Paper - The morphology of the forebrain vesicle in vertebrates

From Embryology
Embryology - 19 Apr 2024    Facebook link Pinterest link Twitter link  Expand to Translate  
Google Translate - select your language from the list shown below (this will open a new external page)

العربية | català | 中文 | 中國傳統的 | français | Deutsche | עִברִית | हिंदी | bahasa Indonesia | italiano | 日本語 | 한국어 | မြန်မာ | Pilipino | Polskie | português | ਪੰਜਾਬੀ ਦੇ | Română | русский | Español | Swahili | Svensk | ไทย | Türkçe | اردو | ייִדיש | Tiếng Việt    These external translations are automated and may not be accurate. (More? About Translations)

Johnston JB. The morphology of the forebrain vesicle in vertebrates. (1909) J. Comp. Neurol. and Psychol. 19: 457-539.

Online Editor  
Mark Hill.jpg
This historic 1909 paper by Johnston describes development of the forebrain in many different species, including human.

See also his later paper: Johnston JB. Further contributions to the study of the evolution of the forebrain V. survey of forebrain morphology. (1923) J. Comp. Neurol. : 143-
Our understanding of forebrain development has improved significantly over the last 100 years, some current research looks at early molecular patterning mechanisms as well as the development of connections between different regions.

See the links below for the current notes pages.

Neural Links: ectoderm | neural | neural crest | ventricular | sensory | Stage 22 | gliogenesis | neural fetal | Medicine Lecture - Neural | Lecture - Ectoderm | Lecture - Neural Crest | Lab - Early Neural | neural abnormalities | folic acid | iodine deficiency | Fetal Alcohol Syndrome | neural postnatal | neural examination | Histology | Historic Neural | Category:Neural
Historic Neural Embryology  
1883 Nervous System | 1893 Brain Structure | 1892 Nervous System Development | 1900 fourth ventricle | 1905 Brain Blood-Vessels | 1909 corpus ponto-bulbare | 1912 nuclei pontis - nucleus arcuatus | 1912 Diencephalon | 1921 Neural Development | 1921 Anencephaly | 1921 Brain Weight | 1921 Brain Vascular System | 1921 Cerebellum | 1922 Brain Plan | 1923 Neural Folds | 1904 Brain and Mind | 1904 Brain Structure | 1909 Forebrain Vesicle | 1922 Hippocampal Fissure | 1923 Forebrain | 1927 Anencephaly | 1934 Anencephaly | 1937 Anencephaly | 1945 Spinal Cord | 1945 cerebral cortex | Santiago Ramón y Cajal | Ziegler Neural Models | Historic Embryology Papers | Historic Disclaimer
Historic Disclaimer - information about historic embryology pages 
Mark Hill.jpg
Pages where the terms "Historic" (textbooks, papers, people, recommendations) appear on this site, and sections within pages where this disclaimer appears, indicate that the content and scientific understanding are specific to the time of publication. This means that while some scientific descriptions are still accurate, the terminology and interpretation of the developmental mechanisms reflect the understanding at the time of original publication and those of the preceding periods, these terms, interpretations and recommendations may not reflect our current scientific understanding.     (More? Embryology History | Historic Embryology Papers)

The Morphology Of The Forebrain Vesicle In Vertebrates


J. B. Johnston.

University Of Minnesota.

With Forty-five figures.


In all classes of vertebrates the forebrain vesicle of the early embryo gives rise to two secondary brain segments, the dienceph_alon and telencephalon. The morphological features of the diencephalon are remarkably constant throughout the vertebrate series, although this segment is the most peculiar and irregular in form. There is always a membraneous roof, tela chorioidea and one or two epiphyses. The dorsal border of‘ the lateral wall presents a paired enlargement known as the ganglion habenulae, to which comes from the telencephalon a tract of fibers which often appears as a gross featurethe stria medullaris. The massive lateral walls bound a third ventricle which is narrow from side to side. In these lateral walls develop a considerable number of special centers of which the lateral genieulate bodies are recognizable microscopically in fishes. The ventral wall of the diencephalon is depressed and variously shaped according to the degree of development of the olfactory and gustatory organs. From in front the optic tracts enter through the base and run up in the lateral walls. Behind the chiasma is found always in the embryo and usually in the adult (fishes, amphibians, reptiles) a transverse groove or depression known as the recessus postopticus. Behind this is a wider depression which in mammals is provided with thick walls, the tuber cinereum, but in fishes and amphibians is a broad bilobed structure with thinner walls, concave within, known as the Zobi inferiores. Between these and continuing backward and downward from them is a thin walled sac, the sacicus vasculosus. This forms the neural part of the hypophysis in mammals and projects more or less directly downward from the tuber cinereum. The connection of the sac with the tuber cinereum is the infundibulum and its cavity may be called the infundibular cavity or recess. The posterior part of the ventral wall forms a second bilobed structure, the corpora mammillaria.

The telencephalon shows greater changes of form and size in different vertebrates than any other segment of the brain. At the same time the essential morphological relations are completely preserved throughout the series from cyclostomes to man. It is part of the purpose of this paper to make this fact more clear and explicit with regard to certain features of the forebrain, but those features upon which there are no difierences of opinion may first be sketched. The telencephalon is very deeply bilobed—bifurcated—in front. Each forward prolongation receives an olfactory nerve and is known as the olfactory bulb. This may be closely applied to the body of the telencephalon as in cyclostomes and amphibians, or it may be removed. by a longer or shorter distance as in most other classes. In the latter case there is an obvious olfactory tract. The body of the telencephalon is always bilobed, the lateral halves being joined. by thinner portions which are usually membraneous- or thickened only by nerve fibers. Only in selachians is there a considerable thickening of the median part of the roof by gray matter. The ventricle which continues forward from that of the diencephalon is always bifurcated, each lateral division extending into the olfactory bulb (except in some adult mammals where it becomes obliterated during development). A membraneous roof continuous with the tela chorioidea of the third ventricle extends over the whole length of the telencephalon except the olfactory tracts and bulbs. In all classes the lateral halves of the telencephalon are connected at the rostral end by the lamina terminalis, a membrane formed .by.the closing of the anterior neuropore of the embryo. This lamina is always thickened by transverse fibers of two or more systems and sometimes divided into two bundles, the commissura anterior. At the lower border of the lamina terminalis is a prominent depression which lies just in front of the optic chiasma and is therefore called the recessus prwopticus. In the walls of the lateral ventricles are found the secondary olfactory centers and the corpora striata. To these in higher forms are added more complex cortical structures which in mammals become the predominant part of the telencephalon.

The above mentioned general features of the forebrain are universally recognized and require no further comment. Other features remain less well understood or their interpretation is in dispute so that they present problems for solution. These problems may be stated here, and will be discussed after the description of certain features in the embryonic and adult brain which contribute to their solution.

1. The anterior end of the brain. It is generally recognized that the lamina terminalis represents the seam of closure of the anterior neuropore, but the location in the adult brain of the lower border of the neuropore is placed by different authors at various points between the anterior oommissure and the infundibulum. The determination of the anterior end of the brain is of fundamental importance for fixing the segmental order of the parts of the forebrain along the brain axis. Thus, if the anterior end of the brain is at the infundibulum, the optic portion of the hypothalamus together with all the telencephalon would be dorsal structures, while if the anterior end lies at the preoptic recess the optic part of the hypothalamus must be considered as belonging to the ventral portion of the brain. In the latter case the olfactory centers would belong to the first segment, in the former they would fall behind the optic chiasma around the dorsal margin.

2. The boundary between the diencephalon and telencephalon. This problem is closely bound up with the first one. Although His placed the anterior end of the base of the adult brain behind the infundibulum and included the pars optica hypothalami in the telencephalon, most authors have referred the pars optica to the diencephalon and this is the obvious meaning of the tables of the BNA. In the roof of the brain of true fishes and amphibians and in the brains of embryos of most vertebrates occurs a well marked transverse fold of the membraneous roof, the velum. trams-versum. This is considered as marking the boundary between di- and telencephalon in the roof. In cyclostomes this is only slightly marked and has been clearly described recently by Sterzi. In mammals the velum has not been recognized hitherto and therefore the boundary line in question is wholly uncertain in mammals.

3. The problem of the membraneous roof and its relations to the massive walls and the nervous pallium. This question naturally presented itself to the earlier neurologists because of the obvious great differences between the mammalian or human brain in which the ventricles were apparently wholly covered by massive nervous structures and that of lower forms, especially teleostean fishes, in which a relatively great expanse of membraneous roof» covers a broad ventricle in the telencephalon. A discussion of this problem naturally involves consideration of the next.

4. The problem of the median and lateral ventricles in the telencephalon and the interventricular foramina. Does the telencephalon possess a median ventricle? One interpretation of the mammalian brain would assign the whole median ventricle in front of the posterior commissure to the diencephalon (third ventricle). The walls bounding this ventricle then belong to the diencephalon and the boundary between di- and telencephalon is assumed to be the interventricular foramina. If, however, the pars optica hypothalami and lamina terminalis belong to the telencephalon, as was held by His, they must carry with them into the telencephalon the adjacent front part of the IIId ventricle. In the brain of a fish or even of an amphibian there appears at first sight to be a large median ventricle belonging to the telencephalon and it is more or less difficult to fix the location of the foramina leading into the lateral ventricles. In the teleostean and ganoid brain only the olfactory tracts and bulbs appear to possess distinct cavities which could be called lateral ventricles and the whole of the olfactory centers and corpora striata are found in the lateral walls of the median ventricle. There is, therefore, need to determine whether a part of the median ventricle belongs to the telencephalon and to fix for the different classes of vertebrates the homologue of the interventricular foramina.

5. Subdivision of the di- and telencephalon with reference to the longitudinal zones laid down by His in other segments of the brain.

The determination of the anterior end of the brain will fix the extent of the floor plate and roof plate of His and will show the point at which the prolongation of the sulcus limitans must end. In the midbrain, hindbrain and spinal cord the sulcus limitans divides the lateral walls into dorsal or alar plate and ventral or basal plate. In the di— and telencephalon this sulcus can scarcely be distinguished with certainty, but the fact that zones of widely different functions are separated by this sulcus when present leads one to seek the equivalent of these zones and the position of the sulcus in the di— and telencephalon.

Other problems concerning the internal morphology and genetic history of the "telencephalon—the origin of the pallium in relation to the primary centers, the evolution of functional localization, etc.,will be discussed in the present paper only so far as necessary in connection with the subject of nomenclature.

I. Descriptive Part

In this part will be brought together under each class of vertebrates the data which bear upon these problems. As will be seen, the new facts to be brought forward are chiefly embryological (selachians, amphibians, mammals). The summaries of facts which have been previously published will be made as brief as possible and the reader is referred to the original papers for more complete accounts.

1. Cyclostomes

In cyclostomes is found a completely bilobed telencephalon, almost as strongly marked as in any class of vertebrates. The reason for taking especial account of this fact in the telencephalon is that the division into symmetrical lateral halves which is present in all parts of the central nervous system is more pronounced in the telencephalon and in mammals and man becomes one of the most striking features of the whole brain. The cause for the decided division of the telencephalon is to be found without doubt in the presence of paired olfactory organs and the importance of the sense of smell. In amphioxus the front end" of the brain is not bilobed because there is no localized olfactory epithelium. In cyclostomes the olfactory organ is paired in the embryo and the mode of development of the mouth, hypophysis and olfactory pit shows that the paired definite olfactory organs are older than the circular mouth. These organs have, therefore, exerted an influence on the brain for a long time previous to the period when cyclostome characters became permanently fixed, and the result is the bilobed telencephalon. This is seen in the presence of massive lateral walls which are connected above, below and in front by relatively thin membranes which are thickened only by transverse fibers and nowhere contain gray matter. It is seen in the forward projection of the olfactory bulbs into which the paired olfactory nerves enter. The olfactory bulbs are sessile; there is no extended olfactory tract or peduncle. It is seen further in the lateral extensions of the ventricle to which attention has been called by Studnicka and the writer (1906). These are rather wide cavities extending into the olfactory bulbs and connected with the median ventricle by narrower openings, the intervenwicular fommina. Owing to the pushing backward of the telencephalon in cyclostomes the lateral ventricle presents a posterior prolongation similar to but not homologous with the posterior horn of the ventricle in mammals. The tela chorioidea of the HM ventricle forms in front of the nuclei habenulae a broad low dorsal sac whose roof in some species is depressed by the epiphysial bodies lying over it. In front of this is a slight transverse fold (better seen in embryos than in the adult) first described by Sterzi (1907) which this author homologizes with the velum tmnsversum of other vertebrates. This velum is continued laterally around the brain by a constriction (groove) without and a fold (ridge) within which marks the boundary between the diencephalon and telencephalon. In front of the velum transversum the roof continues for some distance as a thin membrane, the lamina supraneuroporica of Sterzi, and then is thickened by the fibers of the superior olfactory decussation. According to Sterzi, the point corresponding to the neuroporic recess of other vertebrates is at the front or lower border of this decussation.

The conditions in cyclostome embryos (solid nerve cord, solid connection of brain with ectoderm instead of open neuropore, compactness of structure) are not favorable for the study of the anterior end of the brain, and the discussions of V. Kupffer, Koltzolf and others do not seem to the writer to offer anything of value in the present state of our knowledge. (See note, p. 535.)

It should be noted that the ridge in the brain floor which contains the optic chiasma and other decussating tracts is especially high and prominent in cyclostomes and that the pit behind the ridge, recessus postopticus, is quite as deep and sharply marked as that in front of the ridge, recessus p/roeopticus (commonly called recessus options).

2. Selachiams

Notes on Head Morphology

In the adult selachian brain the diencephalon presents no peculiarities of especial importance to our problems. The limit between the diencephalon and telencephalon is clearly marked dorsally by a deep velum transversum which forms the anterior wall of the dorsal sac belonging to the diencephalon, and the posterior wall of the paraphysis belonging to the teleneephalon (fig. 1). In front of the paraphysis the membraneous roof is complexly infolded to form the plexus chorioideus of the telencephalon. The anterior limb of the plexus is attached to the massive nervous wall which overarches the front part of the median ventricle. In the more primitive selachians such as Heptanchus and in Chimaera the massive roof is smaller and the membraneous roof extends farther forward. As shown in fig. 1, the massive roof is pierced from the dorsocephalic surface by a vascular canal which reaches nearly to the ventricle. This has been interpreted (Johnston, 1906) as the remnant of a deep groove which separated the lateral lobes of the telencephalon earlier in the phylogeny of selachians, and the process by which the present form was reached is indicated by the accompanying diagram (fig. 2, B).

Fig. 1. The mesial surface of the right half of the brain of Mustelus. From Johnston, 1906. For the significance of the abbreviations used on all of the figures, see the list at the end of this article.

Fig. 2. A, an outline of the brain and brain ventricles of Mustelus as seen from above. B, a diagram of one side of the fore-brain to show the primitive relations of the wall and ventricle. From Johnston, 1906.

