Talk:Paper - The development of the cerebral ventricles in the pig (1913)

From Embryology



Harvard Medical School, Boston, Massachusetts


In this study of the cerebral ventricles of pig embryos three methods have been employed — wax reconstruction, dissection, and the making of minute casts. Wax reconstructions of the ventricles have the advantage of large size. The volume of the various cavities of the brain can best be estimated by immersing portions of such reconstructions in water, and observing the displacement. From these observations the actual size of the ventricles can be readily calculated. But dissections and casts are more accurate for showing details, as, for example, the neuromeral grooves; and most of the drawings have been made from such preparations.

Brains of pig embryos measuring from 12 to 45 mm. were dissected under the binocular microscope, with finely ground instruments. Some of the dissections were designed to give a median sagittal view of the right half of the entire brain. The embryo, fixed preferably in Zenker's fluid, was held in a mass of cotton wet with alcohol, between the thumb and index finger of the left hand; and under a small stream of alcohol, a longitudinal cut was made with a sharp safety razor blade. Very perfect cuts can thus be prepared, such as will pass between the hemispheres and thence backwards, sectioning the infundibulum, corpus mamillare and other divisions of the brain almost exactly in the median plane. Mesenchyma which covers the hemisphere, olfactory lobe, or other portions which it is desired to expose, can be easily removed with delicate needles and forceps.



Casts of the ventricles, in embryos measuring from 12 to 22 mm., were made after the following method. Longitudinal sections of the entire embryo, made slightly to the right side of the median plane, were prepared as above described. Then, under the binocular microscope, obstructing portions of the brain walls were dissected off, and the ventricles were cleaned of all precipitation with a small brush and syringe. Next the embryo was taken from the alcohol and pinned down to a piece of cork and with narrow pointed strips of filter paper all the alcohol was removed from the cavities of the brain. In a large measure the success of the cast depends upon the completeness of the removal of the alcohol. The cast was made by filling the cavities with black wax with the aid of a fine electrically heated needle. A small piece of wax was carried with a cold needle to the proper place and then melted with the electric needle. This process was repeated until the ventricles had been filled. Then if desired, white wax was melted over the upper surface of the dissection for holding the parts more closely together. Moreover, the white background thus provided contrasts sharply with the black cast. Lastly the preparation was turned upside down and the brain wall was carefully dissected away from the wax. For making a permanent preparation of the cast, it is best to mount it in a dry cell on a microscopic slide. Casts thus made show a little more than one-half the ventricles. They can be studied with the binocular microscope, and drawings can be made quite as satisfactorily as from larger models. After embryos exceed 20 mm. in length the interventricular foramen is so small that it is difficult to get the lateral ventricles completely filled, and after 24 mm. the lateral recesses can hardly be cast.

For the electric needle a piece of no. 18 copper wire is coated, except for about 15 mm. at one end, with a thin layer of some insulating paste which will harden and be resistant to heat. The commercial 'caementium' is very satisfactory. The bare end is ground down to a fine point which is the working part of the instrument. Behind this, covering a length of about 20 mm., a piece of no. 32 German silver wire, about 20 cm. long, is wrapped around the insulated copper wire, the coils being well insulated from each other and covered with caementium. Attached to the ends of the German silver wire are small copper wires which run


to a lamp board. The lamp board, with the sockets connected in parallel, serves as a rheostat, and any desired temperature can be obtained by turning on the proper lights. One 50 e.p. globe allows about the desired amount of current to flow. A larger electrically heated smoother for wax models and an electrically heated knife for cutting wax plates, which were used in making the models, will be described at a future date.

Those who study the nervous system from the standpoint of comparative physiology regard the "functional system of neurones" as the real unit of the nervous system. Thus Herrick ('08), tin discussing the morphological subdivisions of the brain, refers to the influence of metamerism as primitive, dominating the subdivision of the nervous system of lower vertebrates. He further states that transverse divisions, such as the diencephalon and mesencephalon, are not ' natural regions' because the primary metamerism has ceased to be an important factor. But the anatomical subdivisions exist no less than the physiological, and although their significance may not be profound, they form a convenient basis for description. In the following account the primary divisions of the brain will be considered in this order — fore-brain, mid-brain, hind-brain. Reference will be made especially to the casts and models of the ventricles, but the wall of the brain must also be studied as the changing structure which modifies the cavities within.


A model of the cerebral ventricles of a 5.1-mm. embryo, the youngest one studied, shows that the three divisions of the brain are distinctly marked off from each other, and that the fore-brain is aheady subdivided into telencephalon and diencephalon. Johnston ('09) states that the boundary between the telencephalon and the diencephalon is determined in mammals, as well as in lower vertebrates, by the velum transversum above and the primitive optic groove or postoptic recess below. This subdivision is distinct in the young pig studied as seen in figure 4. In the model very slight lateral swellings from the telencephalon indicate the cavities of the rudimentary * cerebral hemispheres. The diencephalon shows a division into two parts or neuromeres, the first


or parencephalon, and the second or synencephalon of Kupffer ('03). The relatively large cavities of the optic vesicles extend out from the antero-ventral part of the diencephalon. Almost one-fourth of the length of each vesicle, representing the optic stalk, extends nearly straight out laterally, and the remainder bends sharply toward the mid-brain.

In an embryo of 12 mm. the lateral ventricles have expanded considerably and they now slope outward forward and slightly upward from the interventricular foramen. The foramen is almost circular in outline and has an area of about 0.428 sq. mm. In the cast (fig. 5) the rounded mass of wax filling the lateral ventricle shows a slight concavity below, which has been produced by the developing corpus striatum. This body arises as a thickening of the ventro-lateral wall of the hemisphere, as seen in the dissection, figure 6. This figure, it may be noted, bears a very striking resemblance to the reconstruction of the brain of Anguis fragilis (40-50 somites), published by Kupffer ('03, p. 220, fig. 224). In both, the parencephalon and synencephalon are sharply marked off from each other. Two swellings, which correspond with these subdivisions, are seen in the cast (fig. 5). The parencephalon is larger than the synencephalon, and its cavity produces a more extensive swelling, which however, is quite low.

The mamillary recess and the infundibulum below it, neither of which had appeared in the 5.1-mm. embryo, are now very distinct. The velum transversum, a portion of which is shown in figure 7, is well developed, producing a deep inward bend of the anterior wall of the fore-brain above the hemisphere.

The brain of a 17-mm. pig embryo, as seen in median section (fig. 8), shows a prominent corpus striatum, above which is the interventricular foramen. The foramen is now crescentric, since it has been invaded from below by the corpus striatum, but notwithstanding the growth of the entire embryo, its area has become actually reduced. It measures about 0.315 sq. mm. As seen in the dissection, the lower portion of the corpus striatum is bounded behind by a deep groove which is continuous with the optic recess. This groove appears as a ridge upon the cast (fig. 9), and in front of it the position of the corpus striatum is indicated by an excava


tion. Abo^-e this hollow is seen the inferior horn of the lateral \'entricle. The outline of the ventricle is no longer round, as in the 12-mm. embryo. It is prolonged anteriorly or downward to form the first indication of the olfactory lobe, and posteriorly or above, it is somewhat flattened. This upper portion terminates in the inferior or descending horn. Fraser ('94) has referred to this horn in an abnormal human adult brain as in ' ' reality ascending" and Thompson ('08) similarly states that "from the developmental point of view, this descending horn is in reality an ascending horn." In pig embryos, however, as seen in figures 5 and 9, its primary direction is apparently ventral to the cerebral axis so that the inferior horn may be said to descend, even in early stages. The cavity of the lateral ventricle is seen laid open in figure 10. It differs from that of the 12-mm. pig embryo (fig. 6) since it has been invaded by a prominent chorioid fold. This fold is a lateral extension of the velum transversum, and it consists of vascular mesenchyma covered by the thin wall of the brain. At this stage its ventricular surface is perfectly smooth, but in a 22mm. embryo the chorioid fold has developed a vascular fringe which appears as a plaited frill. This is shown in figure 13.

The walls of the diencephalon in the 17-mm. embryo have thickened considerably so that the external depression between the parencephalon and the synencephalon has become almost imperceptible. Their cavities however, can be distinctly seen when the dissection of the brain is studied (figs. 8 and 9). The mamillary recess now projects out some distance from the third ventricle. Midway between this recess and the infundibulum the floor of the brain has thickened to form the tuber cinereum. The optic thalami, represented by the thickenings of the lateral walls of the diencephalon, have developed to such an extent that they have already partially fused with the corpus striatum. The place of fusion appears as an interruption of the groove between the interventricular foramen and the optic recess.

The median section of a brain of a 22-mm. embryo (fig. 11), as compared with that of the 17-mm. specimen, shows several new features. Along the cut edge dorsally, the pineal body has appeared as a slight elevation. Within the third ventricle,


the thalamus has enlarged, and the corpus striatum appears as if pushed forward before it. Between the corpus striatum and the upper part of the thalamus, the interventricular foramen is seen as a slit, which arches up over the corpus striatum, being widest immediately under the velum transversum. The area of the aperture has become further reduced. Between the corpus striatum and hypothalamus, the optic recess appears as in the 17-mm. specimen. Somewhat further back, and parallel with it, is the recessus infundibuli, and between the two grooves made by these recesses, is the pars optica hypothalami. This is convex toward the ventricle. Between the hypothalamus below and the thalamus above is the sulcus hypothalamicus (B.N.A.) or sulcus limitans (His '93) — a groove which continues from the mid-brain downward and forward becoming somewhat indistinct above the pars optica hypothalami. Reichert ('61) described this groove in the brain of a pig embryo (also in cat and human embryos) as extending from the foramen of Monro to the entrance of the aqueduct of Sylvius, and he named it the sulcus of Monro. In fig. 12 an extension may be traced from the sulcus to the lower end of the interventricular foramen; another extension proceeds toward the recessus postopticus. But the main continuation is probably that which ends in the optic recess. This accords with the opinion of His ('93, p. 177), and also of Johnston ('09, p. 517). The sulcus limitans is indicated in the embryo of 17 mm. and in younger specimens, but it is more distinct at 22 mm.

The median section of the 22-mm. embryo shows the uncut median surface of the right hemisphere, ending below in a rounded olfactory lobe. Extending from the lower border of this lobe toward the notch below the velum transversum, is a groove, which, with the lamina terminalis, bounds a triangular portion of the hemisphere. Smith ('03) named this triangular thickening the ' paraterminal body'. Herrick ('10) proposes to divide this body into a ventral component, the 'corpus precommissural,' and a dorsal component, the 'primordium hippocampi,' but these subdivisions do not appear externally in the 22-mm. embryo.

In the dissection of a 45-mm. embryo (fig. 14) most of the paraterminal body together with the adjacent medial wall of the right


hemisphere have been cut off, thus displaying the structures within the ventricle. The corpus striatum is seen to be divided by a vertical groove into two portions. The groove is deep below but fades out above. Thompson ('09) has described a similar groove in a cat embryo 20 mm. in length, and he refers to the larger lateral and the smaller median subdivisions which are set apart by the groove as 'roots' of the corpus striatum. His ('89) described the corpus striatum as being composed of three limbs. In the 45-mm. pig there are also three limbs, but only two are seen in figure 14. The small median limb, which is not shown, fuses with the lamina terminalis just below the interventricular foramen. The groove is shown at x in figure 18, which is a frontal section through the head of a 45-mm. pig. The plane of this section is indicated in figure 14. The groove is best seen in the right side of the section, where the median subdivision may be observed to pass down toward the paraterminal body. In a section further back (fig. 19) the groove is quite shallow. Above the corpus striatum (figs. 14 and 19) the chorioid plexus is seen projecting into the lateral ventricle. The fold, surmounted by a frill, which was seen in the 22-mm. embryo, has given place to a very thin layer with many reduplications and subdivisions. Villi are nowhere present, but in thin sections the smaller processes extending out from the main folds may simulate them. Under the binocular microscope it is clearly seen that this plexus consists only of folds with secondary subdivisions. The plexus is attached along a fissure measuring 2.1 mm. in length, which extends backward from the interventricular foramen.

The caudal portion of the plexus — about one-fourth of the entire length — is very much attenuated. It forms a short free projection extending 0.18 mm. beyond the end of the fissure, and the free portion shows no secondary folds (fig. 21). The anterior portion of the plexus forms a larger free projection which extends 0.72 mm. past the front end of the fissure.

In the 22. mm. embryo (fig. 13) a slight invagination of the medial wall of the hemisphere, above the chorioid plexus and parallel with it, extends from the interventricular foramen backward into the inferior horn. This represents the hippocampus,


which is better developed in the 45-mm. embryo. It is seen in figures 19, 20, and 21 (marked +).

Considered as a whole, the lateral ventricle has expanded greatly and may readily be divided into three parts — the anterior horn, which is in front of the corpus striatum; the body, which is above it; and the inferior horn, descending behind it. A cast of the left lateral ventricle is shown in side view in figure 16, and in ventral view in figure 17. The latter shows the large excavation made by the corpus striatum. Toward the olfactory lobe this concavity is bisected by a ridge, which lies between the large roots of the corpus striatum, as described in connection with figure 14. Below the concavity, in figure 17, there is a ridge which is fissured, posterior to the interventricular foramen, to receive the chorioid plexus. This ridge separates the hollow for the corpus striatum above, from that for the hippocampus below.

The lateral ventricle communicates with the third ventricle by an interventricular foramen which is smaller than in the preceding stages. The third ventricle has become reduced to a slitlike space, owing to the great thickening of the walls of the diencephalon. A portion of it has become practically obliterated. This is where the thalami have grown against each other (figure 21). They have not yet fused, however, to form the massa intermedia; that is the ependymal layer over each thalamus is still uninterrupted. Along the dorsal margin of the cleft, which represents the ventricle, there is a thin lateral expansion on either side (fig. 15). The expansion begins at the interventricular foramen, where it is broadest, and it diminishes backward, ending a short distance in front of the pineal recess. Thus the expansion is wedge-shaped when seen from above. The brain-wall which overlies this expansion is the tela chorioidea, which has a corrugated surface in relation with the vascular mesench^ma. Below and behind the thalamus is a deep groove — the sulcus limitans — ■ which, as already described, sends prolongations to the interventricular foramen, recesses postopticus and recessus opticus. The sulcus is seen in section in figure 20, with the thalamus above and the pars optica hypothalami below.


The oldest embryo studied measured 260 mm. Its brain was dissected out, embedded in celloidin, and cut into sections 0.2 mm. thick. From these sections a wax model of the ventricles was constructed, as shown in figures 25 and 26. The three subdivisions of the lateral ventricle have become highly developed. The anterior horn, which in the 45-mm. embryo ended in a short and slightly pedunculated olfactory bulb, now extends through the olfactory stalk and terminates in the expanded ventricle of the olfactory bulb. The body of the lateral ventricle is corrugated above, w^here it is in relation with bundles of fibers in the corpus callosum. The inferior horn not only descends, but extends forward as a slender prolongation, which ends blindly in the temporal lobe of the brain. As seen from below, the cast of the lateral ventricle shows the concavities for the corpus striatum and the hippocampus respectively, separated by a ridge into which the chorioid plexus is invaginated. The ridges bounding the corpus striatum correspond with those seen in the 45-mm. embryo.

In the 260-mm. specimen, the interventricular foramen has expanded considerably, and its area was found to be 6.92 sq. mm. Extending backward from the foramen, between the region occupied by the thalamus below and the tela chorioidea above, there is a tubular expansion of the third ventricle, somewhat flattened dorso-ventrally. This corresponds with the much smaller expansion, T-shaped in section, which was described in the 45-inm. embryo. In the 260-mm. embryo there is a long caudal extension of the dorsal part of the third ventricle which curves slightly upward and forms the suprapineal recess. It measures about 5.6 mm. in length and extends backward with quite a uniform diameter of 1.4 mm. This conspicuous feature is not represented in the 45-mm. specimen. The cast of the cavity (fig. 25) is deeply corrugated on all sides by longitudinal folds. These accomodate branches of the medial cerebral vein. A large branch of the vein on either side produces a deep groove along the dorsal expansion of the third ventricle. Between the suprapineal recess above and the pineal recess below there is a well marked lateral compression of the ventricle, produced by the hab


enular ganglion. The conspicuous ridge at the lower margin, of this habenular concavity continues caudally into the pineal recess. Between the pineal recess and the groove for the posterior commissure there are two slight projections of the third ventricle (infrapineal recesses). The very extensive fusion of the thalami, forming the massa intermedia, has produced a hole in the model of the third ventricle. It is oval in outline, measuring about 4.6 mm. by 3 mm.


It is well known that in young embryos the cavity of the midbrain is very large, forming the middle cerebral vesicle. In the adult it is generally described as '^ reduced to a narrow slit— the aqueduct of Sylvius." Cunningham states that the mid-brain "is tunnelled by a narrow passage, the aqueduct of Sylvius, which extends between the fourth and third ventricles, "and Bensley ('10) writes: The mesencephalon is noteworthy in a mammal as lacking a ventricle. Its cavity is a narrow canal, the aqueduct of Sylvius . . . . " But Retzius ('00) described a spindle-shaped expansion in the middle portion of the cavity of the adult human mid-brain, and named it 'ventriculus mesencephali.' It will be shown that a distinct cavity exists in the mid-brain of the adult pig, comparable with that found by Retzius in the human brain.

In the 5.1-mm. pig (fig. 4) a slight constriction separates the caudal part of the diencephalon from the cavity of the mid-brain. Another slight constriction marks the position of the isthmus between the mid-brain and the hind-brain. The angle formed by a line passing through the axis of the diencephalon with another through the isthmus and cavity of the medulla, is very acute (30°) in this stage. In the adult it is very wide.

The cavity of the mid-brain continues to expand and becomes sharply marked off from that of the hind-brain at the isthmus, as shown in the 12-mm. embryo (fig. 5). Between the mid-brain and fore-brain, however, the cavity shows only a slight constriction, which is limited to the lateral surface. As seen in the cast


neither the mid-dorsal nor the mid-ventral line is indented at this point.

Figure 9, of a 17-mm. embryo, shows that the dorsal surface of the mid-brain has become flatter than before and its caudal end projects further toward the hind-brain. This projection, on either side, marks the position of the future inferior colliculus. The ventricle of the mid-brain is somewhat quadrilateral in cross section, and in the cast it shows a prominent longitudinal ridge corresponding with the line of maximum width. The groove in the wall of the brain corresponding with the ridge may be referred to as the lateral sulcus (sulcus lateralis internus). It must not be confounded with the sulcus limitans which, in the cast, is a much less conspicuous ridge, more ventrally placed. The sulcus limitans is more evident in the dissection, figure 8. It separates a small thick-walled ventral zone from an extensive thin-walled dorsal zone. The ventral zones, on either side of the mid-ventral sulcus, form low rounded eminences projecting into the cavity of the ventricle which constitutes the tegmental fold of His. The lateral sulcus passes longitudinally through the dorsal zones, terminating on either side, in the projection which forms the inferior colliculus.

In a slightly older embryo (22 mm.) the dorsal wall of the mid-brain has become relatively thinner, and its ventricle has expanded greatly. At this stage it forms a very conspicuous portion of the whole ventricular system. Its volume is estimated at 4.1 cu. mm. Separate recesses for the inferior colliculi can now be recognized. In earlier stages the posterior overhanging portion of the mid-brain ended abruptly in a straight transverse line. In the 45-mm. embryo this border shows a deep median cleft, and each arm of the bifurcation thus formed represents the cavity of an inferior colliculus. In the 22-mm. specimen the cavities projecting backward from the dorso-lateral corners of the midbrain ventricle have just begun to develop.

