1897 Human Embryology 27

From Embryology
Embryology - 22 Oct 2017    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)

Minot CS. Human Embryology. (1897) London: The Macmillan Company.

Human Embryology: Introduction | The Uterus | General Outline of Human Development | The Genital Products | History of the Genoblasts and the Theory of Sex | The Germ-Layers | Segmentation | Primitive Streak | Mesoderm and the Coelom | Germ-Layers General Remarks | The Embryo | The Medullary Groove, Notochord and Neurenteric Canals | Coelom Divisions; Mesenchyma Origin | Blood, Blood-Vessels and Heart Origin | Urogenital System Origin | The Archenteron and the Gill Clefts | Germinal Area, the Embryo and its Appendages | The Foetal Appendages | Chorion | Amnion and Proamnion | The Yolk Sack, Allantois and Umbilical Cord | Placenta | The Foetus | Growth and External Development Embryo and Foetus | Mesenchymal Tissues | Skeleton and Limbs | Muscular System | Splanchnocoele and Diaphragm | Urogenital System | Transformations of the Heart and Blood-Vessels | The Epidermal System | Mouth Cavity and Face | The Nervous System | Sense Organs | Entodermal Canal | Figures | References | Embryology History

Note - this online text is only at a very early draft stage and contains many errors from the original scanning.
Historic Disclaimer - information about historic embryology pages 
Mark Hill.jpg
Pages where the terms "Historic Textbook" and "Historic Embryology" 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 and interpretations may not reflect our current scientific understanding.     (More? Embryology History | Historic Embryology Papers)

Chapter XXVII. The Nervous System

The formation of the vertebrate cerebrospinal axis has already been treated at length, pp. 173-181. In its first stage it appears as the medullary tube with ectodermal walls. The second stage is the differentiation of the brain from the spinal cord by the enlargement of the anterior end of the tube. The sharp distinction Avhich we have just drawn between the stages does not maintain itself in the amniota. In fact the medullary groove widens at its cephalic end before it closes to form a tube, so that the brain is indicated in the embryo before the medullary tube is formed. Moreover the development of the brain progresses while the groove is closing, so that the brain is already quite advanced before the medullary tube is closed at its caudal end. These irregularities in the development of the central nervous system render it impossible to decide at present whether the simple medullary tube without a brain enlargement, or a (perhaps solid) central nervous system with a brain enlargement, represents the phylogenetically primitive condition. The diflSculty of reaching a decision is still further increased by the fact that the tubular condition of the nervous system was probably acquired within the vertebrate series, see p. 180.

Definition of the Brain

The vertebrate brain is the anterior lx)rtion of the medullary tube, and is characterized by two primary features: 1, the enlargement of the tube; 2, its special associations with higher sense organs (olfactory, visual, and auditor}'). The brain is further characterized in all true vertebrates : 1, by having three principal enlargements separated from one another by two constrictions (H. Ayers, 90.1, claims that the three enlargements can 1x3 traced in Amphioxus also) ; 2, by being bent at the region of the second enlargement (mid-brain) owing to the development of the head-bend of the embryo ; 3, by containing the principal centres for the co-ordination of sensations and movements. All modifications of the brain can be traced back to this primitive type, and it seems probable that the evolution of the brain has been dominated by the advantages of more perfect co-ordinating apparatus, as the special senses on the one hand and the locomotive organization on the other acquired a higher development. ♦

Cerebral Vesicles

The enlargement which produces the brain extends about half the length of the embryo, compare Figs. 114 and 155, and takes place unevenly, so that there are produced three successive lobes, which are known as the primary" cerebral vesicles, Fig. 113 and 114; the second and third vesicles (mid-brain and hindbrain) are often imperfectly divided from one another. The three vesicles subsequently subdivide, so as to form — to follow the traditional description — five secondary vesicles. It has long been customary to describe the medulla as dilating to form the three ajid later 6ve vesicles, but unfortunately the descriptions have been 8i"> much conventionalized in subservience to tradition that they an? misleading in several important respects. The attempt is here made to give an untrammelled objective account.

Optic Evaginations

The first indication of brain formatioa seems to me to be tbe widening of the extreme anterior end of the medullary plate or groove, which can be recognized in all vertebrate embryos at a very early stage. In elasmobranchs it appears to mo evident that the widening is due to the very process of concrescence itself, and is initiated while tbe ectental lines are approaching one another, and ia fully marked before tbe longitudinal axis of the embryo is completed by concrescence. Fig. 317 represents a di^-fish embryo; m is the point at which concrescence has begun ; it will be oh8er\-ed that the embryonic r m curves around this point and in consequence s spread out laterally; in later stages the lateral protrusion, which we see initiated in ,^ P Fig. 317, at )M, becomes still more marke<l and can be followed until it is evidently the optic diverticulum. In mammals we find the medullary groove specially widened at its anterior end — noticeably so in the mole. Fig. 09, op. A cross section through the optic vesicle at this stage offers a very singular appearance. Fig, 100; the entoderm, Eii, haw not closed over, although tbe notochord, nch, B already distinguishable under the medullary groove; the ectoilerm, Ec, is greatly thickened on the dorsal side to form thover\vide medidlary plate, which has a median depression, Mp, corresponding to the medullary grtx)ve proper, and two latend depressions corresponding each to an optic vesicle. If we imagine the medullary plate to bend upward and to (rlose over itself, then tbe two edges of the optic depressions, op, which are outermost in Fig. ion, will meet in the median line, and as soon as the groove, by closng, becomesatnbe, there will lie at this jxiint two lateral diver ticida, liaving the same cliaracteristic«l]y thickened ecto<lerraal lining as the rest of the metlullary tube. These diverticula are the so-called optic vesicles, which "^° are ultimately transfonne<i into the optic I m mj nerve, retina, and choroid of the eye.

n hra Q In the chick the optic vesicles become clear neb Jd "" ly indicated by the twenty-fourth hour, when i Su "a "" there are from five to seven distinct pairs of primitive segments, and the head projects si ghtl o er the proan niotic area. Before the me<lullary groove has clos hI anywhere the optic diverticula are quite distinct. In a chick of twenty-nine hours, Fig. 336, the vesicles, op, are very large, their growth being an important factor in the precocious distentiou of the head.

Widening of the Medullary Tube

While the optic vesicles are developing the medullary tube expands in diameter throughout its cranial or anterior half, without there being at first much change in the structure of its walls or much evidence of subdivision, but very soon the expansion becomes unequal, so that the tube is slightly constricted immediately behind the optic vesicles. Fig. 336, op; then follows a slight dilatation, F', the mid -brain {Mittelhim) , which is separated by a second constriction from the long and large hind-brain, T" {Hinterhirn) , which is widest in front and gradually diminishes in diameter, and merges without distinct boundary into the posterior unexpanded portion of the medullary tube or future spinal cord. Transverse sections show that the widening, by which the brain is differentiated from the cord, is due chiefly to the enlargement of the medullary cavity, and that the walls change but little in thickness until the three vesicles are differentiated, when the walls begin a series of characteristic moditications.

The three primary vesicles {Gehimhliischen, vesiculce cerebralea) were known to Malpighi and Haller according to Tiedemann, 61.1, 0. Btschoff, 45.1, 170, appears to have been the first to observe that they are formed before the medullary groove is entirely closed in the cephalic region. Owing to the fact that the optic vesicles grow out so early and that the remainder of the brain as a whole widens out, we ought, perhaps, to accept A. Goette's view, 76. 1, 280, that a double division precedes the triple. In this case we should have to describe the mid-brain and hind-brain as arising hy the subdivision of the second primary enlargement.

1. The Fore-Brain

As we have seen above, the fore-brain originally includes the optic vesicles, which primitively show no trace of any demarcation from the central portion of the fore-brain. This condition, however, does not last long, for the central l>ortion of the fore-brain soon begins to expand upward and for\vard, making I a separate central enlargement, which a may be designateil as the penmment | fore-hrain. Meanwhile the distal ends i of the optic diverticula al.«o dilate rap- ' idly, whilothepart of each diverticulum nearest the fore -brain proper grows slowly. It is often erroneously statetl that part of the optic vesicle is constricted: in reality it enlarges, though pJii'Biiii. ami npi"ib"\rak.ieS'"of"i relatively slowly. From these modifi }'PF^h4io ""e^ «ti^e™ * ^aS". (■ations there are developed a wide me- iwreof lens ..p oinii vkici^ En ™ dian fore-brain and two lateral optic '*^*™' *"" B^'f""- ""i >'«^i'" vesicles connected by tubular stalks with the ventral side of the brain proper. Fig. .337.* In short, the primitive vesicle is divided into three parts, one merlian and two lateral, an<l it is only the median part that enters into the formation of the brain. The bii^itory of the median division is, therefore, treated in this chapter, while that of the two lateral divisions is dealt with in Chapter XXVIII., on the organs of 8enHi5. It ma3% however, be state^l now, in order to facilitate the comprehension of the figures, that the optic vesicles expand dors-^ 1waru, Fig. 1337, op. It should l>e noted that the walls of the forebrain and optic vesicle are still nearly uniform in thickness, and. W} far as yet oljserve<l, in stnicture. The changes described in this paragraph occur in the chick at alx)ut thirty-two to forty hours, in the rabbit the ninth day, in m:i!i about the eighteenth day.

  • Compare also ngii. 170, ITI, aod ITS.

The next series of changes in the fore-brain lead to the diflFerentiation of the cerebral h(?mispheres. By a long-c<mtinued tradition it has ^HHtftnc ciistomary to describe the process as the subdivision of the primary v(.*sicle into two secondary vesicles, designated as the fonj-brain i)n)])er (Vo rder him ^ prosencephalon) and ^f ween -bra in (Zirischenhirn^ thfilnmenreph(tlon). Such a description, however, seems to me hardly justifie<l either by embryology or comparative anatomy, and to lie esixxrially apt to mislead and confuse. In fact every embryologist must admit that it is scarcely cc^rrect to say that the fore-brain divides into two vesicles, from the anterior of which the cerebral hemispheres grow out. It is more in accord with the actual facts to describe the hemispheres as appendagf^s of the fore-bniin, that is to say, of the so-called Zwischenliim or thalam(»ncephalon. Accordingly I pjeseiit the history of the origin of the c<»rebral hemispheres somewhat differently from usual, though, of courses without changing the facts. Spc t ^ For conv(»nienco I defer mention of the head-lK»nd (set? p. Ooo), which F,«. m-Brainnf K,ni,ry<, N... :«. n «,7 develops while the hemispheres are

large and finshes itself, so to speak, forward and, owing to the hetwl-liend, downward. The flexure is at first slight, but incrciases as development prot^eeds, compare p. fiOO. Th(» (^nlarg(»d end of the medullary tul)e is in no way divideii off from th(» first c(»n?l)ral v<»sicle until the en<l begins to dilate toward eacli side to pnxluce the hemis])heres. The manner in which the homisphf»ri»s grow out can Ik» iK^tter underst(XKl from the Figs. ;J3k, :vM)^ and :J4n, than from any nu^re d(»scription. At first, as just indic^ite<l, they form an undivided coinm<^n anterior enlargement, but th<» lateral ex|Mmsion lK,'gins very early, and with it the anlages of the two h('misph(T(^s are given. If the ])Osition of the hemispheres is ol)st»rv<'<l can»fully. Fig. ;j:58, i/, it will Ik* si»eu at onct» that it is the pnMlu(!t of the dorsal side, luid that the ventral half of the primi

tive fore-brain, an ahown by W. His, 89.4, does not participate in the outgrowth. The coneideraticoi of this important fact demonBtratee that the hemiBpheres cannot be strictly compared with one of the primary vesicles, each of which includes a ventral as well as a donukl portion of the medullary tube The origin of the hemispheres from the dorsal side has so great importance morpholf^cally that i?pecial emphasis ^ mufat bo laid' upon the , fact. The ventral / boundarj'of the hemi k spheres must be placed \ near the optic stalks, ' so that the hemispheres include that portion of the bram wall which unites with

tliP prtrxlprm to fnrm ^"> *>* — Reemunnictlon of Uip Brain o( HlB Embijo Eo

TUe eciouenn lO lOrm ofatkenldnge, lO a mm) Mb Mtdlirsln /B (ore-hrifn: H.