That portion of the massive roof which lies behind the vascular canal is formed by the great growth of the lateral walls and it has for its basis a fiber—decussation which is comparable in position and in a part of its fiber components to the superior olfactory decussation of Petromyzon. In selachians, then, owing to the enormous development of the olfactory centers, that part of the membraneous roof of the telencephalon which contains fibers in cyclostomes is invaded by gray matter as well.

The form of the ventricle is shown in fig. 2, and it is necessary only to point out that quite distinct lateral ventricles are present, that they extend into the olfactory bulbs» and that they are connected with the median ventricle by the interventricular foramina. A triangular prolongation of the median ventricle projects a short distance in front of the interventricular foramina to meet the vascular canal mentioned above. This is the recessus neuroporicus.

The solution of the problems of the anterior end of the brain and of the boundary between diencephalon and telencephalon requires the study of the development. Although all the essential facts were first acquired from the study of amphibian embryos (see below), it was thought best to extend the study to other vertebrates, and as the selachians present the most primitive conditions the description of certain important processes hitherto overlooked will be given first as they are seen in embryos of Squalus acanthias.

For the material for this study I am deeply indebted to Dr. H. V. Neal. When I wrote to Dr. Neal of my findings in Amblystoma and of the desirability of a control study upon selachian embryos he generously sent me from his magnificent collection of mounted preparations of Squalus embryos, the specimens representing the stages from B to M of B-alfour’s notation. These preparations include whole mounts and sections in all three planes prepared by various stains. These are the preparations upon which was based Neal’ s very valuable paper of 1898 on the segmentation of the nervous system. My attention has been directed chiefly to stages from F (15 somites) to K and M. It has been a great pleasure to verify the clear and accurate descriptions given by Neal of these embryos. I have studied the brain, cranial ganglia, the mesectoderm derived from the neural crest, the mesodermic somites, the branchial clefts and arches, and have given especial care to the region about the forebrain and the premandibular and anterior head cavities. I have made plate reconstructions from frontal and sagittal sections of several stages and have had numerous camera drawings made to illustrate points not shown in the models.

I wish to comment briefly, on the basis of these preparations, upon certain problems that have been under discussion in past years which have some bearing on the special problems being considered here.

Upon the question of the value of the pre-otic “head cavities” as dorsal somites, Neal’s statements seem to me very careful and conservative. With regard to the difierentiation of sclerotome and myotome, the preparations seem to me to warrant a positive statement that a sclerotome is clearly differentiated in the second or mandibular somite as well as in the third. I am in perfect agreement with Neal’s interpretation of the preotic head cavities, including the anterior cavity of Platt as dorsal somites.

Upon the question of a lost branchiomere between the mandibular and hyoid arches, which I have discussed elsewhere (1905) I have been surprised to find so little evidence in the preparations of Squalus. While the third somite lies over the hyomandibular cleft and is somewhat constricted or dumb-bell shaped, its two parts enclose a single cavity and the somite extends over the hyoid arch with the mesodermic core of which it is connected. The fourth somite is smaller than the third and shows that it is rudimentary. While it reaches forward to the hyoid arch, it is connected with the mesoderm of the first branchial arch. The fifth somite has practically corresponding relations to the second branchial cleft and the following (second) branchial arch. The sixth somite is clearly connected with the third branchial arch and the seventh somite with the fourth branchial arch. These relations, indeed, may be clearly inferred from 1\Teal’s admirable drawings from cleared specimens, figs. 15, 16, 17. These correspond to the stages from which I have drawn the conclusions stated above. The posterior ends of somites (2) , 3, 4. and 5 are connected with their respective branchial arches.

I must here make reference to a series of papers on the head problem coming from Prof. H. E. Ziegler’s laboratory in Jena (Klinkhardt 1905, Guthke 1906, Ziegler T1908, Brohmer 1909). These authors discuss the fundamental questions of segmentation and the relationship of the cranial nerves. The total material on which the descriptions and conclusions are based consists of ten selachian embryos including four of Torpedo between the stage I-K.and 20 mm. ; four of Spinax in the stages K, L, M-N, and 7.78 mm. ; one of Chlamydoselachus in the stage L-M; and one of Abanthias 22 mm. in length. 4.68 7ourmzl of Comparative Neurolog'y and Pxyc/aology.

One new idea is introduced into the subject of head morpholog , namely that in the mesoderm the two structures in each segment heretofore known as the somite and the branchial arch together constitute the somite, and the junction of the branchial arch with the pericardial cavity is to be compared with the junction of the trunk somite with the peritoneal cavity. This conception is used in support of the view that mesomerism and branchiomerism coincide. It may be questioned whether the new conception is not more in need of support than that for whose support it was called in. The material which these authors have studied is far from adequate for the study of the primitive segmentation of the head or the morphology of the cranial nerves. This is especially noticeable in*their statements regarding the anterior head cavity, which could not be studied in the material on which they worked, and the other head somites, whose development is well advanced in the earliest stage represented in their material. In determining the first segment of the head no value is attached to the first two brain vesicles, the eye, the olfactory nerve, the nervus terminalis, the rudimentary nervus thalamicus, or the epiphyses. The first segment is that of the premandibular somite and the ophthalmicus profundus (to which the name Ciliarganglion is given). This nerve is related to the brain behind the cerebellum in the embryos studied by Ziegler and his colleagues, but whether they would include all the brain in front of the cerebellum in the first or premandibular segment is not stated. Most of the processes upon which an intelligent judgment regarding the primitive segmentation of the head can be based have been completed in selachian embryos prior to the earliest stage studied by these authors,

With regard to the mesectoderm derived from the neural crest which extends down into the mandibular and other visceral arches from the cranial ganglia, I can fully confirm the statements made by Neal and illustrated in his Plate 3 (1898). To these statements one addition requires to be made which is at the same time an addition to the early history of the anterior head cavity and preoral entoderm. The details of the earliest appearance of the anterior head cavity have not been' given by Miss Platt or by Neal, who was concerned chiefiy with neural segmentation, and Dohrn’s 1904) treatment of the anterior head cavity is unsatisfactory.

Hoffmann’s description (1896) is more complete, but he failed to recognize the source of the mesectoderm beneath the terminal ridge. What he describes as a median mass between the lower border of the neuropore and the “infundibulum” connecting the anterior head cavities with one another is the mesectoderm derived from the terminal part of the neural crest to be described below.

Fig. 3. Squalus acanthias, 17 somites, median sagittal section. The notochord is marked by cross lines and the undifferentiated median mass and preoral entoderm by dots.

Fig. 4. Detail of a part of the section drawn in fig 3. X 100.

In embryos of 15 somites the notochord is ‘continuous anteriorly with a thickened mass lying over the anterior end of the archenteron and separated from the entoderm. Laterally the mandibular somite is connected with both the notochord and the median undifferentiated mass. The archenteron is drawn out anteriorly to a very slender pointed cavity (figs. 3, 4, 5) which ends beneath a depression in the floor of the brain usually called by‘ authors the infundibulum. As this is not the infundibulum I shall for the present refer to this part of the brain as the “infundibulum,” using quotation marks. Following a series of transverse sections forward one sees that as the narrow archenteron disappears, the proper entodermal wall fuses with the overlying median mass and thence continues forward beneath the “infundibulum” as a solid mass of preoral entoderm. At this time the neuropore and dorsal seam of the neural tube are not yet closed and the preoral entoderm extends up to the point where the neural plate passes over into ectoderm. These relations are properly shown in Neal’s figs. C and D. For some time the anterior part of the median mass and the preoral entoderm cause a conspicuous folding upward of the neural plate in the middle line so that a prominent keel appears in the floor of the neural tube (fig. 7)

Fig. 5. Squalus ac., 15 somites, medial view of a model of the right half of the head. ** part of neural tube which is still open. X 25.

Fig. 6. Squalus ac., Balfour's stage D, sagittal section of anterior end. x 75.

In embryos of 17 somites the preoral -entoderm comes into slight but actual contact with the bridge between ectoderm and brain at the lower border of the neuropore. This is indeed the condition which exists from the time the head fold begins to form (fig. 6).

Fig. 7. Squalus ac., 18 somites, frontal section through the premandlbular somite and median mass. The right side of the section is a little further dorsal than the left and the cavity of the somite show only on the right. X 150.

In embryos of 18 somites the cavity of the premandibular somite (som 1, v. Wijhe) is first seen in the lateral part of the undifierentiated mass above described (fig. 7). The median part of this mass maintains the relations just described.

In embryo of 19 somites the mesoderm and preoral entoderm show no change but at very noteworthy structure has appeared between ectoderm and brain, surrounding the anterior end of the preoral entoderm. Paired sheets or flaps of cells migrating from the ectoderm at the lips of the neuropore extend in between the ectoderm and brain and surround the anterior part of the preoral entoderm on its under or ectal surface. These sheets of cells resemble in every way the neural crest in other parts of the embryo and should be considered as the terminal part of the neural crests.

In embryos of 24 somites the premandibular cavities are somewhat larger and especially longer. The bodies of these somites are in contact with the mandibular somite, are still continuous with the median mass and converge forward and downward like the arms of a letter V to fuse completely with the median mass. In the lateral part of the median mass directly forward from the premandibular somites and behind the “infundibulum” are found the first indications of the anterior head cavities. Many mitoses appear here, especially in the front wall of this cavity. The preoral entoderm continues forward beneath the “infundibulum.” As compared with earlier stages, before the formation of the anterior cavities the preoral entoderm is very slender and at one point beneath the “infundibulum” it is completely obliterated. In some embryos it is impossible to distinguish any preoral entoderm beneath the brain in front of the “infundibulum,” but in others there is no doubt whatever that a slender rod of cells lying in this position is the remnant of the anterior p-art of the preoral entoderm. This rod of cells is surrounded by cells (mesectoderm) derived from the neural crest above described, which constitute by far the largest part of the cells lying in this position. This mesectoderm forms a considerable mass of cells filling a lens-shaped space between brain and ectoderm and between “infundibulum” and neuropore and also extends as a sheet along the sides of the “infundibulum” beneath the optic vesicle. fig. 9 is a copy of Neal’s fig. 11, Pl. 3, of this stage, with the addition of this mesectoderm.

Fig. 8. Squalus ac., 22 somites, frontal section through the terminal neural crest. X 150.

Fig. 9. Nea1’s fig. 11 modified by addition of the terminal part of the neural crest. The figure represents the lateral view of a cleared embryo of Squalus of 24-25 somite and shows the extent of the mesectoderm derived from the neural crest.

In one embryo of 26 somites (fig. 10) I have found the slender rod of preoral entoderm persisting beneath the “infundibulum,” but from this time on it is impossible to recognize entoderm in this position. The “infundibulum” has become depressed until it is in contact with the ectoderm and at the same time, whether because of pressure from the “infundibulum” or not, the median part of the preoral entoderm becomes obliterated while its lateral portions form themselves into the walls of the anterior head cavities. The mesenchymatous cells beneath the front end of the brain are in overwhelming proportion of ectodermal (neural crest) origin.

In Dohrn’s beautiful plates illustrating his articles on the mandibular and premandibular cavities, this terminal neural crest is clearly shown. See especially Pl. 9, figs. 7-12. Dohrn does not distinguish between premandibular mesoderm and the mesectoderm of neural crest origin. The latter he calls the anterior part of the premandibular mesoderm (Praem. 1) after the “infundibulum” has pressed down to meet the ectoderm. In fig. 8 the origin of this from the neuropore is strongly suggested, especially as the spot marked neurop. is at the upper border of the lamina terminalis, some distance dorsal to the extreme point to which the preoral entoderm or premandibular mesoderm ever reaches. In Dr. Neal’s preparations which I have studied the entoderm has a different tone from the other tissues and there is a decided difference in the form of the cells and in the size of the nuclei between the premandibular mesoderm and the terminal neural crest mesectoderm.

Fig. 10. Squalus $10., 26 somites, frontal section. mf., the so-called infundibulum.

In embryos of 29 and 30 somites (figs. 11 and 12) premandibular and anterior head cavities are small, separate from each other, but both connected with the median mass which is crowded behind the “infundibulum.” The median band connecting the anterior cavities is small and disappears at this stage. The premandibular somites retain the connection with the median mass in which cavities appear in later stages and finally fuse with the cavities of the premandibular somites as fully described by Neal and others. The anterior cavities extend forward at the sides of the “infundibular” region in the position in which they were first described by Miss Platt. Except where they are in contact with the premandibular somite the anterior somites are almost completely ensheathed by the mesectoderm derived from the region of the neuropore. The mesectoderm sheet derived from the thalamic nerve rudiment has now come down behind the eye and met with this which lies beneath the eye. From this stage onward Neal’s figures 12, 13, etc., show the distribution of neural crest mesectoderm almost correctly. The inesectoderm lying in the region between the optic vesicle and somite 1 should be continued as a thin sheet between the brain and ectoderm beneath the neuropore and the origin of all this part from the terminal part of the neural crest should be taken into account.

Fig. 11. Squalus ac., 30 somites, frontal section through premandibular and anterior head cavities. X 150.

Fig. 12. Squalus ac., 29 somites, frontal section through the premandibular and anterior head cavities. These cavities are slightly more advanced in development than in the 30 somite embryo shown in fig. 11. X 150.

Fig. 13. Squalus ac., 39 somites, frontal section through the anterior head cavities, showing their relation to the premandibular somite (left side) and the mesectoderm derived from the terminal neural crest. The connection between n. tlialamicus and terminal neural crest shows on the right side. X 150.

The relations of the anterior head cavities to somite 1 and to the mesectoderm remain essentially unchanged up to embryos of 50 somites. figs. 12, 11, 13, 14 show these relations in frontal sections of embryos with 29, 30 and 39 somites. The brain, mesoderm and mesectoderm with the cranial ganglia have been modelled in an embryo with 42 somites and fig. 15 shows these relations as seen in the model.

Fig. 14. Section from same series as fig. 13, passing through the tip of the anterior head cavity on the left side and in front of this cavity on the right side. X 150.

Fig. 15. Squalus ac., 42 somites. Left view of a model of the brain, mesoderm, ganglia and mesectoderm derived from the neural crest. The auditory pit with the ectoderm bordering it are also shown. )( 50. The boundaries of the mesoderm are marked in black lines. 3, 1;, 5, somites of Van Wi;|he’s series.

Development of the Forebrain

a. The Optic Vesicle and the Primitive Optic Groove

The early appearance of the optic vesicles on the open neural plate has been described by Locy (1895), and the writer has elsewhere (1905, 1906, 1909) brought together the evidence that the optic vesicle is derived from the alar plate of the embryonic neural tube. The study of its relations for the present purpose begins with embryos of about 15 somites. The medial surface of the right half of a model of the head of such an embryo is shown in fig. 5. The neural tube has a cephalic flexure and is open at the neuropore and for a short distance along the dorsal surface. The lateral wall presents two concavities, a broad shallow one for the midbrain and a deeper one for the forebrain. The depth of this concavity is due chiefly to the formation of the optic vesicle, which seems to involve the greater part of the lateral wall of the forebrain vesicle. In the floor of the brain the optic vesicles of the two sides are connected with one another by a transverse groove or depression which has heretofore been called the “infundibulum.” That it does not become the infundibulum will be shown farther on. Its relation to the optic vesicles and its later history suggest that it be called the primitive optic groove. Its relations to the preoral entoderm as above described and to the neuropore as seen in figs. 3, 4 and 5 show that this groove is in the floor-plate of the brain some distance behind the anterior end of the entoderm.