In the 45-mm. embryo the third ventricle is connected with the cavity of the mid-brain by a more elongated constricted portion than in previous stages. At the anterior end of the midbrain ventricle there is a median dorsal extension 1.2 mm. caudad


to the pineal recess. This outgrowth (figs. 14 and 15) which Ues just behind the posterior commissure is the incisura postcommissurahs. It is not present in the 20-mm. pig but is very distinct in the 45-mm. embryo. The ependymal and mantle layers but not the ectoglial layer of the brain-wall make a distinct outward bend above it. Caudad to the incisure, the dorsal surface of the mid-brain cavity slopes gradually upward to a height of 0.6 mm. and then extends nearly straight backward over the body of the cavity. The inferior collicular recesses are very much more distinct than in younger pigs.

The length of the mid-brain, including its collicular recesses is 5.0 mm. and- the width, in the widest part is 2.6 mm. The volume of the cavity is very much greater than in preceding stages and is now about 6.9 cu. mm.

A cast of the cavities of the mid-brain in an embryo measuring 110 mm. is of an irregular prismatic form as shown in figure 24. It presents a median dorsal ridge which extends from the incisura postcommissuralis posteriorly, ending in a median depression between the eminences representing respectively^ the cavities of the right and left inferior colliculi. The outer margin of the cavity of each inferior colliculus is continued forward as a prominent ridge, corresponding with the sulcus lateralis. This ridge flattens out rather sharply and completely a short distance behind the posterior commissure. At the base of the cavity of the inferior colliculus, a prominent ridge ascends from the isthmus, but it can be followed only a short distance behind the posterior commissure into the territory of the mid-brain, where it terminates. This ridge represents the sulcus limitans of the isthmus. Ventrally there is another ridge which may be regarded as the sulcus limitans of the mid-brain, although it is not continuous with the structure just described as the sulcus limitans of the isthmus. It arises beneath the latter in or near the median ventral sulcus. It then passes laterallj', diverging from its fellow on the opposite side. The point of widest separation is soon reached, after which the two sulci gradually converge, coming together in the mid-ventral line beneath the incisura postcommissuralis. The surfaces between the ridges on the cast are


all concave, and in the central part of the mid-brain there are five surfaces — two dorso-lateral, bounded by the median dorsal ridge and the lateral ridges; two ventro-lateral, extending downward to the sulci liniitantes; and a ventral surface between these sulci.

The form of the mid-brain ventricle in the embryo measuring 110 mm. has been described at length, since subsequently it undergoes only slight modifications. This was determined by modelling the ventricles in the 260-mm. pig (figs. 25 and 26) and by making dissections of the adult. The ventricle of the mid-brain continues to increase in size but it does not keep pace with the growth of the adjoining cavities. The grooves remain as described, and the ventral surfaces, between the sulci liniitantes, are always subdi^dded posteriorly by a forward extension of the median ventral sulcus.


In the 5.1-mm. embryo the cavity of the hind-brain or third cerebral vesicle is elongated and quite straight. Behind the ventricle of the mid-brain (fig. 4) a well marked constriction indicates the future isthmus. Caudad to this, the cavity, or fourth ventricle, widens somewhat, and then slopes gradually through the medulla to the spinal cord. The low ridge which follows the line of greatest width corresponds at this stage with the line of attachment of the thin roof-plate, and not with the sulcus limitans. Along the ventral surface of the cast there is a sharp median ridge which represents the sulcus medianus of the rhomboid fossa.

Seven neuromeral grooves can be identified. The first produces a low but broad ridge on the cast anterior to the widest part of the rhombencephalon. Dorsally it flattens out before reaching the line of maximum width and ventrally it does not extend quite to the median sulcus. The second neuromeral groove is situated opposite the widest portion of the fourth ventricle. The next four are about equalh' spaced. They are very prominent in the ^-entral zone and some of them appear to reach upward a short distance into the dorsal zone. The last one is



represented on the cast by a prominent mound, but the elevation does not extend as far dorsally or ventrally as do those immediately preceding it. Caudally it blends with the sulcus limitans. Bradley ('04 and '05) described seven neuromeres in a nineteen-day embryo. ^

The pontine flexure soon becomes definitely established as shown in figure 5, which is a cast from a 12-mm. embryo. In this embryo the median ventral edge of the cast is very sharp around the flexure and backward toward the spinal cord. At the pontine flexure there are six neuromeral grooves. The relation of the grooves to the median sulcus is clearly shown in the dissection, figure 7, The former seventh groove has been taken up by the sulcus limitans so that it is no longer recognizable. Above the grooves the A^entricle has become quite wide. In dorsal view the body of the ventricle slopes out laterally quite rapidly behind the isthmus, attaining its maximum width in the region of the future lateral recesses, which become distinct in embrj^os of about 15 mm.

In the 17-mm. embryo the isthmus leading from the midbrain to the body of the fourth ventricle is diamond-shaped in cross section; the ventral zones are here more extensive than the dorsal, which is the reverse of the condition in the mid-brain. The ventral median sulcus is narrow and deep, so that in the cast there is a sharp edge on the ventral surface of the fourth i^entricle extending around the pontine flexure, and onward caudally to the spinal cord. There are now five neuromeres in the hind-brain. The body of the fourth ventricle has expanded a great deal since the last stage described, and the lateral recesses are now well indicated. Slight concavities on either side of the body of the ventricle between the isthmus and the lateral recesses, in the dorsal zones, are caused by the thickening of the brain in this region to form the lateral portions of the cerebellum. Behind the lateral recesses the thin roof of the hind-brain has become deeply invaginated to form the chorioid plexus. About 100 villi are developed on the fold.

1 According to Keibel's Normentafel, pig embryos of nineteen days measure from 4.5 to 8 mm.


In a 22-nim. embryo (figs. 11 and 12) the fundamental arrangement of parts has not been altered. The cerebellar thickenings have enlarged considerably, producing correspondingly greater depressions in the dorsal zones along the body of the fourth ventricle. The sulcus limitans, for the same reason, has been made verj^ much more evident. The lateral recesses extend further to each side and are comparatively much narrower than in the younger stage considered. The chorioidal lamina extends out to the ends of the lateral recesses, and the number of villi springing from it has increased to about 150. Caudad to the chorioid plexus the dorsal part of the cavity is slightly distended or puffed up dorsally, thus forming the 'caudal protrusion' (Blake '00). In advanced embryos of dogs, cats, pigs, sheep and also the chick, according to Blake, there is a marked caudal protrusion. He wrote, This protrusion is completely closed and resembles the finger of a glove." It can hardly be described in the pig as 'finger-like' since, as shown in figure 11, it forms a rounded dome.

Taken as a whole, the fourth ventricle in the 45-mm. pig (figs. 14 and 15) has the same parts as seen in the 22-mm. embryo. The cavity of the isthmus has increased in dorso-ventral diameter from 0.75 to 0.85 mm. As seen by comparing figures 11 and 14, it has become relatively low, but laterally it has become further expanded. Its thin lateral edge, as seen in the cast (fig. 15), extends backward and becomes continuous with the body of the ventricle a short distance in front of the lateral recess. Where this thin edge fuses with the body there are three small wrinkles — the lateral remnants of the neuromeral grooves. The two halves of the cerebellum have become very much thicker and meet at an angle of about 130°. The medial part of the cerebellum is also thicker than before, so that all together the broad median fissure of younger stages has been appreciably reduced. The lateral recess, while absolutely larger than before, appears much slimmer. The floor and roof of the recess have almost come together, the cavity being largely filled with the chorioid plexus. In figure 23, in the left recess, is seen a long stretch of the chorioid fold. Sections a short distance caudad


to this one show the fold to be continuous from one recess to the other. The marked villous character of the chorioid plexus is illustrated well in this section (fig. 23). It shows also the rhomboid fold projecting into the cavity of the hind-brain below the chorioidal lamina. The rhomboid fold appears in the model (fig. 15) as a sharp rim, extending along the upper part of the body of the ventricle behind the lateral recess. It extends to the caudal protrusion of the fourth ventricle, which has expanded to a very great size; its volume is more than one-half that of the remainder of the ventricle. This protrusion, from above and from behind, appears almost spherical. The posterior boundary slopes downward making almost a right angle with the narrow cavity of the caudal part of the medulla. When the body of the fourth ventricle is viewed from below, the ventral median sulcus and the two sulci limitantes are seen converging towards the cervical flexure. The ventral median groove is nearly straight and produces with the slim cavity of the cord an angle of about 125°.

In the 260-mm. embryo (figs. 25 and 26) the cavity of the isthmus is very broad and flat. On the ventral surface a deep ridge — the median ventral sulcus — is continuous with the similar ridge of the mid-brain ventricle. On account of the extensive growth of the cerebellum the dorsal surface of the body of the ventricle has become deeply concave. Only a very slight caudal protrusion is now present. Bradley ('05) described the first appearance of openings in the lateral recesses — foramina of Luschka — ^in embryos of about 100 mm. Because of these openings and on account of the growth of the chorioid plexus the cavities of the lateral recesses, as seen in the model, have been reduced to slim irregular bodies.


In order to show the change in the volume of the cavities during development, measurements of wax models were made by the immersion method. The results of these measurements, reduced to cubic milhmeters, are given in table 1.




Volume of the cerebral ventricles in cubic millimeters








mm. 5.1








































From table 1 it is seen that each part of the entire ventricular system increases in size continually during development. The increase is relatively greater at first, and as would be expected, it becomes gradually less toward the adult. For a time — until embryos measure between 12 and 20 mm. — the fourth ventricle is the largest part of the ventricular sj^stem. After the lateral ventricles have expanded appreciably the cavity of the fore-brain is the largest part.

The cross-sectional area of the interventricular foramen was determined from wax models by measuring the area of the foramen cut through the narrowest portion.

The cross-sectional area of the interventricular foramen becomes progressively greater up to about the 12-mm. stage after which it becomes both relatively and absolutely smaller for some time. In later stages it again expands considerably; in the adult it is very much larger than in younger stages.

TABLE 2 Area of the foramen interventriculare



Length in mm.

Area in square millimeters









The mid-brain angle in early stages is very acute. For measuring the angle in the different embryos during development it



is difficult to establish lines whose loci will be absolutely comparable throughout the series. The apex of the angle has been located in the middle point of the mid-brain ventricle; one line has been drawn from it to the optic recess, and the other to the middle point in the cavity of the central canal opposite the cervical bend

Fig. 1 Outline from figure 5, showing the position of the mid-brain angle

TABLE 3 Increase in the mid-hrain angle























The angle increases from 30° in the 5.1-mm. embryo to 152° in the adult. The curve of growth of this angle rises up relatively rapidly at first and slowly in later stages to the adult.


The well-known series of models of the developing human brain made by His and the papers which they illustrate make it possible to compare the conditions which have been described in the pig with those in man. Comparisons with the embryonic brain of other mammals cannot be made, since it appears that no one has attempted to extend the work of His here referred to, or even to repeat it critically. The first impression derived


from comparing His's models of the interior of the human embryonic brain with those of the pig is that they are strikingly similar, especially in the earlier stages; but more careful study shows very considerable differences. The 6.9-mm. human embryo (His, '89, Taf. 1, Fig. 2) is somewhat more advanced than the pig embryo of 5.1 mm., as indicated by the greater development of the hemisphere, and of the pontine flexure. In His's model the cavities of the parencephalon and synencephalon are not indicated and the neuromeres of the hind-brain are not shown. It may be questioned however, 'whether these structures are actually absent from the human brain at this stage. On the other hand, the model of the human brain shows a prominent tegmental fold passing from the mid-brain toward the fore-brain, but no such fold occurred in the pig embryo. In a model of a pig embryo of 5 mm. Johnston ('09) has shown the neuromeres of the fore-brain and hind-brain but there is no tegmental fold. This fold is the portion of the tegmentum which projects dorsally into the cavity of the mid-brain; the fold on either side extends from the median ventral sulcus to the sulcus limitans. Altogether Johnston's model of the brain of the 5-mm. pig agrees very closely with the first model described in the present paper, and it is evident that we have more detailed and accurate knowledge of the shape of the brain in the pig embryo than in the human embryo of corresponding stage.

His's model of the human embryo of 10.2 mm. corresponds quite closely with the pig embryo of 12 mm. and his 13.6-mm. specimen is comparable with the pig of 17 mm.; the human embryo of the third month, which completes His's series of models, is the stage intermediate between those dissected in pigs of 22 and 45 mm. In all of His's models, except the oldest, the tegmental fold is very prominent; and in the embryo of three months it is perfectly distinct. The sulcus limitans is accordingly well defined in the mid-brain, but it does not form a distinct continuous longitudinal groove from the spinal cord to the optic recess in any of His's models. In fact. His described the sulcus as becoming leveled off in the territory of the isthmus



by the growth of the walls — being interrupted at the junction of the isthmus with the mid-brain.

In the casts of the ventricles of the pig, in early stages, there is a broad line of maximum width corresponding with the sulcus limitans.


Fig. 2 Sections through the hind-brain of pig embryos, {A, 3.9 mm. ; B, 10 mm. ; C, 23.6 mm.), showing the change in position of the sulcus limitans.

This groove which is lateral in the cord becomes a ventral groove in the medulla as described by His, and at the junction of the isthmus with the mid-brain it comes to an end. On account of the expansion of the roof-plate in the hind-brain the sulcus limitans soon falls below the line of maximum width. Thus in the diagram (fig. 2, B) the sulci are ventrally placed, and the width of the fourth ventricle between them is less than it is slightly further dorsally. In an older embryo (fig. 2, C) the sulci are close together and on the floor of the ventricle. In the mid-brain there is early a line of greatest width, which becomes shifted so as to form the outer border of the tegmental fold, as shown in the diagram, figure 3.

■ S. I.

Fig. 3 Sections through the mid-brain of pig embryos, (A, 3.9 mm.; B, 23.6 mm.; C, 110 mm.); showing the change in position of the sulcus limitans.


This shifting in the mid-brain is brought about by the expansion in the dorsal zone to form the cavities in the colHcuh. The line of maximum width begins behind the constriction between the mid-brain and fore-brain in the superior collicular cavity and terminates at the caudal end of the mid-brain in the inferior collicular recess. Anteriorly the hypothalamic sulcus is seen to connect with the tegmental sulcus. This connection is most distinct in the 22. mm. embryo.

His first called attention to the dorsal and ventral zones of the brain and cord. A demarcation of the sensory from the motor portion of the central nervous system is obviously of fundamental importance and Johnston ('09) has described the sulcus limitans as the most important landmark in the brain. Diagrams have appeared which show an uninterrupted line, extending from the optic recess backward through the brain and spinal cord. However, there is no real picture which shows a sulcus continuous throughout the brain. It is distinct in each division of the brain, but at the cephalic end of the isthmus it is interrupted.

The hemisphere of course is very much larger in the human embryos than in approximately corresponding stages of the pig. The thalamus and habenular region are also further advanced in the human stages. The tegmental folds never form conspicuous projections into the ventricle of the mid-brain as they do in early human embryos. The olfactory lobe becomes progressively more conspicuous in the pig series — ^the reverse of the condition in human embiyos, of later stages especially. In the latter, the olfactory lobe is covered over by the rapidly expanding pallium. The olfactory lobe in the pig always extends some distance in front of the end of the hemisphere. The roof of the fourth ventricle is not present in aii}^ of the His models, hence there is no indication of a caudal protrusion of the fourth ventricle. This protrusion is first well indicated in pig embryos about 22 mm. long. It expands very rapidly until the cerebellum grows downward into this region. There being no foramen of Majendie in the pig, the posterior medullary velum stretches nearly straight across the caudal part of the fourth ventricle so that the caudal protrusion is just recognizable in the adult.


The form of the ventricles in the adult pig has apparently never been studied bj^ means of casts or models. Figures 25 and 26, representing the cerebral ventricles in an embryo of 260 mm., are however very similar to those which would be obtained from an adult and they ma}^ be compared with Dexler's figure of the cavity of the brain in the horse, and with any one of several figures of the ventricles in the human brain. The first of these were published by Welcker (78) who filled the ventricles with wax, injecting through the infundibulum, and published figures and a brief description of the model thus obtained. Testut ('97) figured and described a plaster cast of the ventricles. Barratt ('02) constructed a wooden model from measurements obtained from thick sections (12.5 mm.). From the method employed accuracy of detail could scarcely be expected. By far the most delicate and satisfactory' figures are those of Retzius ('00) who made casts by the use of Wood's metal. Harvey ('10) has recently used Wood's metal for the same purpose with similar results.

In the human brain immediately above the pineal body there is a backward extension of the cavity of the third ventricle which forms the supra-pineal recess. This term was introduced by Reichert ('61, Bd. 2, S. 69) who described the structure and showed its relations in median sections of the brain. In his figures the recess measures nearly 5 mm. in length. In Welcker's plaster cast it appears to be somewhat larger and in the cast by Retzius it measures 10 mm. But Testut and Harvey show onl}^ very slight protrusions. The recess may be variable in the human brain. In Dexler's model of the ventricles in the horse it is a very conspicuous object, nearly 25 mm. in length. In the adult pig it measures 5.5 mm. In proportion to the size of the third ventricle it is very much longer than in man but much shorter than in the horse. The significance of this recess has not been determined.

A very conspicuous feature of the casts of the third ventricle is the aperture made by the massa intermedia. In the several models of the human brain this aperture varies considerably in size, but in no case is it as large as in the pig or the horse. In


the pig it appears to be somewhat smaller than in the horse, but in both these animals the opening is very large. The area of fusion in the 260-mm. pig measures about 8.4 sq. mm., and in the adult pig about 61.6 sq. mm.

The lateral ventricles in pig, horse and man differ very greatly. In man there is a posterior horn which is absent in the horse and pig. But the latter both show an extension of the ventricle into the olfactory lobe, whereas in man, the anterior horn ends bluntly. In the horse the pedicle of the cavity of the olfactory stalk is very long and markedly concave dorsally, it ends in an elongated irregular ventricle. The pedicle in the pig is nearly straight and much shorter. It terminates in a flattened expansion which in the anterior end is compressed dorso-ventrally.

In connection with the lateral ventricle it may be noted that its chorioid plexus in the pig is not provided with villi. Meek ('07) in his discussion of the general morphology of the plexus writes: "Villi are scarce in the chicken, duck and pigeon, but more abundant in the hog, while they reach a considerable development in the, horse, ox, and especially among porpoises, crocodiles, and some of the selachians (Pettit '02-'03)." Findlay ('99) writes similarly of this plexus: The surface of the chorioid plexus is beset with a large number of highly vascular villous projections. These are of all sizes, and the largest may branch and subdivide many times before the ultimate villi are formed." The study of the brain of the pig has shown that the chorioid plexus of the lateral ventricle first develops as an extension of the velum transversum into the lateral ventricle. The free border rapidly becomes much longer than the attached part so that it is early thrown into folds. These primary folds increase in number and give rise to secondary folds, but as shown by means of the binocular microscope, villi are not developed in this plexus.