the olfactory plate, al- hemUphwe « olUctorj lobe Op craUo nerve hy hypophy readydescnbed,p5:5 •'» ™ '-•'•n^mamn.llare Af.er^'-Hi.L

The cerebral hemispheres grow more rapidly than any other part of the brain, see Fig. 3:)9, H, but liieir growth is principally in their distal parts, so that, like the optic vesicles, they become large pouches connected by relatively small hollow stalks with the fore-brain. The stalk is short. The passage through the stalk is called the foramen of Munroe, Fig. 340, f.m. As this foramen enlarges but little, while the brain increases enormously, it appears in the adult as a small opening in proportion to the size of the whole brain. Althoi^h the foramen enlarges absolutely, it is sometimes described erroneously as becoming smaller during development. While the hemi^ _ ^ spheres are expanding the olfactory .plate,

Fig. 339, 01, acquiree more marked differentiation beneath them ; vnd shows traces of division into a dorsal or anterior, and ventral or posterior, lobe. Even at the stage of Fig. :)39, it can still be _ recc^^ized that the ol '"'^ factory region corre Fig. 8*),— Rn.iiiiirtnn-i«! Hedlu View of the Fore-Brnia ot ajvvnAa tn ^vhnt \PaH

His' Enibiyo Ko (.V«rt™(diifle, 10,3 inm.>, H. Hemlaphe: fPOnUS TO wnat WHS,

/.iH, foramfiiofMiiiiroe: if.o.PBoemusoptlciu: (.ctuberclne- before the brain WaS

n..um;m.corp™m«D.n,n«re:«6.mid.bmio. After W. HU.

treme anterior wall of the fore-brain. But the olfactory region ia already paired, and is associated in its development with the hemispheres. This leaves a part of the wall of the fore-brain in the median line, Fig. 340 (between the reference lines f.m and R.o), which is known as the /amino terminalis and represents throughout life the extreme anterior wall of the fore-brain. As seen in Fig. 340, it extends from the level of the foramen of Munroe to the level of the optic stalks. In the same tigure it can also be seen that the hemispheres and olfactory lobe project further forward than the lamina. The hemispheres expand, not only upward and forward in regard to the longitudinal axis of the fore-brain, but also backward, as can be well seen in Figs. 339 and 341. The history of the hemispheres is given more fully and for later stages below, p. 690.

The primary differentiations of the floor or ventral wall of the fore-brain are also clearly indicated in a human embryo of 10-12 mm. {Nackenldnge) y Figs. 339 and 341. The lower part of the fore-brain has expanded, forming, as it were, a hanging pouch. Fig. 339, from which pass off the optic stalks. Op, Following the median wall of the pouch around from the mid- brain to the level of the foramen of Monroe, Fig. 340, /.///, we find, firsts a protuberance, m^ which extends nearly half-way to the optic stalk, and indicates the future mammillary bodies; second^ a slight swelling, t.c, which marks the future tuber cinereum; thirds the future apex of the infundibulum ; fourth^ the area of the brain wall united with the hypophysis; midjiffh, the lamina terminalis, just beyond the recessus opticus, Ii,o,

2. The Mid-Brain

The second cerebral vesicle undergoes less modification than the first and third. Its walls are at first of nearly uniform thickness, see Duval, ** Atlas, Fig. 255. It is oval or round in transverse section. It is situateil at the ix)int where the head-bend takes place ( compare p. 000) , and by the head-bend its shape is profoundly altered, its dorsal 8urfiic*e becoming more arched and expanded. Fig. 338, Jl/ft, while its ventral wall as seen in profile becomes concave; further, the dorsal wall becomes relatively much thinner than the ventral wall. The cavity of the mid-brain remains very large, and during the early expansion of the brain the communication between the foro-brain and mid-brain enlarges more than does the passage l)etwe<»n the mid- and hind-brain. This is commonly expressed by saying that the constriction between the first and second cenjbral vesicles is much less marked than between the second and third.

In the lower vertebrates the fore-brain and hind-brain do not advance either in growth or complication as in the amniota. In binis and reptiles the mid-brain develops to a greater extent than in mammals, and in the embr3'o early acquires great size, see Fig. 390, II. In mammals, on the other hand, the mid-brain grows more slowly. Roughly speaking, then, we may say that the imix>rtance of the midbrain diminishes as we ascend the vertebrate series, and that it does not participate in the advance of organization which characterizes the first and third cerebral vesicles.

3. Hind-Brain

The third cerebral vesicle is especiall.v characterized b}' the great exj^msion of its very thin dorsal wall, by the thickening of the doi-sal wall imme<liateh' l)ehind the constriction sepsirating the second from the third vesicle, and by the great and prominent bend formeil by the ventral wall of the hind-brain. Fig. 341, Hb. The thin dorsal wall corresponds to the epithelial ependyina of the adult ; its morphological significance is explained in the section on the zonea of His, p. fiOO. The dorsal thickening is the anlage of the cerebellum and corresponds to a commissure found in the lower vertebrates. The apex of the ventral flexure is the anlage of the pons Varolii of the adult. The thickened floor of the hind -bra in, between the pons and the spinal cord, sp.c, gives rise to the medulla oblongata. We thus have the four chief structures, which develop from the hind-brain, definitely mapped out by the earliest changes. The modifications w^hich result in this four-fold differentiation all take place simultaneously iuid are interdependent. They are the result of two factors; 1, the unequal development of different regions of the medullary walls; 2, the appearance of the Varolian bend (Briickenkriimmung). These factors are considered later.

It is usually stated that the hind-brain subdivides into two vesicles, for which the names secondary hind-brain and after-brain (yachhini) have been employed; the Nachhirn is the part nearest the spinal cord. In fact, it ia convenient to designate the anterior part of the hind-brain, out of which the cerebellum and pons Varolii arise, as the hind-brain proper (meteiicephalon) and the posterior part as the Nachhirn (epencephalon or jn telencephalon) or, better, as the medulla oblongata. On the other hand, it is incorrect to speak of the primitive hind-brain as forming two secondary vesicles. This error goes back to the time of Von Baer, II., 10(1, who ol>serve<l such division in the chicken embryo. It has also been described and figured by Mihalkovics, 77.1, 25, Taf. IV., Fig. ;t3, in a chick of fiftyeight hours. These authors, and most otiiers who have written on the subject, assumed that their observations were upon a constant and typical condition. In reality there is great irregularity in the growthofthe walls of the hind-brain, and sometimes in birds and perliapK in reiitiles the third cerebral vesicle is temporarily more dilated at its anterior end than elsewhere. The dilatation soon disappears, and no i)roof lias been brought yet, to my knowledge, to establish an identitj" between it and the region corresponding to the cerebellum and pons — it seems to take in more than the cerebellum, less than the pons. In chicken embryos the separate dilatation is usually wanting, and it has, so far as I know, never been observed in any mammalian or ichthj-opsidan embrjo. It is interesting to note that Balfour, "Comp. Embryol,." II., 424, though he does not expresshr mention the error of the traditional description, yet skilfully avoitu adopting it in his account of the hind-brain.

The shape of the hind-brain requires more detailed description. As seen in Fig. XiS, the hind-brain at the time of the development of the head-bend is more than equal to all the rest of the brain tn length. It begins with the constriction or isthmus behind the mid

Fig, BU,— Hlnd-Brain

brain and at first widens rapidly, then gradually tapers to the neckbend, where it passes into the spinal cord. Viewed from the dorsal side, Fig. 343, the anterior constriction or isthmus ia still more noticeable, and we can also see the kite-shaped outline ~^ of the thin roof. Comparison of the ngure with

the following, Fig. 34:j, representing a slightly older stage, affords an idea of the widening of the medulla, while comparison of Figs. 'S'-iS and 341 will indicate its motiifications as seen iu proBle. It is important to observe that there is, as yet, no cerebellum, but only a thickening of the dorsal ivalt close to the isthmus. This thickening is the anlage of the cerebellum, and is to be lioraologized with the commissure found in the corresponding position in Ichthyopsida.

Cerebral Flexures

-. liu'Ltt), of the neuron may bo described as B™inof«Huiii«iEiii ■een from the dorwl i hiyo ot one Mruiih

side, nib. MId-bnIa: Straight, for it IS actually ver\' (rflH' Ru). lib, Hid faiirthv<-ntriclF;.4|i.<'. nean\ SO, Up lO U\0 Stage wnen i„^. iv. fourth veu giiDsUiirri. Afiprw. the optic vesicles begin to be con- "-icj": .^'i,,^""' 1«. Compare FlR.SM. . • J^ j a- c- im i r,.,,. I'ord, After W. HIk.

stnctedoff — sec Figs. HO and 3J(i. While the dilatation to form the second cerebral vesicle or mid-brain is taking place, the primary head-lwnd of the embryo ia established, involving the brain. The bend of the brain takes place at the level ot the mid-brain : the fore-brain is bent over ventralward until it forms a right angle with the hind-brain, Fig. .'(:i8, the actual flexure being almost confined to the mid-braiu, in which, as can be seen in the figure, the cerebral axis curves very much, while in the hindbrain it remains nearly straight, and in the fore-braiu is slightly bowed only. This bend may bo called the mid-brain or primary flexure.* During the early st^es of the hemispheral outgrowths the flexure increases until the axis of the fore-brain forms an acute angle with that of the hind-brain, Fig. :i'2n. Mihalkovics, 77.1, 30, proposes to distinguish the right-angled stage as the Hakeiiliriini)iiunf/. and the later acute-angled stage as the Kofpbeuge. Such a distinction is entirely arbitrary', and the suggestion has nut been adopted. The angle becomes ultimately so sharp that the floor of the fore-brain becomes nearly parallel with that of the hind-brain. The 8eci>nd bend to appear is at the junction of the hind-bmin (meilulla oblongata) and spinal cord. Fig. :j:!S, and is termed the neck-bend (yttckenkriimmiinff). Like the primary- bend it affects the whole head ; the summit of its angle appears in the embryo when seen in profile, compare Figs. "220 imd 22.1, during several early stages as a projection (His' Nackenhocker), which is, however, soon obliterated. The neck-ltend develops later than the head-bend, not appearing in mammals until the hemisphere anlages have begun to grow out separately. It is very slight in the Ichthyopsida ; in the reptiles and birds it is more developed, but it attains its maximum only in the mammalia, and notably in man. In human embryos the neck-bend increases from the third to the end of the fifth week, when it reaches its maximum, the hind-brain then forming nearly a right angle with the spinal cord, Fig. 341. Later the bend becomes less again, owing to the gradu^ erection of the head as already described and illustrated in Chapter XVIII. for the human embryo.

The third cerebral flexure is known as the Varolian bend (KoUiker's Briickenkrilmmung) and is essentially different from the two flexures just described, for it is not a bend of the whole medullary tute, as are they, but a bend of the ventral side of the hind-brain. Fig. 341, the dorsal side remaining as seen in profile, nearly straight. As already mentioned, the greater part of the dorsal wall of the hindbrain is a thin membrane, and this membrane takes no part in the formation of the Varolian bend, which depends on the growth of the thick walls of the floor of the hind-brain, and with this growth the bend increases, its formation being accompanied by the lateral exi)ansion of the hind-brain at its anterior or cerebellar end, Fig. 343.

The cause of all the cerebral flexures is, of course, the unequal gro^vth of the various parts. Herein the growth of the brain is certainly the principal factor in determining the result. The general conception of the influence of the unequal growth of the brain dates back to Von Baer, and was revived by Rathke. W. His was the first to attempt an analysis of the mechanical conditions, and to demonstrate that the shaping of the brain depends to a large degree upon these conditions, which are many of them relatively obvious and simple. His has given in his semi-popular work, "Unsere Korperform," 74.1, pp. 93-118, an admirable presentation of his results, which have not yet received from embryologists the attention which their exceptional importance demands.