In the embryos with 17 somites, figs. 3 and 4, the neuropore is closed except at a single point which shows as a thin place in one section of 62/3 microns thickness. The distance from the primitive optic groove to this point is greater than the distance to the lower border of the neuropore in the embryo of 15 somites. The arrange ment of the cells in the lower lip of the closing neuropore as seen in transverse and frontal sections shows that there is a process of fusion of the lips of the neuropore from below upward and therefore the last point of the neuropore to remain open is a point in the dorsal seam of the neural tube some distance removed from the anterior border of the neural plate. 'This part of the seam of closure which represents the length of the neuropore is what is called the lamina terminalis. This is not very long in the embryo of 17 somites, but grows distinctly in length as the forebrain expands. The apparent great thickness of the lamina terminalis in fig. 3 is of course due to the persistence of the fusion of ectoderm and brain wall at this time. The primitive optic groove is deeper and sharper in outline than in the last stage.

The relations in an embryo of 24 somites are shown in fig. 16 representing the medial View of a model of the right half of the head. In this is seen the depression of the primitive optic groove until its wall comes into contact with the ectoderm and cuts off the front part of the preoral entoderm as before described. The lensshaped space in front of the primitive optic groove is filled chiefly by mesectoderm from the terminal part of the neural crest. The lamina terminalis is marked by a small triangular pit above, the neuroporic recess, and by a rounded shallow pit below. This pit occupies the lower portion of the neuropore space and may therefore be called the terminal pit. The optic vesicle is now stalked and the cavity of the stalk is seen near the base of the brain. It is noticeable that this cavity no longer communicates directly with the primitive optic groove. It seems equally closely related to the terminal pit. Between the two pits the brain wall is somewhat thickened.

fiG. 16. Squalus ac, 24 somites, medial aspect of a model of the right half of the head. )< 75.

This thickening corresponds to the anterior border of the neural plate and‘ the lower border of the neuropore. It is the terminal ridge of the early embryo with openyneural plate, figs. 3, 4, 5, 6. From this terminal thickening a ridge is seen in fig. 16 running obliquely caudo-laterad behind the optic vesicle and cutting it off from communication with the primitive optic groove. This is the beginning of a change in the relations of the optic vesicle which the following stages show completed.

An embryo of 42 somites is shown in fig. 17 drawn from a model of the right half of the head and a corresponding model of an embryo of Balfour’s stage K is shown in fig. 18. In these it is clear that the optic vesicle is no longer connected with the primitive optic groove, but the two vesicles are now connected with one another by the terminal pit at the lower border of the lamina terminalis. This is the pit which remains connected with the hollow optic stalk as long as that persists and is known in the adult as the recessus opticus (His) ; better called recessus prceopticus.

fiG. 17. Squalus ac., 42 somites, medial aspect of a model of the right half of the head. This is the same embryo as the one shown in fig. 15.

In the latter part of Balfour’s stage K the optic tract fibers begin to appear in the chiasma. This lies immediately behind the terminal pit and in front of the primitive optic groove (figs. 22, A and B), and therefore lies in the terminal ridge. The lateral prolongation of this ridge, which has been described as running obliquely across the primitive optic groove (fig. 16), furnishes a pathway for the optic tract fibers as they grow in from the retina to the optic centers in the thalamus and tectum opticum. The ridge may therefore be called the optic ridge. The primitive optic groove from this stage on is to be seen just behind the transverse ridge occupied by the optic chiasma and other decussations. Referring to the general description at the beginning of this article it will be seen that what I have called the terminal pit in the early embryo is the same as that called by His the optic recess, and that to which I have in all my previous

fiG. 18. Squalus ac., late stage Ix of Balfour, medial aspect of a model of the right half of the head. x 25.

papers given the name of preoptic recess. The primitive optic groove I have heretofore called the postoptic recess. These terms have been used by some other authors (Mrs. Gage, Sterzi, and others), but not by all. I wish to emphasize the necessity of recognizing the two pits and applying to them clearly distinctive names, because they are both related to the optic vesicle and because the postoptic recess has heretofore been confused with the infundibulum.

b. Remainder of the Floor of the Diencephalon

As soon as the head-bend of the brain tube appears, a broad depression of the floor of the forebrain vesicle can be seen which corresponds to the future inferior lobes (fig. 16). The anterior part of this is the relatively deep and sharply marked primitive optic groove. The posterior boundary is less definitely marked by a slight projection of the brain floor into the ventricle. This is the tuberculum posterius. As the broad inferior lobes involve nearly the whole floor of the diencephalon, neither the infundibulum nor the mammillary bodies are to be recognized at this time. Both are developed later within the general area of the primitive inferior lobes, as specialized portions of their walls.

fiG. 19. Squalus ac., 60 somites, parasagittal section near the middle line.

The mouth is open. Primitive inferior lobes, epiphysis and velum transversum. x 33.

The mammillary bodies are indicated by a rounded caudal projection of the depressed floor of the diencephalon in embryos with about 80 somites (figs. 20, 21, 22).

The infundibular recess is not found until after the completion of the changes described in connection with the optic recesses. As in man the infundibulum is the funnel—shaped depression leading from the floor of the tuber cinereum into the neural part of the pituitary body, so in fishes the infundibular recess must be that somewhat funnel-shaped or tr0ugh—shaped depression in the floor of the inferior lobes which leads into the saccus vasculosus, this being the neural portion of the pituitary body in fishes.

The angle of ectoderm from which the hypophysis will be developed can be readily recognized by the stage when the depression of the primitive optic groove has pushed the preoral entoderm out of the way and come into contact with the ectoderm. The hypophysial ectoderm is in contact with the posterior surface of this depressed part of the brain floor. As the hypophysis pushes in (figs. 18, 19, and following) its anterior limb remains in contact with the posterior wall of the primitive optic groove. The hypophysis grows back in contact with the rounded surface of the inferior lobes, insinuating itself between the brain floor and the median mass connecting the premandibular somites. As this mass in early stages connects the anterior head cavities also with one another, it may be said that the in—pushing hypophysis remains always in front of the entoderni and mesoderm except so much of the preoral entoderm as is cut OH by the primitive optic groove and aborted beneath the terminal ridge. By st-age M (fig. 22) the hypophysis begins _to be constricted ofi‘ from the ectoderm. It presents a shorter anterior lobe which is directed toward the optic chiasma and a longer posterior lobe which is directed toward the mammillary recess. Opposite the tip of the posterior lobe a special outgrowth of the brain wall represents the beginning of the saccus vasculosus. Although the term infundibulum scarcely applies to anything in the fish brain, yet the depression from which the saccus grows out is the region which corresponds to the infundibulum of man. The relative position of all these structures is shown in fig. 22, A and B, from a sagittal series of an embryo in stage M.

fiG. 20. Squalus ac, S0 somites, median sagittal section. X 33. Pre- and potoptic recesses, mamillary recess, eplphysis and velum transversum.

fiG. 21. Squalus ac., stage M. sagittal section. X 33. Pre- and postoptic, infundibular and mammillary recesses, epiphysis and velum transversum. The fibers of the optic chiasma appear in the terminal ridge.

The study of sections can be controlled in these young stages by the study of cleared whole embryos. Neal has given a most instructive series of figures from such cleared embryos and I can attest the accuracy and faithfulness of these figures. If figs. 7 to 11 of Neal’s Plate 3 be examined, it will be seen that the optic Vesicle shifts from the “infundibulum” to a point in front of the anterior head cavity. This agrees with what I have described above. I have carefully studied these whole embryos with the Braus—Driiner binocular and find that all the facts with regard to the form and position of parts in the optic region of the brain derived from the study of sections and models can be seen with perfect clearness in the whole embryos.

In Dohrn’s papers on the mandibular and premandibular somites (1904:) the relations of neural plate and preoral entoderm discussed in this section and the last are beautifully illustrated. Plates 1, 4, 6, 7, 9 and 11 show the early stages in the formation of the primitive optic groove and terminal ridge and the relations of the preoral entoderm and premandibular mesoderm derived from it. As is well known, the anterior head cavity of Miss Platt is found only in the Squalidae and Dohrn does not regard it as a somite. I am forced to believe, however, that he has not analyzed the conditions in Squalus acanthias with sufficient care, and that to this is to be attributed his attitude toward the anterior head cavity as well as his failure to recognize the terminal neural crest and the mesectoderm derived from it. In his figures the primitive optic groove is labelled “infundibulum,” but it is perfectly clear to me that it is the groove related to the optic vesicles. See Pl. 9, fig. 9, where the groove marked Inf. is the base of the optic stalk. In Pl. 1, fig. 15, the reference line Ent. Z w. passes across the primitive optic groove at the front and the true infundibular recess near the deep end of the hypophysis. Compare figs. 23, 24, and 25 of this paper. The terminal ridge is especially clear in Dohrn’s P1. 11, where early stages show its form as Well as the early stages of Amblystoma (see below).

c. Roof of the Diencephalon

In the roof of the interbrain the development of the velum transversum, dorsal sac, epiphysis and paraphysis has been Well described (Minot, Sterzi) so that I have nothing new to add. I wish only to note that a careful consideration of Dr. Neal’s sections and of the models made from them leads me to believe that the segmental position of the optic vesicle is practically the same as that of the transverse velum. The velum does not become prominent until after the optic Vesicle is well formed (stage , but in some specimens a slight fold representing it can be recognized in models of embryos as early as 24 somites or earlier. I am inclined to think that the velum represents an infolding of the brain wall which is begun early on account of the with drawal of material from the alar plate to form the optic vesicle. It is the second neuromere whose dorsal half thus gives rise to the retina, while its ventral half becomes depressed and bulged ventrocaudally to form the primitive inferior lobes above referred to.

fiG. 22. Squalus ac., stage M, two sagittal sections near the median plane. X 25. In addition to the features shown in earlier figures, the optic chiasma, anterior, posterior and habenular commlssures are seen.

3. Gamoids and Teleosts

The diencephalon presents no features of especial importance. There is a greater development of the inferior lobes than in selachians, although the olfactory apparatus is of less importance. This is doubtless to be attributed to the much greater importance of the gustatory apparatus in ganoids and teleosts. The saccus vasculosus reaches a very great development in some of these forms and it has been shown that there is an intermingling of the epithelial sacs of the saccus with those of the hypophysis. The optic tracts form a chiasma in the floor of the brain in some forms and in others cross at some distance from the brain.

The telencephalon presents certain great peculiarities. It is usually somewhat more elongated than that of most selachians and resembles that of Chimaera or Heptanchus. Also the telencephalon has no massive roof, but only a broad membraneous tela continuous with that of the diencephalon. The boundary between the di- and telencephalon in the roof is marked by a velum transversum which forms the front wall of the dorsal sac of the diencephalon.

The membraneous roof of the telencephalon is much more extensive than in selachians or other vertebrates. In many cases the lateral walls of the telencephalon are rolled outward (laterad) so that the morphological dorsal border is directed laterad or latero—ve11trad. This makes the membraneous roof in these forms exceedingly broad. The ventricle is correspondingly extensive and toward its anterior end divides into lateral ventricles which extend into the olfactory bulbs. These relations have been described and figured by several authors. (See Kappers 1906, 1907, Johnston, 1906, fig. 151.)

4. Amphibians

In the adult amphibian brain the large size of the telencephalon and the form and relations of its nervous and membraneous portions are of interest for our problems. The large telencephalon has

thinner walls and larger lateral ventricles than are found in selachians. The lateral ventricles are connected With the median ventricles by wide but Well defined interventricular foramina (Johnston 1906, fig. 150, 151). The lateral ventricles extend forward into the olfactory bulbs and also have a caudal prolongation into the so-called occipital pole of the hemisphere.

fiG. 23. Amblystoina punctatum, neural plate stage. Sagittal section of head end. Ectoderm dark, neural plate medium, entoderm light. X 25.

fiG. 24. Amblystoma p., neuropore tage. Sagittal ection. The section falls to one side of the median plane in the dorsal region and shows the mesoderm lateral -to the notochord. Its cephalic limit is the same as that of the notochord. X 25.

The tela of the diencephalon is separated from the membraneous roof of the telencephalon by a prominent velum transversum which in the adult becomes complexly folded in connection with the chorioid plexus. In front of the velum is a highly developed paraphysis.

Practically the whole of the membraneous roof of the telencephalon is involved in the two complex structures, the chorioid plexuses which extend into the lateral ventricles and the paraphysis which projects upward between the lateral lobes. -The essential fact is that the membraneous roof extends forward over the interventricular foramina to meet the lamina terminalis. The roof of the telencephalon near the middle line is membraneous for its whole length.

The questions regarding the anterior end of the brain and the boundary between diencephalon and telencephalon have been most carefully studied in embryos of Amblystoma punctatum. In these embryos the entoderm, mesoderm and notochord present essentially the same features and the same relations to the brain as in selachians. In particular, the premandibular somites, the median undifierentiated mass in which the notochord ends anteriorly, and the preoral entoderm have the same disposition as in selachians. The only difference is that all the structures are more compact in amphibians and the preoral entoderm is shorter. As figs. 23, 24, 25 show, the preoral entoderm fills the angle between the floor of the neural tube (neural plate) and the ectoderm and there is a short prolongation of the archenteric cavity into it in front of the site of the future mouth.

fiG. 25. Amblystoma p., after closure of neuropore, model of the right half of the head, viewed from the medial surface. X 25.

The neural plate is bounded by neural folds which meet in front in the transverse temninal ridge. This terminal ridge marks the line along which ectoderm and neural plate meet and, when the neural plate rolls up into a tube, the ridge forms the lower border of the neuropore. These relations are as simple and clear in Amblystoma as in Squalus (figs. 23, 24). Even after the neuropore has closed the arrangement of cells and nuclei in this region shows the outline of the terminal ridge. After the brain is separated from the ectoderm the terminal ridge forms a distinct fold, convex toward the ventricle (fig. 26), which in later stages is occupied by the fibers of the optic tracts in the chiasma (fig. 33). No neuroporic recess is to be seen in A. punctatum in early stages following the closure

fiG. 26. Amblystoma p., invagination of hypophysis beginning; primitive inferior lobes. Sagittal section of head. X 25.

fiG. 27. Amblystoma p., hypophysial invagination at its height; velum transversum and epiphysis; median sagittal section reconstructed from several sections. X 25.

of the neuropore, but in later stages a slight pit is found which may correspond to the neuroporic recess described in other forms (fig. 33).

