manx cat. Bull. Soc. Anthr., p. 303. Bateson, W. 1894 ^Materials for the study of variation. Macmillan and
Company, London and New York. Darwin, C. 1883 The variation of animals and plants under domestication.
Second edition, vol. 1, p. 281; vol. 2, p. 4; D. Appleton and Co., New
York. Davezs^poet, C. B. 1905 Details in regard to cats. Report on the work of the
Station for Exp. Evol., Cold Spring Harbor; Fourth year-book, Carnegie Institute.
1906 Inheritance in poultry. Publ. Carnegie Inst., Xo. 52. Flower, W. H. 1885 An introduction to the osteology of the ]\Iammalia.
Macmillan and Company, London. GoDRON, D. A. 1865 De la suppression congeniale de 1' appendice caudal.
Observee sur une famille de chiens. !Mem. Ac. Stanislas. Hind, W. 1889 Taillessness in the Manx cat. Ann. Rep. N. Staffs. Field
Club, p. 81. Jayne, H. 1898 jNIammalian anatomy. J. B. Lippincott Co., Philadelphia;
London. Kennel, J. 1901 Ueber eine stummelschwanzige Hauskatze und ihre Nach kommenschaft. Zool. .lahrb., Syst., vol. 15, p. 219.
AN ANOMALOUS ORIGIN OF THE SUBCLAVIAN
ARTERY
GEORGE BEVIER
From the Department of Anatomy of Stanford University
THREE FIGURES
The subject containing this variation is that of a large Caucasian female, aged probably 40 or 45 years, with a good muscular development and no signs of emaciation. The right subclavian arises from the aortic arch distal to the left subclavian, and the right common carotid comes directly from the aortic arch in the position usually occupied by the anonyma, which is absent in this case. Hence, instead of arising from the anonyma, the a. subclavia dextra arose directly from the aortic arch, at a point 1 cm. distal to the origin of the normal a. subclavia sinistra. From here it took a com'se toward the right and cephalad, across the ventral surface of the second thoracic vertebra passing dorsal to the esophagus and trachea, and making an angle of about forty degrees with the latter. Its branches are (1) an a. vertebrahs dextra which arises 4 cm. from the origin of the right subclavian; (2) an a. cervicalis profunda, arising from the dorsal aspect of the subclavian 7 mm. lateral to the vertebral; (3) an a. cervicalis ascendens coming from its cranial margin 12 mm. distal to the vertebral; (4) an a. mammaria interna, from the caudal border directly below the ascending cervical; and (5) an a. transversahs colli, in the normal position, 18 mm. distal to the vertebral. After passing dorsal to the m. scalenus anterior the anomalous vessel follows a normal course into the axila. The truncus thyreo-cervicalus, the a. thyreoideus inferior, and the truncus costocervicalis were absent.
The a. subclavia sinistra arose in the usual place, but presented abnormalities in its branches similar to those on the right side. There are no truncus thyreocervicahs, an a. thyreocervicahs
785
786 GEORGE BEVIEE,
inferior, or a truncus costocervicalis. The following rami arise in this region: (1) the a. vertebrahs sinistra; (2) a small a. intercostalis suprema; (3) an a. cervicaHs ascendens; (4) an a. mammaria interna; (5) an a. cervicalis profmida; and (6) an a. transversalis colli.
The aa. thyreoidea inferiores were absent and the a. thyreoidea ima was also missing. However, the a. thyreoidae superior on the right side w^as miusually large and divided into two branches as high as the level of the hyoid bone.
The right lobe of the thyroid gland was also much larger than the left and was formed by two partly distinct lobes. It is possible, but unhkely, that the upper of these two dextral lobes represents the pyramidal process, which in this case has been developed as an extra dextral lobe. A well-developed levator glandulae thyreoidea was present, connecting the capsule of the gland to the hyoid bone.