In pig embryos of 35 to 40 mm. and in all older stages, there is a median dorsal recess behind the posterior commissure. This postcommissural recess is shown by Retzius who labels it the 'incisura postcommissuralis.' It is shown also in Harvey's lateral view. None of the other authors have an indication of it in their figures. In the adult pig the recess is very prominent.


Finally, it should be noted that the aqueduct of Sylvius remains as a distinct ventricle of the mid-brain in the adult pig. One of Retzius's figures alone gives a good lateral view of the midbrain of man. This he described as 'ventriculus mesencephali.' The cavity of the mid-brain in the horse stands out also as a distinct ventricle. In the pig the mid-brain has throughout development and in the adult a well defined ventricle which constantly increases in size as long as the brain grows.


Barratt, J. O. W. 1902 The form and form relations of the human cerebral

ventricular cavity. Journ. Anat. and Physiol., vol. 36, pp. 106-125. Bensley, B. a. 1910 Practical anatomy of the rabbit. Philadelphia. Blake, J. A. 1900 The roof and lateral recesses of the fourth ventricle, considered morphologically and embryologically. Jour. Comp. Neur., vol.

10, pp. 79-108. Bradley, O. C. 1904 Neuromeres of the rhombencephalon of the pig. Rev.

Neurol, and Psych., vol. 2, pp. 625-635. Dexler, H. 1904 Beitrage zur Kenntnis des feineren Baues der Zentralnerven systems der Ungulaten. Morphol. Jahrb., Bd., 32, s. 288-389. FiNDLAY, J. W. 1899 The choroid plexus of the lateral ventricle of the brain.

Brain, vol. 22, pp. 161-202. Harvey, R. W. 1910 A cast of the ventricles of the human brain. Anat. Rec,

vol. 4, pp. 369-384. Herrick, C. J. 1908 The morphological subdivisions of the brain. Jour. Comp.

Neur., vol. 18, pp. 393-408.

1910 The morphology of the fore-brain in Amphibia and Reptilia.

Jour. Comp. Neur., vol. 20, pp. 413-547. His, W. 1889 Die Formentwickelung des menschlichen Vorderhirns vom Ende

des ersten bis zum Beginn des dritten Monats. Abh. math.-phys.

Classe Kgl. Sachs. Gesellsch. Wiss. Nr. 8, Bd.'lo, S. 675-736.

1890 Die Entwickelung des menschlichen Rautenhirns vom Ende dea

ersten bis zum Beginn des dritten Monats. I. Verlungertes Mark. Abh.

math.-phys. Classe Kgl. Siichs. Gesellsch. Wiss., Bd. 17, S. 1-74.

1892 Zur allgemeinen Morphologie des Gehirns. Archiv. f. Anat. u. Physiol. Anat. Abt., S. 346-383.

1893 Vorschlage zur Eintheilung des Gehirns. Archiv. f. Anat. u. Physiol. Anat. Abt., S. 172-179.

1904 Die Entwicklung des menschlichen Gehirn wahrend der ersten

Monate. Leipzig. Johnston, J. B. 1909 The morphology of the fore-brain vesicle in vertebrates.

Jour. Comp. Neur., vol. 19, pp. 457-539. Keibel, F. 1897 Normentafel zur Entwicklungsgeschichte des Schweines.



VON KuPFFER, C. 1!)03 Die Moiphogenie clos Xorvensj'stems. In Oskar Hert wig's Handbueh des vorgleichenden ii. experimentellen Eutwicklungs lehre der Wirbelthiere, Bd. 2, Th. 3, S. 1-272. Jena. Le"V\'IS, F. T. 1903 The gross anatomj' of a r2-mm. pig. Amer. Jour. .\nat., vol.

2, pp. 211-225. .Meek, \V. J. 1907 A study of the choroid plexus. Jour. C'omp. Near., vol. 17,

pp. 286-306. MiHALKOVics, V. 1877 Entwicklungsgeschichte des Gchiriis. Leijjzig. MiNOT, C. S. 1892 Human embr3^ology. New York. Norman, C, and Fraser, A. 1894 A ease of porencephaly. Journ. Mental

Science, vol. 40, pp. 649—665. Reichert, C. B. 1859-1861 Der Bau des Menschlichen Gehirns. Leipzig. Retzius, G. 1900 Die Gestalt der Hirnventrikel des Menschen, nach Metallaus gussen dargestellt. Biol. Untersuch. von G. Retzius. Bd. 9. N:o.

3 u. 4. Jena. SissoN, S. 1910 A text-book of veterinary anatomy. Philadelphia. S.MiTH, G. E. 1903 On the morphologj^ of the cerebral commissures in the verte brata, with special reference to an aberrant commissure found in the

fore-brain of certain reptiles. Trans. Linn. Soc. London. July, pp.

455-500. Testut, L. 1897 Traite d'anatomie humaine. Paris. Thompson, P. 1909 Description of a model of the brain of a foetal cat, 20 nuu. in

length. Journ. Anat. and Physiol., vol. 43, pp. 134-145. Welcker, H. 1878 Wachsausguss der Gehirnventrikel. Virchow's Archiv. f.

Pathol. Anat., Bd. 74, S. 500-503.



B.olf., bulbus olfactorius, caudal protrusion

CbL, cerebellum

Ch.op., chiasma opticum

Coll. %., colliculus inferior

Coll.s., colliculus superior

C.inf., cornu inferius

C.par., corpus paraterminale, corpus pineale

C.str., corpus striatum

For., foramen interventriculare

F.plx., fissure for plexus chorioideus

Hah., habenula

Hip., hippocampus

Inc.pc, incisura postcommissuralis

/., infundibulum

Is., isthmus, lamina chorioidea epithelialis

Mes., mesencephalon

A'^i — Nt, neuromeres

N.op., nervus opticus

N^.tr., nervus trigeminus

Pa7\, parencephalon

Plx., plexus chorioideus

R.i., recessus infundibuli

R.I., recessus lateralis

R.m., recessus mamillare, recessus opticus, recessus pinealis, recessus postopticus

R.sup., recessus suprapinealis

Rf., rhomboid fold

S.I., sulcus limitans

Syn., synencephalon, tela chorioidea

Th., thalamus

V.b.olf., ventriculus bulbi olfactorii

V.I., ventriculus lateralis, velum transversum

V.op., vesicula optica



4 Wax model of the cerebral ventricles of a 5.1-mm. pig. H.E.C. series 1904. X 20 diam.

5 Cast of the cerebral ventricles of a 12-mm. pig. X 15 diam.

6 Dissection of the brain of a 12-mm. pig. X 15 diam.

7 Dissection of the brain of a 12-mm. pig. X 15 diam. The plane of the cut edge of this dissection is indicated by arrows in figure 5.




Is., Mes.^


Fi.i. n

Fig. 7




8 Dissection of the brain of a 17-mm. pig. X 10 diam.

9 Cast of the cerebral ventricles of a 17-mm. pig X 10 diam. 10 Dissection of the brain of a 17-mm. pig. X 10 diam.






C. str.





11 Dissection of the brain of a 22-mm. pig. X 10 diam.

12 Cast of the cerebral ventricles of a 22-mm. pig. X 10 diam.

13 Dissection of the brain of a 22-mm. pig. X 10 diam.






Vel. tr. ^


C. pa r.

Fig. 11

pi n.



14 Dissection of the brain of a 45-mm. pig. X 6.5 diam.

15 Wax model of the cerebral ventricles of a 45-mm. pig. X 6.5 diam.

16 Left lateral ventricle of the model shown in Fig. 15.

17 Ventral view of the same ventricle.




T. ch.

C. ))ln. .Mt^sS^Z




Jnc. pc.

V.holf. a inf.





18 to 23 Alicrophotographs of frontal sections through the head of a 45-mm. pig. H.E.C. series 1826. X 5 diam. The position of section 563 is indicated by

arrows in figure 14.


Section 563.


Section 629.


Section 687.


Section 755.


Section 1067.


Section 1115.




Fig. 18

Fig. 19

Fig. 21

Fiq. 21

Fig. 23




24 Wax model of the mid-brain ventricle of a 110-mm. pig, viewed slightly from below and behind. X 15 diam.

25 Wax model of the cerebral ventricles of a 260-mm. pig. X 3 diam. 20 Ventral view of the same model.




Colli .





V. b.olf.


F. plx.

V. b. olf.





Department of Anatomy, Cornell University Medical College, Neiv York City



Numerous investigations have recently been directed towards an analysis of the processes concerned in the development of the vertebrate eye. Both mechanical and chemical methods of experiment have been employed and at the present time most of the results seem open to explanation. Various theories and speculations, however, have been advanced which will probably form a source of contention until the experimental results are better or more uniformly interpreted.

By way of introduction to the experiments recorded in the present paper and a discussion of their significance, it may be well to outline briefly the status of the main problems concerned.

The writer has previously recorded the results of experiments bearing upon an analysis of the manner of origin and development of the optic cup and the optic lens. It was demonstrated that in the- species studied the crystalline lens can arise entirely independent of any influence from other eye parts. It seems also equally clear from the many cases observed that the optic vesicle or cup can at some stage during its development induce the formation of a lens from the ectoderm with which it comes in contact.




even though this ectoderm is not exactly that from which the normal lens would have arisen. Lenses were induced to arise only from head ectoderm and independent lenses invariably arose in the head region, it was thus concluded that the head ectoderm principally possessed the independent lens forming power.

A consideration of numerous eye abnormalities which occurred in these experiments seemed to throw light on the earlier conditions of origin of the optic anlage from the medullary tissues. A number of the abnormal eyes appeared to lend themselves to a common interpretation and for this reason I advanced in an entirely hypothetical manner the view that these conditions might be considered as arrests in eye development.

Spemann in two more recent contributions to the development of the eye has confirmed all of my observations that were touched upon by his experiments. In an equally general manner he has disagreed with most of the deductions which were drawn from my experimental results. ' An attempt to satisfy these disagreements will be made in the body of the present paper.

It is now concluded from a study of abnormal eyes and the experiments below that the eye anlage in the medullary plate is primarily median and single and normally separates or spreads into two almost equal growth regions which develop in lateral directions reaching further and further out until finally the optic vesicles come in contact with the ectoderm at the sides of the head. Provided this view is correct cyclopia is then an arrest in eye formation.

Spemann., on the contrary, holds that the eye anlagen originally arise lateral in position along the borders of the medullary plate. The cyclopean defect according to him is due to a failure of central medullary tissue to develop so that the lateral eye anlagen slump towards the median plane, fuse and form a single cyclopean eye. Spemann, however, presents no experimental evidence to show that the eye anlagen do occupy lateral positions since all of his operations included the median medullary tissue as well as the lateral.


The only evidence to draw upon is the interesting experiments of Lewis, in which it was found that cyclopia sometimes resulted in Fundulus embryos when the anterior end of the embryonic shield was injured by pricking. Lewis also interpreted these defects as due to a fusion of the optic anlagen, and had suggested, as Spemann now does, that the chemicals used in my experiments suppressed the development of niedian tissue in the medullary plate and thus caused the eye anlagen to come together, fuse and produce cyclopia. There are a great many strong objections to this hypothesis of Lewis and Spemann which have been enumerated in my previous papers and to which I shall take occasion to refer briefly in the following discussion.

An objection of primary importance to the idea of cyclopia as a result of the coming together of lateral anlagen through a failure of intermediate tissue to form is the fact that cyclopean eyes are rarely in size and extent equal to the sum of the two normal eyes combined. A cyclopean eye is, as a rule, little if any larger than one normal lateral eye and in fact is often much reduced or actually minute in size as compared with a normal eye. This fact indicates most decidedly that eye material, as such, has been, injured or arrested in its development and differentiation. One is then scarcely warranted in assuming that the defect is solely due to a failure in formation of material between the eyes,

Spemann has found, although he locates the anlagen in the wrong place, that not only is the eye anlage definitely localized in the open medullary plate but actually the tapetum nigrum is distinct from the other retinal layers. How then could absence of material between the lateral eye anlagen cause less eye material to arise?

The differentiation of these prelocalized anlagen require definite amounts of energy. Any treatment that weakens the developing embryo at certain periods in a definite way renders the eye anlagen incapable of differentiation so that they do not arise from the brain.

The entire problem is readily open to experimental test. The contention may resolve itself into the question: Where are the


optic anlagen originally located in the medullary plate or tube? The present communication will present the results of experiments aimed towards an answer to this question. Certain interpretations put forward by Spemann regarding the origin and development of the primary components of the eye will also be considered.


The material used in all the experiments has been the developing embryos of the salamander, Amblystoma punctatum. Amblystoma eggs are surrounded by masses of jelly-like material from which they may readily be separated. The eggs live perfectly without the jelly mass, provided they are well covered with fresh water.

These eggs are of sufficient size to render it possible to cut out with fair accuracy definite regions or groups of cells from the medullary plate or groove.

The method of experiment has been entirely by mechanical operation. This method is particularly useful for the problem in hand since definite areas may be removed and the results studied. The operations were made under a binocular microscope. Fine steel needles and the smallest scissors were used as instruments.

The embryos were kept from three to five days after the operation in water which had been previously boiled. They were killed in a mixture of one part formalin to three parts saturated corrosive sublimate, left for three hours, rinsed in water and put into 70 per cent alcohol with iodine. Others were fixed for three hours in Bouin's fluid, formalin and picroacetic. The sectional embryos were stained with Delafield's hematoxylin and eosin.



1. Sticking the future brain portion of the medullary plate with


The anterior portion of the medullary plate was stuck with steel needles in such a manner as to disturb the cells over an area extending from the anterior border of the medullary fold back to the constricted portion of the plate and laterally from fold to fold. The needle was inserted below the outer layer of cells and raised so as to push the cells apart; this was done a number of times with each specimen. The needles were also swung to the right and left in the medullary tissue until the cells were considered to be disturbed to quite an extent. The object in such an experiment was to determine how severe an injury to the cells was necessary in order to prevent the development of the optic vesicles.

Twenty-three embryos were treated in this manner, and all were killed four days after the operation. Under a high power binocular microscope most of them distinctly showed that the optic cups were well pushed out laterally and in contact with the ectoderm at the sides of the head. Both eyes were clearly seen in seventeen of the individuals, while six seemed to have eyes yet not so well developed. These six doubtful specimens were sectioned and studied microscopically.

Both eyes were present and apparently normal in structure in five of the six embryos. The sixth individual showed eyes which were slightly irregular in form and poorly developed, yet both eyes were distinctly present.

The experiment would indicate that a disturbance, of the type employed, of the cells constituting the eye anlage in the medullary plate was not sufficient to prevent the normal development of the eyes in these embryos.

Another point of interest might have been attacked, by such an experiment provided the embryos had been allowed to develop sufficiently long after the operation. That is, whether or not the cells destined to form the tapetum nigrum layer might be


intermixed with those cells destined to form other retinal layers, ^nd so produce an eye with the pigment cells scattered irregularly through the retina. Spemann's recent experiments on cutting out and reversing certain areas of the open medullary plate indicate that the tapetum cells are fully localized and separate from the other retinal cells in certain amphibian embryos. The embryos in my experiments had not differentiated the retinal layers sufficiently far to determine with certainty whether there was a persistent disarrangement of the cells, yet the general appearance of the eyes seemed perfectly normal.

2. The median region of the anterior part of the 7nedullary plate cut out, reversed and transplanted in the medullary plate

This experiment is similar to 'fehose performed b}^ Spemann on several amphibian embryos. Spemann found that pieces of the medullary plate when cut out and turned around continued to develop with their original orientation undisturbed, thus indicating the early prelocalization of certain future parts of the brain and eyes. When the operation chanced to cut the eye anlage so that part of the future eye material was anterior to the cut and remained in position, while part was contained in the cut-out piece which was then turned around and transplanted, carrying the future eye cells to a more posterior position, two eye regions developed on each side. One arose from the anterior undisturbed cells and the other from the transplanted posterior cells.

The reversed pieces in the present experiment were not long enough to carry the eye back to a distant posterior position, and the cut extended so far foreward that the eye anlage was not divided transversely as in Spemann's operations. The operations were done chiefly to test whether the eye anlage in Amblystoma was well localized and would develop after such reversal of tissues.

Eight embryos were studied after having had antero-median pieces of the medullary plate cut out, reversed and transplanted. Seven of the eight developed both eyes, many of which showed


indications of their misplacement. One individual showed one abnormal eye while the other was probably indicated by a doubtful eye structure situated within the archenteron. The general structure of the eyes seemed normal and only slight indication of their reversed origin was shown by a study of these early stages.

Thus if the eye anlage was contained within these cut-out pieces the reversal and transplantation of the pieces did not show any very detrimental effects on the future development of the eyes, though in one of the eight cases one eye was abnormal and the other was greatly misplaced and indistinctly indicated.

The experiments of sticking and disturbing the cells in the anterior end of the medullary plate without actually removing the cells does not prevent the subsequent development of the optic vesicles in an apparently normal manner. Cutting out rectangular pieces of the medullary plate which contain the eye anlage, reversing and replanting them merely cause the eyes to develop in misplaced positions. These two experiments demonstrate the fact that unless the future eye material is well removed by the operation the optic vesicles may later form; this is important in estimating the value of the results in the experiments to follow.

3. Lateral regions cut from the anterior part of the medullary -plate

Four embryos were operated upon, as shown in figure 1. The indicated region of the right lateral portion of the flat medullary plate was cut entirely away with fine scissors. The area removed did not extend quite to the median line except that in one case it may have or probably did remove tissue beyond the median line. The embryos were killed three days later, cut into sections and studied microscopically.

One of the four lacked both eyes or had eyes so small and poorly formed that they could not be recognized, and the brain showed on one side indications of the operation. Both eyes had thus been removed by an operation which extended at most


slightly beyond the median line and certainly did not include the left lateral medullary tissue.

The other three embryos had both eyes present, though one eye in one specimen and both eyes in another were abnormal or defective. In these cases the cut did not extend close enough to the median line to remove all of the eye substance of that side; 3^et if the eye anlagen at this time had been lateral in position, one would have been completely removed.

A similar experiment was performed on seven somewhat earlier embryos. In these cases where the medullary plate was younger and wider in extent the removed area was confined to a lateral position and did not extend close to the median line, yet it extended laterally into the medullary fold (fig. 2).

Three days later the embryos were killed. On studying the sections of the head region six of the seven embryos were found to possess two perfect eyes of normal proportions. The seventh individual had only one well defined eye while the other eye was absent.

The removal of this lateral area of tissue which has been considered the position of the early eye anlage by certain investigators gives no effect on the development of the future eye unless the cut be made to extend very close to the median line. In eleven operations of this type nine individuals did not suffer the loss of either eye, while one specimen lacked both eyes and another one had only one eye. The case in which both eyes were absent might have been due to the fact that the cut extended a little way beyond the median line and removed cells destined to form the material of both eyes. Experiments recorded below would indicate that this was the case. A second possibility is that the operation so weakened the individual — • which may have been below normal in vigor, though apparently perfect specimens were selected for all the operations — that it lacked the power to differentiate the eyes from the medullary tissue.