The early appearance of the retinal areas on the neural plate was first described by Eycleshymer (1890) and the fact that the optic vesicles are formed from the lateral parts of the neural plate has been pointed out by the_ writer (1905, 1906). While the neural plate is still open the retinal pits are connected with one another by a shallow groove running just behind the terminal ridge (fig. 23, r. po). As the neural plate rolls up and the optic vesicles are evaginated this groove grows deeper (figs. 24, 25) and by the time the neuropore is closed the front end of the neural tube presents a prominent depression resting against the preoral entoderm and connecting the optic vesicles (fig. 25). This is the primitive optic groove as described in Squalus. The model of this stage shows a second angle at the front of the neural tube, separated from the primitive optic groove by the terminal ridge. This is a pit formed in the lower part of the neuropore and is the terminal pit (fig. 25).

fiG. 28. Amblystoma p., a little more advanced than the one shown in fig. 27. Parasagittal section through the optic ridge. X 40.

fiG. 29. Ambiystoma p., of about the same stage as that in fig. 28. Model or the right half of the brain seen from the medial surface. X 40.

From the earliest stages after the formation of the neural plate and folds, the region from which the hypophysis will be formed can be accurately located. In a median sagittal section of any stage up to the time when the hypophysis is invaginated, a slight reentrant angle is seen between the terminal ridge and the preoral entoderm. The ectoderm of this angle will form the hypophysis. VVhen the neuropore closes this hypophysial ectoderm is slightly thickened and is continuous with the lower border of the thick plate formed by the fusion of ectoderm and neural tube in the neuropore. When the hypophysis begins to push in it presses on the anterior surface of the preoral entoderm and as the primitive optic groove becomes depressed the brain wall presses on the preoral entoderm from above. In this way the preoral entoderm is pushed back and the hypophysis insinuates itself between the entoderm and the posterior wall of the primitive optic groove as in Squalus. The preoral entoderm becomes shorter and blunter, but none of it is cut oi} as in Squalus.

fiG. 30. Ainblystonia 1).. stage when the paraphysis is formed. Model of the right half of the head seen from the medial surface. The model was made from a series of sagittal sections which were oblique to the longitudinal axis, so that the surface of the model lie in the median plane at the forebrain but passes to the left of the middle at the hindbrain.‘ X 40.

The only important diiference between Amblystoma and Squalus is that anterior head cavities are not formed in Amblystoma. In later stages the preoral entoderm and median mass proliferate as mesenchyme, so that essentially the same end is reached as in Squalus. In some embryos are found indications of a connection of the archentoderm with the hypophysis through the preoral entoderm, but this and the details of the formation of the hypophysis will be described in another paper.

As development proceeds the same shifting in the relations of the optic vesicles is seen as has been described in Squalus. In the lateral wall of the forebrain vesicle a thickening is formed which runs from the terminal ridge in the middle line obliquely latero—caudad across the primitive optic groove. This thickening is formed in anticipation of the ingrowth of optic tract fibers and may be called the optic ridge. It separates the optic vesicles from the primitive optic groove and causes them to be connected by the terminal pit. figs. 25, 28, 29, 30 show this in sections and models.

fiG. 31. Amblystoma p., nearly the same stage as that shown in fig. 30, median sagittal section of forebrain and midbrain. X 40.

fiG. 32. Amblystolna p., later stage, median sagittal section. Note the ex ‘creme curvature of the brain in this and following stages. )( 40.

By the time the optic ridge is formed the floor of the forebrain vesicle has become depressed to form broad primitive inferior lobes and in the caudal wall of this a mammillary recess marks the beginning of the mammillary bodies (figs. 26 and 27). This is bounded caudally by the tuberculum posterius. Later, when the hypophysis has reached its definitive position a saccus outgrowth from the inferior lobes appears and the region at which it is connected with the brain may be called the infundibulum. There are therefore in the floor of the forebrain vesicle four recesses formed as in selachians: terminal pit or preoptic recess, primitive optic groove or postoptic recess, infundibulum and mammillary recess.

The formation of the velum transversum, epiphysis and paraphysis need not be described as they are already known (Minot and others). These structures are represented in figs. 30, 31, 32, 33.

fiG. 33. Amblystoma p., all the chief features of the ‘forebrain developed. Median sagital section. X40.

5. Reptiles and Birds

I have to say here only that the study of whole mounts of chick embryos between 20 and 40 hours of incubation shows that the same relations exist in the region of the optic chiasma as in Squalus and Amblystoma. In early embryos the optic vesicles are connected by a primitive optic groove behind the terminal ridge. Later, the optic ridge is formed, the terminal pit becomes connected with the cavities of the optic stalks, and the optic chiasma occupies the terminal ridge.

6. Mammals

The chief peculiarity of the mammalian brain is the great size of the cerebral hemispheres. In the adult, as is well known, there is a membraneous tela over the median ventricle and this is continued as the roof of the interventricular foramen into the wall of each hemisphere. In front of the foramina the tela meets the lamina terminalis, so that as in amphibians there is a membraneous tela in the median region of the telencephalon for its whole length. For the morphological relations of this tela, the ventricles and the chorioid plexuses it is necessary to study the embryology.

While in all lower classes, except cyclostomes, a prominent velum transversum marks the boundary between diencephalon and telencephalon, in mammals the velum has not heretofore been recognized. The large collection of pig embryos in this laboratory gives excellent opportunity for comparison With the lower classes described above.

Fig. 34. Pig embryo of 5 mm. Model of right half of head seen from medial surface. The optic vesicles are still connected with the primitive optic groove. The Roman numerals indicate the brain neuromeres. X 25.

The earliest stage available is a 5 mm, pig cut in transverse series from which a model of the brain has been made (fig. 34). From the figure it will be seen that this brain agrees very closely with that of the Squalus embryo of about 20 somites. The cavity of the optic stalk is continuous with a groove which traverses the median line, the primitive optic groove. Behind this is the primitive inferior lobe, a ventral expansion bounded caudally by the tuberculum posterius. In front of the primitive optic groove is a transverse ridge whose cross section in the median plane presents the form of an arch.

This is the terminal ridge. In front of this is a median pit, the terminal pit. Following the lamina terminalis around the front end of the brain there is found in the dorsal wall a distinct transverse fold, followed by an arched portion and a second more shallow fold. In this fold and for some distance behind it are seen in later embryos the fibers of the posterior commissure and the decussation of the tectum mesencephali. The more anterior and deeper of the two folds is the velum transversum, as the further description will make clear. In embryos of 6 and 7 mm. the same change of relations in the optic region takes place as has been described above for Squalus and Amblystoma. The terminal ridge becomes prolonged candolaterad as the optic ridge and in these later appear the optic chiasma and tracts. The identity of these structures in the classes of animals studied is absolutely clear.

Fig. 35. Pig embryo, 6 mm. A., median sagittal section. B and C, parasagittal sections. Neuromeres numbered in Roman. X 20.

The time of development of the mammillary recesses and of the neural part of the hypophysis (saccus vasculosus) varies somewhat in pig embryos. In most 6 mm, embryos the mammillary recess is already clearly recognizable as a caudal expansion of the primitive inferior lobe (fig. 35, A) whose border above and behind is the tuberculum posterius. From now on the mammillary recess is always clear. Between it and the primitive optic groove the hypophysis lies against the under surface of the primitive inferior lobe. In some 6 mm. embryos there is to be seen just behind the tip of the hypophysis in sagittal section a slight, but definite, depression and thickening in the brain floor, the beginning of the evagination of the neural part of the hypophysis. This shows a variable development in embryos between 6 mm. and 9 mm., but is always clearly present in 9 mm. embryos (fig. 36). After this time the sac grows out and enwraps the tip of the hypophysis and the relations familiar to all embryologists are established. The outgrowth of this sac is the earliest mark of the position of the infundibulum and this is a considerable distance behind the primitive optic groove, as in selachians and amphibians. In later development the floor of the primitive inferior lobe becomes drawn down in funnel shape and the lateral portions become thickened as the tuber cinereum. It is one of the peculiarities of the mammalian and human brain that the infundibulum is drawn down close behind the chiasma and forms a deeper and narrower funnel than in lower vertebrates. There is no fundamental difference of relations.

Fig. 36. Pig embryo, 9 mm., median sagittal section. X 20.

The evidence that the fold in the roof which has been called the velum transversum is correctly identified may be seen in figs. 34'

Fig. 37. Pig embryo, 12 mm. Median sagittal section, recontructed from several sections. X 15.

to 39. The posterior commissure is a point about which there is no dispute. In the arch (neuromere) in front of it appears later the epiphysis (fig. 38). In front of the epiphysis and in the same arch appears the commdssum ha.ben-ulamis. This arch then is the roof of the diencephalon. It "is bounded in front in all other classes of vertebrates by the velum transversum. The fold to which this name has been given lies in the proper segmental position. This is further supported by its relations to the structures in front of it. Immediately in front of the velum the roof is raised in a distinct arch.

This corresponds in position to the paraphysis of lower forms (cf. figs. 30 to Since there is no glandular development known in mammals, it may be called the pamaphysal arch, a name which Minot (1901) applies to the corresponding structure in birds. In front of the paraphysal arch a membrane continues forward to meet the lamina terminalis. When the lateral cerebral vesicles are formed it is seen (figs, 37, 38, 39) that this membrane lies over the ventricle between the interventricular foramina.

Fig. 38. Pig embryo, 15 mm. Median sagittal section of the forebrain. The dotted outline of the hemisphere is reconstructed from several sections. X 15.

As development proceeds the velum transversum becomes a fold with a sharper angle but less deep in proportion to the size of the brain. The paraphysal arch remains a distinct median pouch until the lateral vesicles are well formed. In sagittal sections to one side of the median plane the lateral ventricle appears as a dorsal cavity opening by way of the interventricular foramen in front of the ‘paraphysal arch into the median ventricle. (fig. 38, 39.) These simple relations persist up to the 17 mm. stage or later. By the 15 to 17 mm. embryo the chorioid plexus of the lateral ventricle is forming. Its position with reference to the velum transversum is shown in a parasagittal section of a 17 mm. embryo (fig. 40) and in three frontal sections of the brain of a 15 mm. embryo (fig. 41). The velum transversum is not only a dorsal fold, but is continued around the lateral wall of the brain as the constriction (external fur fiG. 39. Pig embryo, 17 mm. Median sagittal section reconstructed from several sections. The outline of the hemisphere lateral to the median plane is shown in dotted lines. X 12%.

row and internal ridge) between the diencephalon and the ‘selencephalon. By the stage mentioned the cerebral vesicle is sufliciently expanded to push back past the boundary line. In the angle between the vesicle and the dieneephalon appears the chorioid plexus pushing into the lateral ventricle. It appears as a folding of the anterior Wall or limb of the velum transversum and its lateral prolongation.

In this way appears the chorioid fissure whose further history need not be traced. Near the median plane the plexus appears as a fold projecting into the interventricular forarnen and separated from the velum transversum by the paraphysal arch. From this stage on the plexus grows rapidly and becomes very large and in the median region both the velum and the paraphysal arch become involved in the plexus and their identity is lost. If the paraphysis is to be found in adult mammals it should be looked for in the chorioid plexus in the middle line between the interventricular foramina. The membraneous roof extending forward from the paraphysal arch to meet the lamina terminalis is relatively long in the embryo, spanning the wide opening into the lateral ventricles. In later development these interventricular foramina grow much less rapidly than the hemispheres and the roof in question becomes of insignificant length. This is due in part also to the expansion of the lamina terminalis by the commissures which develop in it.

Fig. 40. A parasagittal section from the embryo drawn in fig. 39.

Fig. 41. Pig embryo, 15 mm. Three sections from a frontal series showing the relations of the chorioid fissure and plexus to the lateral hemisphere and the thalamus. The section to the left is the most ventral, that to the right the most dorsal.

The neuroporic recess can be located with certainty in the pig embryos. In young stages there appears a slight ridge in each lateral wall just rostral to the preoptic recess. This is the beginning of the corpus striatum. As the two ridges converge toward the middle line they cause a thickening of the lamina terminalis above the preoptic recess (figs. 38 and 39, between 1*. p. and r. This is the location of the anterior commissure in later stages. Above this thickening is another pit which in 15 mm. embryos is a smooth rounded concavity in the middle line, not a transverse groove (fig. 38). That this is the recessus neuroporicus seems clear from the fact that it is above the corpus striatum and below the interventricular foramina. As the striatum and lateral vesicles grow this pit becomes deeper and more pointed.

Three of the young human embryos described in recent years seem to the writer to give clear evidence that the relations in the optic region are the same in man as in fishes, amphibians and other mammals. The embryo described by Low (1908) shows optic pits on the open neural plate. These are apparently connected with one another by a groove running across the middle line parallel with the terminal ridge. The embryo described by Broman (1896) is 3 mm. long and has the neuropore closed. It shows the terminal ridge essentially like that of Squalus or of Amblystoma. The optic vesicles are connected with the primitive optic groove caudal to the terminal ridge. Broman noticed the terminal ridge and gave to its ectal, concave surface the name “Fossa interocularis,” but did not speak of its significance. The embryo described by Mrs. Gage (1905) is a trifle larger and considerably more advanced in development. In it the optic vesicles are connected with the pit formed in the neuropore-space. This pit seems to correspond to the preoptic recess above described. From these three embryos it appears very probable that the course of development in man is the same as in the pig and lower forms.

II. Discussion

1. The Anterior End of the Head and Brain

Owing chiefly to the lack of certain essential facts, an extensive literature _has grown up about the question of the anterior end of the brain. Since the facts which were wanting are supplied in the preceding pages, a detailed review of the discussions from v. Kupfier and His onward would be unprofitable. The determination of the anterior end of the brain is a matter of direct observation. In the study of successive stages in the early development of a- selachian, an amphibian and a mammal the essential facts are found to be perfectly clear, and the three forms agree in the form changes of the brain and in the relations of the brain to the ectoderm, entoderm and mesoderm.

The anterior boundary of the neural plate is formed by a transverse ridge, the te7'm.ina.l ridge, which is continuous with the neural folds bounding the neural plate laterally. This terminal ridge is clearly seen in successive stages and is readily followed up to the time when the optic chiasma is formed in it. The optic chiasma therefore occupies the anterior border of the floor plate of the brain. This is a matter of fact, not of interpretation.

Behind the terminal ridge lies a transverse groove which laterally becomes continuous with the optic pits or vesicles. This is to be seen from the earliest stages when the optic pits are recognizable in the neural tube or, indeed, on the open neural plate. This groove is the depression which His (1892, 1893) called the recessus infundibuli and which V. Kupfier (1893) called the sinus postopticus. Later workers have followed His, although his own numerous figures (1892) show that there are two distinct depressions in the region between the chiasma and the mammillary recess. It is clear also (His, 1892, fig. 2 and. p. 162) that the opening into the saccus vasculosus represents the infundibulum and the depression which he calls recessus infundibuli must be something else. This depression I have called the recessus postopticus (1902, 1906). Because of its relation to the optic pits I have called this in the early embryo the primitive optic groove.