The a. carotis commimis dextra is longer than usual for it also takes the place of the anonyma. As shown in figures 1 2, it follows the normal course of these two arteries, passing anteriorly and to the right as far as the right margin of the trachea and then turns almost vertically cephalad. As there is no a. thyreoidea ima, the only vessel given off below the a. thyreoidea superior is a small branch arising 13 mm. superior to the aortic arch on the ventral surface. This supplies the pericardium and the remnants of the thymus.
The n. recurrens laryngis forms a short loop above the subclavian and does not hook around it, as is usually the case. As is well known, the nn. recurrentes hook about the fourth aortic arches and as these develop into the anonyma and subclavian on the right side, and the aortic ai'ch on the left, the nerves are dragged down to a lower level, thus establishing the recurrent condition. The fact that the right recurrent nerve does not hook around the subclavian artery may indicate that the latter was not developed from the fourth arch. An interesting recent article on the embryological development of such variations may be found in connection with a report of a sinnlar anomaly by Cobey (;14).
ANOMALOrS ORIGIN OF SUBCLAVIAN ARTERY
787
N. vagus
A. vertebraiis- — — A. cervicalis profunda A. cervicalis ascenden A. transversa
A. carotis commune
A. vertebraiis
A. cervicalis profunda A. cervicalis ascendens
A. transversa colli
Fig. 1 Diagrammatic sketch of the actual arrangement.
Fig. 2 Trachea and esophagus pulled forward, heart and aortic arch thrown toward the left; 1, transverse portion of the aortic arch; 2, descending aorta; 3, trachea and esophagus; 4, a. carotis communis dextra; 5, a. carotis communis sinistra; 6 a. subclavius sinistra; 7, a. subclavius dextra; 8, a. mammaria interna dextra; 9, v. azygos (retracted); 10. v. cava superior.
788
GEORGE BEVIER
There was no noticeable aortic impression on the vertebral column, but the anomalous subclavian (which arose from the aortic arch pointing directly toward the vertebral column and which then turned to the right at an angle of nearly ninety degrees to cross the body of the second thoracic vertebra) produced a distinct groo^'e on the left and also a flattening across the ventral aspect of the body of that vertebra. However, if the current in the artery was impeded because of its position and origin the obstruction was not sufficient to affect the development of the right arm, for both arms seemed equalty well-developed
A. vertebralis A. subclavius dextra
Ductus arteriosus A. subclavius sinistra
Fig. 3 Diagram indicating possible development of this anomaly (modified from Piersol).
and were covered with a good panniculus adiposus. Strangely enough, the right ulna was 1 cm., and the right radius 0.9 cm. longer than the corresponding bones of the left arm although the humeri were the same length. Although the arteria volaris superficialis originated abnormally from the radial artery about midway between the origin of the radial and the wrist no other vascular anomalies were present in the arm. The mm. palmares longi and extensores carpi ulnaris were completely absent. The veins also exibited slight abnormalities in the region under discussion. The vena azygos was unusually large, having a diameter of nearly 1 cm. where it joined the aorta. The v. hemiazygos crossed the vertebral colunm at the level of the
ANOMALOLS ORIGIN OF SUBCLAVIAN ARTERY 789
8th and the v. hemiazj^gos accessorius at the level of the 6th thoracic vertebra. It anastomosed with the v. mtercostalis suprema, which joined the v. anonyma sinistra dorsal to the internal niainmary and together they drained the first seven intercostal spaces.
The accompanying diagram (fig. 3), modified from Piersol, indicates the probable developmental explanation of this anomaly.
In conclusion, I wish to express my thanks to Professor ]\Ieyer for his assistance in the preparation of these notes.
LITERATURE CITED
Brodie, G. 1888-89 Abnormality of the aortic arch. Proc. See. Anat. Great
Britain and Ireland, London, p. 7. CoBEY, J. F. 1914 An anomalous right subclavian arterj-. Anat. Rec. vol. 8. Deaver, J. B. 1888-89 Anomalies of the arch of the aorta. I'niv. M. ^lag.,
Philadelphia, vol. 1. Smith, W. R. 1890-91 An abnormal arrangement of the right subcla\'ian
artery in a rabbit. Journ. Anat. and Phsyiol., London, vol. 25. Walsh, J. J. 1897-98 Right subclavian arising from the lower part of the
descending arch of the aorta. Univ. ^L Mag., Philadelphia, vol. 10.