Six of the nine specimens which possessed both eyes showed no effects of the removed material in either the size or form of their eyes. The other three embryos had one or both eyes some




what smaller than usual. The cases with only one eye defective are perhaps clearly understood on the ground that the cut encroached upon the material destined to take part in the development of this eye. In the cases with both e3^es defective it scarcely seems possible that the opposite eye could have been injured by these cuts without entirely removing the eye on the side where the cut was so extensive. The only plausible possibility is that the material for both eyes was mediallj^ located and the two eye anlagen closely connected. Any injury to this region might later be shown by both eyes if the injury were at all extensive, and if less extensive and confined entirely to one side of the median line only one eye might show the effect.

I realize that this explanation might be interpreted in part as opposed to Spemann's experuuents in which he found the future eye materials so consistently definite in the medullary plate. When the eye anlage was cut transversely and the posterior portion reversed and planted in a distant position it was found that if the forward part of one eye was small the part which had been cut from this and placed posteriorly was large. This fact does not indicate entirely that the eye stuffs are separate; only that when the eye anlage, even if it be median and single, is cut across obliquely and the posterior part removed and transplanted it continues development in a normal consistent manner, the small eye part coming off on the side where less material exists and vice versa.

4. The anterior lateral part of the 7ieural fold and a portion of the medullary wall cut from one side

These experiments were performed on embryos older than those used above in order to test whether the eye anlagen shortly before the coming together of the neural folds might be situated laterally along the border of the folds. In other words, do the eye anlagen primarily arise in the median line or region of the medullary plate and later occupy more lateral positions?

The operation consisted in cutting away one side of the anterior part of the medullary fold and also including in the removed


part some of the material constituting the lateral wall of the medullar}^ groove (fig. 3).

Fom- embryos were operated upon in this manner. Three days after the operation they were killed and prepared for study. A careful examination showed that three of the four specimens possessed both eyes, while one had no eye, or a questionable formation, on the operated side although a perfect eye was present on the unoperated side. Thus in three-fourths of the few cases used, neither of the future eyes had been injured by the operation.

Six other embryos were operated upon in a similar fashion. These were killed two days after the operation. Studying the six embryos in cross section it was found that three possessed both eyes in perfect condition, one individual had two eyes, yet one eye was small and defective on the operated side, and two of the specimens had no eye on the operated side. Thus four of the six specimens had two eyes each, and two of the six had only one eye each.

Combining the results of both experiments it is shown that, of the ten operated embryos seven possessed both eyes and three lacked an eye on the operated side. One of the seven which had both eyes present showed a defective eye on the operated side.

A comparison of these experiments with those recorded in the preceding section in which the operation was performed on earlier embryos would indicate that the eye anlage has possibly widened out so as to extend into more lateral regions in these later stages. The weight of evidence, however, would indicate that the position is not directly lateral along the edge or border ol the medullary plate, as Spemann has assumed.

It must be remembered that in the folding process which converts the medullary groove into a tube the original borders or edges of the medullary plate come to meet in a middorsal line. After the medullary tube is thus formed the optic vesicles push out from a lateral or really ventro-lateral region and certainly do not in any sense come from the original borders of the medullary plate which are now dorso-median in position.


5. Both anterior lateral parts of the neural fold cut away just before the folds close together

The operation consisted in clipping away with scissors the lateral wall of the neural groove from both sides; that is, the raised folded parts and the dorsal portion of the lateral neural wall indicated in figure 4. The actual crest of the neural folds is of course not future lateral material as in the closure of the furrow to form the neural tube the crest becomes dorsomedian as just mentioned.

Four embryos were successfully operated upon in the given manner. Three days after the operation they were killed and prepared for study. All of the embryos were found to possess both eyes in well developed conditions.

Five other embryos in a similar stage of development were operated upon so as to remove the lateral neural folds as indicated in figure 4.

Four days after the operation the embryos were killed. Studying these embryos in section showed that four of the five possessed both eyes in a well formed condition giving no indication of the injury while the fifth individual possessed only one eye. This last specimen was evidently cut too near the median line on one side so that the future eye material of that side was destroyed.

Thus in nine specimens with both sides operated upon, eighteen operations, only a single eye was missing. The other seventeen operations did not cause smy noticeable effect in either the development or nature of the resulting eye.

These experiments clearly demonstrate that the eye anlage is not located along the lateral edge or border of the medullary plate or groove, as Spemann holds. The importance of such facts in connection with opposing explanations of certain eye abnormalities will be fully considered in a following section of this paper.

The next experiments to be presented were performed in order further to substantiate the fact that the eye parts do occupy a


median position in the medullary plate or groo\'e. I have been led to think from a study of a number of eye abnormalities that the eye anlage is more or less median in its original position.

6. The removal of anterior median strips of cells from the wide

medullary plate

Tn these operations a narrow median strip was cut from the anterior region of the wide medullary plate. The strip in all cases was narrow, being in fact only about one-third or one-fourth the entire width of that portion of the medullary plate lying between the medullary ridges, as shown in figure 5.

Four embryos with wide medullary plates were operated upon in the above manner. Four days later they were preserved for study.

Three of these four specimens were eyeless, showing that the operation had removed the entire eye anlage from the middle of the medullary plate. The remaining one of the four embryos had one poorly formed eye while the other was only slightly indicated, so that in this case almost all of the ability to develop eyes had been lost as a result of the operation.

Five other embryos were subjected to a similar operation to that indicated in figure 5 except that the median strip removed was still narrower than in the preceding experiment. These embryos were of different stages but all showed the medullary folds far apart so that the medullary groove was wide open. Three days after the operation the embryos were prepared and studied.

One of the specimens was eyeless, one had two poorly developed eyes, two had one well formed eye while the other eye was questionable or absent, one had both eyes present. Thus the removal of a very narrow median strip gave less decided effects than the removal of somewhat wider strips in the experiment above. Yet four out of the five embryos showed eye defects, two having both eyes affected and two having one normal eye and one eye questionable or absent.


Combining the results of the two experiments it is found that the removal of narrow median strips from wide medullary plates exerts the following influence on the future development of the eyes: Of nine embryos thus operated upon four failed entirely to develop either eye. Two showed two defective eyes. Tivo individuals developed one perfect and one defective or questionable eye. Only one of the nine embryos showed two apparently normal eyes.

Since six of the nine embryos had the development of both eyes either entirely suppressed or decidedly affected, and two of the remaining three had one eye affected, it seems most certain that cells destined to take part in eye formation are located in the median region of the medullary plate and are removed by the operation employed. One must conclude that the median optic anlage occupied at least one-fourth or one-third of the width of the medullary plate in the anterior region.

A general statement of the results of certain of the experiments described above may be expressed as follows: Thirty embryos studied after various operations in which lateral portions of the medullary plate were removed at slightly different developmental stages (sections 3, 4, 5) showed in twenty-four individuals, or 80 per cent of the cases, subsequent development of both eyes, while only six specimens or 20 per cent of the cases, showed absence of the eye. In one case the presence of the eye is questionable, in five cases one eye and in one case both eyes were absent. The absence of eyes in the latter cases is possibly due to the cut having been made in a more median position than was intended.

Nine embryos studied after having been operated upon so as to remove a narrow median strip of cells from the anterior portion of the medullary plate (section 6) showed in four cases, or about 45 per cent of the specimens, entire absence of eyes. In four other individuals the eyes were highly defective, one specimen having one poorly formed eye while the other was questionably present. In only one of the nine embryos did both eyes approach the normal condition, from this specimen an extremely narrow piece had been cut away. The optic anlage in this case might have been sufficiently wide at the time of the


operation to allow its median portion to be removed and yet enough material remain on either side of the cut to give origin to the two eyes. According to the views of several investigators the removal of this median material should have caused cyclopia, yet it did not. I shall presently attempt to show that cyclopia is not due to the coming together of lateral material in the median line but to a failure of median material to spread laterally.

When the results of these two classes of experiments are contrasted one must conclude that: The eye anlage in the medullary plate occupies an antero-median position as shown by the various abnormalities incurred when this region is cut away. The failure to injure the development of the eyes in the great majority of cases when the lateral portions of the medullary plate are removed by operation indicates further that the eye anlagen do 7iot occupy lateral positions during this stage of development.



There has lately been considerable discussion regarding the way in which the cyclopean defect occurs. The experiments described above may serve to concentrate the case on a definite developmental period and, to my mind, settle the question as far as the medullary plate stage is concerned. Whether it is possible to carry any question of vertebrate eye abnormalities, as such, further back in development than this stage seems doubtful, since here it is that the localization of the eye anlage is first knOwn to exist. It is certain, of course, that this anlage does come from cells which are present before the medullary plate has formed. Whether these cells are localized and are entirely future eye anlagen cells, and not indifferent ectoderm cells which might have the power of forming any portion of the brain or central nervous system, one is at present unable to state.

As the case stands, it seems possible to explain the entire genesis of the cyclopean defect from the earliest time at which the optic anlage is definitely known to be localized. This as


sumption is not made solely from the material presented in the present paper, but from the facts furnished by these experiments together with the observations made upon the large number of Cyclopean eyes and brains which the writer has studied during the past several years.

Various authors have at different times thought that cyclopia was due to a fusion of the eyes after they had arisen from the brain. The earlier in development the fusion occurred the more intimately associated the two eye components became. This view has been proven incorrect by actual observation on Cyclopean monsters where it is found that the cyclopean condition of the eye, whether large and hour-glass shaped or of small size resembling a normal eye, is present from the earliest appearance of the optic vesicle from the brain. ' In other words, the several degrees of the cyclopean eye come off from the brain in their final conditions.

The idea of the fusion of the eye parts was deep rooted, however, and now exists in the recent views of Spemann in a refined form. Spemann believes, as several others had previously suggested, that cyclopia is due to an absence of non-ophthalmic tissue in the median region of the medullary plate or groove. This lack of median tissue allows the eye anlagen which he holds to be lateral in position, near the borders of the medullary plate, to come together and fuse iii the median plane and later give rise to a cyclopean eye. Cyclopia, according to this idea, occurs in a more or less passive manner, and is, after all, actually a fusion of the eye anlagen of the two sides during development.

I am certain that this fusion explanation which has now been forced entirely back into the medullary plate, is as false as* its bolder predecessor which assumed the fusion to take place outside of the brain tissues after the optic vesicles or cups had arisen. Spemann did not advocate this late-fusion view, but claimed from his beautiful experiments on Triton that the cyclopean eye arose out of the medullary tissues in its final condition. He now, however, assumes the role of a most ardent supporter of the fusion of the optic anlagen within the medullary plate.


The present writer's opinion of the cause of cyclopia, which was advanced in 1909, was to the effect that this deformity is the result of a weakened development. Spemann has termed this view the 'laming: hypothesis,' since my assumption was that the A^arious chemical substances employed in producing Cyclopean defects in fish embryos have a tendency to lower the developmental energies and so reduce the power necessary to accomplish the processes involved in the outpushing of the optic vesicles from the brain.

Considering the probable manner in which the cyclopean defect occurs, Adami ('08, p. 241) has theoretically concluded that it is due to developmental arrest or lack of vigor. While I am unable to agree with the details in Adami 's argument of a primary growth point at the anterior tip from which is budded off successively the paired parts of the two sides, the anterior ones necessarily arising last after the other parts had been left in more posterior positions, the final conclusion that a weakening of particular developmental processes results in cyclopia is confirmed by all my experiments.

The different . types or degrees of the cyclopean defect depend upon the stage in development at which the arrests occur as well as upon the strength or severity of the treatment employed. I shall now attempt to defend this position with the evidence at hand, and in so doing shall as decidedly prove the mistake in considering the defects as the result of any failure to arise of median medullary tissue (other than future eye tissue) and the subsequent fusion of the lateral optic anlagen. There is no median tissue between the eye anlagen. The median tissue is the eye anlage itself and will subsequently go to form some portion of the eye, either optic cup or optic stalk, depending largely upon its position and the extent of normal development attained.

The writer had ('07, '09, '10) recorded a number of eye conditions which are considered to be different degrees of cyclopia using the term in a general sense. At any rate, these several conditions differ only in degree and grade perfectly into a continuous series. There is no qualitative difference between them. Spemann has objected to including among these defects certain



modified types which are at least closely related to the series in manner of origin, though they may not actually intergrade. (We are now considering only the cyclopean series, not the monster monophthalmica asymmetrica which will be dealt with later.)

One finds on referring to my paper of 1909 (p. 293) that the cyclopean series, A to G, passes from the normal individual through different degrees of association of the two eyes to a median cyclopean eye only as large as one normal lateral eye, then to a cyclopean eye of smaller dimensions until it is extremely minute and may finally be deeply buried beneath the brain as a small pigmented vesicle, as is shown in figure 52, page 321. Only one step further and the eyes fail to arise entirely so that eyeless individuals exist which with a slightly greater power of differentiation or more developmental energy might have given cyclopean monsters. The last assumption is warranted since these eyeless specimens actually resemble the cyclopean monsters in other structures; for instance, the mouth is a narrow proboscis similar to that in the cyclopean monster instead of the usual laterally spread mouth of the normal embryo.

Why should every step and gradation in this series exist if several, or any of the conditions are of a different quality or type? It seems certain that one examining the large number of cyclopean fish on which* my study was based would be forced to admit the correctness of the statement that these individuals exhibit different degrees of one and- the same defect.

The question then follows, if cyclopia were due, as Lewis, Spemann and others assume, to a failure to develop of median medullary tissue so allowing the eye anlagen to come together in the median plane and fuse, why is not e^^ery cyclopean eye equal in mass to the two normal eyes fused? Spemann does not suggest in any place that eye material also fails to arise. He shows in his recent experiments that the future eye is fully laid down in the medullary plate. Not only is the eye present in the medullary plate but the cells destined to form different layers are distinct. Spemann found that certain cells cut out of the medullary plate and planted in more posterior positions formed only the tapetum nigrum layer. If the eye is thus so


definitely predetermined in the medullary plate and Spemann believes that cyclopia is due to failure of median medullary cells other than future eye cells then cyclopean eyes ought always to be large or double in size.

The fact is that these predetermined eye cells in the medullary plate in most cases of cyclopia are incapable of perfect differentiation on account of insufficient energy, so they remain in the brain, or only part of them is capable of difTerentiation and thus small defective cyclopean eyes result.

The actual 'Lahmungs' or suppression is of the eye material itself. Cyclopia is an eye defect, and an injury of the eye forming material is the cause. The brain may also be defective as an accompaning abnormality, although in some cases the brain, with the exception of the optic tracts and parts, may be structurally and functionally perfect, as is indicated by the normal life and behavior of many of the cyclopean Fundulus embryos as well as by the existence of the huge cyclopean ray, Myliobates noctula, reported by Paolucci in 1874.

The chemical substances employed by the author in producing the cyclopean defect and a number of others with which McClendon has obtained similar results, all tend to suppress or arrest the development of the eye material in the brain. This future eye material is assumed to occupy a median position. When the arrest is complete, and necessarily taking place at early stages, no eye parts arise from the brain nor are any differentiated within the brain substance itself. Thus a completely eyeless individual is produced.

Spemann states that the eye is capable of differentiation even though it be contained within the brain substance as he has found in amphibia and as MencI recorded in a teleost. It must be realized that these are exceptional cases. A certain amount of energy is necessary for differentiation of the eye to take place even within the brain and when only this amount of energy is present the eye may differentiate within the brain, but when the required energy for any reason is not available the eyes are incapable of any differentiation. Many eyeless individuals have been observed in my experiments which have no indication what


ever of eye parts within the brain. Could any one ask, whethei this be due to the failure of non-ophthalmic parts of the medullar}^ plate to arise or, on the other hand, to a failure of the future eye forming cells to arise, or to differentiate after they have arisen?

It is not meant to convey the idea that all eyeless brains are related to the cyclopean series, as this is not the case. The future eye forming cells may in some cases have been absent from the start. In other specimens the future optic vesicles might have been in positions to arise normally and laterally and for some reason were incapable of outpushing or of differentiation.

Certain eyeless brains, however, such as those in individuals having the proboscis-shaped mouth, do belong to the series and must be caused in the same way as are the various cyclopean conditions. They have merely responded to a more exaggerated degree.

The most extreme cases of cyclopia with actual eye structures are those in which a small pigmented vesicle arises from a ventromedian part of the brain, as is shown in my figure 52, 1909. This pigment la^^er of the retina seems to be its most persistent portion, as it may appear when all other recognizable retinal parts fail to arise. There is a possibility that the small tapetum nigrum groups of cells which in some cases formed from Spemann's transplanted portions of eye anlagen may not be due to the fact that only the anlagen qf such cells were transplanted, but that all other retinal cells except these were incapable of differentiation when so small a piece of future eye tissue was isolated by the operation.

The next degree of cyclopia is exhibited by an individual having a median eye that is much smaller than one normal lateral eye. We then have a median cyclopean eye of about the same size as one normal lateral eye. The latter case has been termed the 'perfect' cyclopean condition.

All these abnormalities are best explained as follows: The future eye forming cells occupy a median position in the medullary plate and the cells destined to form the two eyes are arranged in one group. This median group of future eye cells


normally widens or spreads laterally while two centers of active growth become established which gradually assume more lateral positions until they push out as the two optic vesicles. In the degrees of cyclopia mentioned in the preceding paragraphs the median eye anlage does not widen or spread laterally but is arrested in its primary condition; thus the two growth centers are not sufficiently separated and only a single center exists, and even more than this, the arrest is to such an extent that the entire or normal amount of optic material does not differentiate. Hence, one finds a median cyclopean eye consisting of an amount of eye material far below that normally present.

Other individuals are found in which greater masses of eye material have succeeded in differentiating, and development has been vigorous enough to allow the early eye anlagen to spread to a greater or less degree and establish the two eye forming centers. Such specimens finally present a cyclopean eye showing distinctly its double composition, the two retinae are more or less distinct and the eye large consisting of a greater amount of material than a normal lateral eye, yet less bulky as a rule than the sum of the two normal eyes of the species.

The hour-glass eye or incomplete cyclopiq, is commonly observed in the experiments. This case is due to a later or less complete arrest in development than those mentioned above. Both eyes have differentiated out of the medullary tissue but the embryo was not vigorous enough to permit their normal separation and outpushing. Thus the eyes come off from their original ventro-median position and remain in close contact or actual union.

Finally one observes individuals in which the eyes are separate and distinct yet unusually close together. Such embryos are able to differentiate their eye material and this material is capable of pushing out from the brain but a slight weakness or arrest has occurred on account of which the optic stalks are short and the optic cups are unable to take a normal positionso that they remain unusually near together and look in an .abnormal direction.


In the ordinary individual the future eye forming material is first located in a median position in the medulla^-y plate. This material becomes more extensive or widens laterally, and two growth centers are established, the material between the centers finally becomes the median ventral layer of cells of the optic stalks. Later the incipient optic vesicles begin to evaginate or push in a ventro-lateral direction and finally turn dorsally and laterally to reach their usual places at the sides of the head. The optic stalks, however, still lead back to the ventro-median position and there in the fish the optic fibers, following the optic stalks as paths, cross and in higher vertebrates form the optic chiasma always in the ventro-median plane below the brain floor and from here the optic tracts proceed to their centers in the brain.