This primitive optic groove forms a ventral projection of the brain floor which outwardly might be considered as a transverse ridge and it was to this external ridge that His gave the name Basilarleiste; To the groove within he gave the name recessus infundibuli s. basilamis. The Basilarleiste of His is not a thickening of the brain floor but a fold which appears Within as the primitive optic groove. It is important to make this clear because of the incomparable value of His’s Work in matters of general morphology. The Basilarleiste in the embryos studied by His meets the front end of the notochord and of the entoderm (Seessel’s sac). Upon this relation of the brain to the notochord and entoderm His based the conclusion that the Basilarleiste or the recessus infundibuli formed the anterior end of the brain floor. “In der vorderen Endflache endigt die Gehirnaxe und es bedarf einiger Vertsandigung dariiber, wohin dies Ende zu yerlegen sei. In Anschlusse an v. Baer und andere habe ich selber in friiheren Arbeiten dies Ende in das Infundibulum oder richtiger ausgedriickt, in die Mitte der Basilarleiste verlegt. Andere, wie neuerdings Keibel, lassen die Gehirnaxe im Chiasma opticum auslaufen. Die Discussion dariiber, wer mit seiner Behauptung im Recht sei, hat nur dann einen Sinn, vvenn man zuvor festgestellt hat, was unter Gehirnaxe fiir eine Linie zu verstehen sei, Ich selber habe darunter stets die Mittellinie des Hirnbodens verstanden. Das heisst die Linie, welche, wenigstens auf friiheren Stufen léings der Chorda, als der anerkannten Kiirperaxe verliiuft. Diese basilare Axe endigt unzweifelhaft in der Basilarleiste. Versteht man dagegen unter Grehirnaxe eine Linie, welche der Mitte der Rohrenlichtung folgt, so wird diese mittlere Axe in einer Ebene liegen, welche die Grund- und die Fliigelplatte des Grehirns von einander scheidet, und ihr Endpunkt trifit die vordere Endflache im Recessus opticus, bez. dicht vor dem Ort des Chiasma opticum. Wollen Wir zur basilaren und zur mittleren Axe noch eine dritte dorsale Langsaxe oder Langslinie annehmen, so haben wir deren Ende am oberen Rande der Lamina terminalis zu suchen, Vor der Stelle, wo die fissura chorioidea ihren Anfang nimmt.”

Now it must be noticed that more recent work (Platt 1891, Hoffmann 1896, Neal 1898, and others) has shown that the entoderm does not end anteriorly in contact with the Basilarleiste, but extends forward beneath the terminal ridge. His did not study sulficiently early stages to see this. Early stages show clearly that the basal axis of the brain ends not in the Basilarleiste but in the terminal ridge in which later the optic chiasma appears. Furthermore, this is equally true of selachians, amphibians, birds and mammals. It is altogether probable that the same is true of petromyzonts also, for the depression called by Koltzoff “infundibulum” is doubtless the same as the primitive optic groove of other forms. In all vertebrates studied by the writer the entoderm extends forward beneath the transverse ridge which afterward becomes the optic chiasma. The definition of the anterior end of the head previously given (Johnston, 1905) may now be simplified to read: in all vertebrates the anterior end of the head is the point at which the brain plate meets the general ectoderm at the same time that it comes into contact with the anterior end of the entoderm. This point is marked in the adult by the optic chiasma.

It has been shown in this paper that the depression in front of the optic chiasma which has been known to His and other authors as the recessus opticus, is related to the optic vesicles only secondarily and is primarily a pit in the basal part of the neuropore (lamina terminalis).

2. The Homology of the Saccus Vasculosus

Here I wish only to point out the homology of the saccus vasculosus of lower vertebrates with the neural part of the pituitary body in man. The saccus vasculosus is an evagination from the floor of the diencephalon which is more or less branched, is lined by ependymal cells and sensory cells, and is supplied by nerve fibers ending in its epithelial lining. In all lower forms it comes into close relations with the hypophysis. In many cases the subdivisions of the two structures become intermingled or interlaced. In man the neural part of the pituitary body has the same relations but the cavity becomes obliterated. During early development the evagination appears in identical manner and relations in all vertebrates and the writer can see no ground for doubting the complete homology of the structure in all vertebrates.

This homology was understood by His but seems not to be universally accepted. Edinger (1908,lp, 203) publishes a schematic figure of a sagittal section of the vertebrate brain in which he shows an infundibulum in contact with the hypophysis and behind it a wholly separate evagination of the brain floor which he calls the saccus vasculosus. This diagram stands in contradiction to the drawings from actual specimens in the same book (fig. 167, Ohimaera; fig. 175, 176, Varanus; fig. 178, Ammocoetes; fig. 181, Siredon; fig. 219, Hexanchus). In all of these there is only one evagination of the brain floor between the optic chiasma and the mammillary bodies and it comes into relation with the hypophysis. -The writer does not know of any vertebrate in which the condition shown in Edinger’s diagram is found.

3. Segmentation of the Neural Tube in Front of the Cerebellum

In the hindbrain the neuromeres are generally recognized as brain segments corresponding to the segments of the organs in the head. In front of the cerebellum there is no such unanimity of opinion. The Writer has discussed this subject at length (1905) and has found nothing in the studies here reported to change any of the conclusions there expressed. On the contrary, the conclusions there based on indirect evidence from other authors are confirmed by direct observation. The segments in the mes-, di- and telencephalon are clearly indicated in fig. 42, representing parasagittal sections of the brain of a pig of 7 mm’. The optic vesicle is here seen somewhat out of line with the other neuromeres but no one would doubt that it represents one brain segment. In front of it is the first segment, from which the telencephalon is formed. Opposite the optic vesicles in the median region is the velum transversum. Behind the optic vesicle are clearly seen in the figure three segments. In connection with the first of these (neuromere iii) appears later the epiphysis. The other two (iv and v) obviously enter into the mesencephalon. In se1achians these two segments have connected with them respectively the thalamic nerve of Miss Platt which probably forms the ciliary ganglion, and the part of the neural crest which forms the ophthalmic division of the trigeminus. The terminal part of the neural crest in close relation with the neuropore presumably gives rise to the ganglion of the nervus terminalis in selachians. If this be true, every neuroinere of the embryonic brain has connected with it in one class of vertebrates or another some sensory nerve or sense organ (including the optic vesicle and epiphysis. The five brain segments are equally clearly to be seen in figs. 34 and 35.

4. Boundary between Diencephalon and Telencephialon

The posterior boundary of the diencephalon has never been in dispute. It is the constriction between the forebrain and midbrain vesicles and is later marked dorsally by the posterior commissure and ventrally by the tuberculum posterius. When the forebrain vesicle becomes divided into diencephalon and telencephalon the exact location of the boundary between them has not been entirely clear. In all vertebrates in which a definite velum transversum is recognizable this is considered as the mark of the boundary. The existence of a paraphysis and lateral plexus chorioideus in front of the velum and of a‘ dorsal sac and one or two epiphyses behind it is now so thoroughly understood as to need no further comment (Gaupp 1898, Minot 1901, Johnston 1905, 1906).

Fig. 42. Pig embryo, 7 mm. Two parasagittal sections to show the segments of the forebrain and mid-brain. Compare figs. 34, 35 and 18.

The velum transversum has been described in cyclostome embryos by Sterzi (1908) and in mammals in the foregoing pages, so that the boundary line sought for is now clear in the brain roof in all classes of vertebrates. From the velum transversum a groove or constriction continues around the sides of the brain. Owing to the early evagination of the optic vesicles this constriction in the dorsal half of the brain occupies the space left vacant, so to speak, by the withdrawal of the retinal tissue. Ventrally the groove is to be thought of as lying in front of the neuromere to which the optic vesicle belongs. The diencephalon consists in its dorsal half of but one neuromere after the withdrawal of the optic vesicle ; in its ventral half it includes two neuromeres, the more posterior of which is narrow while the more anterior one forms the depression of the brain floor which I have called the primitive inferior lobe. The boundary between the diencephalon and the telencephalon in the brain floor has been in dispute because of the obscurity which has existed over the optic recesses and the anterior end of the brain.

His placed the boundary behind the infundibulum and assigned the pars optica hypothalami to the telencephalon. He was led to this by his conviction that the telencephalon consisted of a complete brain ring or segment and by his belief that the end of the brain axis was in the Basilarleiste or infundibular recess. As shown above, the optic chiasma is formed in the terminal ridge and therefore occupies the extreme anterior border of the floor plate of the neural tube. If the telencephalon is a complete transverse segment of the brain, as His always insisted, there is no alternative but to include the optic chiasma within it. The primitive optic groove which bounds the optic chiasma behind belongs to the same neuromere with the optic vesicles and therefore is included in the diencephalon. The telencephalon can include no more thanpthe optic chiasma and the associated decussations in the brain floor which lie in the terminal ridge. The boundary between the diencephalon and telencephalon is marked by the velum transversum above and by the primitive optic groove or postoptic recess below. In adult mammals, in which both these landmarks have disappeared, the boundary can be defined by a line drawn just behind the interventricular foramen and meeting the posterior surface of the chiasma ridge (fig. 44).

The external groove which separates the diencephalon and telencephalon is usually well marked and in the brains of amphibians, reptiles and mammals increases in depth and prominence with the enlargement of the cerebral hemispheres. The description of the early development has shown that the lateral chorioid plexus is formed in mammals immediately in front of the velum transversum and of the groove which continues from the velum transversum around the lateral wall of the brain. From this it results that in the adult the chorioid fissure is found at the bottom of the very deep groove between the hemispheres and the brain stem and that the boundary between diencephalon and teleneephalon runs just along the posterior (thalamic) border of the fissure. These relations are clearly set forth by Gr. Elliot Smith in a recent paper (1908) on the forebrain of Lepidosiren, which agrees in essentials with that of mammals.

Fig. 43. Sketches to illustrate the boundary line between the diencephalon and the telencephalon. The brains of a selachian (A) and an amphibian (B) are outlined as seen from the medial surface and the boundary set by His is indicated by a dotted line, that fixed in this paper by a heavy continuous line.

Fig. 44. Sketch of the human brain for comparison with fig. 43.

When the internal structure of the brain is taken into account it is seen that the boundary line indicated by the development separates centers of different significance. Before it lie the primary and secondary olfactory centers, behind it in the nucleus habenulae and inferior lobes (tuber cinereum) lie the tertiary olfactory centers with reflex functions. This is not true of the boundary line laid down by His which placed the region of the infundibulum (pars optica hypothalami) in the telencephalon. His was apparently not followed in this by the Basle nomenclature commission and the tables of neurological terms adopted by the commission contradict His’s explanatory notes in that the tables place the pars optica hypothalami in the diencephalon while His states that it belongs to the telencephalon. (See His, 1895, pp. 161, 162.) This is perhaps because anatomists generally have felt the incongruity of assigning the tuber cinereum, infundibulum and hypophysis to the telencephalon. This objection fails when only the chiasma and the fiber decussations adjacent to it are included in the telcncephalon.

The usage adopted by the BNA goes to the other extreme and involves at least as bad consequences. The BNA includes the lamina terminalis in the pars optica hypothalami, and implies that the lamina terminalis is the front wall of the diencephalon. The discussion of this usage, which is widely followed by anatomists, will come best in the next section, but here it may be pointed out that it implies the inclusion in the diencephalon of various structures which certainly can not be so interpreted.

a. "The lamina terminalis contains the anterior commissure, and according to the researches of G. Elliot Smith the corpus callosum and hippocampal commissure develop in it also. These commissures would then all fall in the anterior wall of the diencephalon. This is obviously impractical and confusing and would lead to endless difficulties in fixing an arbitrary boundary.

b. The gray matter in the wall of the preoptic recess constitutes generally in vertebrates an important secondary olfactory center which, unless there are strong reasons for assigning it to the diencephalon, should be placed with the other secondary olfactory centers in the telencephalon. All the facts of development and general morphology, however, favor the retention of this center in the telencephalon.

c. In many fish-like vertebrates the larger part of the telencephalon (corpus striatum and olfactory lobe) lies lateral to the lamina terminalis and forms the wall of the median ventricle. These structures in fishes would be included in the diencephalon and there would be endless confusion as to the boundary line in various classes. No such confusio and no p.ractical diffcuties in the description of the adult brain arise from the recognition of the boundary suggested above which is clearly marked in the development.

5. The Ventricles and the Tela

An essential part of the question of the boundary between diencephalon and telencephalon is the problem of the median ventricle; specifically, does any part of the median ventricle belong to the telencephalon? The view held by His was that the anterior part of the median ventricle belonged, with the pars optica hypothalami, to the telencephalon. The View which makes the lamina terminalis the anterior wall of the diencephalon assigns the whole of the median ventricle in front of the aqueduct to the diencephalon. It is obvious that the writer must agree with His in recognizing a median ventricle in the telencephalon, although a shorter part of the median ventricle is included than was included by His. The above diagrams (figs. 43 and 44) show the boundary line of His and that adopted by the writer and it is clear that the short part of the median ventricle between this line and the lamina terminalis belongs to the telencephalon and makes communication with the lateral ventricle through the interventricular foramina.

The view which regards the lamina terminalis as the anterior wall of the diencephalon and of its ventricle denies the existence of any median portion in the telencephalon. This means one of two things: either the diencephalon is the terminal segment of the brain and the telencephalon lies lateral to it as two hemispheres, or the diencephalon is terminal and the telencephalon consists of ultra—terminal hemispheres. Neither of these is true. Aside from the fact that the latter View involves a contradiction in terms, it cannot be considered, because all the evidence shows that the hemispheres are lateral structures. (a,) In the ontogeny of all vertebrates the hemispheres arise as evaginations or expansions of the lateral brain wall behind the lamina terminalis; (b) the lateral ventricles thus formed remain always as lateral prolongations of the ventricle and the median ventricle always extends forward beyond the interventricular foramina ; (0) when the hemispheres by their great growth push forward beyond the lamina terminalis, as they do in most vertebrates, they are still connected with the lateral wall of the brain stem and in the middle line the lamina terminalis is alway the most antler or structure of the brain in the adult as in the embryo. It is no more true to say that the telencephalon is ultra-terminal than to say that it is postoptic or post—velar. The occipital lobe extends as far behind the velum transversum as the frontal lobe extends in front of the lamina terminalis. The Whole hemisphere is a great expansion of a part; of the lateral wall of the brain between the lamina terminalis in front and the optic vesicles and primitive optic grooves behind. The point in dispute is whether that portion of the preoptic brain segment which is not carried out in the hemispheres belongs in the diencephalon or the telencephalon.