APPLICATION OF THE CAJAL IMETHOD TO TISSUE PREVIOUSLY SECTIONED
EDWARD F. r^IALONE
From the Anatomical Laboratory of the University of Cincinnati
While engaged in a stiiclj' of the mammahan brain-stem it became necessary for me to prepare serial sections of certain regions by the Cajal method. In order to obtain a series of a large portion of the brain-stem, and in order to stain anj' individual section by the method of Nissl instead of that of Cajal, the problem was presented of applying the Cajal method to mounted sections. Since the Cajal method can be applied to tissue fixed in strong alcohol (the best fixative also in case of the Xissl method) the problem appeared to resolve itself into the establishment of certain conditions, mechanical rather than chemical, which would insure a uniform and sharp reduction of the silver in the cell-bodies and cell processes. Not only has this expectation proved justified but it has been possible to obtain satisfactory Cajal preparations from sections previous!}' stained by the Nissl method; the value of applying the Nissl and Cajal methods successively to the same section needs no comment. Aly problem involves the demonstration of all portions of the neurones rather than that of the internal structure of the cell-body, and although the internal structure of many cells is well shown I have made no effort to adapt conditions to obtain this end. Of course the method should be varied according to the animal, region of the nervous system, or the especial picture desired. I shall now describe a method which fulfilled the requirements of my problem, namely, the correlation of the internal structure of the cell bodj' (Nissl) with the connections of the cell processes (Cajal).
I shall assume that the tissue to be prepared is an adult human brainstem. The entire brain-stem is placed in two liters of 95 per cent alcohol and kept in the ice-box, but the tissue should not freeze. After two hours change alcohol and leave in cold overnight; thereafter the alcohol should be changed once every day. After two or three days the tissue should be cut into pieces 1 cm. in thickness, and should remain in 95 per cent alcohol two or three days longer (always at low temperature). Dehydrate several days in absolute alcohol; clear in chloroform of good quality for two days or longer, changing twice. The pieces of tissue are then placed in a mixture of equal parts of chloroform (fresh) and 42° paraffin at 35°C. for at least twelve hours and at 50°C. for four hours. To remove chloroform place in 42° paraffin at 50° C. for at least twelve hours (four to six changes). Embed in 50 to 55° paraffin. The essentials of this technique, which is the best
791
792 EDWARD F. MALONE
one for the Nissl method also, are the avoidance of acid, quick and thorough removal of water (keeping tissue in the cold until dehydrated) and thorough impregnation with paraffin. Prolonged stay in alcohol and prolonged heating are necessary, and poor results are due not to these causes but, on the contrary, to maceration in weak alcohol and to incomplete dehydration and impregnation.
The sections are then cut; for the purpose of following the course of cell-processes 24 ix is not too thick. The sections to be prepared by the Cajal method are mounted on large slides (2 X 3 inches), while every sixth and every seventh section is mounted on a separate slide (one section on each slide) to be stained by the Nissl method. Of the two Nissl preparations one is retained permanently, while the other, after being studied and drawn, is restrained h\ the Cajal method. Mounting the sections demands certain precautions: The slides are covered with a thin film of fresh egg albumen, without glycerine or preservative, flooded with distilled water, and the albumen uniformly distributed by rubbing with a brush. The slides are then placed on a warm bar or water l)ath and the sections allowed to flatten out. Drain off excess of water and after the slide begins to cool (about ten seconds) express the water from beneath each section by means of a small brush. The brush must be moist but not wet, and must be rotated so that it passes over the section as a roller. The direction of the movement is indicated by the nature of the section; in the case of large sections it is often necessary to begin at the center of each section and work outward radially. The water pressed, out must be removed from the slide. If the sections are not too warm and if the brush passes over the sections as a roller, a fair amount of pressure may be employed; but at first the pressure should be light and increased after most of the water has been expelled. I recommend this method for mounting large paraffin sections of the brain (especially if the sections be rather thick) , and have never observed any bad results. The sections dry rapidly. The essentials in mounting are to obtain a uniform mixture of distilled water and fresh egg albumen, and to express all water from beneath the sections by rolling them with a brush. The brush should be about 4 mm. in diameter and about 1 cm. long.