The median position of the optic chiasma outside and below the brain is an important structural fact in the present consideration. Figure 6 is a diagram of a transverse section through an early brain with the optic cups in their usual position. The optic stalks connect with the brain and the median ventral cell layer is actually part of the stalks. As development proceeds the optic fibers arising from the cells in the retina follow along the optic stalks to reach the brain. The investigations on the development of the optic nerve have shown that the optic stalks become solid and form the supporting paths or neuroglial scaffolding along which the optic nerve fibers grow. The fibers from one retina meet those from the other in the median plane below the brain and in the fish the fibers from the two retinae cross directly while in higher forms partial crossing takes place and the optic chiasma is formed outside the brain. Figure 7 is a sketch representing a cross section of the actual condition of the early optic nerves in a fish embryo. This position of the optic cross is only possible if the median tissue be optic stalk tissue.

• Suppose, on the other hand, that the eyes primarily originate from lateral medullary tissues and between the two eyes other brain tissue is present. The optic stalks are then attached to. the lateral regions of the brain from which the optic vesicles



Fig. 6 A diagrammatic cross section through the brain and optic cups of an early embryo, with all of the median ventral medullary tissue represented as being part of the optic stalks. The future fibers of the optic nerves, shown in red, follow the optic stalks to the median plane where they cross and afterward enter the brain to pursue their course as the optic tracts. The optic cross is entirely beneath and outside the brain.

Fig. 7 A sketch representing the actual outlines of a cross section through the brain and eyes of a late fish embryo. The optic nerves cross below and outside the brain and are surrounded by cells derived from the original optic stalks.



pushed out. Figures 8 illustrates diagrammatically a cross section of this condition. In the course of development the fibers of the optic nerve following the stalk reach the lateral position and must enter the brain and continue within its tissue in order to meet the nerve of the opposite side and form the cross or chiasma. Brain tissue would lie beneath the optic chiasma and the chiasma would necessarily be within the brain. This condition is never found in any normal vertebrate.

Fig. 8 A diagrammatic cross section through the brain and optic cups in an imaginary case in which the optic connections are lateral with median brain tissue originally lying between the eye anlagen. The future optic fibers gr-owing along the optic stalks as paths reach the lateral points on the brain from which the stalks arose and then enter the brain tissue before having formed the cross. Continuing to grow, the optic nerves meet and cross in the median plane. The cross is within the brain itself and lies above the median mass of tissue which has always existed between the eyes. No vertebrate brain exhibits such a condition.

One might claim that the optic fibers on reaching the brain ran ventrally and formed the cross beneath, but no such change in direction is seen at any stage of their development. The structural relationships seem to depend upon a median origin and connection of the optic stalks.

The fact that a cyclopean eye may have no well formed optic stalk and is entirely median in position not necessitating the


absence of any other brain part is in entire accord with the above facts.

It is believed, therefore, that the various degrees of the typical Cyclopean condition from eyes unusually close together, to median double or hour-glass eyes, to the large oval eye, to a median round eye of usual size, to the median eye smaller than one normal eye, and finally to complete failure of eye material to arise from the brain are probably all due to developmental arrest. The arrest in development is the result of some influence which has reduced the developmental vigor below the normal so that the energy is not available to carry out the usual processes of differentiation and growth.

The author has figures and described other cases of eye defects which do not fall exactly into the series of cyclopia as considered above; yet these cases are modified conditions which are closely related to the Cyclopean series both in point of origin and in their final condition. The curtain-like eyes which face the median plane and often have a single lens between them might be considered as delayed cases of cyclopia. The eye anlage widened, the eyes became separated and continued their diff"erentiation, yet they were unable to turn out and assume their normal lateral positions so that they faced in a ventromedian dkection with their anterior walls closely approximated as is shown in figure 38 ('09) and figures 1, 4, 5, 6, 13, 14 ('10 a). Such eyes often excite a single stimulus upon a more or less ventral ectodermal region which responds by forming a single lens lying between the two eyes.

The writer has illustrated by a diagram (fig. 15, A and B '10 a,) the difference between these eyes with their choroid surfaces against the lateral ectoderm of the head and their concave retinal surfaces facing medially, and the eyes of a normal individual ('10 a p. 380). As was then stated, the experiments did not give a definite clue to indicate the position of the optic anlagen in the early brain, thus my explanation, or 'laming hypothesis,' referred mainly to the pushing out and lateral development of the optic vesicle and cup. All these cases are decidedly


of a type which would suggest lack of developmental energy necessary to attain the normal.

Another group of eye anomalies were extremely common under the same experimental conditions which caused the cyclopean series. These individuals possessed one normal eye in the usual lateral position while the eye of the opposite side in the numerous specimens showed various degrees of imperfection from a condition slightly below the normal in size to complete absence of the eye. Such anomalies were termed 'monophthalmica asymmetrica' in contrast to the symmetrical one-eyed monsters with a median cyclopean eye.i

The genesis of the asymmetrical defects is not entirely clear, yet they also are probably due to developmental arrest or suppression of the one eye. The growth centers representing the two future eyes of an individual are rarely equally vigorous and it is frequently noticed that one eye arises slightly before its mate and develops at a little faster rate. It might be that at some critical point in development one of the future eye centers is affected after the growth centers had begun to localize in more or less lateral positions.

In treating the eggs with alcohol a number of embryos occurred in which both eyes were small and defective even though they arose from the brain and attained more or less lateral positions. One might assume this to occur as a result of an arrest in development which affected both eye forming centers after the centers became separate or distinct from one another. Part of the eye forming material is suppressed or its differentiation is prevented so that each eye is decidedly under size and defective. In some of the cases of cyclopia mentioned above the eye was also very small and defective, in these cases the two growth centers which would give rise to the future lateral eyes did not become sufficiently separated so that only a single median eye arose and the reduced vigor permitted this to form only as a small and poorly differentiated structure.

In some instances where an embryo possessed only one member of the normal eye pair this eye was unable to attain its usual


lateral position so that it faced the median plane, and if the lips or peripher}' of the optic cup failed to reach the ectoderm the eye was without a lens.

In none of the specimens, although many were old with the central nervous tissue highly differentiated, was I able to detect any material within the brain which might be considered as representing the missing eye.

Eye conditions such as these are doubtless due to the action of some inhibitory influence which prevents the complete origin and differentiation of eye material, or when it does arise allows it to develop onlj^ in a weakened or defective manner. Since monophthalmica asymmetrica occurs so persistently in the same experiments with cyclopean individuals as Lewis, McClendon and I have all found, it is not improbable that the cause is the same in the two cases.

The above considerations have been entirely from the standpoint that the ophthalmic defects under discussion originate during the medullary plate stage. The cause of cyclopia or the tendency to produce such a defect might of course have its origin much earlier in development. It might, in fact, go back to the germ cells themselves or finally it might possibly occur as an hereditary variation. All experimental cyclopia, however, furnishes evidence directly contrary to the latter possibilities. This point I have considered in a previous paper and have clearly demonstrated as have several other investigators that the condition is not due to a germinal variation but may be induced as the result of external stimuli applied during the early development of the eggs.

The cyclopean abnormality may be caused in Fundulus embryos by subjecting the eggs to various chemical stimuli after they have developed normally for as long as fifteen hours. A fifteen-hour Fundulus embryo has the germ ring beginning to form and descend over the yolk sphere, the embryonic shield is scarcely indicated but appears very soon afterwards. Embryos of later stages subjected to the same treatment develop normally, or do not show cyclopia, while stages younger than fifteen hours and


as early as the first cleavage are much more readily affected in such a manner as to cause the cyclopean defect. The optic vesicles appear at about thirty hours after fertilization, but the stimulus must be applied at a time sufficiently long before this process occurs, since a number of important steps in eye formation are doubtless taking place before the visible signs of optic vesicles are present.

The fact that cyclopia may be produced after the beginning of the germ ring and embryonic shield in the teleost embryo indicates directly that the defect may occur in what would be the medullary plate stage of amphibian embryos. Thus explanations of the cause of cyclopia must consider it as occurring in the medullary plate stage.

Spemann ('12 b, pp. 38-39) has taken exception to my statement regarding the non-occurrence of cyclopia when eggs are treated later than fifteen hours after fertilization (although the optic vesicles do not arise until about the thirtieth hour) "since insufficient time exists for the substances to act on the eye anlagen.^' The paragraph following this statement in my paper (p. 388, '10a) is : The solutions are effective up to a stage in development preceding the formation of the germ ring and embryonic shield, and the action of the Mg on the eye anlagen probably takes place while the embryonic shield and outline of the embryo are forming J^^

This statement seems to me perfectly direct and clear; Yet Spemann intimates in one sentence that I mean to infer that fifteen hours are necessary for the substances used to penetrate the egg membrane! He then states that I showed in 1907 that KCl would penetrate within a few minutes and stop the embryonic heart beat. The Mg and other solutions used may pass through the membrane equally as rapidly, this I have not fully tested. The fact that it does has no bearing on the fact that stimuli do not have sufficient time to act upon the optic anlagen to induce cyclopia when applied to the eggs at later periods in normal development than fifteen hours after fertilization since they must act on the anlagen in the embryonic shield, and to act later is

1 The last clause was not originally in italics.


too late. The reason for the nonoccurrence of cyclopia when eggs were treated at periods later than fifteen hours after fertilization or fifteen hours before the appearance of the optic vesicles is that the optic anlagen and all other embryonic parts are constantly changing during development and have passed beyond the critical stage. It is not only a question of how quickly one may stimulate but in what condition the anlagen are at the time of stimulation. It is safe to say that cyclopia can not be produced after the optic anlagen have proceeded to some definite stage in their normal development. Even if Mg should penetrate all the membranes within a few seconds its action could never induce cyclopia in an embryo with the two optic vesicles visibly formed. There was no question of the time necessary for penetration, but an important question of the embryonic condition to be acted upon, the critical condition of the optic anlagen.

In the above connection it may be mentioned that such substances as alcohol and ether, when administered to an early embryo may cause it to develop into a decided monster. A late embryo or foetus may respond to the same treatment by producing an individual exhibiting no gross morphological defects yet showing decidedly abnormal nervous reactions, while a similar treatment might exert little or no effect upon a fully formed individual. A stimulus could not cause the mature individual to change into a structural monster. The developmental period of administration is of as high importance in determining the result as is the nature of the stimulus used, unless of course the stimulus be entirely destructive.

Again Spemann misinterprets a statement regarding the action of Mg. It was remarked that the cyclopean embryos developing in the Mg solutions were, except for the eye defects, more perfect than those arising from treatments with alcohol, ether, and so forth. The Mg cyclops often had apparently normal brains, could swim in normal fashion, took food and reacted to stimuli much as normal embryos did, while those embryos treated with alcohol and ether had various defects of the brain and cord. In this connection it was cited as of interest that Mayer had re


cently found in studying nerve muscle preparations that Mg salts seemed to prevent activity by affecting the muscle directly without apparently affecting the nerve. There is, of course, no direct connection between these facts and cyclopia; the mention was made merely in a general way, and most decidedly did not intend to convey the notion that muscle contractibility and the outpushing of the optic vesicles were phenomena of similar nature. They are similar only in that both are dynamic processes and require energy for their accomplishment.

There is no necessity for further discussing the fact that a number of eggs when subjected to the same solution do not all respond in a like manner (Spemann '12 b, p. 37)'. This is a typical case of differences in individual resistance and vigor which is observed among any one hundred individuals of any living species. It is equally true that the two sides of a so-called bilateral individual are rarely, if ever, identical.

Spemann is no doubt correct in stating that the relationship between cause and effect in my chemical experiments on cyclopia is not clear. Yet it seems to me that it is not entirely dark, the entire relationship between cause and effect in biological experiments is rarely if ever clear step for step.

I should like, however, to point out that the chemical experiments did one thing in proving that cyclopia could be caused from normal embryos through the action of the environment. This fact did away with all theories of germinal origin of the defect, one of which was strongly presented by Wilder about the same time. The experiments also make clear the stage in development at which cylopia may occur, and they further supply the richest amount of material yet available for the study of this defect. Finally, they prove to my mind that cyclopia is a developmental arrest and may be due to any cause which lowers developmental vigor at certain critical stages in the formation of the eye anlagen. These important points in the study of this defect have certainly not been made clear by the mechanical experiments though I do not deny that they might possibly have been.

To quote again from Spemann:


Gegen diese Defekthypothese erhebt Stockard (1909b, p. 172) die Fragen: warum sollte bei den ]\Iagnesiumembryonen gerade das Gewebe zwischen den Augen ausf alien, und keine anderen Gewebe?; warum sind die Riechgruben bei Cyclopie manchmal verschmolzen und manchmal getrennt?; . . . . ist bei den asymmetrisch-einaugigen Missbildungen der einseitige Augendefekt etwa auf die Abwesenheit der einen ersten Augenanlage zuriickzufiihren? — ^Darauf mochte ich mit der Gegenfrage antworten : sind denn all diese Tatsachen verstandlicher bei Zugrundelegung der Stockard 'schen Lahmungshypothese?

I should answer now, as it was inferred above, that they most decidedly are, and an attempt to show this fact in some detail has been made in the preceding pages.

The title of the recent paper by Spemann Zur Entwicklung des Wirbeltierauges" might better have been '^The development of the vertebrate lens," as little or almost no attention is given to other parts of the eye. The problem of the lens formation is very fully considered.

"Cyclopia of the lens" is discussed from the same standpoint as cyclopia of the optic cups. In the case of the lens median ectodermal cells may be missing so that the two normally lateral lenses fuse in the median line. It seems to me that this is the one straw too much for the 'Defekthypothese.' The imagination which pictures the falling out of just the exact amount of median ectodermal tissue which would allow the primary lens forming cells of the ectoderm to fuse towards the center and keep proper pace with the movements and final position of the various cyclopean eyes which myFundulus material presents must be most vividly active. An even more plausible possibility out of this imaginary dilemma is to consider the lens anlage as a single median group of cells that divides into two parts which come later to lie in lateral positions. This view would at least have the advantage that in cyclopia of the eye cyclopia of the lens" would maintain the lens forming cells in a fairly median region, and in normal development the lens cells would have a more or less definite path to follow and place to reach. The pineal eye in many forms possesses a fairly definite lens which must have arisen medially and the present lens may have been somewhat more anterior yet also median. These suggestions are of the


most speculative nature and are intended merely as such. In my experiments many of the embryos which possess supernumerary lenses show, as Spemann has called attention to, that the accessory lens may actually lie more anterior than the eye. Nevertheless, others, show free lenses in lateral positions.

The presence of a lens in the cyclopean eye is explained by the fact, well established for several species, that an optic vesicle or cup possesses the power to stimulate lens formation from any region of the head ectoderm with which it comes in contact. There would be no necessity of imagining a condition of cyclopia of the lens" even though a median lens should be observed in an anophthalmous monster. Normal lateral lenses have been observed in anophthalmic monsters (see figs. 1 and 5, and fig. 3, plate II, 1910 b).

The embryo from which my figure 4 ('10 b), was taken is of new interest in connection with a hypothetical case called for by Spemann (p. 81 '12 a). He states as a possibility that

der cyclopische Defekt nur die Epidermis betrifft, so dass Riechgruben undprimare Linsenbildungszellen median zusammenriicken, wahrend die Augenbecher, wie normal seitlich gelegen, sich ihre Linsen aus der dortigen Epidermis bilden. Durch solche Falle wiirden in der Tat an einem und demselben Kopf beide Fahigkeiten demonstriert, die der Linse zur Selbstdifferenzierung und die des Auges zur Linsenerzeugung. Bis jetzt liegen aber derartige Falle nicht vor.

The case however, was, recorded at the time and seems to. fill the requirements set forth. The independent origin and differentiation of the lens is demonstrated in a median position slightly more anterior than the eyes, and the more or less lateral eyes have derived lenses from the ectoderm with which they came in contact as is shown by figs. 9 and 10, 1910 b, yet this ectoderm is part of the usual region from which an optic cup has the power to derive a lens. The power to form lenses, as Speman and I have claimed, is possessed by the ectoderm of the head.

According to Spemann's assumption, the embryo (fig. 4) presents true cyclopia of the lens." The condition, however, may better be interpreted as an illustration of the high lens-forming



capacity possessed by the ectoderm at the anterior tip of the head.

Lens-forming power seems to diminish from the anterior tip of the head backward until trunk ectoderm no longer possesses the capacity to form a lens, as Spemann found in transplantation experiments. For this reason independent lenses arise, as a rule, far anterior to and often in front of the more or less lateral eyes (my figs. 3, 4, 7, 9, 10, 11, 12, 21, etc. '10 b). In fewer cases independent lenses are found in more posterior lateral regions approximately in the normal lateral eye position (figs. 1 and 2

Fig. 9 Two diagrams indicating the primary lens-forming power of various portions of the head ectoderm. The lens-forming tendency is considered to be greatest at the anterior end and gradually decreases towards the trunk ectoderm until the ability to form a lens is lost where the trunk begins. The rows of circles indicate the magnitudes of lens-forming tendency in different regions and do not signify the size of the lenses. Posterior lenses may be as large as anterior ones, yet they occur less frequently as independent structures. Free lenses usually occur near the anterior tip.

and 3, plates I and II '10 b). Figure 9 may serve to illustrate diagrammatically the extent and gradation of the lens-forming power possessed by the head ectoderm. This rather diffuse localization of lens-forming cells in the general head ectoderm as demonstrated by numerous experiments seems sufficient to account for all phenomena of lens formation in cyclopia, as well as the supernumerary lenses which the writer has reported.

The median position in cyclopia of normally bilateral organs involves one other part. The nose or nasal pits in the cyclopean



fish are sometimes median and single, or occasionally bilateral and more or less normal in position. I might also add that the nasal pits are often absent. Again the 'defect hypothesis' must expect median cells to fail and so allow the two nasal pits to fuse medially. In this case the evidence is still stronger for the primary median origin of the ectodermal anlage. Dohrn's studies on Ammococtes have shown the median position and relationship of the nasal and hypophyseal invaginations. The question of monorhiny in the cyclostomes is not fully determined but evidence is certainly available to indicate a median nose anlage. These, however, are phylogenetic considerations which would only serve to prolong the present discussion.

The experimental results presented by Spemann relative to questions of lens formation agree almost entirely with the conclusions which were presented in my paper of 1910 b. He disagrees, however, with many of my interpretations, yet I believe the disagreement is not as complete as it often seems.

I had suggested that the power of the ectoderm to form a lens without the presence of an optic cup was less vigorous or efficient than when the optic cup combined its stimulus with the tendency to lens formation possessed by the ectoderm. For this reason when the ectoderm was injured by many of the mechanical operations which have been employed in the study of lens formation the injured ectoderm was unable to form independent lenses although normally it would have had such power. Attention was called to the different results of Lewis and Miss King on Rana palustris.

Spemann ('12 a, p. 49) rejects this idea although he produces evidence in his paper to prove its very probable correctness. Compare the results he obtained in the origin of independent lenses after cutting out the optic anlagen from medullary plates with glass needles, in which case the ectoderm in the primary lens-forming region was uninjured, with the results following a burning out of the eye anlagen with hot needles, in which case neighboring tissues were necessarily injured. After the latter operation only one well formed lens occurred in five cases. A


further comparison between the glass needle operations on earl}^ stages with open medullary plates and later stages where the ectoderm was raised and the optic vesicles cut out from beneath it show a difference in the response of the ectoderm in producing free lenses in favor of the less injured or less disturbed ectoderm.