The first step in answering this question is to see clearly that in the early embryo the lateral hemispheres and the median portion exist together undiiferentiated as a simple ring or segment in front of the optic Vesicles. This segment is bounded from the earliest stages, even before the neural tube is closed, by the sharply marked primitive optic groove and the optic vesicles. It is only some time after the formation of the optic vesicles that the dorsal part of this simple segment bulges out at either side to form the lateral hemispheres. If the segmentis simple at the start, is there any ground for separating the ventral part and adding it to the diencephalon which lies behind the primitive optic groove? The only thing to give support to this View is the connection of the hollow optic stalk with the preoptic recess. Since the optic vesicles have always been referred to the diencephalon, their close relation to the lamina termi nalis through the preoptic recess suggests the inclusion of the lamina terminalis in the diencephalon. Now, however, it is shown that the optic vesicles are primarily connected with the post-optic recess and are only secondarily related to the preoptic recess.

In view of this there remains no ground for separating the median and lateral structures which develop from this primitive first segment. The embryological facts leave only one course open, namely, to consider the lateral hemispheres as the dorsal portion, the region of the optic chiasma as the ventral portion of one segment.

Finally, it is impossible to harmonize the relations of the velum transversum in lower vertebrates and in all embryos with the view that the lamina terminalis bounds the dieneephalon. That the velum transversum marks this boundary dorsally is universally agreed. It stands at some distance from the lamina terminalis, however, and if the boundary line is to follow the latter it must run along the roof of the brain and then around its front wall, an obvious absurdity.

A recent writer on the development of the forebrain (Fann.y Fuchs, 1908) states that it is only for practical convenience in describing early stages that the term telencephalon should be used at all. Whatever is left after the development of the hemispheres (she recognizes tacitly that there is something left) should be reckoned with the diencephalon. She states that in Rana the telencephalon has no roof because the di- telencephalic groove meets the upper end of the lamina terminalis. Since she has not studied early stages, has not recognized the velum transversum and gives no figures to show what she includes in the lamina terminalis, her conclusions on this point can have little value. Such figures as Fraulein Fuchs gives show a long median ventricle extending far forward beyond the interventricular foramina. The roof of this, as the writer’s own preparations show, is the same as the roof of the similar median ventricle in all other vertebrates, namely a membrane reaching from the velum transversum to the upper border of the lamina terminalis. Fraulein Fuchs has simply included this roof in what she calls the lamina terminalis. This author has studied only the obvious features in the later stages of the development of the forebrain and these give no suflicient ground for any conclusions regarding the morphological value of the telencephalon. There may be quoted here the conclusion of His in his paper on the general morphology of the brain (1892, p. 383) : “dass eine solche allgemeine Morphologie nur dann endgiiltig zu gewinnen ist, wenn wir auf die allerersten Entwicklungsstufen zuriickgreifen.”

The tela chorioidea of the third ventricle and lateral ventricles requires some comment. It must be noted first that in all vertebrates, embryo and adult, at membraneous tela extends over both diencephalon and telencephalon from the habenular commissure to the dorsal border of the lamina terminalis. In all vertebrates except adults of higher forms there is an obvious narrow place in the nervous brain wall between the diencephalon and telencephalon, and the tela is widest here. This seems to the writer to be due to the withdrawal of nervous material from the dorsal part of the lateral brain walls to form the optic vesicles. This withdrawal of retina-substance leaves a gap which is filled by membraneous tela only. The tela in the median region of the telencephalon is perfectly evident in any brain, embryonic or adult, so far as the writer is acquainted. From the posterior part of this tela next to the velum transversum arises the paraphysis or the rudimentary paraphysal arch, and forward from that the tela stretches across the median ventricle between the interventricular foramina. It is equally evident that at all stages of development these foramina are roofed by lateral prolongations of the tela. This is true of all forms with the apparent exception of cyclostomes, teleosts and ganoids. In cyclostomes this is probably due to the compression of the forebrain by the oral funnel and olfactory organ. In ganoids and teleosts the interventricular foramina have been widened beyond recognition by the eversion of the lateral walls. In amphibians and higher forms the prolongation of the tela over the interventricular foramina to become the roof of the lateral ventricles has great importance for the formation of the lateral plexuses. The beginning of these has been described and the only further comment which the writer wishes to make is that the complexity and mystery which the text—books throw around the relations of the velum interpositum and the lateral plexuses should be brushed aside for the sake of the student, who finds the subject difficult enough without artificial stumbling blocks being put in his way. The student should be told simply that the median tela extends laterally as the roof of the lateral ventricle and this becomes infolded to form the lateral plexus. This process continues around the side wall just in front of the junction of the hemisphere and thalamus.

6. Dorsal and Ventral zones in the Diencephalon and Telencephalon

The sulcus limitans of His marks the boundary between alar plate (dorsal zone) and basal plate (ventral zone). The dorsal zone throughout the central nervous system is sensory, the ventral zone motor. Both zones include gray matter and fiber tracts belonging to the correlating mechanisms, and in those segments in which the primary sensory or motor centers are reduced or wanting owing to reduction or absence of the peripheral organs, the correlating mechanisms constitute the whole of the zone concerned. The writer has repeatedly (1902, 1905, 1906, 1909) emphasized the fact that the longitudinal zones constitute the most fundamental divisions of the brain and hence the sulcus limitans is the most important landmark in the brain. The two sulci converge at the anterior end of the brain to meet in the lamina terminalis and this meeting—point marks the anterior end of the central axis of the brain. The end of this axis His placed at about the middle of his lamina terminalis, namely in the recessus przeopticus. The facts set forth in this paper show that the chiasma region must be taken from the lamina «terminalis and added to the brain floor. Still, in the writer’s opinion, the central brain axis has its ending in the recessus prwopticus (figs. 43 and 44). The reasons for this are to be found in the following facts: (a) the ventral zone of the brain becomes greatly reduced in volume in front of the third nerve by the absence of all motor centers; (b) it is further reduced by the distribution of fiber tracts to various parts of the diencephalon and telencephalon; (c) the sensory centers are represented in the telencephalon by the large olfactory apparatus; (d) the correlating mechanism of the dorsal zone is greatly hypertrophied in connection with the olfactory centers and in higher forms in connection with the somatic cortical centers (neopallium). In other words the ventral zone at the front end of the brain is represented chiefly by the decussating fibers (optic chiasma and commissures of Gudden and Meynert) of the ventral commissural system, while the dorsal zone contains both sensory and correlating mechanisms which arevery large. These facts account for the bending down of the sulci limitantes to meet near the ventral border of the lamina terminalis. There is no evidence known to the writer tending to show that the recessus neuroporicus has any significance in this connection. It is only a convenient practical mark of the dorsal border of the lamina terminalis and the anterior end of the brain roof.

In the diencephalon the location of the sulcus limitans is still more diflicult. The typical formation of the ventral zone extends no farther forward than the nucleus of the III nerve, or at most the nucleus of origin of descending fibers in the medial longitudinal fasciculus. The ventral commissural system is interrupted by the downgrowth of the substantia reticularis to form the inferior lobes and mammillary bodies. This downgrowth has so completely altered the relations of parts in the diencephalon that it is practically impossible to trace a boundary line between dorsal and ventral zones. The inferior lobes themselves doubtless represent a part of the correlating substance of the dorsal zones (Johnston 1906, p. 277 and fol.).

7. Pallium of the Telencephalon

A long discussion has been waged over the subject of the general morphology of the pallium since the discovery by Rabl—Riickhard (1882) of the forebrain roof of teleosts. It would not be profitable to enter into the details of this discussion. The hypothesis of Rabl-Riickhard and Edinger was to the effect that lower forms possessed no true or nervous pallium, but that the membraneous pallium as seen in teleosts and other fishes has been transformed into a massive pallium by the development of nervous elements in it. The hypothesis of Ahlborn was that the anlage or beginnings of the pallium of higher forms must be found in the massive portions of the brain of lower forms, that a membraneous (ependymal) roof can never be transformed into a nervous pallium. For many years the Rabl-Riickhard—Edinger hypothesis dominated the field of forebrain morphology by sheer force of the authority of its sponsors. Studnicka made an effort to show the truth of the Ahlborn thesis, but for the time was overborne by Edinger and his followers. Although in his first work the present writer accepted Edinger’s views, a wider study of the subject led him to a treatment in 1906 much more nearly in accord with the view of Ahlborn and Studnicka. Kappers and others have added to the discussion and with the fuller knowledge of the comparative anatomical and embryological facts the general morphology of the membraneous and nervous portions of the forebrain may be regarded as a closed subject. Much of the discussion has been due to misunderstanding and difierences in the use of terms and it will be suificient here to define the terms applied to the parts of the forebrain and indicate briefly the differences in form in various classes of vertebrates.

The term hernxisphere is applied in the BNA to each half of the telencephalon. It would therefore include the right or left half of all that lies in front of a plane passing behind the interventricular foramina and the chiasma—ridge. It is well known that this portion of the brain is not hemispherical in form in all classes. It is somewhat so in cyclostomes, many selachians, amphibians, reptiles, birds and mammals. In Heptanchus, Hexanchus, Chimaera the hemi~ sphere is more elongated and the membraneous roof is more extensive. In ganoids and teleosts the width of the membraneous roof is greatly exaggerated. The nervous walls are rolled outward so that the membraneous roof is attached along the lateral or even latero—ventral aspect and arches up over the ventricle. This eversion of the forebrain Walls in teleosts has made the term hemisphere inapplicable in the descriptive sense. However, most of the organs which make up the hemisphere in other forms are present in the teleostean telencephalon and these organs hold the same fundamental morphological relations to one another and to other parts of the brain. Therefore the term hemisphere may be employed throughout the vertebrate series, although in no two classes does the telencephalon approach in the same degree the form of a sphere.

In each hemisphere are represented nervous and membraneous portions. The membraneous portions include the lamina terminalis and the tela chorioidea. The lamina terminalis is supposed to be coextensive with the anterior neuropore, but there is no neuropore in cyclostomes and teleosts and in some other vertebrates (some amphibians at least) the upper border of the neuropore is not marked in the early embryos. Where an unambiguous recessus neuroporicus exists it is the clear mark of the dorsal border of the lamina terminalis. Where this landmark is not clear an arbitrary border for the lamina terminalis must be placed at some distance in front of the interventricular fora/rn.r'na.. The tela forms the roof from the lamina terminalis to the tela of the diencephalon, from which it is separated by the infolded velum transversum. The term pars supraneuroporica of the lamina terminalis which was used by Burckhardt (1894) and is used by Edinger for this portion of the brain roof is wholly without justification.

The nervous portion of the hemisphere includes numerous structures the arrangement of which will be spoken of in the next section on nomenclature. The term pallium has been loosely used by various authors for the membraneous roof of the telencephalon, the dorsal part of the nervous portion and the ‘superficial cell layers in the nervous portion. Edinger uses it in all these senses and in the last edition of his textbook (1908, Bd. 2, p. 249) he distinctly states that the epithelial roof of the teleostean telencephalon is the pallium. “Dieses Dach der Hirnblase heisst Hirnmantel, Pallium cerebri. Dazu gehort auch der auf fig. 220 noch rein epithelial gebliebene Abschnitt, derselbe, welcher schon bei den Selachiern und Amphibien aus eigentlicher Gehirnsubstanz besteht.” This ambiguity is very unfortunate. Since we have the convenient term tela for the membraneous roof of the forebrain, the term pallium should be reserved for the cerebral cortex. The question, then, whether teleosts (or other forms) possess a pallium should be answered, not with RablRiickhard by pointing to the membraneous roof, but by ascertaining whether there is present any nervous substance whose fiber connections and functions warrant its being compared with the cortex of higher forms.

It is still too early to define in an exact way what is meant by cerebral cortex. It is not sufficient to define it as superficial layers of cells in the telencephalon because in all classes of vertebrates and in man, superficial gray matter is found in the forebrain whose fiber connections and functions are very different from those of the true cortex. To say that the cortex consists of superficial gray in the roof or dorsal wall of the forebrain gives no means of determining its extent or boundaries. Although the term cortex implies and was first used for superficial layers, it has come in recent years to signify the brain substance which constitutes certain functional mechanisms, whether superficial or not. It is necessary to define the cortex by its fiber connections and from the functional point of view. In mammals the general cortex is understood to be a collection of highly complex centers which exercise functions of correlation and control over bodily movements, etc., through lower sensory and motor centers. The sensory impressions coming to these cortical centers usually pass over chains of three neurones. ‘The existence of neurone chains of only two links connecting the peripheral sensory surface with the cortex is somewhat in dispute, but it is certain that such chains are relatively few in number. The general cortex provides in its structure the means for association and correlation between the areas concerned with various modes of sense impressions. This general cortex has its efferent pathway over the cortico-spinal tract and other bundles descending through the cerebral peduncle.

In lower vertebrates, in which the general cortex is not yet known, the telencephalon seems to consist of olfactory centers and corpus striatum and it" is generally believed that the first cortex to appear was olfactory in function. The writer was the first to attempt a definition of the olfactory cortex (1901, p. 239). It was pointed out that the cortical center receives olfactory fibers of the third order, not of the second order. The olfactory pathway consists of: fila olfactoria——bulbus and tractus olfactorius—lobus olfactorius and its efferent fibers—cortex. This definition of the cortex has since been adopted and further developed by Kappers (Kappers and Thennissen 1908, Kappers 1908). However, olfactory fibers of the third order run to other centers in addition to the cortex. In the diencephalon the nucleus habenulae and hypothalamus, and in the telencephalon itself, the epistriatum (nucleus amygdalae in mammals), receive olfactory fibers of the third order (Edinger 1896, Johnston 1898, 1901, Kappers 1906, 1908). The epistriatum fulfills this definition in selachians, ganoids and perhaps teleosts when there is no other part of the forebrain that does meet the conditions. However, in higher vertebrates (Kappers 1908) a part of the epistriatum becomes gradually pushed back until it finally occupies a position in immediate proximity to the pyriform lobe (nucleus amygdalae), while a true cortical formation appears in the roof of the hemisphere in dipnoans (Elliot Smith 1908) and all higher classes. Now the epistriatum of forms above fishes, whose history has been so beautifully traced by Kappers (1908) does not represent all of the formation to which he gives the name epistriatum in selachians. The writer has shown that the epistriatum in Petromyzon (1902) and Acipenser (1898, 1901) receives an ascending tract from the hypothalamus and Kappers (1906, 1908) has recognized this tract also in selachians. I have interpreted this as an ascending gustatory tract (1906, p. 304:). The center into which the tract enters at first (Petromyzon) receives secondary olfactory fibers, but in most fishes, especially the selachians in which the olfactory apparatus is highly developed, receives tertiary fibers as well. The entrance of an ascending, presumably gustatory tract, into a tertiary olfactory center in fishes creates a condition analogous to that found in the general cortex of mammals; namely, a center serving for the correlation of two sorts of sense impressions which are received over neurone chains of three links. We seem, therefore, to have in the so—called epistriatum of fishes a primitive olfactory cortex. The gray matter does not consist of superficial layers of cells, but forms part of the wall of the ventricle.