Origin of the Sensory Ganglia

To fully understand the historj'- of the ganglia the reader should consult the section on the ganglionic sense-organs in the following chapter. The origin of the ganglia has been carefully traced in a human embryo * with thii*teen segments, by M. von Lenhossek, 91.1, three of whose figures I reproduce. Fig. 344. As seen in A, the ectodermal cells, GZ, which immediately adjoin the medullary plate, differ in size and by their rounded form from the cells of the neighboring ectoderm and of the medulla. These cells constitute two bands, which unite in a single median band when the medullary groove closes. The median band has l)een termed the Zwischenstravg in the chick embryo by His, but is more usually termed the neural crest or ridge (Neuralleiste)^ as proposed by Balfour. In B, the cells are about to unite in the median line. In C they have united, and though incorporated in the medulla and separated entirely from the external ectoderm are readily distinguished from the cells of the medullary plate proper. The cells are also growing out on each side, GZ, toward the myotome. The emigration continues until all the cells are transferred from the median line to the lateral masses, GZ, which are the anlages of the sensory ganglia. As the cells depart from the neural crest the medullary plate proper closes over in the median dorsal line. Tho ganglionic lateral masses exhibit a segmental arrangement very early, so that the cells appear in clusters, each cluster on the inner side of a myotome. According to Chiarugi, 90. 1 , ^ese clusters, at least in the post -auditory region of the head, are bud -like growths from the neural crest ; between the clusters the crest persists for a short time like a commissure. These clusters are found in older stages to enlarge rapidly and to move farther down toward the not«<-liord. They are the rudimentary ganglia. The ganglia are alwaya strictly segmental in position, l)«th when first formed and later. This is especially noticeable when - they attain their maxinium uffarTS™Tn \b relative size, for thev then

  • This embryo is the one designfttect •• No. 18, and described p. 295.

jEt, Ectoderm; (/i, ginpllonic anlngr; mil, medulla; mri, a segment.

dered it highly probable that the celts which form the anlages of the spinal ganglia emigrate singly from the ect<Hierm ; these cells bear an obvious resemblance to the germinating cells, which become the neuroblasts of the nie<lullary tube: see also the account of the olfactorj- ganglion. Chapter XXVIII.

In all vertebrates the ganglia are develoi>e<l essentially as in man, but the process varies considerably in dettiil. Thus in Petromyzon according to Kupffer, 90. 1, 4S6, Taf. XXVIII, Figs. 33, -Hi, 24, and 36, the medidlary cord is completely formed, and afterward the cells are dilTerentiated to form the dorsal median neural crest (Gaiiglienleiste, yervenleisfe). The account given by Kujiffer differs from that given by Sagemehl, 82.1, which has Ijeen accepted by Shipley, 88.1, and Scott. 87.1. If Kupffer is right, then the lamprey "is characterized by a veri- late differentiation of the neural crest. This is true also of elasmobranchs, see Balfour, "'Comp. Embrvol.," 11., 4411, Rabl, 89.2, 223, Taf. X., Figs. 34 and -i^, also Kastschenko, 88.1, 4ti3; in this class the medullary cjmal is completely foi-med, and the neural crest appears afterward, and inoni>ver without any marked differentiation of its cells from those of the mednllarj- tissue proper. In the axnlotl, Lenhossek, 91.1, 19-21, finds the neural crest early sei>arated from the medullary canid. which closes ilorsally by a single row of cells, each of which stretches completely across, see his Fig.lO. In birds, as first observed by W. His, 68. 1,78, the neural crest is a separate distinct thickening of the ectoderm, which can be seen, at least in the cephalic region, while the medullary groove is still open. Fig. 147, Gl; it is readily distinguished by the larger size of its cells from the tissue of the medullary plate. This band was termed by His the intermediate cord {Zwischenstrang) and he was the first not onl}' to demonstrate the existence of a neural crest, but also its genetic relations to the ganglia.

In certain cases there appears, while the medullary groove is still open, a slight groove in the ectoderm close to, and parallel with, the edge of the medullary plate. This groove has been named by His, 68.1, the Zwischenrivne. It apparently results, as suggested by Chiarugi, 91.1, from the effort of the ectoderm to fit in between the edge of the metlulla and the myotomes. The ectoderm, even when there is no groove, is thickened along this line, and this thickening was formerly thought to be connected with the development of the neural crest. This appears, as Beard, 88.3, 160, has correctly maintained, not to be the case. Beard has adopted with this correction His' view of the origin of ganglia, but, without giving his reasons for so doing, advances it as a new conception.

In regard to the early history of the ganglia the following points deserve special mention: 1, the ridge appears first in the region of the hind-brain, and thence its development progresses forward and tailward; the same law governs the appearance of the separate ganglionic anlages; 2, the ganglia arise near the dorsal sununit of the neuron, as seen in cross sections, but rapidly migi*ate toward the notochord until they reach their permanent level alongside the medullary tube; 3, as they descend the ganglion anlages lose all connection, so far as can be observed, with the medullar}' tube. Kolliker, however, expressly states (" Grundriss," 2te Aufl., 267) that the ganglia always remain connected dorsally with the medullary tube; 4, the continuity of the neural crest is preserved, it remaining as a slender band connecting on each side of the body the dorsal parts of the ganglia with one another longitudinally. The connecting band maybe called the ganqlionic commissure. It has been observed by Kolliker (" Grundriss," 2te Aufl., 268) in a human embryo of the fourth week.

The ganglionic commissure is undoubtedly a very important morphological structure, as insisted upon by Balfour, " Comp. Embryol.," II. 450-451. There are a number of valuable observations upon it scattered in various articles, but until these shall have been collated or considerably extended, it will remain impossible to give a satisfactory account of the commissure, its significance or its fate. A special investigation of this problem is much to be desired.

Cephalic Ganglia

As the ganglia of the head differ somewhat in their primitive arrangement from those of the rump, I add a brief description of them.

As long ago as 1847 Remak described in chick embryos of sixty hours the four ganglia of the head to which the neural crest primarily gives rise, at least in amniota. W. His, 68.1, 106, 1G8, gave a fuller description and studied also earlier stages. No study of the ganglia corresponding to the present requirements and resources of embryology has yet been attempted. The four ganglia to be seen in the chick before the head-bend appears are thus described by His, 88.2, 417: There are two ganglionic masses in front of and two behind the auditory vesicle; the foremost of these is the trigeminal ganglion, which is very long, occupying nearly half the length of the head ; it begins in front of the optic vesicle, perhaps even at the olfactory pit, passes along the dorsal side of the optic vesicle, alongside the mid-brain, and ends a short distance after the beginning of the hind-brain. Later this large ganglion separates into the ciliarjganglion and the trigeminal ganglion proper, the former arising from that part of the original anlage which is near the optic vesicle. A. Froriep, 91.2, has observed that in torpedo embryos of G mm. the trigeminal ganglion also sends a large branch, which runs straight to the dorsal side of the isthmus to the point where the trochlear nerve arises later; this branch may be called the trochlear arm; in embryos of mm. the arm is represented only by a few groups of cells; and in embryos of 1(] mm. one of these groups still persists as a small ganglion appended to the trochlear nerve. In embryos of 20 mm. even this remnant of the trochlear arm had disapjx^ared. The second ganglion lies between the trigeminal and the auditor^' vesicle, and is known from the nerves with which it becomes connected as the acustico-facialis. The third and fourth ganglia lie behind the otocyst, and are concerned in the development of the glosso-pharyngeal and vagus nerves respectively. The second, third, and fourth ganglia are much smaller than the trigeminal, and in a chick at sixty hours are of about the same size as the otocyst and primitive segments at the same stage.

The form of the four cephalic ganglia as seen in cross sections (of the human embryo at least) is very characteristic. His, 82.3, 371. The trigeminal appears oval; the acustico-facial subdivided by diverging bundles of fibres; the glosso-pharyngeal is almost circular; the vagus is like a long spindle.


The entire medullary tube undergoes a segmentation by a series of alternating slight enlargements and constrictions. Each enlargement is supposed to give rise typically to a pair of ventral nerve-roots and is joined by the corresponding dorsal (or ganglionic) rootS. In certain neuromeres of the brain this relation to the nerve-roots is modified and even obliterated. The neuromeres are most distinct in amniota at the stage when the hemispheres are just beginning to grow out from the fore-brain, and, after persisting for a short time distinctly marked, are gradually, but rapidly, obliterated. They appear first in the hind-brain and cervical region, and from thence they appear progressively toward the fore-brain and the tail. Their appearance seems to depend upon the development of the primitive segments of the mesothelium (compare p. 102). When the segments are fully formed, and before their inner w^all has changed into mesenchymal tissue, they press against the medullary tube, and oppose its enlargement ; at least one sees that the tube becomes slightly constricted between each pair of segments and slightly enlarge<l opposite each intersegmentiil space. Each intersegmental dilation is a neuromere, and later produces the nerve for the segment (? behind it).

A caution ia here necessary. Each neuromere produces a pair of nerves, but when the first trace of roots appears, they are seen to spring from the constriction between the neuromere, but later from the neuromere. The origin from the neuromere is therefore secondary, as pointed out by Julia B. Piatt, 89.1, who, however, has ignored the difference between the ganglionic and medullary nerve fibres. I deem it probable that the neuromeres, as here described, really comprise each a half of two adjacent trve neuromeres.

Each neuromere is separated from its fellows by an external dorsal -ventral constriction and opposite this an internal slmrp dorsal-ventral ridge, Fig. :)45, aa, so that in a longitudinal horizontal section. Fig. ;tJ5, each half of a neuromere forms a small arc of a circle. So far as at present known, the constrictions are confined to toe sides of the medullarj' tube and do not crc^s either the dorsal or the ventral plateof the neuron. Fig. a45sliows the arrangement of the cells in the neuromeres at a very early stage. The elongated cells are placed radially to the inner curved surface of the neuromere. The nuclei are generally nearer the outer surface, and approach the inner surface only toward the apex of the dividing ridge. On the line between the ajwx of the internal ridge and the pit of the external depression the nuclei are crowded together, but the cells of one neuromere do not extend into another neuromere. Often a light space marks the boundary between the adjacent neural segments. fio, ws.

As to the number of neuromeres our knowledge ^Ih" wJSi is still defective. It is not impossible that the Brain of b number in the head, especially in the fore-brain ,sS^(." . and vagus region of the hind-brain, is less in the ^^'b^i! amniota than in primitive vertebrates, for there is r.iioer*»: i evidence that the number of meso<iermic segments Horr! ° has been reduced in the head, and it is probable that the formation of the neuromeres is conditional upon the presence of the mesodeniiic segments. In the spinal cord there is evidently a neuromere for each jtair of nerves; for example, in chicken embryos of the second day the neuromeres are readily seen to correspond exactlv, as do later the nerves, with the number of segments; compare Duval's Atlas, Figs. 84, 9'J, 93, 98. 100. 102. In the hindbrain of a lizanl (Aiiolis) and of the chick. McClure, 90.1, finds six neuromeres (but in amblj'stoma five only); these six he assigns to the following nerves,* beginning in front, trigeminal, abducens. facial, auditorj. glosso- pharyngeal, and vagus; he believes that the abducens neuromere is wanting in the newt. In the mid-brain we find as yet no evidence of neuromeres among the amniota. but Kupflfer, 86. 1, states that in teleosts two can be distinguished, and Vt .

B. Scott, 87.1, PI. IX., Fig. 15, has given a figure which suggests the existence of two or possibly three neuromeres in the mid-brain of Petromyzon. The fact that two nerves — the oculo-motor and trochlear — arise from the mid-brain, renders it probable that there are corresponding neural se^ents. In the fore-brain McClure has observed two neuromeres in Amblystoma, Anolis, and the chick ; in Anolis these are seen in the region of the Zwischenhim {thalamencephalon) after the optic vesicles have become stalked and the hemisphere anlages have appeared. McClure calls the anterior of these the olfactory neuromere, and says it is connected with the olfactory nerves. I question the existence of such a connection, of which no evidence is given, because the olfactory nerves do not arise from the Zwischenhim. The second neuromere is called the optic by McClure, and is stated by him to produce no nerve.

The total number of neuromeres in the head, exclusive of those belonging to the hypoglossus, is fixed at ten by McClure, 90. 1,51.

Historical Note. — The neuromeres were observed bv C. E. von Baer, 28.2, 64, 74, in chicken embryos of the third and fourth day, and were figured in a dog embryo by Bischoflf, and were noticed bv Remak, 50.1, §28, 67, Dursy, 69.1, A. Goette, 75.1, Taf. VIII.\ Fig. 151, Mihalkovics, 77.1, 49, Beraneck, 84.1,W. B. Scott, 87.1, 273, Michael v. Lenhossek (in man), 91.1, 5. Foster and Balfour (" Embryology," Ist ed., 137) were the first to suggest their segmental value, and this suggestion was adopted by Anton Dohm, 75.2. C. Kupffer, 86.1, definitely asserted that they indicate a '* primary metamerism" (segmentation) of the medullary tube. Orr, 87.1, 335, was the first to clearly demonstrate their relations to the nerves, and these relations were specially studied by McClure, 90.1. The term ** neuromere " was introduced by Orr.