A few of the early lenses figured by Spemann are at least questionable. For example, figure 44 ('12 a) shows on the operated side an irregular thickening within a single cell layer. Such a thickening might readily have resulted from the disturbance to which the portion of ectoderm had been subjected. This specimen is cited as evidence that the accessory or supernumerary lenses in my experiments may have arisen as buds from a common origin. The possibility of early lenses constricting or budding into two or more is freely admitted. Figures 14, 17, 18 and 20 ('10 b) show constricted or double lenses. If the constriction had been carried further two lenses might have resulted. Yet the relative positions occupied by several other of the lenses figured are difficult to account for on the above basis.

Finally, there is no question of the fact that in numbers of the fish monsters which I have figured and described small and ill-formed eye vesicles or fragments are associated with large well-formed lenses. The fact of the constant association of such optic structures with lenses whenever the optic parts chance to lie near the ectoderm makes it practically certain that the defective e3'"es have stimulated the lenses to arise in these positions. The size of the lens in Fundulus is not regulated by the size of the optic cup. This is further proven* by the small lenses in large eyes and by large protruding lenses in rather well-formed but small eyes. The error in logic which Spemann ('12 a, p. 81) claims to exist on page 405 of my 1910 b paper in discussing these eye fragments I am unable to detect and several other embryologists have been unsuccessful in pointing it out.



1 . Experiments in which certain regions are removed by mechanical operations from the medullary plate of Ambly stoma punctatum seem to show that the earliest optic anlage is median in position.

Thirty embryos from which lateral portions of the medullary plate and the anterior lateral part of the medullary fold were removed at slightly different stages gave in twenty-four cases, or in 80 per cent of the individuals, subsequent development of both eyes. In five individuals one eye was absent and in one specimen both eyes failed to arise. The absence of eyes in the latter cases was probably due to the cut having been made in a more median position than was intended.

Nine individuals were operated upon so as to remove narrow strips of cells from the anterior median portion of the medullary plate. Four of these cases, or about 45 per cent of the specimens, failed entirely to develop eyes. According to Spemann and others they should have given some degree of cyclopia. Four other individuals possessed highly defective Qyes, one embryo having one eye poorly formed while the other was questionably present. Only one of the nine specimens so operated upon was capable of developing both eyes to an extent approaching the normal.

2. When the cells in the anterior portion of the open medullary plate are disturbed by being stuck and scraped in various ways with steel needles they do not loose their power of giving rise to optic vesicles and cups which are normal in appearance during the early stages, later stages were not studied.

3. If the optic anlage be cut out of the medullary plate and reversed in position and then transplanted in the medullary plate it still retains the power to give rise to optic vesicles and cups which are abnormal in position to an extent depending upon the distance the anlagen were shifted by the operation.

The facts furnished by these experiments are considered in connection with recent views regarding the genesis of certain ophthalmic defects.



Adami, J. G. 1908 Principles of pathology, vol. 1, p. 241, Lea and Febiger, New York.

DoHRN, A. 1883 Studien zur urgeschichtc des Wirbelthierkorpers III. Die Entstehung und Bedeutung der Hypophysis dei Petromyzon planeri. Mitt. Zool. Stat. Neapel., Bd. 4, pp. 172-189.

King, II. D. 1905 Experimental studies on the eye of the frog. Archiv f. Entw.-¥ech., Bd. 19.

Lewis, VV. H. 1909 The experimental production of cyclopia in the fish embryo (Fundulus heteroclitus). Anat. Rec, vol. 3, pp. 175-181.

Mayer, A. G. 1909 Rhythmical pulsation in Scyphomedusae. Carnegie Institution Pub., no. 102, pp. 113-131.

McClendon, J. F. 1912 An attempt toward the physical chemistry of the production of one-eyed monstrosities. Am. Jour. Physiol., vol. 29, pp. 289-297.

Mencl, E. 1903 Ein Fall von beiderseitiger Augenlinsenausbildung wiirhrend der Abwesenheit von Augenblasen. Arch. f. Antw.-IMech., Bd. 16.

1908 Neue Tatsachen zur SelbstdifTerenzierung der Augenlinse. Arch, f. Entw.-Mech., Bd. 25.

Paolucci, L. 1874 Sopra una forma mostruosa della myliobatis noctula. Atti

della societa Italiana di Sc. Naturali, tom. 17, pp. 60-63. Spemann, H. 1912 a Zur Entwicklung des Wirbeltierauges. Zool. Jahrb., Bd.

32. Abt. f. allg. Zool. u. Physiol., pp. 1-98.

1912 b tJber die Entwicklung umgedrehter Hirnteile bei Amphibien embryonen. Zool. Jahrb., Suppl. 15. (Festschrift fiir J. W. Spengel,

Bd. 3) pp. 1-48. Stockard, C. R. 1907 The artificial production of a single median cyclopean

eye in the fish embryo bj 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, pp. 285-338.

1910 a The influence of alcohol and other anaesthetics on embryonic development. Am. Jour. Anat., vol. 10, pp. 369-392.

1910 b The independent origin and development of the crystalline lens. Am. Jour. Anat., vol. 10, pp. 393-423. Wilder, H. II. 1908 The morphology of cosmobia. Am. Jour. Anat., vol. 8, pp. 355-440.



Sheffield Biological Laboratory, Yale University



The present work was started by the senior author under the supervision of Professor W. R. Coe in the spring 1907, and in 1908 Professor Coe published in Science a brief statement of what had been found. In the summer of 1911 the junior member, Mr. Burr, took up the work. In the interim the literature of the subject had been enriched by three papers, and since then two additional ones have appeared. Lantz in 1910 contributed to a United States government report, on the economic importance of the rat, a short paper on the natural history of the animal. The author describes the different species of rats, their distribution, and general habits, but pays little attention to the details of their reproduction.

Sobotta and Burckhard ('10) made a careful study of the maturation and fertilization of the egg of the albino rat, and they describe and figure the ovarian egg in the stages of the first polar spindle, and first polar body with the second polar spindle, and the tube egg in the stages of the second polar spindle, fertilization, second polar body, and the pronuclei. Ovulation is stated to occur independent of pairing within thirty-six hours after the birth of a litter, and the eggs fertilized nine to twelve hours after copulation. Sobotta and Burckhard found the mature rat egg, in the ovary, to measure in preserved material 0.06 to 0.065 mm. in diameter; practically the same as the



mouse egg. These investigators never saw a definite first polar body associated with an egg in the tube.

Newton Miller's paper ('11) on reproduction in the brown rat is based solely upon observations of the living animals. He found that both sexes become sexually mature at least by the end of the fourth month," that the litters contain from six to nineteen young apiece, and that these animals breed the year, round.

Mark and Long ('12) devote most of their contribution to an extended description of the elaborate warm chamber they have devised for the study of living mammalian eggs. When it comes to the results obtained with their apparatus they have, at present, little to say. Living eggs of rats and mice obtained in a manner similar to that described by one of us (Kirkham '07) were placed on the stage of the microscope in the warm chamber and spermatozoa added, the mouse eggs underwent no change, but the rat eggs within five minutes to two hours began the formation of the second polar cell. Cleavage has never been observed, and after twelve hours the eggs begin to degenerate.

The latest contribution to the literature on the subject of rat breeding is by Helen Dean King ('13) who records for the albino rat somewhat the same phenomena previously observed by Daniels ('10) in mice. The normal period of gestation for the albino rat, according to Miss King, is twenty-one to twentythree days. If six or more young are being carried while a previous litter of five or less are still suckling the period of gestation may be prolonged, while if more than six young are suckling the period is always prolonged, regardless of the number being carried. Unlike the mouse, the albino rat appears not to exhibit any exact relation between the number of young either suckling or borne and the extent of prolongation of the gestation period. This paper also contains evidence that the eggs of a given oestrus cycle in the albino rat may be discharged from the ovaries in two sets, with an interval of two to three days, and also that in very rare instances this interval may be extended to two weeks. Miss King would like to interpret the latter cases as instances of a distinct oestrus cycle occurring during pregnancy.


The present paper has as its object the filling in, as far as possible, of such stages as have not previously been described, and the presentation of evidence regarding the time relations in the development of individual eggs. The authors' thanks are due to Prof. W. R. Coe for the use of the notes and drawings of the eggs of the brown rat, and to Dr. T. B. Osborn of the Connecticut Agricultural Experiment Station for the designs of the cages used and for the animals with which the work was started.


About 150 albino rats were under observation at different times during the investigation. One' large cage was used for all rats not at the time under special care. For individuals two types of cages were employed, one, a cylindrical cage of wire netting of sufficient size to accommodate two rats at a time, and the other a much larger, rectangular cage of galvanized iron, with wire netting only on the front and bottom. This second type of cage was designed primarily as a breeding cage and was large enough to house a mother rat and a litter of the largest size until the latter were sexually mature.

The food of the animals consisted of oats, corn, wheat, sunflower seeds, and dog-biscuit, together with bits of lettuce, string beans, bread, and various kinds of cooked meat and fish.

All cages were kept as clean as possible, but except when absolutely necessary litters less than two weeks old were never disturbed. At the times when we were inspecting them the rats were encouraged to come out of their cages and run about the room, and to this faixdliarity with us as well as to the additional exercise thus secured we attribute much of our success in rearing large litters without their being maimed or eaten by the parents.

Usually females were isolated in breeding cages as soon as they were seen to be pregnant, but in the few instances when males were left with such females until several days after the birth of the litter no mortalitv occurred. This fact leads us to


agree with Miller ('11) and King ('13) that mother rats, unless they are in an unhealthy condition, or have been frightened in some way, rarely if ever kill or maim their young.

Albino rats give birth to young in all seasons of the year, but it is only from April to October that ovulation as a rule occurs within 48 hours after parturition; during the remaining months they are apt to skip oestrus cycles, ovulation not occurring until some three weeks after parturition.

The senior author showed in 1910 that the albino rat o\tilates regardless of whether pairing has previouslv taken place, and when males are continuously present copulation may occur before the ripest eggs in the ovaries have formed the first polar spindles. On several different occasions we have seen the actual pairing. It differs markedly from the condition described by Sobotta ('95) for the mouse, since the male albino rat is not prostrated by the sexual act, but walks slowly away. When a previously isolated female who is in heat is placed in a cage wdth several males they will all pair with her in rapid succession.

The period of gestation in the albino rat is twenty-two days when the female is not nursing a previous litter, in which event the period may be lengthened as found by King ('13). The litters varied in number from four to twelve and the birth usually took place in the late afternoon or the early evening, although probably it may occur at any hour of the day, since we have observed it at noon. The process of parturition is briefly as follows : The female in order to aid in the expulsion of the foetus flattens herself against the bottom of the cage while a series of wave-like muscular movements pass posteriorly along the body starting just behind the shoulder. As soon as the young rat is free from her body, the female rises up on her haunches, seizes in her forepaws the button-like placenta? which is still attached to the offspring by the umbilical cord, and devours first it and then the cord, cutting off the latter as close to the body of the young animal as she can get with her teeth. The female then again flattens herself out against the bottom of the cage preparatory to the appearance of the next young rat. The process is repeated until all have been brought forth. Then, and not


before, does the mother assemble the young, cleaning them up with her tongue, after which they lie close together under her to keep warm. From this time on until the young are able to crawl around by themselves the mother never leaves the nest until she has carefully covered her litter. On returning she always looks around for any that may have rolled or crawled out in her absence, and such offenders are quickly seized in her jaws and hauled back into the nest.

The albino rat becomes sexually mature, at least in some cases, as early as fifty-five days after birth, since in one instance a litter was born to rats that were only seventy-seven days old.


The paper by Sobotta and Burckhard ('10) on the maturation and fertilization of the albino rat is by far the most complete account of the subject that has so far been published. However this report left a number of things to be cleared up.

Working with material from 81 rats we have attempted to investigate and make clear the following points: (1) the early development of the egg previous to the formation of the first polar spindle, (2) the formation of the first polar body, (3) the condition of the egg at o\nilation, (4) the process of fertilization and second polar body formation.

At first the rats were watched and killed at short intervals up to forty-eight hours after pairing. This gave no data that could be depended upon for determining the stage which either the ovarian or the tube eggs had reached. However, by relating the time of killing the female to the time of parturition it was found that tiie approximate development of the egg could be predicted without much difficulty. We say approximate because even though the exact hour of parturition be known it is impossible to say that at a given interval of time the eggs are in a given stage of development.

The parturition of a female caged with a male having been observed, she was killed twenty-four hours later. This female yielded unfertilized tube eggs, indicating that ovulation had re


centh^ occurred. Cases such as this show that ovulation usually occurs about twenty-four hours after parturition. The individual variation is so great that any complicated apparatus for determining the exact date of parturition is valueless. We have obtained the best results by killing the females at half hour intervals, beginning in the later afternoon and continuing through the early evening. By doing this practically all the stages of maturation can be obtained.

In a number of instances the senior author dissected out the Fallopian tubes, and after placing them in warm salt solution, by slitting the tubes he was enabled to obtain two eggs fertilized but unsegmented, three eggs in the two cell stage and three eggs so obscured by follicle cells as to prevent any exact information as to their condition. The technique of this operation is so simple, requiring only a binocular microscope, two needles, some warm physiological salt solution, and a female rat that has given birth to a litter at least twenty-four hours before and not more than five days previously, that we recommend the rat as highly as the mouse for obtaining live mammalian eggs for class demonstration.

In all other instances the ovaries (and also the tubes, wherever ovulation was thought to have occurred) were fixed in either Zenker's 'fluid or in a strong solution of Flemming, imbedded in paraffin, cut serially into sections 0.010 mm, thick and stained in Delafield's haematoxylin. Such sections as were found on subsequent examination to be worthy of detailed study were later decolorized with acid alcohol and restained with Heidenhain's iron-haematoxylin.

A study of the ovaries of the above rats showed that there is a progressive development of the egg until it is ready to leave the ovary at ovulation. The developing eggs of any adult ovary can be readily divided into six groups. The first of these (fig. 1) includes all those eggs that are in the resting condition. These vary considerably in size, as do also their follicles. The earlier stages show a small egg with a follicle consisting of from one to three layers of radially arranged follicle cells with scattered cells lying between the layers, the later stages lie in larger follicles



with many more layers of cells. The egg nucleus presents a constant appearance, a clearly defined nuclear membrane, scattered chromatin and a deeply staining nucleolus.

The second group includes those eggs which differ from those of Group I only in their size and in the fact that they lie in much larger follicles, the latter consisting of a large number of cells closely packed but showing no radial arrangement except in the layer immediately surrounding the egg. Such an egg is shown in figure 7.

The third group includes a much smaller number of eggs which lie in follicles similar to the preceding, except that the cells lying

Fig. 1 Normal resting follicle. X 630

in or near the center of the follicle show a marked tendency to separate, leaving a clear space. This condition may, however, be found in follicles belonging to eggs of Group II, for the factors governing the growth of the follicle are not, according to our observation, constant, since growth may set in when the egg has reached the stage of development included in either Groups II, III or IV (figs. 2, 3 and 4). The nuclei of the eggs of this third group show a marked change. The nuclear membrane is still distinct, but the chromatin is less scattered and the nucleolus has become partially vacuolated, since it shows much less affinity for the stain. Figure 8 shows an egg of this group.

The fourth group shows further modifications. It is at this point that the maximum growth in the size of the follicle takes



Fig. 2 Earliest observed maturation phenomena — increase in size of follicle. X 90.

Fig. 3 Follicle of egg with first polar spindle shown in figure 10 of Plate II. X90.

Fig. 4 Follicle of egg with first polar body and second polar spindle shown in figure 11 of Plate III. X 90.


place. While growth may have started in either of the two preceding groups, the greatest growth occurs with the egg in this stage of development. Eggs of this group have been observed with follicles similar to those of the two preceding groups, and also with follicles of nearly the maximum size. The nuclei of these eggs show a diminution in the amount of chromatin present and a complete vacuolization of the nucleoli, the latter showing no affinity whatever for the stain. Such an egg is shown in figure 9.

The fifth group consists of the eggs with first polar spindles. The follicles here are typical, showing a slight tendency to be thinner in the region where the follicle is nearest to the surface of the ovary. The nucleus of the egg has disappeared, and in its place lies the first polar spindle (fig. 10).

The sixth group shows no change in the size of the follicle. The first polar body has been extruded and a second polar spindle formed (fig. 11).

In all the above divisions, with the exception of the sixth group, wherever a distinct zona radiata can be seen, very fine protoplasmic bridges can readily be distinguished crossing froni the follicle cells to the egg. The presence of these very distinct filamentous processes of the follicle cells seems to have been entirely overlooked by previous investigators.

One striking thing is to be noted with regard to the above divisions — never were all six found together in one ovary at a given time. As was to be expected. Group I, since it included all the resting eggs, was present in all ovaries. Group II, on the other hand, was seen to drop out on the appearance of Group

IV and to reappear on the disappearance of the latter. Group

V also appeared on the disappearance of Group IV. When Group V dropped out, Group VI appeared. Group III was found in all ovaries.

From the fact that perfectly normal eggs of Groups II and III were found in the ovary just at, and also just subsequent to ovulation, it was evident that more than one oestrus cycle was necessary for the development of the egg from the resting stage to the stage of the first polar body and second polar spindle.


at which stage the egg leaves the ovary, for, if the above changes occurred in one oestrus cycle, ail normal eggs in the above condition would go out of the ovary at ovulation, leaving only Group I eggs in the ovary. This condition was not seen. Hence we were forced to find some other explanation of the facts.

Figure 5 shows in the form of a table the facts described above. The vertical readings show the groups of eggs. The horizontal readings show the periods into which the oestrus cycle is divided. Period a is the division of the oestrus cycle extending from ovulation to the twenty-first subsequent day and covers a period of time in which there is little change in the personnel of the ovary. Period h covers the succeeding six hours; period c, the next six, and period d, the last six hours remaining before ovulation. The above figures are only approximate, as the individual variation is too great to permit of any exact data.

By studying the figure it will be seen that Group IV disappears at period c. At the same time the ovary contains Groups I, II, III and V, IV and VI being absent. During the interval between periods c and d Groups I, II and III remain unchanged, but Group V disappears and Group VI appears.

After ovulation we find in period a, Groups I, II and III only. But in period h, II disappears and IV appears.

From the above data we were led to believe that the development of an egg follows the arrows in the diagram. That is, that Group II comes from I in period c, remains unchanged through d and a, becomes transformed into III during period h, remains unchanged through c, d and a, grows to IV in 6, to V in c and to VI in d, and so out at ovulation.

The above explanation of the facts rests on the assumption that the normal rate of development is approximately the same for all eggs. This assumption we think is warranted, for if II developed into IV during period h instead of remaining unchanged until the next oestrus cycle, the number of Group III eggs found should be very small, since the change would be a rapid one. On the other hand, if the development involved a longer period of time — that is, if Group III became a second resting stage— one would expect to find a comparatively large number



I n m ivv VI

Fig. 5 Diagram showing probable development of an egg through ovulation. Roman numerals I-VI indicate successive ment of eggs; a-d indicate periods in oestrus cycle. Arrows course of development of individual eggs.

from resting stage stages in developindicate probable





Shotving the relative number of eggs in the various stages of development at different ■ periods of the oestrus cycle, as found in individual ovaries









Period a

168 145 / 171 \ 167 j 152.2 \ 152





22 24




7 11

4 24


3 14



5 5

3 2


Period b


Period c


Period d

35 120


of such eggs in an ovary at any given time. This, however, v^^as not the case, the number of Group III eggs found being relatively close to the number of Group I eggs.