This primitive epistriatum, as seen in Petromyzon and Selachians, is not all accounted for in the history of what Edinger and Kappers call the epistriatum in higher, forms. The primitive epistriatum lies in the side wall (floor and roof) of the selachian forebrain (Kappers). The fiber connections which I have worked out in the greatest detail in Petromyzon and Acipenser, show that the epistriatal formation in the side wall continues. caudad to the border of the diencephalon, '5. e., nearly to the nucleus habenulae. This is the region called by Kappers the dorsal part of the praethalamus. That this is telencephalic territory is readily shown by the fact that the velum transversum is attached to the lateral wall of the brain ust in front of the ganglion habenulae and behind the peculiar structure here being considered. The primitive epistriatum therefore consists of (1) a dorsal or roof portion, (2) an epistriatum in the narrow sense resting upon the striatum (palaeostriatum, Kappers) and (3) a caudal portion forming part of the wall of the median ventricle of the telencephalon. I have shown (1906, Chap. 18) that the caudal portion is of greatest size and importance in Petromyzon, is still of considerable importance in Necturus, and in mammals is represented by a small structure called by older authors the paraphysis but shown by Elliot Smith (1896) to be a nervous structure. The caudal portion decreases in size and importance in the vertebrate series. The second portion is the true epistriatum which Kappers has traced through the phylogenetic series up to the nucleus amygdalae of mammals.

The dorsal portion of the primitive epistriatum is seen in the roof in Petromyzon and typical selachians, probably in the short roof overhanging a shallow lateral ventricle in Chimaera, Heptanchus and Hexanchus, and possibly in a corresponding structure in ganoids at the anterior end of the olfactory lobe. This structure has generally been wholly lost sight of in ganoids and teleosts and when it reappears in dipnoans and amphibians in exactly the same position and relations as in selachians it has been treated as a new structure, the pallial formation or hippocampus. It must be recognized that the ganoids and teleosts have no other significance than that of a side branch of the phyletic line. The dipnoan brain represents the next step in advance from the selachian, and in the dipnoans the pallial formation appears just where the dorsal part of the primitive epistriatum is found in selachians. The further history of this olfactory pallium has been so clearly worked out by Elliot Smith and others that no further comment on it is needed.

When all fishes are taken into account it is seen that all three parts of the primitive epistriatum receive olfactory fibers and ascending fibers from the hypothalamus. Only the dorsal portion develops into what is universally recognized as olfactory cortex in higher forms (hippocampal formation). Now if it be shown that the ascending tract (gustatory) from the hypothalamus enters the hippocampus we could say that throughout the Whole vertebrate series the archipallium (Elliot Smith) is a correlating center for olfactory and gustatory impulses. If it should prove true that the gustatory center is in the hippocampal formation, all parts of the cortex can be defined as correlating centers; the archipallium for olfactory and gustatory impulses, the neopallium for impulses coming from the eye, ear, skin, muscles and joints. General visceral sensation would be represented also in the archipallium.

I have long felt that the term epistriatum is an unfortunate one. In only the smaller number of "forms is it descriptive of the structure to which it is applied. The view expressed here and in 1906 is that the body which Edinger called epistriatum is a part of a more extensive formation which in primitive forms has essentially the same structure and connections in all of its parts. This statement of fact is subject to revision if further studies show it to be incorrect. With regard to the name, however, I find that the extension of the term epistriatum to include all of this formation has led to misconceptions of my meaning. This formation may be described as the. visceral correlating center of the telencephalon, or as the correlating substance of the visceral sensory zone of the telencephalon. Instead of the term primitive epistriatum used above,‘ this might be called the primitive visceral cortex. The dorsal portion of it becomes the true visceral cortex (archipallium) when it receives tertiary olfactory and gustatory fibers.

Some of the factors which enter into the definition of the term cortex cerebri may be indicated as follows:

a. The term is applied only to structures in the telencephalon (excludes lobi inferiores, etc.) ;

b. The afferent paths of the cortex are predominantly of the third order (excludes the secondary olfactory centers; the cortex shows an uncertain grade of development in the more primitive forms) ;

c. The cortex serves functions of correlation for afferent impulses of two or more kinds (olfactory, gustatory, optic, auditory, etc.; excludes the epistriatum sensu stricto or nucleus amygdulae) ;

Whether such correlating centers are superficial in position is not of essential importance. The general cortex of mammals is separated from the ventricle only by fibers related to the cortex itself, i. e., by its own white matter. ‘The question of superficial position is of much less importance in the case of the cerebral cortex than in that of the inferior olives, the medial and lateral geniculate bodies, and other centers which are separated from the ventricle by voluminous fiber bundles and gray masses which have no direct relation to themselves.

The point of view of the writer stands in contrast to that of Kappers who in his recent paper (1909) extends the concept of cortex to the lobus olfaetorius (“palseocortex”) although the centers concerned are simple secondary olfactory centers throughout the vertebrate series. His reason for this is that these centers occupy a superficial position in mammals (e. g., cortex lobi pyriformis). Elsewhere Kappers insists upon tertiary afferent pathways as the essential criterion of the cortex. The use of two different criteria at difierent times leads to confusion of thought. I would not apply the term cortex or palaeocortex to these secondary olfactory cenaters, but would use the simple terms lobus olfactorius, lobus pyriformis, etc. I would apply the term cortex to certain functional mechanisms. The above suggestions toward the definition of these mechanisms are of necessity incomplete and in part hpyothetical. If such a term as palaeocortex were used it should be applied to the morphological forerunner (homologue) of the true cortex.

8. Divisions and Nomenclature

The nomenclature of the brain adopted by the Basle commission is still the best that we have, largely because it embodied the results of the indefatigable work and keen insight of His. Before suggesting certain changes in the BNA tables to bring them into accord with the facts I wish to examine briefly the nomenclature offered by some recent authors.

Edinger has shown great fertility and enterprise in the production of new names in brain anatomy. Edinger’s terms have arisen from his comparative studies of adult brains and are the expression of his effort to present large and obvious relationships in attractive form. He considers the lamina terminalis as the anterior boundary of the diencephalon, agreeing with the BNA. The narrow portion of the brain extending forward from the optic chiasma (very long in Chimaera) he calls the praethalamus (1908, p. 194). When he comes to describe the telencephalon (p. 251) he describes the lamina terminalis as the plate which unites the two halves of the telencephalon. He treats the anterior commissure system as belonging to the te1encephalon and even speaks of the “recessus praechiasmaticus” as one feature of the telencephalon. Here is a contradiction for which there is no remedy in Edinger’s mode of treatment. The difficulty is augmented by Edinger’s definition of the primitive basal portion of the telencephalon (hyposphaerium): the primary and secondary olfactory centers and the corpus striatum. Now the floor of what Edinger calls prsethalamus is a secondary olfactory center which I have called the nucleus praeopticus. It receives fibers from the bulbus olfactorius and gives rise to some fibers of the tractus olfactohabenularis. Considerations of practical convenience and clearness would dictate that this secondary olfactory center be included in the telencephalon, not in the diencephalon.

Edinger distinguishes a neencephalon from a palaeencephalon. His palaeencephalon includes the lower segments of the brain together with that portion of the telencephalon which he calls the hyposphaerium. The neencephalon is the same as his episphaerium and includes the tertiary olfactory centers (Elliot Smith’s archipallium) and the general cortex (Grrosshirn, Elliot Smith’s neopallium). Is it true that the whole of the lower segments of the brain are to be set in contrast to that part of the telencephalon to which the name episphaerium is given? Are the centers for the cochlear nerve in the medulla oblongata, the inferior olives, the nucleus dentatus in the cerebellum and the auditory centers in the inferior colliculus and metathalamus older than the tertiary olfactory centers or the general cortex? Or is it true only that our knowledge of them is older? The terms palaeencephalon and neencephalon are undoubtedly useful as expressions of the functional evolution and growth organization of the whole brain; but as descriptive terms for the topographical features of the brain they would not be useful in the lower brain segments and are decidedly misleading when applied to the forebrain alone.

The terms hyposphaerium and episphaerium seem to apply fairly well in mammals, but I see no advantage in introducing new terms which will not apply to the brains of lower vertebrates as well.

In describing the minor divisions of the telencephalon Edinger is neither consistent with himself nor with the majority of authors. His description of the olfactory centers is quite confusing and contains several self-contradictions. The body into which the olfactory nerve enters he calls (1908, p. 252) the lobus olfactorius. Almost all recent authors have agreed to use the name bulbus olfactorius for this, while the term lobus olfactorius is given to the collection of secondary centers which make up a greater or less part of the body of the forebrain. To this posterior part Edinger proposes to give the name lobus parolfactorius. This term replaces the term area olfactoria which Edinger used earlier (1896, p. 141). The present use of lobus parolfactorius is likely to lead to confusion with the area parolfactoria Brocoe used by the BNA, which is only a specific part of the whole group of secondary olfactory centers. Edinger uses the term area parolfactoria in the BNA sense in figs. 24:7, 275, 279, which are old figures reproduced in this edition Without. revision. This disregard by Edinger of the usage of the majority of other authors is responsible largely for confusion which arises in the Work of younger authors or those who are not thoroughly familiar with the internal structure of the brain. For example, Fuchs (1908) call-s the bulbar formation in the frog larva “lobus olfactorius” and applies the term “hemisphere” to the rest of the telencephalon. The confusion of other authors who attempt to follow Edinger’s work would be less if Edinger himself always used his terms in the same sense. On p. 260 of the same book he uses the term lobus parolfactorius as synonymous with tuberculum olfactorium at least in reptiles, birds and mammals. This center, Edinger thinks, is a special center for the oral sense, an interpretation which G. Elliot Smith (1909) shows to be wholly improbable. To the secondary olfactory center which covers the lateral and ventral surface of the striatum Edinger gives the names lobus olfactorius (figs. 230, 231, 234), area olfactoria (figs. 247, 273, 274), cortex olfactorius (figs. 240, 265, 280), and nucleus taeniae (fig. 239).

Professor 0. J. Herrick in the course of a very valuable paper on the subdivision of the brain (1908) gives expression to the current idea of the telencephalon in the following sentences. “The telencephalon is well named. It is terminal, not only in position but also in point of time, having been added relatively late in the phylogeny to the rostral end of the original neural tube. The BNA has done well to omit from it the pars optica hypothalami which was originally tabulated as part of this region by Professor His. Originally developed as primary and secondary olfactory centers, it has added successively more and more complexity during the whole course of phylogenetic history.” I quote this not for the sake of criticising I-Ierrick’s work—for the whole spirit of his paper and most of the details of it are in perfect harmony with my own views but because it brings out clearly the difference between the results of the comparative study of adult brains and the results. of a complete genetic method in which embryology contributes its just share. The early development shows that as matter of fact the telencephalon is not added late in the phylogeny but is actually the first segment of the original neural tube in all classes of vertebrates. A certain part of this segment expands and grows in complexity with the increasing complexity of the vertebrate organism and of its mode’ of life. Further, it is clear that the telenoephalon originally contained more than primary and secondary olfactory centers. The existence of the nervus terminalis is evidence of this; the existence of preoral entoderm and of a well developed neural crest in the telencephalic segment of the embryo is evidence of it; the existence of a well developed correlating center, the corpus striatum, in the brains of all vertebrates, is further evidence. Of all the authors who have represented the telencephalon as purely olfactory in function, not one has shown or attempted to show that the corpus striatum is accounted for by its relations to the olfactory centers alone. The present writer is the only one who has given facts to show the pathway of impulses both to and from the epistriatum and the striatum in lower vertebrates. In my description of the brains of Acipenser (1898, 1901) and Petromyzon (1902) I showed that olfactory tract fibers ended in the epistriatum and that fibers arising from the cells of the epistriatum ended in the striatum. From the striatum the well known basal bundle (Edinger, Van Gehuchten) passed backward. These results have been confirmed by Kappers (1906, 1908) but Edinger has persistently disregarded the fact that in his descriptions of the forebrain no fiber tracts are mentioned which would enable either the epistriatum or the striatum to carry out any functions whatsoever. In the last edition of Edinger’s textbook the epistriatum is represented as an end-station for olfactory tract fibers but no fibers are described in lower vertebrates which go from the epistriatum to any other part of the brain. The striatum, on the other hand, gives rise to the tractus strio-thalamicus, but no fibers are described which come to end in the striatum. The epistriatum receives olfactory impulses but has no way of giving out any impulses; the striatum has an efl"erent pathway but receives no impulses. Neither of these important forebrain centers is provided with the means of carrying on any function.

As a further indication that the primitive forebrain has some functions in addition to the olfactory sense, the writer has described two ascending tracts to the forebrain. One of these, the tractus loboepistriaticus, is believed to carry up gustatory impulses to the epistriatum from the tertiary gustatory center in the hypothalamus (fishes and amphibia 1898, 1901, 1902, 1906). If this hypothesis is correct the epistriatum must be regarded as a correlating center for smell and taste and so a forerunner of the smell-taste cortex. A second tract has been traced in Acipenser (1901) from the tectum opticum only as far forward as the optic chiasma where it enters the telencephalon. Whether it ends in the corpus striatum or in some other part of the forebrain remains to be seen. In my textbook (1906, p. 336) I have pointed out that the entrance of such a tract as this into the telencephalon constitutes evidence of the beginning of the correlating centers which in higher vertebrates we call the neopallium. The writer has been convinced for some years that the elements or beginnings of all the chief parts of the telencephalon of mammals and man are to be found in the telencephalon of primitive vertebrates.

Herrick’s revision of the nomenclature of the diencephalon and mesencephalon contains two new terms, ophthalmencephalon, and medithalamus. As a pedogogic term based on function, “ophthalmencephalon” has my hearty approval. As a morphological subdivision of the brain it is open to the objection that the regions included—retina, chiasma, lateral geniculate bodies, pulvinar and tectum opticum—do not have sufficient morphological unity. The term medithalamus is ofiered by,Herrick provisionally for the things left over after the ophthalmencephalon has been set apart. It thus includes the central gray and a number of nuclei of diverse functions. The fact that it must include the medial geniculate body on the lateral surface of the diencephalon seems to the writer a fatal objection to the term medithalamus.

In the diencephalon the epithalamus and hypothalamus are fairly clearly marked both functionally and morphologically. The hypothalamus requires new definition both toward the thalamus and toward the telencephalon. The latter is furnished in the new facts brought out in this paper; the proper boundary between thalamus and hypothalamus can be determined only after We have fuller knowledge of the internal structure. The metathalamus and thalamus each presents morphological unity and cannot well be improved upon at present. The chief changes needed in the BNA at present are such as are required by the new facts regarding the telencephalon and the boundary between it and the diencephalon. These will be indicated below.

Fig. 45. A, transverse section of the telencephalon of Petromyzon.