The Zones of His. — By this name I propose to designate the four thickenings which run the entire length of the medullary cord, and the morphological significance of which was first fully recognized and elucidated by W. His, 88.3, 350. In this connection we have also to consider the thin portions of the medullary walls on the dorsal and ventral sides of the neuron. These ]:)ortions are termed by His, 86.1, 483, respectively Deckplatte" and Bodenplatte.^^ L. Ldwe, 80.2, had insisted upon the importance of the thickenings running lengthwise of the neuron, but faileil to discover their relations to the nerves. These relations have been made clear by His.

The wall of the medullary tube is of uneven thickness even in the earliest stages. As seen in cross section, Figs. 101 and 103, the external outline is oval in amniota (more nearly round, however, in ichthyopsida) while the outline of the cavity of the tube is compressed from side to side. In other words, the walls are thin on the median line dorsally and ventrally, and much thicker on each side. We have then from the start two thickened bands, which can l)e traced back, as described in Chapter VIII., to the double thickening of the medullary plate. In the brain the thickenings can also bo traced without difficulty, although in early sUiges they are less sharply marked than in the spinal cord, Fig. KJl.

The next stage is reached by the subdivision of each lateral thickening into a dorsal and a ventral thickening. The change is most readily studied in the spinal cord, to which, therefore, the following description primarily ref era.* The central canal widens out in its dorsal part. Fig. 34G, but so that it remains in its extreme uppermost part a slit, as it does also through most of its ventral part. The widening of the canal cuts into the lateral wall of the medulla, leaving a smaller upper thickening, which I propose to call the dorsal zone of His^ Z), and a larger ventral thickening, which I shall name the ventral zone of His^ V. The dorsal zone forms in cross sections a high rounded prominence into the central canal, and carries in its outermost layer the longitudinal bundles of nerve fibres, which enter the cord from the ganglia through the dorsal roots, D.By and constitute the anlage of the posterior horn; the zone is connected by means of the thin deck-plate, d.p/, with its fellow of the opposite side. The ventral zone, T^, exceeds the dorsal in both height and width ; its boundary toward the central canal is convex ; externally it gives off the fibres which constitute the ventral or motor nerve-root. Between it and the dorsal col-

Fig 84«.-Diagrammatic secumn there is, at least in the hiunan embryo, tion of the Embryonic spinai cord, d.pl^ Deck-plate: A dor a temporary external groove, but the con- saizone; ov. 6, ovai bundle :/>./?, nection between the dorsal and ventral zone ^'i^it'iS^oS^^vl^S.S^SitJ on the same side remains broad. The ven- ^'*«. ventral nwt; 6, Bodentral zone is connected with its fellow by the ?i5in^yinai layen*^ *^*°* ' *^ thin Bodenplatte, 6.

His at fu*st, 86.1, 497, termed the dorsal and ventral zones respectively ArwYere.s Markprisma and vorderen Markcylinder^ but later, 88.3, 350, named them respectiyely Fliigelj^latte and Grundplatte. The external groove, which in man separates the two zones, has an upper angle near the dorsal root ; this angle corresponds to His' Randfurche; and it has also an angle next the ventral column; this lower angle corresponds to His' Cylinderfurche. As the groove and its angles are temporary, it seems to me unnecessary to give them special names.

We distinguish, then, six longitudinal zones in the embryonic cord. These are:

1. Deck-plate.

2,3. Dorsal zones of His.

4,5. Ventral zones of His.

6. Bodenplatte.

The six zones appear in each division of the brain with characteristic modifications, which have been thoroughly studied by His, 88.3, 89.4, 90.2, and must now be passed in review.

  • Further details are given in the section on the spinal cord, p. 058.

1. Medulla Oblongata

The course of development differs from that of the spinal cord somewhat, owing chiefly to the precocious widening of the region and the accompanying expansion of the deckplate to form a largo rhomboid epithelial membrane, as already described and figured, p. GOO, Fig. 343. Owing to the expansion of the deck-plate the lateral walls Hare outward, and consequently the zones of Mis, which are developed from those walls, are changed in position. We may distinguish five regions in the medulla oblongata (His, 90.2, 5), as follows:

  1. The transitional region, next the neck-bend and adjoining the spinal cord.
  2. The region of the calamus scriptorius, which is imperfectly separated from the transitional region in the embryo, althougn perfectly distinct from it in the adult.
  3. The region of greatest width, which includes the part nearest the auditory vesicle and about the origin of the trigeminal nerve.
  4. The region of the cerebellum and pons Varolii.
  5. The isthmus or narrow connection with the mid-brain.

The widening begins (in human embryos during the third week), as indicated in Fig. 347, in the headward part of the medulla, the ventral part of the central canal remaining very narrow ; the change suggests the differentiation of the dorsal and ventral zones. As the widening continues (human embryos of the fourth week), the lumen becomes more triangular, and later five-sided in section. Fig. 348. The largest side is dorsal and is constituted by the widened deck-plate; the other four correspond to the zones of His ; the dorsal zones form a decided angle with the ventral ones ; each zone as seen in section projects toward the interior, appearing concave on the outer, convex on the inner side. The assumption of the five-sided form is not simultaneous throughout the medulla oblongata. The widening of the medullary tube continues, and becomes so extreme in the thiiS region that the zones of His are brought by the enormous expansion of the deck-plate into one plane. Fig. 350. While this is being accomplished there appears, along the morphologically dorsal edge of the dorsal zone of His, a fold by which that edge is bent over outward and downward, Fig. 349. This everted, edge has been named by His, 88.3, 35G, the Rautenlippe; it extends through the recHionS ^e HInd-braIn of His' Embryo o. §6, (

deck-plate : FU dorsal zone; ^;?r. ventral

.-IV., ana m iresn human Bodenplatte. After W. His. xaodlams.

embryos of five weeks may be seen as a bright border around the edge of the rhomboid sinus formed by the deck-plate. The Rautenlippe is simply a fold, and is accordingly separated by an external groove from the rest of the dorsal column, while internally there is another groove, Figs. 349 and 350, C, which is bounded on one side by the bent-over edge of the dorsal zone, on the other by the lateral margin of the deck-plate. The

Fig. 847.— Section of the Medulla and Otocysts, ^6, of a human Embryo (His' BB) of 8.2 mm. After W. His. x TOdiams.

Fig. 8i8.— Sections through the Regions 3 and 5 of

ibryo «. G6. Otocfst; D, zone; Bd^

grooves are designated by His rGSpectively outer and inner lip-groove \Lippenfurche) . The junction of the Rautenlimw and the deck-plate in dititiiiguiBhed by His ae the Tcenia. The Rautenlippe plays eui important role in the differentiation of both the medulla oblongata and of the cerebellum. By the end of the fifth week in the human embryo the expansion is carried so far in the region of greatest width that l£e dorsal zones I are forced over so as to in a lower plane ventral of the plane of the dorsal zones.

Later the process of bendinCT ^^- MB.— Sectton throuKb the BeKion S of Uie Hlnd Hli- Embryo A (eompuB Fig. !1B). 06,

down the dorsal zones occurs owcj*. Afwr w. rfiB.

also in the region of the calamus, though it ia not carried so far as in the region of greatest width. In the r^ion of the cerebellum, on the contrary, the medullary wall constituting the dorsal zone does not bend over, but remains nearly in a vertical plane.

In human embryos of the latter part of the second month, His found, 90.2, 20, the following relations: The Rautenlippe begins as a small band in the transitional region and runs forward, increasing in width until it reaches the lower half of the region of the calamus Bcriptorius, then diminishes in width throughout the region of greatest width, and finally attains it« maximum size in the cerebellar region; at its anterior extremity the lippe tapers oflf to end in a point. The external groove between the Rautenlippe and rest of the dorsal column of His becomes obliterated by the walls of the groove growing together. The union of the walla does not take place simultaneously throughout; it occurs very early in the region of the calamus, much later in the cerebellar region, where the groove becomes especially deep. In the region of the medulla oblongata

raoper Tbe unioD is io part temporarr. while in that of the cerebellum It is pennaiieDt.

The isthmus. Fig. 350, B, or part connecting with the mid-brain, is characterized br remaining smaller than the rest of the third cerebral vesicle and bv the absence of the expansion of its deck-plate. As seen in cross section. Fig. '-ioi'. B, in a human embryo of five weeks, it appears somewhat compressed from side to side, and the deck-plate and Bodenplatte project somewhat, producing each a slight external median ridge (His, 66.3, :)■>?).

The expanded deck-plate in man, up to the middle of the second month, arches over the wide ca\"ity of the medulla oblongata; in older human embryos, owing to the gnjwth of the cerebeUum, it becomes bent, so as to form a transverse fold, the plica chon'oidea, which is situated close behind the cerebellum and projects inwarti toward the tloor of the fourth ventricle. The fold is anlage ot the choroid plexus (His, 90.2, 2(i).

2. Mid-Brain

In the embryonic mid-brain, Fig. 350, A, the transverse diameter exceeds the vertical. The deck-plate projecta as in the isthmus, but the Bodenplatte is broadened and thickened, and having become convex towartl the interior, concave toward the exterior, constitutes an internal ridge and external longitudinal groove. The ventral zones of His are well defined and are much narrower in extent tlum the dorsal ».ines, which constitute the largest part of the wall of the mid-brain, and which merge Avithout anv distinct boundarv into the deck-plate (His. '68.3, 357), Later, that is to say, by the time the oculo-motorius has grown forth from the mid-brain, the Ixiundariesnf the six primarv' longitudinal zones are almost obliterated, compare Fig. 3r.7. and still later they entirely disappear. *

[. Fore-Brais. — The

7/>nes of Hirt are less distinct

-oU^^^Mi,rH.,7r^-^,^oyiii-^\UT^i^i-U:;i^ i" the fore-brain than in

licnrire; £-l«nili«il«Tuibiill8:7> infuiidlbulum:/lj,. the hmd-braiu. but maV be tfiBlamii-a: n. ilulamiu: Z. imWn' of pin»l Klaii.l; traCCtl m vmuig CmbrVOS

lora quadriKMnlua: ilh

lora quadr


iiii; fh crrhciiuHi; ';. v.-B- in sections at right angles to the axis of the fore-brain.

The ventral zone tends to arch inward, while the larger dorsal zone tends to arch outward. His, 69.4, lirO-dS-'i. has endeavored to trace out the exact course of the zones in the fore-brain, a most difficult task, owing to the flexures and to the outgrowth of the optic vesicles and of the hemispheres. He concludes that the two ventral zones extend primitively to the optic chiasma and include at least a part (the retinal) or the whole of uie optic evaginations. As shown in the diagram, Fig. 351, this makes the regio sub-thalamica, P.Sy the mammilary process, Ma^ tuber cinereum, To, and recessus infundibuli, Bi, derivatives of the ventral zone (Orundplatte) ; while on the other hand, the optic thalami, The hemispheres, Hs^ corpus striatum, Cs^ and olfactory lobe, iJZ, are derivatives of the dorsal column (Flilgelplatte) .

The Boden^atte loses its individuality in the fore-brain, but the deck-plate becomes much specialized, as described in connection with the history of the fore-brain, given later.

The division between the ventral and dorsal zones is readily traced in the wall of the third ventricle of the adult; it is the sulcus Munroi of Reichert, and extends from the lower edge of the foramen of Munroe to the aqueductus Sylvii ; this groove is figured in Obersteiner's " Handbook" and elsewhere, but is often omitted in anatomical figures in which it should be represented. As a morphological division it is, of course, of fundamental importance.

Origin of Nerve-Cells. — The first step in the histological differentiation of the medullary walls is the separation of the cells into two classes: 1, ihe spongioblasts^ or young neuroglia cells; 2, the germinating cells, which are the parents of the young nerve cells or neuroblasts. This section deals with the germinating cells and their transformation into neuroblasts. The history of the spongioblasts is sketched in the two following sections.