When the sections are drj' the paraffin is cautiously melted on a hot bar and the slides placed successively in xylol, absolute alcohol and 95 per cent alcohol; in any of these reagents the slides may remain for days without injury (this applies also to the Nissl method), but they must not be placed, even for a short time, in weak alcohol or water, and of course all traces of acid must be avoided. From 95 per cent alcohol the slides are placed (sepai'ated by intervals of at least five minutes) in a 1.5 per cent solution of silver nitrate in distillcMl water; this solution is kept at 55 to 60°C'. A beaker holding KM) to 150 cc. is satisfactory, and this amount of solution will serve for al)out (>ight 2X3 inch slides; if used too long the solution ])econies discolored and after reduction the slides show a diffuse reduction of the silver and poor contrast. In the silver bath the shdes remain 10 to 15 minutes.
CAJAL METHOD FOR SECTIONED TISSUE 793
The most important step in the technique is the reduction. This is accomphshed by a 1.5 per cent solution of pyrogalhc; acid in distilled water to which is added 5 per cent of formalin. This solution should be made up fresh, and should not be kept longer than two hours. The sublimed pyrogallic acid should be used, and I cannot recommend the crystalline form, which remains colorless in solution. Into a dish 8 cm. in diameter enough of the reducing solution is poured to make a depth of about 1 cm.; this solution must be renewed for each slide (2X3 inch). Before placing the slide into the reducing solution three conditions must be observed, namely, the excess of silver solution must be drained off, the hot (55 to 60°C.) sections must not be allowed to dry, and finally, the silver solution must be uniformly distributed over the slide (or at least over the sections and in their immediate neighborhood). To avoid drying one may pour some cool silver solution on the slide and allow slide to cool before draining off excess. A uniform distribution of the silver solution may be obtained by tilting the slide; with the aid of a bit of filter paper isolated drops may be removed or distributed. Remove excess of silver also from under surface of slide. As soon as this is accomplished the slide is quickly placed, in a horizontal position and with the sections on the upper surface, in the reducing solution. The slide is left in the reducing solution absolutely undistiu-bed for about Ij minute. The slide must undergo once more the silvering and reducing processes, and to accomplish this successfully proceed as follows : As soon as slide has remained the proper time in the reducing solution remove and flood it two or three times with distilled water; the object of this procedure is to remove some, but not all, of the pyrogallic acid. The slide is then placed on a hot bar or water bath at about 60°C. and flooded with a 1.5 per cent solution of silver nitrate (not previously used) and allowed to remain one to two minutes; during this process the sections usualh' l)ecome darker. If under the microscope the reduction appears sufficient the slide should be very quickly washed with distilled water (two seconds) before reducing (but in the great majority of cases the slide should not be washed before the second reduction) ; if the first reduction appears unusually successful the second silver bath should be employed half strength, followed by reduction without previous washing. After second silver bath flood slide with cold silver nitrate solution to avoid any local concentration, due to evaporation, drain off excess, distribute evenly and reduce 1 j minute as before; usually the same solution may be used for both reductions, but should then be thrown away. The sections are washed for a few minutes or longer in several changes of distilled water (or tap water) , placed in 95 per cent alcohol, absolute, xylol, and mounted under cover in balsam. Sections may remain for hours in water, alcohol or xylol
The principal defects to be guarded against are insufficient deposit of silver, unequal deposit of silver over slide (due to unequal distribution of silver solution before reduction), and most troublesome of all, diffuse deposit of silver (often in a finely granular state) which seriousty injures the contrast. This last difficulty is due either to the deterioration of
794 EDWARD F. MALO^E
the first silver bath (replace with fresh), to the deterioration of the reducing solution (not good for more than two hours) , or to the presence of too much reducing solution in the sections when they undergo the second silvering (wash out somewhat more of the reducing solution before placing in silver bath). A third silvering and reduction is usually not desirable. Sections not too heavily stained may be conterstained with toluidin-blue as follows: Stain in a 1 percent aqueous solution of toluidin-blue (Gruliler) for two hours or longer (or for a few minutes if heat be employed), wash cjuickly in several jars of 95 per cent alcohol, absolute, xydol, mount; the stain washes out in alcohol more readily than in the case of a primary toluidin-blue stain.