Table I is compiled from a count of all the follicles in six ovaries, representing each of the four periods. It shows the relative number of eggs in each group present at the same time in a given ovary. The count can only be an approximation, owing to the occasional loss of a critical section and the frequent difficulty in determining with accuracy whether or not an egg was normal, but is sufficiently exact for this purpose.

We were unable to obtain any stages that intervene between the eggs of Group IV and those with the first polar spindle, so we cannot say whether the nuclear membrance disappears before or after the first appearance of the first polar spindle. With regard to this spindle, however, there are a number of details worthy of attention. It is short and broad, with well defined fibers which do not come to a sharp focus (fig. 10). The possibility of centrioles being present was mentioned by Coe ('08), but these are apparently lacking in polar spindles of the albino rat. The chromosomes are numerous, crowded, and never found in a definite equatorial plate. Moat of the first polar spindles seen are parallel to the surface of the egg, and this appears to be the position in which the spindle waits for the stimulus that leads to the formation of the first polar body (fig. 6 a) . When this stimulus comes the spindle rotates on its long axis, coming to lie more or less radially (fig. 6 b and 6 c) .



The next stage we were able to obtain is shown in figure 11. This is an ovarian egg with the first polar body and the second polar spindle. As in the case of the mouse, the nuclear material is never gathered into a resting stage between the time of exti-usion of the first polar body and the formation of the second polar spindle.

The first polar body is rarely seen in eggs outside of the ovary, but there is absolutely no reason to doubt that it is always formed, since it is almost invariably present beside normal ova

Fig. 6. Reconstructions of three spindles showing gradual rotation from the paratangential position (a), through (6), to the radial position (c).

rian eggs, possessing a se'cond polar spindle. Even in the ovary, however, its protoplasm displays its characteristic tendency to uiidergo rapid disintegration. In such fully matured eggs as have failed to escape from the ovary and are just starting to degenerate, as well as in those about to be discharged, the second polar spindle may be sharply defined, yet a careful search fails to reveal a trace of the first polar body. The chromatin in the first polar body is always scattered, and when first formed this


polar body is, in all probability, always larger than the second, though disintegration may set in immediately upon its formation. The second polar spindle as seen in the ovary is much longer and narrower than the first, but resembles the first polar spindle in having open ends and no centrioles. The chromosomes in the second polar spindle are almost always spherical.


The living unsegmented egg of the albino rat measures about 0.079 mm. in diameter (the exact size varies a few thousandths of a millimeter in different specimens), and is surrounded by a zona of transparent jelly about 0.022 mm. in thickness. The two unsegmented rat eggs that were obtained sufficiently free from follicle cells to be available for detailed study, both possessed two polar bodies, measuring in one specimen 0.019 and 0.0132 mm. in diameter respectively, and in the other specimen 0.008 and 0.0065 mm. These eggs while translucent were filled with highly refracting globules scattered through the protoplasm. In one egg there was a clear area near the center, where we thought we could distinguish the two pronuclei lying side by side.

The rare occurrence of the first polar body associated with the egg in the tube is to be attributed to its rapid disintegration, which, as already stated, begins almost as soon as it is formed, and may lead to its complete disappearance before ovulation occurs. A stained and sectioned tube egg, accompanied by the first polar body, is shown in figure 12. This polar body is very small, contains only a little stainable chromatin scattered through it, and its protoplasm is much denser than that of the egg.

Until after fertilization, and if this fails to take place until it degenerates, the chromatin of the second -polar spindle remains in a clearly defined equatorial plate, but in the egg in the Fallopian tubes, this spindle always appears much longer and thinner than in the ovarian eggs.

The rat spermatozoon has an exceedingly long tail (fig. 16 a), and like that of the mouse carries more or less of its tail with it


when it enters the egg, a fact mentioned by Coe, and by Sobotta and Burckhard. As soon as the sperm head begins to penetrate the cytoplasm of the egg the formation of the second polar body is started.

In the albino rat the second polar body is characterized by having the chromatin content massed, while the chromatin of the first polar body is always scattered through the cytoplasm. This distinction, however, does not hold for the Norwegian rat, of which two eggs are shown in figures 17 and 18. The chromatin left in the egg after the formation of the second polar body rounds itself up and becomes surrounded by a membrane, thus forming the female pronucleus. The sperm head on its entrance swells up and likewise assumes a rounded form with a nuclear membrane, as is shown in figure 16.


1. Male albino rats rarely, if ever, are responsible for the killing or maiming of their young. Diseased condition or fright are probably the chief causes of the destruction or injury of their offspring by the females.

2. Albino rats give birth to young the year round, but only from April to October do the females regularly ovulate twenty to forty-eight hours after parturition.

3. Albino rats of both sexes are sexually mature when less than two months old.

4. Living rat eggs are easily obtainable during the four days following ovulation by dissection of the Fallopian tubes.

5.. The maturing eggs in the ovary are joined to the surrounding follicle cells by very definite cell bridges.

6. The development of eggs can be traced in the ovary through two oestrus cycles preceding their discharge.

7. The first polar spindle is short and broad, and is usually formed less than twenty-four hours after parturition.

8. The first polar body is always formed in rat eggs, but its protoplasm is very unstable, and disintegrative processes often bring about its complete disappearance about the time the egg reaches the Fallopian tube.


9. The second polar spindle is long and narrow. Its appearance marks the end of maturation phenomena in the ovary, and the termination of all development of the egg unless fertilization occurs.

10. In albino rats the chromatin of the first polar body is scattered, that of the second polar body is massed.

11. The very long middle piece of the sperm tail follows the head into the cytoplasm of the egg.

June 1913


CoE, W. R. 1908 The maturation of the egg of the rat. Science, N. S., vol.

27, no. 690. Daniel, J. F. 1910 Observations on the period of gestation in white mice.

Jour. Exper. Zool., vol. 9. Donaldson, H. H. 1912 The history and zoological position of the Albino rat.

Jour. Acad. Nat. Sci., Philadelphia, vol. 15, 2d series. King, H. D. 1913 Some anomalies in the gestation of the albino rat (Mus

norvegicus albinus). Biol. Bull., vol. 24. Lantz, D. E. 1910 Natural history of the rat. Treas. Dept., Pub. Health and

Marine Hosp. Service U. S., Washington. Mark, E. L.,and Long, J. A. 1912 Studies on early stages of development in

rats and mice. No. 3. Univ. California Pub. in Zool., vol.9, no. 3. Miller, N. 1911 Reproduction in the brown rat (Mus norvegicus). Amer.

Naturalist, vol. 45. SoBOTTA, J. 1895 Die Befruchtung und Furchung des Eies der Maus. Arch.

f. Mikr. Anat., Bd. 45. SoBOTTA, J., u. BuRCKHARD, G. 1910 Reifuug und Befruchtung des Eies der

weissen Ratte. Anat. Hefte, Bd. 42.


Figures 1 to 4 and 7 to 18 were drawn with the camera lucida. All figures, except 17 and 18 were drawn with Zeiss no. 4 oc. and xV oil immersion obj. giving a magnification of 1000 diameters. Figures 17 and 18 were drawn with a no. 6 oc. and tV oil imm. obj., giving a magnification of 1760 diameters. These figures are reduced one-third, giving a magnification in the finished plate of 1174. All other figures are reproduced at the size drawn.

PLATE 1 explanation of figures

7 Shows an ovarian egg in the resting stage. The egg (Group II) has attained approximately its greatest diameter. Nucleolus solid and deeply staining. Protoplasmic bridges well marked. A follicle cell is shown dividing at^the left.

8 An ovarian egg (Group III) at a slightly later stage showing the vacuolization of the nucleolus well started.








9 A later stage than fig. 8, showing the complete vacuolization of the nucleolus (Group IV).

10 A radial first polar spindle (Group V) showing the blunt ends.








11 An ovarian egg (Group VI) showing the first polar body with a spindle, and the early type of short, thick second polar spindle within the egg. The protoplasmic bridges have at this stage disappeared.

12 A tube egg, showing the first polar body at the right, and the long slender type of second polar spindle at the lower left-hand margin of the egg.









13 A tube egg, showing the second polar body in the process of formation. Th^ enveloping follicle cells still retain their continuity.

14 A tube egg, showing the first polar body at the top, the second polar body in the process of formation at the lower left-hand margin of the egg, and the sperm head at the right, with a portion of the tail.














15 A tube egg, showing the second polar body at the left and the sperm head at the right.

16 A series of drawings showing the changes in the sperm head: a, a spermatozoon in the tube. The head was drawn from a stained section — the tail was added from data of a living spermatozoon, b to g, various stages in the transformation of the sperm head within the egg.









17 A tube egg of a brown rat, showing the second polar body at the top, below the deeply staining female pronucleus, the entrance cone to the right, the male pronucleus with the sperm aster in the lower left-hand margin and the sperm tail extending diagonally through the egg.

18 A somewhat later stage of the- egg of the brown rat, showing in addition to the above-mentioned points the nucleolus vacuolated in the female pronucleus.

Figures 17 and 18 were drawn by Dr. W. R. Coe, and are here published with his kind permission.









Anatomical Laboratory, Bellevue Hospital Medical College, New York City


The embryo which forms the basis of this work was given to me by Dr. Rudolph Boencke in the spring of 1911.' It has been placed in the collection of the Department of Anatomy at the University and Bellevue Hospital Medical College and is called embryo no. 4.

The embryo was aborted two weeks after the last menstrual period. There was no record of coitus. After fixation and with the amnion intact the embryo measured 2.3 mm. in length. It was cut into transverse sections 5 m in thickness, and stained with iron haematoxylin. The embryo yielded 287 sections.

Wax plate reconstructions were made of the complete embryo, the heart, the foregut, also of the caudal part of the medullary tube with the hind-gut and the belly stalk vessels. A graphic reconstruction was made representing the embryo cut in the mid-sagittal plane. All the reconstructions were made at a magnification of 200.

The embryo appears to be normal in every respect and the following points of structure have been determined.


In its general configuration this embryo is very similar to Pfannenstiel III described by Low ('08). The body has a regular dorso-ventral curve and has a slight twist so that the head is situated to the right of the mid-sagittal plane. The yolk sac communicates with the primitive gut by means of an extensive yolk stalk. The latter has its greatest diameter in the cephalo 319


caudal direction and its lateral width is greatest at the cephalic end. Caudal and to the right of the yolk stalk the belly stalk leaves the embryo passing ventrally and curving to the right and caudad. Lateral to the yolk stalk the embryonic coelom has an extensive communication with the extra-embryonic coelom.

The heart produces a prominent bulging of the right side of the bod}^ immediately caudad to the head. The most prominent part of the bulging marks the flexure in the heart tube between the bulbus cordis and the ventricle. The neck flexure has not advanced to any prominent degree. There are two prominences on the dorsal surface of the head region, one at the cephalic end of the mid-brain and the other at the cephalic end of the hindbrain. Caudally the bod}^ curves gradually in a ventral direction. There is no distinct caudal flexure.

The medullary tube is open to the exterior at both ends. The cephalic neuropore exhibits an unusual appearance for an embryo of this age. It is very wide and gives a great breadth to the head when viewed from the ventral aspect. The lateral lips of this neuropore curve dorsally and form the ventral boundary of a deep groove which is directed cephalo-caudally. The caudal end of this groove nuis into the stomodeum. This part of the nervous system which represents the forebrain has not kept apace with the development of the remainder of the tube. It apparently is a persistence of the condition which is present in an earlier stage of development. Eternod's ('95) embryo of eight somites and the embryo of seven somites described by Dandy ('10) exhibit cephalic neuropores which appear to be in about the same stage of development.

There are no indications of otic invaginations. Two pairs of entodermal pouches are in contact with the ectoderm. The points of contact are indicated on the surface by shallow depressions. In figure 2 their positions have been indicated on the surface by broken lines.

The amnion lies close on to the body of the embryo. The head fold crosses the ventral aspect of the heart at about its middle. The lateral folds follow the lateral lips of the coelom. The tail fold is situated on the dorsal aspect of the belly stalk.



The nervous system has not proceeded very far in its differentiation. The brain flexures do not agree with the His models of this stage, but correspond more to the older embryos described by Thompson ('07) and van den Broek ('11). The most distal portion representing the forebrain is still open and is bent almost at right angles to the mid-brain. The long axis of the forebrain lies in a cephalo-caudal plane and almost parallel with the long axis of the hind-brain. The most cephalic point of the nervous system is thus represented by the junction of the forebrain and the mid-brain. Near the caudal extremity of the forebrain there is a thickening together with an evagination of the brain ectoderm. This evagination is almost in contact with the ectoderm of the stomodeum and undoubtedly represents the infundibulum. Cephalad to the infundibulum and about in the middle of the lateral expansions of the cephalic neuropore there is a slight depression of the ectoderm on each side which represents the beginning of the optic vesicles.

The mid-brain is quite extensive as is apparent from an examination of figure 3. Its floor is smooth and exhibits a thickening at the cephalic end. Caudally there is a flexure of the floor between the mid-brain and the hind-brain. The floor of the mesencephalon is thickened at its cephalic end. The trigeminal ganglion is present as a distinct mass of cells. Its position is represented in figure 3 by a broken circle. The hind-brain passes gradually into the spinal cord. A distinct neck flexure is not present.

The medullary tube has its greatest diameter at the cephalic extremity. It diminishes gradually in size caudally. At the caudal neuropore it exhibits a slight enlargement.


The stomodeum is a broad and deep invagination of the ectoderm between the heart bulging and the head. It touches the entoderm of the pharynx and forms with it the beginning of an oral plate. There is no indication of an hypophysis. The ectoderm lining the stomodeum is thickened especially in the roof.


The cephalic extremity of the pharynx projects beyond the oral plate and nearly reaches the floor of the forebrain, a small amount of mesoderm intervening.

The median thyreoid anlage is a very prominent evagination of the entoderm of the floor of the pharynx. It projects between the layers of splanchnic mesoderm at the arterial end of the heart immediately caudad to the endothelial aorta and the first aortic arches. The cephalic wall of the evagination is considerably thicker than the caudal. Cephalad to the thyreoid anlage the first branchial pouches are evaginated from the lateral wall of the pharynx and immediately caudad to the thyreoid the second pair of pouches are present. The first pair of pouches are the larger. Their long axes are directed laterally, cephalad. and slightly dorsal. Opposite the venous opening of the heart the liver anlage is present as a thickening of the gut entoderm. Lung buds have not developed in this stage.

A cross section of the foregut has a crescentic outline with the concavity directed dorsally. The tube is widest at the point where the first pair of branchial pouches is developed. The cephalic part of the foregut is flattened dorso-ventrally. Caudally the dorso-ventral diameter increases gradually to the end of the foregut where it becomes greater than the lateral diamter.

The gut entoderm extending out into the yolk stalk retains its thickness only a short distance (fig. 3).

The hind-gut is shorter than the foregut. Its dorso-ventral diameter is comparatively large while its lateral diameter is small. The allantois is evaginated from the ventral wall. The lumen of the diverticulum is very small at its proximal end, but throughout the rest of its extent it is distinct. At first the allantois lies between the aUantoic arteries. At its distal end it comes to lie between the venous and arterial trunks or sinuses of the belly stalk. The end of the allantois is not recurved as found by Lewis ('12) but ends as a straight tube. The hind-gut exhibits a dorso-ventral constriction immediately cephalad to the allantoic diverticulum. Caudal to the allantois the hind-gut widens out to form the cloaca. The entoderm of the ventral wall



Fig. 1 Wax plate reconstruction of complete embryo seen from left side. The broken lines indicate the points where the entodermal pouches touch the ectoderm. X 100.



Fig. 2 Wax plate reconstruction of complete embryo seen from the ventral aspect. X 100.



Fig. 3 Graphic reconstruction representing tiie embryo cut in the mid-sagittal plane. The brokeTi circle above the letter H represents the position of the trigeminal ganglion. X 100.

A, Atrium Cl.M, Cloacal membrane L, Liver anlage

Al., AUantois F, Forebrain M, Mid-brain

ALA, Allantoic artery FG, Foregut P, Pharynx

Al.V , Allantoic vein //, Hind-brain Th, Thyreoid

BC, Bulbus cordis HG, Hind-gut V, Ventricle

/, Infundibulum

Fig. 4 Wax plate reconstruction of caudal end of the medullary tube and hind-gut with the belly stalk vessels viewed from the side. X 100.

Fig. 5 Wax plate reconstruction of a section of the heart with the endothelial tube in position viewed from the cephalic aspect. X 100.

Fig. 6 Wax plate reconstruction of the heart viewed from the cephalic aspect.

X 100.

Fig. 7 X 100.

Wax plate reconstruction of the heart viewed from the caudal aspect.



of the cloaca is fused with the body ectoderm and forms a thick cloacal membrane. At the most caudal part of the cloaca there is a thickening of the entoderm together with a slight evagination which is suggestive of a post anal gut.


The notochord is about in the same stage of development as the one described in a 2.5 mm. embryo by Kollmann ('90). The notochord is intimately connected with the gut entoderm throughout its length with the exception of the caudal end. The caudal end, or tail bud, is cut off from the entoderm and lies imbedded in the mesoderm between the neural ectoderm and the gut tube. There is no distinct notochordal canal as described by Mall ('91), Eterriod ('99) and Grosser ('13). In places the cells of the notochord are vacuolated and apparently in a stage of developing a canal. The relationship of the notochord to the gut entoderm is a very intimate one. In the region of the mid-gut the notochord is composed of but a single layer of cells which appear to be a modified part of the gut entoderm. Where the notochord is composed of more than a single layer of cells the basal layer is directly continuous with the single layer of cells forming the gut entoderm. It is impossible to give any other interpretation than that the notochord is developed from the gut entoderm. In places the cells of the notochord are arranged in two lateral masses giving the appearance of bilateral symmetr}^ This condition is undoubtedly accounted for by the arched nature of the original notochordal plate. In the subsequent proliferation of cells they would grow laterally and unless there were an especially active growth of cells in the central part a gap would naturally intervene between the two lateral groups of cells. At the cephalic end the notochord has more the appearance of a rod and is almost pinched off from the entoderm. On account of the plane of the sections it is not possible to determine with certainty the cephalic limit of the notochord.

328 ivan e. wallin

mesoder:\ial structure

There are thirteen pairs of mesodermal somites. These are hardly discernible on the surface. The first pair is situated at a level of about midway between the neck flexure and the hindbrain flexure. The last pair is opposite the poiAt where the allantois leaves the hind-gut. A myocoele may be observed in most of the somites. The cells of the somite are arranged in a radial manner.

The pleuro-peritoneal coelom communicates with the extraembryonic coelom on the two sides of the yolk stalk. In its cephalic portion it communicates with the pericardial coelom. The lateral lips bounding the open part of the pleuro-peritoneal coelom have a thickened edge produced by the allantoic veins which 11U1 cephalad in this position.

The septum transversum is present as a single layer composed of the cephalic wall of the yolk stalk fused with the caudal part of the pericardium.

The excretory system is represented by pronephric tubules. The morphological details of these, as far as I have studied them, agree with the description given by Fehx ('12).