The medullary tube is at first composed of a single layer of simple epithelial cells of a nearly imiform character — a fact which was discovered by Victor Hensen, 76.1, 383; the discovery has since been verified for all classes of vertebral^. There soon appear special cells of a rounded form in the medullary epithelium on the side of the epithelium toward the central cavity. These cells divide actively and have been named the germinating cells. The germinating cells (Keimzellendes Markes, His, 89.1, 314) are the only ones which undergo division, and as their nuclei divide indirectly, we can readily determine the distribution of these cells by that of the karyokinetic figures in the embryonic neuron. Altmann, in 1881, first pointed out that the figures of nuclear division in parts of the central nervous system of the embryo are found next the central canal, and that, therefore, the pericentral stratum is the growing layer. These observations have since been confirmed and extended by Uskoff, 82.1, Ranter, 86.1, Merk, 85.1, and W. Vignal, 84.1, 208-210, who appears to have been unacquainted with the earlier German observ-ations. In his last paper, 87.1, Merk points out that there is much greater variety in the distribution of karyokinetic figures in the medullary canal than appeared from previous researches, and that each region has its characteristics. Thus in the retina the growing layer is external * or next the mesoderm ; in the corpus striatum and thalamus opticus the proliferation takes place through the whole thickness of the wall, etc. Special stress is laid by Merk apon the difference between cell malti{^canon, which does not neceattarily mean synchronous increase of substance, and cell grow^, which does mean increase of substance. The growth of the nervous system depends chiefly on the enlargonent of the cells, as Boll, 73. 1, and Eichhorst, 76.1, maintained long ago, and it is incorrect to follow the custom of using the terms proliferation and growth a:s synonymous.

  • I feel much doubt as to Merk*8 accuracy in regard to this point.

The typical germinating cells (His, 89.1, 315) are round or slightly oval, and measure frcnn 10 to 14 ;& in diameter. The nuclei measure from 4 to 8/^ ; in the resting stage they are oval, with a distinct outline, and scattered chnxnatine granules; but most of the nuclei in young embryos are in some stage'of indirect division and therefore have no distinct outline, while their chromatine granules are large, conspicuous, and variously grouped according to the stage of karyokinesis, Fig. 352. The protoplasm forms a clear, broad cellbody, and with higher powers can be seen to form a granular endoplaffln and a non-granular ectoplasm. The cells lie between the processes of the neuroglia cells, and lie typicaUy as in Fig. 352, in the rounded spaces between those processes, close to the thin membrana limitans interna^ which is described in the next section on the neuroglia. The number of the germinating cells is very large in the human embryo at four weeks, so that in places they seem to form an almost continuous layer. Later they gradualh' diminish in number, and the spaces occupied by them persist empty for a time. As to the disappearance of the cells our information is incomplete, but it is probable that they are all changed into neuroblasts. That most of them do so change has been proved by His, 89.1, 318; see the sections on the origin of the neuroglia, below, and of the nerve fibres, p. 616.

Origin of the Neuroglia. — The following account refers especially to the human embryo, and is based on His' observations. The cells of the medullary tube have at first a distinctly epithelial character, and in very thin sections (yI^^ - ^|^ mm. ) of well-preserved specimens each cell can be seen to extend radially through the entire thickness of the wall. So long as the epithelial character is preserved, there is an outer and an inner zone without nuclei with a middle laj^er containing all the nuclei, which increase in number as the development progresses. There next appear cells in the inner non-nucleate layer ; these are the germinating-cells ; they differ from the other cells of the cord, and according to His, 89. 1, 321, give rise to the young ner\'e-cells. All the remaining cells, the nuclei of which remain in the middle zone, give rise to the neurogha, and are accordingly named spongioblasts by His. The change of the epithelial cells into spongioblasts can be particularly well studied in elasmobranch embr}'os (e. q.y Pristiurus of 4i mm.). The elongated cells acquire a vacuolated appearance; the cell boundaries become indistinct ; the substance of the cell-body takes on more and more of a trabecular character, and there results a network of metamorphosed cell material instead of a layer of discrete epithelial cells (His, 89. 1 , 350) . While the spongioblast network (myelo-spongium, neurospongium, neuroglia) is developing, the protoplasm alters into a substance which is more homogeneous, more highly refractile, and more readily stained than protoplasm. In other vertebrates the conversion of the epithelial cell into a spongioblast takes place in a similar manner, as has been demonstrated by His' observations on mammals, birds, amphibians, and fishes. Each spongioblast has (His, 89.1, 327) two main processes, an outer and an inner, and several smaller lateral processes. The inner processes run to the inner boundary, where their ends unite to form the memhrana limitaiis interna: the character of these processes calls for further study, because, though they usually run without dividing, yet in certain cases they have been found giving off branches ; the ends of the fibres break up into fine branches which unite to make a close network, and this network is the membrana limitans. The outer processes always branch, their branches being most developed in the outer non-nucleated layer. Fig.

Fig. 352.— Neuroglia of the Dorsal Zone of the Spinal Cord of a Human Embrsro of about three and one-half WeekR. N^ Young oeuroblast; l.tn, limitans interna; Rtch^ Randschleier. After W. His. X 880 diams.

352, Rsch (His' Randschleier) ; the branches form a network, which appears most distinctly in the outer layer. The nuclei are oval, and, as just remarked, lie at various levels ; around each nucleus is an accumulation of protoplasm, which may for convenience be distinguished as the cell-body ; the cell-bodies also give off processes, which anastomose with one another. The cells become still more elongated as the embryo advances, and tend to gather more or less in little groups, as may be seen in human embryos of the sixth and seventh week.

It is to be noted that neither the ventral plate (His' Bodenplatte) nor the dorsal plate (His' Deckplatte) undergo the same histological differentiation as the lateral zones of His. Neither plate develops any young nerve-cells (neuroblasts) ; the ventral plate changes entirely into neuroglia, into which the nerve fibres penetrate secondarily to make the anterior commissure. The dorsal plate retains its primitive, simple, epithelial character wherever there is an ependyma, but elsewhere its cells also become spongioblasts.

The history of the neuroglia shows that it is in no wise related to the mesenchyma or true connective tissue. This relationship was for a long time and generally assumed. Golgi was the first t<j discover the ectodermal cliaracter of the neuroglia (** Studi s. f. anat. d. organi centr. syst. Nerv\," p. 178). Without reference to Golgi 's discovery, Gierke, 85.1, 4*J8, upon a somewhat imperfect basis of observation positively asserted the exclusively ectodermal origin, and the question wfis definitely settled by W. His. Since then the neuroglia in the embryo has lxH?n asserted by Lachi, 90.1, to be partly at lejist, immigrant connective tissue, but that Lachi 's view is erroneous has been more than sufficiently demonstrated by Cajal, 90.1, and M. von Lenhossok, 91.2.

Specialization of the Neuroglia. — Wo know, chiefly through Gierke's researches, 85.1, that the neuroglia assumes various characteristic modifications in the different regions of the adult central nervous system. Gierke, /.r., 400-505, gives some observations on the differentiation of the neuroglia in the embryo, but I have been able to find little in these pages sufficiently definite for use. Gierke held that the matrix of the neuroglia wi\s a nKxlification of the peripheral l)arts of the embryonic cells, an opinion which I deem erroneous.

All the spongioblasts in the embrj-o stretch through the entire thickness of the medullary wall and have a correspondingly elongated

fonn. When treated by Gulgi's chromic-osmium method a portion of the si>ongiobhists are found in two ti> four days to be coloreil, and may be easily followed, as in the same length of time the nerve-cells are not colored, though the blood-vessels are. The metluxl has been applied by Qolgi liimself, 90.3, and by Ramon y Cajal, 90.1, to the chick embryo, by M. von Lenhossek, 91.2, to the human embrvo, and bv Xansen* to Myxine. From these investigations we have learneil that the spongioblasts l>et*<nne very much elongated and remain very slender; where the nucltHis is situatetl, the cell is thickentnl. At first the nuclei are confined to the gray matter, but as development progivsses the nuclei appe^ir to migrate, so that gradually their numIrt thiHHij^h the gray matter dimin a the

Fig. 353.— Cross Section of the Spinal

neuroglia layer (Raialsrhleier). Many of the cells now appear to lose their central ends. Fig. 353, so that only the prolongation of the nucleated b<xly toward the outer surface of the medulla is preserved. Loiter the distal prolongation is also lost, and the secondary branches, which have been meanwhile developed, convert the elongated cell into the so-called Deiters' or spider cell, Fig. 354. Leiihossek, 91.2, has traced the changes hoth in tne gray and white matter, and found typical modifications in each part of each layer. W. Vignal, 66. 1, 320, though he failed to reci^nize the neuroglia until advanced stages, reports some observationa on the later differentiation of the neuroglia in the cerebrum of the human fcetus; at seven months the cells have transparent bodies with numerous granules which can be seen when the cells are examined in water, but not when they are mounted in glycerin ; they have numerous processes and a round or oval nucleolated nucleua. At eight " months the cerebral neuroglia cells vary in size, but some are much enlarged and their procestieB show traces of the change into a homogeneous refringent substance, see Vignal, I.e., PI. XL, Fig a, «.

  • Nauseu's pufHT wus iiublisti(.*d by the Mutieum at Berij^n in 1^^. I have Dot ween Ic.

While some of the spongiobliists have their ^hS^!?!'" ,";?'«id (t main nucleated bodies retained in the gray or white matter, others have their bodies placed close aroimd the central canal, Fig. 354, where they form the socalled epithelium of the central canal, or ependyma, as it may be l>etter called. These ependjrmal cells stretch out through the entire diameter of the medullary wall, there being a fine radial process, Fig. 'ih'A, which passes outward through the gray and white matter, lis was tirfit recorded by Golgi, and later by Gierke, 85.1, 49i), and has since l)een more fully described by Cajal and Lenhossek. The cilia on the inner ends of the ependyma cells appear in the human cmbrj-o about the end of the fifth week (His, 89.1, 330). EichIiorst, 76.1, records that in a three months' embryo the cilia are pi-esent on some of the cells around the canal but not on others, uiwe, 83.1, observed that the ependyma cells resemble spongioblasts, but failed to recognize their identity.

One sees readily in embrj-os of mammals, when about 10 mm. long, a broad layer of nuclei close to the cavity of the medullary tube ; later, where the canal obliterates no trace of this layer is preserved, but where the lumen of the canal is permanent there persists a narrow layer of crowded ependyma! nuclei. This is because many of the cells have changed into true neuroglia cells, and the broad layer has been in part annexed to the gray matter or neuroblast layer.

Fla. 35,— Part of a Tra Human Emhryo nf hodfchroi eDjyinA and Ihu D«ll«rs'

Layers of the Medullary Wall. — By the time the neuroblasts are differentiated we can distinguish three primary layers which persist throughout life with sundry secondary modifications. There IS an outer layer of neuroglia network, Fig. 352, which is the anlage (or homologous with the anlage) of the white matter of the spinal cord; it has been named the JRandschleier by His, and this name I have adopted, although it might be better named the outer neuroglia layer. There is a middle layer, in which all the neuroblasts are situated, and which is the anlage of the gray matter of the central nervous system throughout ; it has been named the mantle layer. Finally there is an innermost layer, in which at first germinating cells are situated, but which, after their emigration, consists merely of spongioblasts ; this is the Innenschicht of His, and may be defined as the ependymal layer; during development it is reduced by the encroachment of the mantle or neuroblastic layer.

We distinguish then :

1. Randschleier, or outer neuroglia IsLyev (white matter).

2. Mantle layei% or middle layer in which all the neuroblasts are situated (gray matter) .

3. Inner layer, or inner neuroglia laj-er (ependyma).

Origin of Nerve Fibres. — We know through the researches of Wm. His, 86.2, 88. 1, 88.3, that there are two sets of nerve-fibres developed in the vertebrate embryo — one set from the medulla, and another from the ganglia. Each medullary fibre arises as an outgrowth from one pole of a nerve-cell, situated in the wall of the medullary tube; each ganglionic fibre arises, on the contrary, by the outgrowth of two opposed poles of a nerve-cell. The cell of the medullary fibre is terminal, while that of the ganglionic fibre is interpolated in the course of the fibre. There is, in fact, a profound morphological difference between the two classes of nerve-fiores, and it is necessary to consider their development separately.