To restain toluidin-blue sections by the Cajal method: Dissolve off cover, and thoroughly remove balsam in xylol followed by absolute. The sections are then washed in 95 per cent alcohol until as much of the toluidin-blue as possible has been i*emoved ; it is well to leave the sections in 95 per cent alcohol overnight. When the slides are placed in the silver solution the rest of the toluidin-blue comes out rapidly upon agitating the slide, but an excess of toluidin-l)lue causes the formation of a coarse precipitate. Accordingly, after washing the slides one or two minutes in hot silver nitrate solution they should be placed in a clean silver bath. Thereafter, the technique is the same as for unstained sections, except that the silver bath deteriorates more quickly.
I recommend this modification of the Cajal method for the human central nervous system, where I have employed it with success both on unstained material and on material previously stained with toluidinblue. It has been used with success also on the brain of the lemur. In case of the rat's brain it failed, but just as poor results were obtained by silvering and reducing en hloc. The advantages of the method are: the possibility of obtaining a series of a large piece of tissue with uniform stain from center to periphery of section, the possibility of staining any section of the series by either the Cajal or Nissl method, and the conversion of a Nissl into a Cajal preparation. It is especially to be recommended when the tissue is veiy valuable or when the pieces of tissue are so large as to involve much labor in their preparation, since the partial failure in the case of one slide does not prevent the successful preparation of the rest. Finally, this method has an important advantage over other methods which demonstrate neurofibrils in previously sectioned material, since with the use of only a few simple reagents a paraffin section may be completed and in balsam within twenty to thirty minutes.
Below is a summary of the method, but I wish to emphasize the necessity^ of strict adherence to the details prcn'iously described.
1. Thorough fixation in cold in 95 per cent alcohol. After several days cut into pieces 1 cm. thick and leave in alcohol several days longer.
2. Thoroughly dehydrate in absolute (in cold).
3. Chlorofoiin about two days.
4. Thorough impregnation with paraffin.
CAJAL METHOD FOR SECTIONED TISSUE 795
5. Sections mounted by means of water and pure egg albumen; water expressed as described.
6. Xylol, absolute, 95 per cent alcohol.
7. Silver nitrate (l.o per cent solution in distilled water) at 55 to 60° C. for 10 to 15 minutes.
8. Drain off excess; careful distribution of solution over slide; avoid drj-ing.
9. Reducing solution (fresh j about 1 j minute; slide must lie horizontally and remain undisturbed.
10. Flood two or three times with distilled water.
11. Place on hot bar at 60° C. and cover with silver solution; leave one or two minutes.
12. Drain off excess and distribute solution uniformly.
13. Reduce for second time 1| minute.
14. Wash thoroughly.
15. Ninety-five per cent alcohol, absolute xylol, balsam, cover.
16. To restain toluidin-blue preparations, dissolve out stain in 95 per cent alcohol and proceed as above, but note that first silver bath soon becomes contaminated.
In conclusion, I wish to remove any impression of being too enthusiastic over the results of this method; man}' slides will be unsatisfactory. But wdth practice nearly all slides may be rendered satisfactory for tracing cell processes, and an ever increasing number (at least for human material) show really beautiful pictures.