The heart tube is composed of three parts, bulbus cordis, ventricle and atrium. The atrium is situated in the mid-line of the body immediately cephalad to the septum transversum. Its greatest diameter is transverse. From the left extremity of the atrium the atrial canal runs to the left, ventral and cephalad to the ventricle. The ventricle pursues a course from the left to right, ventrally and somewhat caudad. At the right extremity of the ventricle the heart tube makes a sharp bend so that its. continuation, the bulbus cordis, comes to lie parallel with the ventricle. The bulbus cordis has a fairly constant size up to its cephalic end where it diminishes slightly. It ends in the mid-line of the body.

The ventral wall of the cephalic end of the bulbus cordis is continuous with the pericardium as is also the case with the


caudal wall of the venous end of the heart (figs. 3, 6, 7). The dorsal wall of the bulbus cordis has a distinct mesentery connecting it with the dorsal pericardium. Near the point where the atrial canal joins the ventricle the ventricle has a mesentery which joins the pericardium at the place where the mesentery of the bulbus cordis joins it. Caudal to the junction of these two mesenteries there is a small space dorsal to the atrium which is free from mesentery and represents the future transverse sinus of the pericardium.

From the dorsal wall of the bulbus cordis a tube-like diverticulum is present (fig. 5). I have been unable to find any references in literature to anything similar to this. The tube runs in the mesentery of the bulbus cordis and at its distal end it comes into close proximity to the ventricle. It is probable that this tube represents a vestige of the space between the two laminae in the closing up of the heart tube and the formation of the mesocardium. Two other tubular spaces of a similar appearance may be seen in the mesentery. They have no communication with the cavity of the myo-epicardium. I observed a similar diverticulum from the bulbus cordis in a 4.06 mm. embryo belonging to the collection of the Department of Anatomy of Syracuse University. It may be noted that this bulbus cordis diverticulum does not contain any endothelium. The endothelial fibrillae, however, appear to extend into it.

The endothelium in no place approximates the walls of the myo-epicardium. The caliber of the endothelial tube varies in the different chambers of the heart, being quite constant in the bulbus cordis, enlarged in the ventricle, and greatly reduced in the atrial canal. In the atrium it widens out into the right and left lateral expansions of the atrium. At its cephalic end the endocardium is continued by the ventral aorta which immediately divides to form the first pair of aortic arches. At the venous end of the heart the most distal part of the endocardium represents the sinus venosus. There is no constriction between the sinus venosus and the atrial part of the endocardium. The endothelial fibrillae which have been observed by various authors


and to which Mall ('12) ascribes the source of the intima may be seen in connection with the endocardium in its entire length.

The blood vessels are collapsed in places so that it is not possible to trace them in their entire extent. The communication between the first pair of aortic arches and the dorsal aortae could not be seen. The dorsal aortae are distinct throughout their course lying dorsal to the gut tube. There is no indication of a second pair of aortic arches. The first pair come off at a point cephalad to the first mesodermal somites. Vitelline vessels containing blood are easily discernible in the wall of the yolk sac and yolk stalk. Vitelline veins run dorsally in the cephalic part of the yolk stalk to gain the caudo-ventral aspect of the sinus venosus opposite the fourth pair of somites. The allantoic veins (fig. 4) begin in the belly stalk as a single trunk or sinus. As the sinus approaches the body of the embryo it bifurcates to fonn the two allantoic veins which diverge and run laterally and cephalad to gain the lateral lips of the coelom. In this -position they run in a cephalad direction to the septum transversum where they enter the caudo-dorsal part of the sinus venosus. The allantoic arteries leave the dorsal aortae at a point opposite the place where the allantois is evaginated from the hind-gut and caudal to the last pair of somites. The arteries run ventrally on either side of the allantois in the belly stalk. At a point more distal than the bifurcation of the allantoic venous trunk the allantoic arteries anastomose to form a single trunk. I have been unable to find any trace of the anterior and posterior cardinal veins. At the cephalo-dorsal aspect of the sinus venosus on the left side there is a short bud-like diverticulum which may represent the future ductus Cuvieri.

I wish to take this opportunity to thank Dr. Boencke for this valuable embryo and Profs. H. D. Senior and F. W. Thyng for assistance and advice in connection with this piece of work.



Van der Broek, A. J. P. 1911 Zur Kasuistik Junger Menschlichen Embryonen. Anatomische Hefte, 44.2, 275-302.

Dandy, W. 1910 A human embryo with seven pairs of somites. Amer. Jour, of Anat., 10, 85-109.

Eternod, a. C. F. 1895 Communication sur un oeuf humain avec embryon excessivement jeune. Arch. Ital. de Biologie, 22

1899 II y a un canal notochordal dans I'embryon humain. Anat. Ana., 16, 131-143.

Felix, W. 1912 The development of the urinogenital organs. Keibel and Mall. Human embryology, Philadelphia.

Grosser, O. 1913 Ein Menschlicher Embryo mit Chordakanal. Anat. Hefte. 47, 653-686.

His, W. 1880 Anatomic Menschliche Embryonen, Leipzig.

KoLLMANN, J. 1890 Die Entwickelung der Chorda dorsalis bei dem Menschen. Anat. Anz., 5, 308-321.

Lewis, F. T. 1912 The early development of the entodermal tract and the formation of its subdivisions. Keibel and Mall, Human Embryology, Philadelphia.

Low, Alexander 1908 Description of a human embryo of 13-14 mesodermal somites. Jour, of Anat. and Physiol., 42, 237-251.

Mall, F. P. 1891 A human embryo 26 days old. Journ. of Morph., 5, 459-480. 1912 On the development of the human heart. Amer. Jour, of Anat., 13, 249-298.

Thompson, P. 1907 Description of a human embrj'o of twenty-three paired somites. Jour, of Anat. and Physiol., 41, 159-171.



Hunterian Laboratory of Experimental Medicine The Johns Hopkins University


It is but natural that neglect of an organ itself should yield a proportional lack of interest in its more detailed structure and even more so, in its less important adjuncts — the blood and nerve supply. Such has been true of the pituitary body.

The recent tremendous stimulus produced by Paulesco's (1) sudden transformation of the hypophysis from a structure of vestigial curiosity to a vitally essential organ, has borne its fruit in the rapid accumulation of co-working histological, (2) experimental (3) (4) (5) (6) and clinical (7) (8) observations. Though still very meager our information is now sufficient to have established a hypophyseal clinical entity, amenable in many cases to medical and surgical treatment.

Forming as it does a link in the chain of internal secreting glands, the hypophysis, essentially of hormone action, must be regulated as other glands in this system, by an autonomic nervous mechanism.

Recent studies from the Hunterian Laboratory (5) by Goetsch, Gushing and Jacobson gave evidences of hypophyseal influence over carbohydrate metabolism. It has been shown that sugar tolerance is dependent upon the functional activity of the posterior lobe of the pituitary body. It was later shown by Dandy and Fitz Simmons (observations unpublished) that a piqure of the hypophyseal region in rabbits produced a heavy glycosuria, therefore giving results similar to a piqure of the so-called Bernard's sugar center in the floor of the fourth ventricle. These results have been amplified by Weed, Gushing and Jacobson (6).




The combination of g;Iandular or hormone activity and the results of mechanical stimuli (presumably of nervous origin) has suggested the possibility of a neuro-hypophyseal sugar center.

The rational interpretation of this and other physiological data has been handicapped by the uncertainty and meager evidence of the regulatory autonomic nervous mechanism. Accordingly at the suggestion of Dr. Gushing under whose direction the experimental hypophyseal investigations have been conducted, the determination of the source and distribution of the nerve supply was undertaken.

Lying as does the hypophysis in such close proximity to the carotid arteries with their abundant superimposed plexus of sympathetic nerve fibers, it is but natural to assume that this is the source of the hypophyseal nerve supply. Indeed evidence of this is found in the infrequent passing reference to a nerve filament which could be traced from this plexus to the hypophysis.


Probably the earliest reference to a hypophyseal nerve supply is the casual mention by Bourgery ('45) that he observed sympathetic nerve fibers passing to the pituitary body. Further substantiation is subsequently given in similar casual mention by Fontona, Cloquet, Bock, Ribbes, (9) and possibly others.

In his Anatomie des Menschen ('79) Henle (9) devotes a paragraph to the hypophyseal nerve supply and supplements this description by a drawing of the carotid sympathetic system, which includes a cluster of two or three twigs running from each plexus to the pituitary body. This is the most extensive description of the hypophyseal nerve supply extant. He casts doubt upon the previous discovery of nerve fibers to this gland and concludes that on account of the inherent difficulties they have mistaken fibrous filaments of connective tissue for nerve filaments, saying, Ohne Zweifel beruhen diese und manche altere Angaben auf Verwechselung fibroser Balkchen mit Nervenfasern, doch zeigte mir das Mikroskop in dem Netzfoniiigen zwischen Carotis und Hypophyse ausgespannten Gewebe feine Nerven


faserbiindelchen dieselben, von denen Luschka sagt, dass sie zwei bis drei jederseits, in den vorderen Lappen der Hypophyse sich einsenken." It is based upon this paragraph and drawing by Henle that an occasional brief mention of hypophyseal nerve supply is found in the more detailed and comprehensive anatom.ies, the majority, however, passing over the matter in silence.

The internal distribution of the hypophyseal nerves was studied by Berkley ('94) (10) in a series of Golgi stained sections. He observed numerous varicose nerve filaments in the interior of the gland, the lobus anterior and pars intermedia in particular, but some also in the posterior lobe. The external connections of the nerves were not studied. On account of his inability to observe nerve cells in the gland, he presumed they were of extraneous origin and thought they probably come from the sympathetic system.


The purpose of this paper is to consider only the relatively grosser aspects, i.e., the origin, course and distribution of the hypophyseal nerve supply. The histological distribution and relation of the ultimate filaments to the gland cells have not been considered. It is analogous in character to a recent publication (11) dealing with the blood supply of this organ.

The difficulties of deductions and the impossibility of an accurate conception of the nerve supply based upon gross human dissection have been shown (Henle) (9) by the supposedly erroneous observations of early investigators in mistaking connective tissue trabeculae for the very delicate nerve filaments, which are almost beyond the range of naked vision. These observations are based upon the canine and feline gland, the animals used in the experimental investigations in the Hunterian Laboratory. The anatomical environment of the pituitary body in these forms is such that the difficulties of a tightly enclosed, deeply imbedded and adherent gland encountered in man and the ape are obviated. The hypophysis dangles from the brain and is readily removed with the brain after liberation of its single point of dural


attachment posteriorly, so that the entering nerves may be studied in their true relations, without tearing or distortion.

We have used almost exclusively" the specific methylene blue intra vitam method of staining the nerves. For the details of this technique we are greatly indebted to the excellent contribution by J. Gordon Wilson (12). Three essentials are necessary for the successful use of this stain: the exsanguination of the tissues must be thorough in order to get a sharply defined picture of the nerves, since the combination of the methylene blue with blood presents a diffuse, indistinct picture with poorly stained nerves ; the nerves must be superficial or covered only by a thin layer of tissue; the air must come in contact with the nerves, otherwise no differentiation takes place.

During the final stages of bleeding the anaesthetized animal from the femoral arteries, a 2^0 per cent isotonic solution of methylene blue "nach Ehrlich" at body temperature was injected into both carotid arteries and continued until the injecting fluid emanated perfectly clear from the femorals. A tourniquet was then applied around the neck below the point of injection under a pressure sufficiently low to insure filling of the cephalic vessels without danger of diffusion or rupture.

On account of the capricious character of this stain, litters of very young puppies or kittens were injected at the same sitting, so that the defects of some might be supplemented by better staining of others. The total nerve supply then is a summation of results, a reconstruction as it were.

After a few minutes to allow penetration of the stain, the skull was opened and a block of tissue, including the hypophysis with its vessels and nerves in their normal relations, was removed from the base of the brain. The hypophysis was gently retracted so as to allow full exposure of one side to the air. The nerves then assume their differential blue. These specimens were immediately studied under the binocular microscope. The study of fixed specimens with post mortem staining was far less satisfactory, because of the collapse of blood vessels, with which the ner\'es are intimately associated, the more stiffened picture, and the deficient maintenance of the blue in the nerve fibers.



The key to the nerve supply of the pituitary body is the arterial supply to this organ. In a recent publication from this laboratory, it was shown (11) that the anterior lobe received an extensive blood supply from a large number of minute vessels, most of which, even when injected, were beyond the range of naked vision. These vessels radiate from the Willisian circle to the hypophyseal stalk like spokes to the hub of a wheel. The majority of these branches are from the anterior and posterior communicating arteries. The network of sympathetic nerves comprising the carotid plexus is continuous along the three main branches which result from its trifurcation. The distribution, however, is very uneven. A few fibers -continue along the anterior and middle cerebral arteries for a short distance but the great majority are found on the two communicating arteries which supply the hypophysis; the posterior communicating artery is particulary well supplied. From these extensions of the carotid plexus numerous filaments are given off and pass along the blood vessels to the stalk of the hypophysis, from which they delve into the substance of the anterior lobe and are lost to view. Some arterial branches have as many as three or even four small filaments, the majority, however, only one or two. The course of the fibers is fairly direct and very few branches are given off. These filaments frequently entwine the vessels but no minute plexuses or anastomoses are visible after leaving the plexus on the main trunks. No nerves have been observed on the external surface of the anterior lobe. All nerves going to the hypophysis are in contact with the sheaths of minute blood vessels. On reaching the stalk it is of course impossible to trace this relation further. Their distribution in the gland has not been observed.



Only by dissection of the hypophysis can the nerve supply of the pars intermedia be traced. By gently separating and retracting the posterior lobe from the clasping mitten-like anterior lobe, it is often possible to trace a single nerve fiber with its branches passing down the stalk and spreading out over the pars intermedia which envelops the posterior lobe (fig. 3).


It has been shown that the posterior lobe is supphed by a median artery which is formed by the confluence of two branches, one from each carotid artery immediately after its entrance into the cavernous sinus. In the canine this vessel enters the posterior lobe at the only point of dural attachment. Vital nerve staining is somewhat more difficult in this region on account of the relatively thicker dural covering which excludes the action of the air and necessitates a delicate dissection of this vessel. For a long time we were unable to find any trace of a nerve entering the posterior lobe. Several branches were always visible at the origin of the vessels from the carotid but the fibers were lost in the dura before the posterior lobe was reached.

However, it was finally possible to demonstrate nerve fibers actually entering the posterior lobe along the artery. Certainly the disparity between the nerve supply to the posterior and anterior lobes is most striking — -in the anterior lobe almost superabundant, in the posterior lobe very few. This contrast may in some measure be due to the difficulties mentioned above; we are however disinclined to lay much emphasis on them.

A most striking color contrast is demonstrated upon removing the hypophysis after vital staining. The anterior lobe is a yellowish white, the posterior a deep indigo blue, possibly due to the (autogenic?) nervous character of the posterior lobe. The blue is of a homogeneous character, no nerve fibers being differentiable under the higher magnifications of the binocular microscope. The intensity of the blue is even much more marked than that of the adjacent, deeply staining oculomotor nerve.



Fig. 1 S(>iiii-(li:igrammatic representation of one side of the cavernous sympathetic system of a canine, showing the nerves passing to the posterior lobe along its artery. Other branches to the dura and a cluster (No) to the N. oculomotorius. The hypophyseal region is viewed from below with dura intact.


This little 'nubbin' resting in a small depression in the floor of the sella, usually enclosed in dura, is present in over 80 per cent of canines, and is evidently a remnant of the embryonic Rathke's pouch. In some adults it may be traced to the pars



Fig. 2 Spmi-diagraminatic reconstruction of sympathetic nerves passing along* the arterioles to the stalk of the hypophysis to supply the anterior lobe and pars intermedia. Note relative dwindling of nerves away from the hypophyseal region. The view is from below with dura, hypophysis and carotid artery removed.

intermedia; it varies greatly in size and histological character. It has an individual blood supply, a small artery given off' b}' each posterior lobe artery. Frequently it has been possible to trace a nerve some distance along this vessel toward this "bod}^" but never have we been able to observe a definite nerve connection.



Fig. 3 Drawing to show the nerve passing from the plexvis surrounding the posterior communicating artery, down the stalk of the hypophysis to the anterior lobe and the pars intermedia which covers the posterior U)be. The anterior lobe has been dissected from the posterior lobe and gently retracted to permit (his view.


During observations on the hypophyseal nerve supply naturally the distribution of the sj^mpathetic filaments were noted in the immediate \'icinity. The dura of the sella region is exceptionally well supplied with filaments from the carotid plexus. Se\'eral branches run from the carotid plexus direct to the oculomotor nerve. A couple of twigs were also observed entering the


optic nerve ; these branches were from the nerves in the adventitia of the anterior cerebral artery. There is thus afforded a direct nervous autonomic path between the optic and oculomotor nerves and between these and the sympathetic trunk.


The ner\e supply to the pituitary body is from the carotid plexus of the sympathetic system. Numerous branches radiate to the stalk along the hypophyseal vessels and are immediately lost to view in the substance of the anterior lobe.

The posterior lobe nerve supply is very scant, in marked contrast to the extensive innervation of the anterior lobe.

The pars intermedia receives its nerves from the stalk.

There is connection between the carotid sympathetic system and the oculomotor and optic nerves.

The absolute differentiation between secretory and vasomotor nerves is of course a matter of much dispute and is impossible. The impression, however, from the character and course of the nerve fibers their greatly increased number in the region of the hypophysis, and their disappearance at a distance from the hypophysis, the dififerences between the supply of the anterior and posterior lobes, the connections established with the other cranial nerves, leads us to regard them as secretory, in contradistinction to vasomotor, the existence of which in the cranial chamber has not been observed.

It is a pleasure to express my gratitude to Dr. Harvey Cushing for his suggestions during the progress of this problem.


(1) Paulesco, N. C. 1908 L'hypophyse du cerveau. Paris, Vigot Freres.

(2) Herring, P. T. 1908 The histological appearance of the mammalian pitu itary body. Quart. Jour. Exper. Physiol., 1, 121-159.

(3) Howell, W. H. 1898 The physiological effects of extracts of hypophysis

cerebri and infundibular body. Jour. Exper. Med., 3, 245-258.

(4) Reford, L. L. and Gushing, H. 1909 Is the pituitary gland essential to

the maintenance of life? Johns Hopkins Hospital Bull., 20, 105-107.

(5) GoETSCH, E., Gushing, H. and Jacobson, G. 1911 Garbohydrate toler ance and the posterior lobe of the hypophysis cerebri. Johns Hopkins Hospital Bull., 22, 165-190.


(6) Weed, L. H., Gushing, H. and Jacobson, C. 1913 Further studies on

the role of the hypophysis in the metabolism of carbohydrates. The automatic control of the pituitary gland. Johns Hopkins Hospital Bull., 24, 33.

(7) Marie, P. 1886 Sur deux cas d'acromegalie; hypertrophie singuliere non

congenitale des extremites superieures, inferieures et c6phalique. Rev. de Med., vi, 297-333.

(8) Gushing, Harvey 1912 The pituitary body and its disorders. Phila delphia.

(9) Henle 1879 Anatomie des Menschen, 3,

(10) Berkley, Henry, J. 1894 The finer anatomy of the infundibular region

of the cerebrum including the pituitary gland. Brain, 17, 575.

(11) Dandy, Walter E. and Goetsch, Emil 1911 The blood supply of the

pituitary body. Amer. Jour. Anat., 9, 137.

(12) Wilson, J. Gordon 1910 Intravitam staining with methylene blue. Ana tomical Record, 4, 267.