Medullary Fibres. — In the section upon the neuroglia, it was pointed out that, when the medullary tube closes, the cells which form its walls are all similar to one another. About the end of the third week in human embryos the cells lose their imiform character and become differentiated into the neuroglia cells, which form a network, and the nerve-cells, which lie scattered alx)ut and produce nerve-fibres, while the neuroglia is developing. While these changes are going on, the medullary tube grows rapidly and in the nucleated layer of its walls two primary layers become distinguishable: these are the so-called inner layer and the outer or mantle layer, Fig. 377. In the latter are situated all the cells which give rise to nerve-fibres, but later, that is, after the blood-vessels have penetrated the medulla, nerve-cells encroach more and more upon the inner layer also (His, 86.2, 509). It is to be noted that the more superficial position of the nerve-cells, which is permanently maintained in the cerebellum and in the cerebral hemispheres, is originally characteristic of the entire neuron. The deep position which the cells have in certain parts in the adult — as, for example, in the spinal cord — is produced secondarily by the growth of nerve-fibres in the Randschleier. This

change is very early indicated in the spinal cord by the growth of the Randschleier. In the inner layer, the cells and tibeir oval nuclei are crowded, and it is here only that all division goes on; the peculiar position of the karyokinetic figures has been described, p. 611.

The nerve-cells, according to W. His, 89.1, 318-326, are all descended from the germinating cells described above, p. 612, and migrate from the inner layer into the mantle layer. That the nervecells arise near the central canal and migrate into the mantle layer, was discovered in 1884 by Herms, 84.1, in his studies upon the facialis neuroblasts of the lamprey.

The metamorphosis begins by the protoplasm of the germinal cell accumulating on the side of the nucleus away from the cavitv of the medulla and there elongating itself into a point, which in its turn soon elongates into the beginning of the nerve-fibre; the fibre, therefore, always points away from the cavity. Fig. 355, iV. The elongation of the fibre continues apparently at the expense of the protoplasm already accumulated in the cell ; the fibre accordingly grows very rapidly at first, and soon passes beyond the medulla, but later the elongation is much slower, for it then depends upon the actual growth of the fibre itself. When the fibre begins to develop, the cells begin to migrate toward the outer part of -^ the medulla to form the mantle layer and are found a short distance from the membrana limitans interna. The reason that the young nerve-cells migrate only to a Fig. ass.— From a sectioD of the certain point is found apparently in the l^^:'%?^^^^%?i ^l^S"!"^?. structure of the outer non-nucleated zone, ^^^l^^^^lL^jt^^r^^'^^' -^^^ as pomted out by His, 89.1, 330. The

neuroglia network, as can be seen in Fig. 352, is so dense in this zone that it blocks the way for all, or nearly all, the neuroblasts. In later stages the meshes become larger again, and the blood-vessels are able to penetrate it to enter the neuron. The nuclei of the migrating cells are oval and, for the most part, have a single nucleolus; the protoplasm is principally accumulated in a pyramidal mass at the distal end of the nucleus, the apex of the pyramid being prolonged as the axis-cylinder process ; the protoplasm forms only an exceedingly thin layer around the sides and proximal end of the nucleus. At this stage of the cells, the protoplasm stains deeply, and in stained sections the distal ends of the nuclei are often obscured or even hidden ; when this is not the case, the distal ends of the nuclei are pointed — a peculiarity which becomes more marked in a slightly more advanced stage. The cells continue their migration and development until they reach the mantle layer as fully differentiated young nerve-cells, which are characterized by having an oval nonnucleolated nucleus, with only a very thin envelope of protoplasm, the rest of the protoplasm having been converted into the nerve-fibre.

According to W. His, 88.1, 370, 89.1, 316, the mantle layer consists in the human embryo at four weeks almost entirely of young nerve-cells, and contains only very few neuroglia cells. The nerve cells be names neuroblasts; they praeeDt the following characteristics: they have an oval nucleus (9-ll;i long and 4.5-5.5,u wide) at the distal end of which is a small cone of protoplasm, which is continued as the nerve-fibre; the nucleus contains considerable chromatin in the form of scattered granules connected by delicate threads; the envelope of protoplasm is exceedingly thin, so that when the nuclei are cut transversely or obliquely they seem almost without protoplasm, and represent the so-cafled nak«l nuclei of the mantle layer ; the nerve-fiore ie of nearly uniform diameter, and presents, as does also the protoplasmatic cone from which it springs, along!tudinally fibril lated appearance. The neuroblasts often lie in groups; in such cases the fibres from one group unite in a long cone or bundle, and continue their growth in association, Fig. 35(1. The young cells have no other outgrowths, the branching processes, which are so characteristic of the adult nerve-cell, not developing until much later. The paths taken by the rapidly lengthening medullary nerve-fibres have next to be considered. The fibres may be dividetl into two classes, according as they make an immediate exit from the neuron or first grow within it. The latter class include, Jirst, fibres which cross through the Bodenplatte to the opposite side; they constitute the formatio arcuata, Fig. 377; second, fibres which take a longitudinal course in the Randschleier; it may be noted that, according to M. von Lenboe. [- -J. t ■ J -' sek, 91.3, l:i3, the ne^^■e-cells, which give off fibres to run longitudinally, can be first seen in the chick the sixth day, which is later than the other cells; their fibres are, moreover, characterized by their branching (Collateralen of Kiilliker) ; i third, fibres which join the ganglionic or dorsal root. M. %-on Lenhossek, 91.3, has obser\"ed in a four-days' chick medullary fibres which joined the ganglionic root.

The neuroblasts of His have now been found, describe<l, and figured in every class of vertebrates except the dipnoans. They have evcrj-where the same essential character, though presenting minor variations. They are unusually small in Petromyzon ; in amphibians unusually large; in the frog they are pigmentetl; in the trout they are particularly numerous and distinct; the trout is further remarkable for having a few unusually large neuroblasts on the dorsal side of the embryonic spinal cord. For furthur details see His, 89.1, 3:n-35fi.

Elasmobranchs offer the peculiarity that the motor nerve-roots l>ecome invaded by mesenchj-mal cells ver>- soon after the fibres grow out of the medulla, hence the roots contain nuclei at a very early stage. The nuclei were first observed by Balfour, 78.3, 70, who drew the erroneous conclusion that the roots arose as cordB of cells, and that the nerve-fibres were developed later. This error has been kept up by Van Wijhe and J. Beard, 88.3, 193, but discussion of it is passed, compare His, 89.1 344, Eastschenko, 88.1, 4G5, and Dohrn, 91.2, who have proven that in the cartilaginous fishes the motor nerve-fibres grow out from the medulla as in all other vertebrates. Dohm has further maintained, 68. 1, that in elasmobranchs medullar}' cells migrate from the medulla with the nerve-fibres. His, 89.1, questioned the correctness of this opinion, but Dohm, 91.1, has renewed his assertion and offers additional evidence. He ascertained nothing as to the ultimate fate of the emigrant cells.

Ganglionic Fibres. — The exact history of the earliest changes in the cells of the ganglia, p. 601, has still to be worked out. His, however, has shown that in the human einbrj"o they all become bipolar, that is to say, niufh elongated; one end pointing toward the dorsal side of the ganglion and lengthening out as a nerve - fibre, which penetrates the myelon, the other end pointing toward the ventral side of the ganglion and lengthening out as a peripheral nerve-fibre. Fig. 35" represents a group of bipolar cells from a spinal ganglion of a young human embryo. the cells are gathered in groups and the fibres from one group unite in prjmarj' bundles. Whenthe cell is turned so as to be viewed in profile, it is seen that the oval nucleus occupies an eccentric position and is surrounded by a mass of protoplasm, which gives off the uerve-fibre in two opposed directions, so that one might almost say that there is a nerve-fibre with a Fig. a»7. -Bipolar OelU from* BplnjlGanKlIra of mi ,, 1 1 ^ -^ -J EmbTTD <H 8' euibrvo N). AIMr W. U >. X about GOO

cell appended to its side. aiaiS In an embryo of six weeks, the cells are still of this type, and resemble the bipolar cells described by Freud in the ganglia of Petromyzon. In an embryo of seven weeks the mesenchymal cells had begun to grow into the ganglion between the ectodermal cells, which thereafter begin to change into pearshaped appendages of the fibres, with the result of developing the T-joints of Ranvier. On the development of the cells proper see p. 636.

The dorsal processes of the cells oiter the myelon as sensory roots. The namber of entering fibres is at first smdl, but gradually increases. Witliin the myelon the fibres at first all take a longitudinal coarse in the outer layer {Randackleier ot Hia), some of tne fibres passing headward, others tailward, but later fibres course within the myelon directly toward the nerve-cells of the mantle layer. According to Ramon y Cajal, 90.1, 92, the ganglionic fibres penetrate the medulla and there fork ; each branch curves around and becomes a longitudinal fibre, but the two branches run in opposite directions as fibres in the Randschleier ; these fibresgive off fine branches nearly at right angles, which penetrate the gray matter and there ramify, but without forming a true network, compare Fig. 35S ; the branches running to the gray matter our author names "collaterals;" their ramifications, at least at first, are confined to the gray matter of the dorsal zone of His. Ramon y Cajal's important discovenr has been confirmed by Kfilliker, who has also made important additions to our knowledge of the distribution of the sensory fibres within the cord.

The distal or ventral processes extend in onft group from each ganglion as the sensory root.

The formation of the nerve roots may be superbly demonstrated, as discovered bv Ramon y Cajal, 90. 1, by the application of Golgi's bichromate - silver method to embryos (chicks of four to sixteen days, and mammals of correspond Fio asa-TiMSTOTM Bectton of the Doiml CoiM «nd Can- ing Stages) , 866 Fig. 358.

Such preparations de bTMchw, ff. mpdulldrj iieurobl««W irlth deDdriUn and monstrate further the

«i.^iud«r. Af«r Cajal. development of the dendrites of the medullary nerve-cells, and the abundant intra-myelic ramifications of the " collaterals" of the ganglionic fibres.

Historical Note. — The first suggestion that all sensory nervefibres arise from the^nglia and, grow centrifugally and c«>ntripetaUy, was made by W, His (His' Archiv, 1881, p. 477) who brought positive proof of the correctness of his view in 188fi, 86.2, 490, and later showed that it was true not only of the spinal, but aJso of the cephalic, ganglia. 88.1, 374, 88.3, 368.

Mbdullary Sheaths. — AH the nerve-fibres are at first simple processes of the nerve or ganglion cells, and they persist in that condition for a long time, but finally there is developed around those fibres, which are destined to form medullated fibres in the adult, a covering of mesenchymal cells. No trace of this covering can be seen in mammals until after the nerves have grown out and ramified through the entire embryo. We have at this stage, as well as later, to distinguiBh between the bundleB of fibres or nerves proper, and the fibres running singly or terminal branches of the nerves.

When a nerve consiBts of several or more fibres, the mesenchyma forms an envelope around it (Vignal, 83.1, 518), which in certain cases at least, and perhaps always, is very distinct and sharply defined (KoUiker, " Gewebelehre, " 6te Aufl., p. 152, Fig. 113). There next follows the penetration of the nerve by the mesenchymal cells, which make their way in between the fibres. In the case of very small nerves and of single fibres, the cells of the connective tissue have direct access to the single fibres.

Whether the cells reach the fibres directly or not, has no influence on their further differentiation. They lay themselves against the nerve-fibre, from place to place, and grow around it so intimately that it becomes difficult to distinguish the boundary between the original fibre and its accessory envelope, and one is inclined, at first sight, to conclude that the fibre has merely become thicker and nucleated.* In reality, the mesenchymal cells close around the fibre, which they cover like a chain of elongated beads. Each cell is the aniage of a medullary segment; the junction of two adjacent cells is the aniage of a node ofRanvier; the nucleus becomes the intemodal nucleus of Schwann's sheath. Each cell is at first short and protoplasmatic. The cells multiply; KoUiker, 86.2, has observed them dividing in amphibians; W. Vignal maintains that new cells are interpolated in mammalian embrj-os between those already enveloping a fibre. It seems possible that the cells may increase in number by both means. They also grow quite rapidly in both length and diameter.