Transverse Section of the Telencephalon. A true transverse section of any part of the central nervous system must cut across both dorsal and ventral zones of the neural tube and must cut roof plate and floor plate as nearly as possible at the same antero—posterior level. A little reflection will show that while such a section is readily obtained in any of the lower segments of the brain and cord, a true transverse section is seldom cut in the telencephalon. Such a section would pass through the optic chiasma and the interventricular fo-ramina. No other plane would out both roof plate and floor plate at the same level. Owing to the reduction of the ventral zone and the enormous expansion of the dorsal zone of the telencephalon in all vertebrates, only a small part of the sections of a series can out both zones, but a section in the plane mentioned may be taken as the true or standard transverse section of the telencephalon. figures 45 A, B, C, D show such sections of the fish, amphibian and human brain.

Fig. 45. B, transvere section of the telencephalon of Necturus; C, of human embryo ; D, of human adult brain.

Summary and Conclusions=

1. The neural plate is bounded by neural folds which meet in front at the terminal ridge.

2. The optic vesicles are evaginated from the dorsal part of the neural tube and are connected with one another by the p/rimritive optic groove.

3. The optic chiasma is formed in the terminal ridge and therefore occupies the anterior border of the brain floor.

4. The lamina terminalis is coextensive with the neuropore and in most‘ vertebrate embryos and many adults its upper border is indicated by a recessus neuroporicus. This is always in front of the interventricular foramina. At the lower border of the lamina is the recessus praeopticus.

5. The roof of the telencephalon is always a tela chorioidea.

6. The formation of the optic ridge in anticipation of the optic tract separates the hollow optic stalk from the primitive optic groove and it becomes connected secondarily with the preoptic recess.

7. The velum transversum is clearly present in mammalian embryos as in all other classes of vertebrates.

8. Just in front of the velum in mammalian embryos is a paraphysal arch.

9. The plexus chorioideus of the lateral ventricles forms immediately in front of the velum transversum and in mammals so involves the latter that its identity is lost.

10. The telencephalon is a complete segment or ring of the brain as His believed.

11. The telencephalon is bounded behind in young embryos by the optic vesicles and primitive optic groove; in adults by the velum transversum and the recessus postopticus; in mammals by a line or plane passing immediately behind the interventricular foramina and the chiasma ridge.

12. Externally this boundary is indicated by a furrow which in mammals is very deep and has at its bottom the fissura chorioidea.

13. The basal portion of this segment is reduced in volume owing to the absence of motor nuclei and other structures. It is represented by decussating tracts (optic chiasma, commissures of Gudden and Meynert) and perhaps by a certain amount of gray matter and some longitudinal tracts.

14. The dorsal portion of this segment is greatly enlarged and increases in size and complexity in the vertebrate series. Its increase is believed by the writer to be due to the development of the structures already present in this first brain segment in primitive vertebrates.

15. The sulcus limitans ends in the recessus praeopticus. All the olfactory centers and the corpus striatum belong in the dorsal zone.

16. The dorsal zone consists in other brain segments primitively of visceral sensory and somatic sensory columns together with central gray or correlating substance. The olfactory centers constitute the visceral sensory portion of the telencephalon. The somatic portion is represented in higher forms by the general cortex (neopallium), in fishes possibly by the beginnings of this cortex and by the sensory center for the nervus terminalis. The corpus striatum and epistriatuin seem to contain the correlating material from which the archipallium and neopallium have developed.

17. The revision of the BNA terms made necessary by the new facts brought out in this paper is indicated in the following table:

Mesencephalon (Pars ventralis—) pedunculus cerebri BNA. (Pars dorsalis——) corpora quadrigemina BNA. Diencephalon (Ventral and dorsal portions not clearly definable; four divisions based on purely topographical features).

Ventriculus tertius, pars diencephalica.

Velum transversum.

Epithalamus BNA. Metathalamus BNA. Thalamus BNA; requires more careful definition. Hypothalamus, modified BN A. Pars mammillaris hypothalami BNA. Pars infundibularis hypothalami Tuber cinereum BNA, Infundibulum BN A. Hypophysis BN A. Recessus postopticus. Telencephalon Ventriculus tertius, pars telencephalica. Foramen interventriculare BNA. Recessus praeopticus (to replace Rec. opticus BN A). Recessus triangularis BNA. I-Iemisphaerium. Pars ventralis hemisphaerii. Chiasma opticum BNA. Commissura superior (Meynerti) BNA. Commissura inferior (Guddeni) BNA. Pars dorsalis hemisphaerii. Lamina terminalis BNA. Commissura anterior (cerebri) BNA. Paraphysis. Corpus striatum BNA. Rhinencephalon BNA. Pallium BNA. Archipallium (including hippocampus, fornix, ebc.). Neopallium (including general cortex, corpus callosum, etc.).

Only so much of the BNA tables is included in the above as is necessary to show the changes that should be made to bring those tables into conformity with the new facts. Certain terms are added to those already in the BNA because they are necessary if the BNA is to be used in comparative neurology: velum transversum, recessus postopticus, recessus praeopticus to replace recessus options, and paraphysis. The chief changes proposed consist in the shifting of certain terms from the diencephalon to the telencephalon in accordance with the new boundary laid down, and a more complete tabulation of terms under the telencephalon in accordance with the results of a genetic and functional analysis of that segment. The changes are such that it will require very little effort for anatomists and neurologists to adjust themselves to the usage proposed. The advantages are that the tabulation and definitions proposed express accurately the actual relationships, harmonizing the facts of ontogeny and phylogeny with those of adult mammalian and human anatomy. The aim is to adjust neurological terms to the needs of comparative as well as human neurology and to avoid confusion arising from apparent discrepancies between embryology and anatomy,wbetween comparative and human anatomy. - These discrepancies now require much time and patient eflort in explanations to students. A constant efiort is needed to revise our terms to bring them into accord with the facts. Such is the object of the present suggestion.

  • Note to Pages 464 And 506. In a paper which has appeared since this article went to press, Hatschek (Morph. Jah/rb., vol. 39, 1909) reaches conclusions with regard to the anterior end of the head in cyclostomes almost identical with my own. His Basilarecke corresponds to my primitive optic groove, his Basila/rlippe to my terminal ridge. He states that the anterior pole- of the craniate body is marked by the H1/popm/sevneclce, where the floor of the neural tube and the roof of the archenteron end forward. I cannot agree with Hatschek’s statement (p. 519) beginning “Die Basilarlippe stellt den primitiven Vorderwall des Medullarrohres dar.” The Basilarlippe or terminal ridge belongs without doubt to the floor of the neural tube and is occupied by bundles of the ventral fiber decussations.


a.h.c., anterior head cavity. arch., archenteron.

au., auditory pit.

c.a., commissura anterior. cbl., cerebellum.

c.h. colnmissura habenularis. ch.op., chiasma opticum. 0.1)., commissura posterior. c.s., corpus striatum.

dicnc., diencephalon.

ec., ectoderm.

en., entoderm.

ep., epiphysis.

f.i., foramen interventriculare. hy., hypophysis.

l.t., lamina terminalis.

771., month.

md.. mandibular somite and arch. mes., mesoderm.

mesena, mesencephalon. m.m., median mass connecting the p1'ema1idibu1:1r somites. n., neuropore.

nch., notochord., nervus thalamicus.

0., bulbus olfactorius.

op.r., optic ridge.

012.42., optic vesicle.

1)., paraphysis., plexus chorioideus. m‘.en., preoral entoderm. pr.m., premandibular somite. r.i., recessus infundibuli. r.m., recessus mammillaris. 7'.n., recessu neuroporicus. r.p., recessus praeopticus. r.po., recessus postopticus.

1)., terminal ridge.

tel., tela chorioidea.

telenc., telencephalon.

thaZ., thalamus.

t.m-ec., mesectoderm derived from the terminal part of the neural crest. t.p., tuberculum posterius. v.l., ventriculus lateralis., velum transversum.

List of Papers Cited

AHLBORN 1883. Untersuchungen fiber das Gehirn der Petromyzonten. Zeit. f. wiss. Zool., V01. 39.

BROHMER 1909. Der Kopf eines Embryos von Chlamydoselachus und die Segmentierung des Selachierkopfes. Jenaische Zeit. f. N atu/rwiss., vol. 44.

BROMAN 1896. Beschreibung eines menschlichen Embryos von beinahe 3 mm. Léinge mit specieller Bemerkung iiber die bei demselben befindlichen Hirnfalten. Morphol. Arbeiflen, vol. 5.

DOHRN 1904. Studien zur Urgeschichte des Wirbe1thierk'o'rpers. 23. Die Mandibularhiihle der Selachier. 24. Die Prfimandibularhfihle. ‘Mittheil. Zool. Sta. zu Neapel, vol. 17.

EDINGER 1896. Vorlesungen iiber den Bau der nervtisen Centralorgane des Menschen und der Thiere. 1908. Vorlesungen u.s.W., vol. 2.

EYCLESHYMER 1890. The Development of the Optic Vesicles in Amphibia. Jour. Morph, vol. 8.

FUCHS, FANNY 1908. Ueber die Entwicklung des Vorderhirns bei niedern Vertebraten. Z001. Jahrb., Anat. u. 0n.tog., vol. 25.

GAGE, S. P. 1905. A three weeks’ Human Embryo, With especial reference to the Brain and the Nephric System. Amer. Jowr. Anat., vol. 4.

GAUPP 1898. Zirbel, Parietalorgan und Paraphysis. ‘Market u. Bennett; Ergeb., vol. 7.

GUTHKE 1906. Embryologische Studien fiber die Ganglien und Nerven des Kopfes von Torpedo ocellata. Jenaische Zeit. f. Natm‘w., vol. 42. Hmuzxcx, O.

JUDsoN 1908. The Morphological Subdivision of the Brain. Jam‘. of Comp. Neurol. and Psuch., vol. 18. Hrs 1892. Zur Allgemeinen Morphologie des Gehirns. Archiv f. Anat. u. Entw4okl., 1892. 1893. Ueber das frontale Ende des Gehirnrohres. Arch?‘/v f. Anat. u. Entwiokl, 1893. 1895. Die anatomische Nomenclatur. Archie f. Anat. u. Phys., 1895.

HOFFMAN 1896. Beitrlige zur Entwlckelung der Selachii. Morph. Jahrb., vol. 24.

1898. The Olfactory Lobes, Fore_Brain, and Habenular Tracts of Acipenser. Zool. Bulk, vol. 1.

1901. The Brain of Acipenser. Zool. Jwlwbeicher, vol. 15.

1902. The Brain of Petromyzon. Jour. of Comp. Neurol., vol. 12.

1902a. An Attempt to Define the Primitive Functional Divisions of the Central Nervous System. Jour. of Comp. Neural, vol. 12.

1905. The Morphology of the Vertebrate Head from the Viewpoint of the Functional Divisions of- the Nervous System. Jour. of Comp. Neurol. and Psych, vol. 15.

1906. The Nervous System of Vertebrates. Philadelphia, P. Blalcistonhs Son «:6 00.

1909. The Central Nervous System of Vertebrates. Ergeb. u. Fortsohr. der Zoologie (Spengel).


1906. The Structure of the Teleostean and Selachian Brain. Jour. of Comp, Neurol. and I’s1/ch., vol. 16.

1908. Weitere Mitteilungen fiber die Phylogenese des Corpus Striatum und des Thalamus. Anat. Anz., vol. 33.

1909. The Phylogenesis of the Palaeo-cortex and Archi-cortex compared with the Evolution of the Visual Ne0—corteX. Arch. of Neural. and Psycmatiry, vol. 4.


1907. Zur vergleichenden Anatomie des Vorderhirnes der Vertebraten. Anatmmscher Anz., vol. 30.

1908. Die Phylogenese des Rhinencephalons, des Corpus striatum und der Vorderhirn-kommissuren. Folia N eurabiologica, vol. 1.


1905. Beitrage zur Entwickelungsgeschichte der Kopfganglien und Sin neslinien bei Selachiern. Jenaische Zeitsch. f. N aturw., vol. 39. Kourzorr.

1902. Entwlckelungsgeschichte des Kopfes von Petromyzon planeri. Bull. de la Soc. Impe/r. d. N atwral. de Moscou, Année, 1901.

VON Kurmmz.

1893. Studien zur vergleichenden Entwickelungsgeschichte des Kopfes der Cranioten. Heft 1. Die Entwickelung des Kopfes von Aclpenser sturio an Medianschnitten untersucht. Mimchen u. Leipzig. Locr.

1895. A Contribution to the‘ Structure and Development of the Vertebrate Head. Jour. Morph., vol. 2.

Low A. Description of a human embryo of 13-14 mesodermic somites. (1908) J Anat Physiol. 42(3): 237-51. PMID 17232769 | PMC1289161 Low

1908. Descriptiox of a Human Embryo of 13-14 Meodermic Somltes. Journal of Anat. a/nd PM/8., vol. 42.

MINOT. 1901. On the Morphology of the Pineal Region, based upon its Development in Acanthias. Am. Jaw‘. of Anat., vol. 1.

NEAL. 1898. The Segmentation of the Nervous System in the Squalus acanthias. A Contribution to the Morphology of the Vertebrate Head. Bull. Mus. Comp. Zool. Harvard 0011., No. 31. PLATT.

1891. A Contribution to the Morphology of the Vertebrate Head based on a Study of Acanthias vulgaris. Jour. Morph, vol. 5.

RABL-BficKHAm>. 1882. Zur Deutung und Entwicklung des Gehirns der Knochenflsche. Archie: Anat, Phz/s., 1882.

SMITH, G. ELLIOT. 1896. The Brain of a Foetal Ornithorhynchus. Quart, Jour. Mic. Sci, vol. 39.

1908. The Cerebral Cortex in Lepidosiren, with comparative Notes on the Interpretation of certain Features of the Forebrain in other Vertebrates. Anat. Anz., vol. 33.

1909. The Tuberculum olfactorius. Anat. Anz., vol. 34.

Suznzl. 1907. I1 sistema nervoso centrale dei vertebrati. Vol. 1. Ciclostomi. Padua.

STUDNICKA. 1895. Beitrfige zur Anatomie und Entwickelungsgeschichte des Vorderhirns der Cranioten. Sitzungsb. d. kim-igl. béhmischen

Gesellsch. d. Wissensch. Math-naturw. Olosse. 1896. Beitréige zur Anatomie und Entwickelunggeschichte des Vorder hirns der Cranioten. Sitzungsb. d. lc¢')'m‘gl. biihmischen Gesellsch. d. Wissensch. Math-naturw. Olasse.

VAN WIJI-IE. 1882. Ueber die Mesodermsegmente und die Entwickelung der Nerven des Selachlerkopfes. Amsterda/m. Zmomn, H. E. 1908. Die phylogenetische Entstehung des Kopfes der Wirbeltiere.. Jenaische Zeitschr. f. N atm-w., vol. 43.

Cite this page: Hill, M.A. (2024, April 19) Embryology Paper - The morphology of the forebrain vesicle in vertebrates. Retrieved from

What Links Here?
© Dr Mark Hill 2024, UNSW Embryology ISBN: 978 0 7334 2609 4 - UNSW CRICOS Provider Code No. 00098G