The differentiation of the cells into the three sheaths of an adult fibre depends upon their forming each a membrane and an internal deposit of myeline. The nucleus takes and keeps its position near the centre of the cell and retains a small quantity of granular protoplasm permanenily about itself. In re^rd to the formation of the membrane, I know of no satisfactory observations, but I think it probable that a membrane is formed over the entire surface of the cell, and that it is this mem- ] brane on the outside of the cell which is known i

)«.— laoUud Nerre am the SclMlc Nerre ID Embryo of IDO mm. !iro<nb ol Ote m«lul.

as the sheath of Schwann, and on the inside UrriiKttth. iu.Mod«otRu] of the cell next the axis-cylinder may be vSi»if'x'neIu^*»d^mV termed the periaxial sheath. This supposition needs to be verified by observation. The medulla or myeline appears quite late — in the cow toward the fourth month, in the sheep at seventy days, according to W. Vignal's observations^ •SoeKOlliker, Zelt wi«. Zool.. XLlfl. , Ttl. L

83.1, 523, on the Kciatic nerve. The myeline begins to appear at the same time in many, but not in all the fibres of a nerve, and it develops later in the peripheral than in the proximal portion of a nerve, and can be earliest observed in the spinal cord. It appears at first as a very thin layer in the mesenchymal cell and next to the axis-cylinder; it is usually deposited simultaneously throughout the entire length of the cell, but sometimes the deposit begins at the centre of the cell; the myeline layer is usually continuous from the start, but sometimes it constitutes a series of separate masses, which grow and unite into a continuous layer ; at this stage one observes that the axis-cylinder is pressed aside by the nucleus of the myeline cell. The deposit of myeline gradually increases, and forms a more regular layer; at the same time the boundary (Ranvier's node) between adjacent cells becomes more distinct and the cells (intemodal segments) elongate.

Historical Note, — Our knowledge of the history of the peripheral nerve-fibres is largely based on the study of the tail of tadpoles, see Rouget, 76.1, W. Vignal, 83.1, 83.2, and Kolliker both for observations and references to the literature. The development of the fibres in mammals has been studied by Vignal, Z.c, and by Axel Key and Retzius {Arch, inikrosk Anat,^ IX., 308).

Origin and Growth of Nerves

There are two sets of nerves, corresponding to the two classes of nerve-fibres. Every nerve consists of a bundle of nerve-fibres. Each ganglion and each lateral half of a neuromere sends out a bundle of nerve-fibres, or a nerve, as we may better say. There are, therefore, typically for everj' segment four primary nerves, two on each side, a dorsal ganglionic and a ventral medullary nerve ; usually the two nerves on the same side of a segment unite at a short distance from the myelon into a single trunk; in this case the ganglionic nerve becomes the dorsal or posterior root of anatomv, and the medullarv ner\'e the ventral or anterior root of the nerve trunk. Nearly all the spinal and several of the cranial nerves conform to this type. In certain cranial nerves, however, we have only ganglionic, in others only medullary fibres. The development of the various nerves is consideretl later; that of the nerve-fibres is described in the preceding section; we shall, therefore, treat here only the general principles of embryonic nerve growth.

As to the mechanical means by which the fibres are first gathered into bundles, we have little positive information. In the case of the medullar}' fibres the iwiths are probably prescribed, as suggested by His, by the structure of the previously developed neuroglia. In the case of the ganglionic fibres they seem to be brought together by the pointeil shape assumed by the ganglion as a whole.

The nerve-fibres, as they grow peripherally, are gathered into short stems (nerve-trunks) . Each stem, whether motor or sensor}', consists (His, 88.1, 375) of a number of fine fibres without nuclei; within the stem the fibres run all in the same general direction, but some of them take partly crooked courses. Paterson, 01.1, 168, has ol>sorved that the nerve-fibres increase in thickness in the spinal nerves of mammals, while they are growing to their destinations; the fibres in these nerves take characteristic wavv courses. Mesoblastic cells penetrate the Btem, which then becomes nucleated ; in the human embryo the number of mesoblastic nuclei in the nerves remaini) small for a long period, during which the nerves appear light and conspicuous in stained sections, owing to their poverty in cells. The ends of the nerves are at first broad and blunt, and it is only by repeated branching that the nerves acquire finer endings. The ends are at first so blunt that the nerves appear as if chop^d oflE, Figs. 86 and 3(10, a peculiarity which formerly misled many observers to conclude that they had not found the end of the nerve at all. All the nervee take a straight course at first and always tend to grow in a straight line representing the prolongation of the direction of the nervefibres. This law, which was discovered by His, applies to all nerves, even to those which take a complicated course in the adult. This is well illustrated by the early stages of the nerves to the eyes, or of the vagus, or of the cervical

Human Embryo iif acconllDK to the umi«l A. recurreot Ikryugial.

nerves, etc. The straight course of a nerve is modified in two ways : by encountering an obstacle, or by a change in the relative positions of parts with which the nerve has become connected. When a ner\'e enaninters an obstacle it is either deflected from its course or forced to divide. The most irapoi-tant obstacles are cartilages, blood-vessels, and cavities lined by epithelium, and it is, therefore, necessary that these tissues be differentiated at the pro|)er points in the embryo, before the nerve arrives, or else the necessarj' mechanical conditions for effecting the normal distribution of the nerve are not established. For example, the third branch of the trigeminus when it strikes Meckel's cartilage divides into the ramus lingualis and the ramus mandibularis, and the h}-poglos3al nen'e when it strikes the wall of the jugular vein divides into its descending and lingual branches {His, 88.1, :di). After a nerve is deflected it grows forward in the direction of the fibres at the growing blunt end of the nerve. Similarly when a nerve is divided each branch tends to grow straight forward in the direction of the fibres at the end of the branch. After a nerve has entered a given part of an embryo it retains a permanent connection with that part, and it is largely owing to the secondarjmigration of organs that the distribution of the nerves becomes so complicated in the adult. It is evident that the migration of the organs must take place after the nerves have reached them. Perhaps the most striking illustration of the translation of an organ with its nerve is affords by the descent of the testis — compare also the recurrent laryngeal, Fig. 300, R.

As the nerves all ^row forth in planes at nearly right angles to the axis of the neuron, it follows that the direction taken by each nerve depends largely upon the cerebral flexures and the curvature of the Bpmal cord. This is admirably illustrated in the human embryo. Fig. 360. The figure also shows that certain of the ner\'eB, as is more fully explained in the section on the spinal nerves, are brought into contact with one another and unite, forming the plexuses.

What has been said suffices to indicate some of the simple and almost self-evident mechanical conditions of nerve development.

Henson has suggested, 76.1, that the nerve fibres have from the start their permanent connections, and that as the cells divide and move apart, the nerve-fibres divide and lengthen out, and he has referred to the filaments seen in the mesoderm of young embryos as being such neire-fibres. This suggestion cannot be adopted, since the outgrowth of the nerve-fibres has been observed; moreover Altmann, 65. 1, has pointed out that the fibres seen in the embryonic mesoderm are really processes of the mesodermic cells, and, as shown in the excellent Fig. 2 of his plate, are quite distinct both from the ectoderm and entoderm; KdUiker also, 86.3, remarks that in the tail of the tadpole the number of nerve-fibres, and of the branches and anastomoses thereof, increases with the age of the animal, they being at first very few in number, so few that it is evident that the innervation of most parts must be developed later, there not being at first branches enough to supply all the terminal organs, which are ultimately furnished with nerves.

Union of Nerves and Muscles. — Trinchese, 88.1, gives a few details as to the changes in the muscle-fibres which precede and coincide with the union of the nerve-fibre with the muscle-fibre, but as he gives no figures, I am unable to follow his description.

Further Development of Nerve-Cells. — The early history of the nerve-cells has already been given, and the final differentiation of the nerve-fibres traced. We ^ have now to consider the histogenetic ^ changes in the main cell bodies, and their nuclei, Jirst in the medullary, second in the ganglionic nerve-cells. 1. Medcllaky Nerve-Cells. — We possess little satisfactory information concerning the phases of the young nerve-cells. The protoplasm of the neuroblast of His is apparently utilized to make the nerve-fibre, bo that very little is left around the nucleus; hence, in sections and in cells rtS^i^Bi^'tiS^piMfciSTaHlSTn isolated % maceration the nucleus Embryo (Mimit foil M) of iflo Day»^^, appears almost naked. The ner\-eceuTf-a/^^neuISKiu nuclei fnu' ^QBiion Cell nuclcus early becomes recognizhll' rS?m™ '!S^°un.ur:™micie?""distinct nucleolus. Fig. TBBwis. X nimin 1,(100 diaing, 3ijl, jiu. The next change consists in a growth of the nucleus and of the court of protoplasm around it. The outline of the cell now becomes irrt'gular and the production of the protoplasmatic process begins,

Fig. 301, A. The first of these processes (dendrites) probably arises during the second month, not, as formerly supposed, during the fourth month. W. His, 90.2, 50, observed that the neuroblasts had one or two short, blunt processes running off from the pole opposite the nervefibres ; in the medulla oblongata of the human embryo these processes were probably the beginning dendritic branches. In later periods (e.f/., sixth month) I hnd various stages at once. In more advanced ganglion cells the nucleus is very much enlarged. Fig. 301, 5, as is also its imcleolus, and the nucleoplasma is vacuolated. The protophism has grown very much, and I find it, at legist in the motor cells of the spinal cord of the human foetus, divided into an inner finely granular layer, and an outer layer with coarser granules, which I have not observed after birth. The further development consists, so far as known, simply in growth of all the jxirts. As to the progress of the dendrites, or protoplasmatic processes, the observations are unsatisfactory', owing chiefly to the failmre of investigators to recognize the difference between the neuroglia and nerve-cells.* In the chick the dendrites arise very early, as shown by Cajal and Lenhossek, 91.3, 118, beginning, namely, during the third and fourth daj's of incubation ; the first motor-cells of the spinal cord have branching dendrites the fifth day. The branches of the nerve-cells become very numerous and extend into the Randschleier of the embryo, and their interlacing causes a largo part of the network appearance which is so characteristic of the embryonic cord. There is no evidence sufficient, I think, to prove that the processes of neighboring nerve-cells unite ; compare W. Vignal, 88.1, 'Z'Zi), and Kolliker, Verb. Anat. Ges.," V., 7, His, 90.1, M. von Lenliossek, 91.3.

The cord and eiwh part of the brain has, as is well known, in the adult its si)ecial and characteristicallj' slia}XHl iier\'e-cells. Concerning the evolution in the foetus of these modifications, we know ver>' little. An isolated motor-cell from the cord of a sheep embr}'o of 10 cm. is figured mid describtnl by W. Vignal, 84.1, 231-233.'^ In older Htages the cells l)ecomo larger, their procc»sses larger and more branclie<l, and fibrillated — Vignal, /.c, 3(»i»-37r), describes the forms in human embryos of six, st^ven, eight, and nine months. In the cerebellum., Vignal, 88.1, 3'21», observed the first trace of the enlargement of the cells of Purkinje in a foetus of five months ; a month later the cells are liirger and conspicuous, and th(»y offer the ix*culiarity that their prot<^plasm is gathered almost wholly on the side of the nucltnis toward the surface of the brain. At six months Vignal could distinguish also the bodies of the small iier\'e-cells of the gnmular layer. In the cerebral hemispheres the enlargement of the nuclei and protoplasm of the large pyramidal cells (Meynert's third layer) lx>gins, according to W. Vignal, 88.1, 250, at five and a-half montlis in the human embryo; the protopbism prest»nts an irrc^gular outline; the nuclei stain more deeply thmi the neighboring ones. During the sixth month the cells elongate toward the exterior and so sissume their characteristic pyi^amidal form; their protoplasm is finely granuhir without very distinct outlines, and their processes or dendrites are neither long nor much branched. At birth the cells

  • ThU iH iiiitAhly the cane with Viiciiul, whu failed to rvcofrnize the oeuroKlia cellH before the fourth uiuntli, W.1, 31tf.

As the nerves all grow forth in planee at nearly right angles to the axis of the neuron, it follows that the direction taken by each nerve depends largely upon the cerebral flemireu and the curvature of the spinal cx>rd. This is admirably illustrated in the human embryo. Fig. 360. The figure also shows that certain of the nerves, as is more fully explained in the section on the spinal nerves, are brought into contact with one another and unite, forming the plexuses.

What has been said suffices to indicate some of liie simple and almost self-evident mechanical conditions of nerve development.

Hensen has suggested, 76.1, that the ner\'e fibres have from the start their permanent connections, and that as the cells divide and move apart, the nerve-fibres divide and lengthen out, and he has referred to the filaments seen in the mesoderm of young embryos as being such nerve-fibres. This su^estion cannot be adopted, since the outgrowth of the nerve-fibres has been obser^-ed ; moreover Altmann, 86.1, has pointed out that the fibres seen in the embryonic mesoderm are really processes of the mesodermic cells, and, as shown in the excellent Fig. 2 of his plate, are quite distinct both from the ectoderm and entoderm; Kollitcer also, 85.3, remarks that in the tail of the tadpole the number of nene-fibres, and of the branches aad anastomoses thereof, increases with the ago of the animal, they being at first very few in number, so few that it is evident that the inne^^■ation of most parts must be developed later, there not being at first branches enough to supply all the terminal organs, which are ultimately furnished with nerves.

Union of Nerves and Muscles.— Trinchese, 88.1, gives a few details as to the changes in the muscle-fibres which precede and coincide with the union of the nerve-fibre with the muscle-fibre, but as he gives no figures, I am unable to follow his description.

Further Development of Nerve-Cells. — The early history of the nerve-cells has already been given, and the final differentiation of the nerve-fibres traced. We ^ ha% enow to consider thehistogenetic ^ changes in the main cell bodies, and their nuclei, ^rsf in the meiliillarj', second in the ganglionic noire-cells. 1 Medullary Nerve-Cells. — We possess little satisfactory information concerning the phases of the young nerve-cells. The protoplasm of the neuroblast of His is apparently utilized to make the nerve-fibre, so that very little is left around the nucleus; hence, in sections and in cells

Fig. Ml.-OelUai TlcalReKinnodhe. solate hy maceration the nucleus

Embryo (iiinot Coll! 06) of ifti dbj-s. A. appears almost naked. The nervecrn'°M9l!'tt?u™«ii" n'ucirif Iiu' KMKulm Cell nucleus early becomes recogn i zh^'?o?m"?^,Slmn^;T"u?tnfX'™ ^^lo by its distinct nucleolus, Fig. rewi«. X aiwut i.nwd'iams. 3li1, HM. The next change consists

in a growth of the nucleus and of the court of protoplasm around it. The outline of the cell now lieoomes irregular and the production of the protoplasmatic process b^ns,

Fig. 3G1, A, The first of these processes (dendrites) probably arises during the second month, not, as formerly supposed, during the fourth month. W. His, 90.2, 50, observed that the neuroblasts had one or two short, blunt processes running off from the pole opposite the nervefibres ; in the medulla oblongata of the human embryo these processes were probably the be^nning dendritic branches. In later periods {e,g,j sixth month) I fmd various stages at once. In more advanced ganglion cells the nucleus is very much enlarged, Fig. 361, B, as is also its nucleolus, and the nucleoplasma is vacuolated. The protoplasm has grown very much, and 1 find it, at least in the motor cells of the spinal cord of the human foetus, divided into an inner finely granular layer, and an outer layer with coarser granules, which I have not observed after birth. The further development consists, so far as known, simply in growth of all the parts. As to the progress of the dendrites, or protoplasmatic processes, the observations are imsatisfactory, owing chiefly to the lailinre of investigators to recognize the difference between the' neuroglia and nerve-cells.* In the chick the dendrites arise very early, as shown by Cajal and Lenhossek, 01.3, 118, beginning, namely, during the third and fourth days of incubation ; the first motor-cells of the spinal cord have branching dendrites the fifth day. The branches of the nerv'e-cells become very numerous and extend into the Randschleier of the embryo, and their interlacing causes a large part of the network appearance which is so characteristic of the embryonic cord. There is no evidence suflScient, I think, to prove that the processes of neighboring nerve-cells imite ; compare W. Vignal, 88.1, 226, and Kolliker, "Verb. Anat. Ges.," V., 7, His, 90.1, M. von Lenhossek, 01.3.

The cord and each part of the brain has, as is well known, in the adult its special and characteristically shaped nerve-cells. Concerning the evolution in the foetus of these modifications, we know very little. An isolated motor-cell from the cord of a sheep embryo of 10 cm. is figured and described by W. Vignal, 84.1, 231-233. In older stages the cells become larger, their processes larger and more branched, and fibrillated — Vignal, /.c, 369-375, describes the forms in human embryos of six, seven, eight, and nine months. In the cerebellum, Vignal, 88.1, 329, observed the first trace of the enlargement of the cells of Purkinje in a foetus of five months ; a month later the cells are larger and conspicuous, and they offer the peculiarity that their protoplasm is gathered almost wholly on the side of the nucleus toward the surface of the brain. At six months Vignal could distinguish also the bodies of the small nerve-cells of the granular layer. In the cerebral hemispheres the enlargement of the nuclei and protoplasm of the large pyramidal cells (Mejnaert's third layer) begins, according to W. Vignal, 88. 1, 250, at five and a-half months in the human embryo; the protoplasm presents an irregular outline; the nuclei stain more deeply than the neighboring ones. During the sixth month the cells elongate toward the exterior and so assume their characteristic pyramidal" form; their protoplasm is finely granular without very distinct outlines, and their processes or dendrites are neither long nor much branched. At birth the cells

  • This is notably the case with Vifniali '^^o failed to recognize the neuroKlia cells before the fourth month, 88.1, 319.

aiv found in rarioua stages, both in the socond and third layer of Me^niert, but the mofit advanced of the large cells differ but little except in size (see Vignal, I.e., PI. IX., Fig. 2, a) from those at seven inontlis. The enlargement of the nerve-cells of the second layer ix-curs during the eighth month. Magini, 88.1, affirms that the cells do not have the pyramidal shape in the fcetal hemispheres, but resemble rather the cerebellar Purkinje's cells, and states that when the cells are colored with Qolgi's osmio-hichromate silver mixture, their processes appear varicose, having scattered nodular thickenings.

As regards the time of development of the nerve-cells. Below, 88. 1 , reiKirts that the cells appear first in the spinal cord and then in the brain in the following order: In the medulla oblongata, cerebellum, mid-brain, cerebrum. He further states that in animals boni helpless (man, dog, cat, rat, mouse, rabbit) the cells are much less developed in the brain than in those animals which are immediately active (horse, cow, pig, sheep. Guinea-pig). Vignal states, 88. 1 , that the Purkinje's cells (nerve-cells of the cerebellum) acquire tlu'ir cell-bodies in man about the sixth month, while the pyramidal cells of the cerebral cortex do not become equally distinct until the eighth month.

'i. Gaxoliosic Nerve-Cells. — These are all spindle-shaped bipolar cells in early stages, as above described; the cell-body and nucleus draw early to one side so as to -J. f'^ -- appear as a lateral appendage to the nerve 'EliS.- ' '" fibres. Fig. 3C.-2. There can be little doubt

^p^^. that the cell-body dniws more and more becomes pear-shaped ; imd that then the pointeil end elongjites

C ' . Vrii until it becomes a nerve-fibre, which joins

if^i' .:i"^>i^ "* fi'i angle the earlier fibre develo|xxl

from the two jxiK'S of the cell. That the

<vl]s thus develop appears probable fn an

the scinty olisorvjitions we poesess. and also because the development would agree tr.™ u^^u.HpuaiuaT^H.'.'r'iionwi with the Series of fonus which have iWn ^iu'viwk " •^'"'"" "' "" tnicwl by G. Retziiis. 80.1, through the vertebnite st.'ries. For example, in the lowest true vertebrate (Petmmyzon) Freud. 78.1, finds that the bi-jxilar form of the cells is ix^nnanent in the adult. The uuipi.>)ar fonn is foiuid in all amphibians and amniota. In a human onibr>-o of the tenth week, I iind the tvUs in various stages tif pr.> gress. Fig. \i'-'i: the nuclei an.' round, as seen in horizonhd sections i>f the ganglia, granular with distinct intra-nudear network: they vary in size: the smaller have so little pn>t*iplasm about them that they apix-ar almi.>st naketl : the amount of pr^rtoplastn increases with the size of the nucleus : the protoplasm lies on one side of the nucleus, and asi^umes n triangubir or (juadrilatef^outline in the sections : iN^tweeii the cells lie the triangular se^flltoC the nerve-fibivs. the tibrilLe of which apitear as (lots.

In the sympathetic gitnglia. the d^ A^ origi^f

discuaeed p. (>:10, retain the bi-j

brought near together, and one pole gives rise to the spiral, the other to the straight m)re of Beale and Arnold. Concerning the development of these cells, I know of no detailed observations.

Spinal Nerves. — It is singular that, although the early history of the spinal nerves up to the period of the unicpi of the nerve-roots has been the object of much investigation, yet their later history has been very little studied. Almost the only observations of importance are those of W. His, 88.3, 380-385. More has been done to elucidate the history of the hypoglossus and spinal accessory nerves, which, though morphologically derived from the spinal cord, have been annexed by the head, and may be conveniently regarded as cephalic nerves.

The results obtained by His, /.c, are as follows : The nerves toward the head develop more rapidly than those toward the tail. The nerve trunk formed by the union of the two roots, p. 622, grows at first in a plane approximately at right angles to the axis of the spinal cord, but owing to changes in the curvature of the cord the cervical and lumbar nerves very early appear oblique, Fig. 363. The obliquity increases especially in the neck, where the neck-bend is gradually lessened as the head of the embryo rises (compare Chap. XVHI.). After the trunk has grown a short distance tne fibres at the distal end are seen to tend to spread apart, and this spreading seems to initiate the branching of the nerve without any special obstacle causing it to divide in the way described on p. 623, for by their spreading the ends of adjacent nerves are brought into contact in the cervical and lumbar regions, and by uniting begin the formation of the brachial and lumbar plexus. A {)ortion of the fibres from one nerve join those of another, and the united portions constitute a new nerve trunk. In an embrj'-o of 7 mm. (His, Z.c, Tab. II., Fig. 4) the anlages of the cervical and brachial plexus are present, that of the lumbar flexure al)out to develop. In an embryo of 10 mm., Fig. 3G3, one can recognize, 1, the N, occipitalis minor arising from the first and second nerve; 2, 3, the N, auricularis magnus and N, cerricalis suprrticialis coming from the second and third nerves; 4, Xu, supraclaviculares, and 5, the N, phrenicus. The phrenic nerve, P, descends steeply past the brachial plexus and the wall of the thorax wliere it is lodged in a small ridge immediately behind the vena cava superior. The brachial plexus is formed by the fifth to eighth cervical and first dorsal nerves. In Fig. 363, the position of the arm anlage is indicatd by a dotted line ; it will be seen that it is such that one branch from the fifth nerve does not enter the arm ; the fibres which enter the arm become grouped in three main stems, but the steps by which the} become so grouped have not been clearly worked out. The second and third dorsal nerves have each an intercosto-humeral branch running toward the brachial plexus. The remaining dorsal nerves at this stage require no special description. Turning to the sacral nerves we find the first gives off two independent brandies, the ileo-hypogastric^ ih, and iTeo-ingualis^ ii, and a third branch, which unites with fibres from the second nerve to form the genito-cruralis, gc. The second to fifth sacral nerves together with the first to third coccygeal nerves unite to form four nerve trunks, which enter the leg, and one which does not. The attachment of the leg is indicated by a dotted line; the four nerves of the extremity are the cutaneus externus, c.e; the cruralis, c.r;

Kio. MS— Peripheral Kei

rouB SvKWni of B Hui ll-Xlf, uuiJhalle nor.

log Irtteni see It

r: I*, phrf'iiic nem*; ffc, onloKH -- ^- ,--. _. Jit) pwitluna ot till) linibB tre ladicaUnl liy d<

' or about I rial's Kinttliiiii

l«lllUf».) AftirW. UK

obtnratorivn, o; and the iscliiodicvs, i.s. The nerve stoin below the leg is the pndeiubis comtnuni.t, pic.

We know verj- little concerning the development of branches of the spinal nerves, other than those resulting from the contact of nerves with one another, imd which are concerned in the production of the plexus. We know from comparative anatomy that a spinal nerve has typically a dorsiil branch, which carries, i, motor fibres to the myotome (or its proiluct the miiscles) and, 2, sensory fibres to the skin of the back, and a ventral branch, which itwelf divides int*> two branches, one running to the somatopleuric wall of the splanchnocoel© and the other running to the splanchnopleure or \nsce«i. This tj-pe, as we know from Paterson's observations, 67.1, 91.1, reappeare in the development of mammals. The tnmk formed by the union of the sensoiy and motor roots grows only a very short distance before it undergoes its first or primary diviBion, one branch running to the primitive B^m.ent, the other continuing obliquely downward and outward. The cause of this division I do not know, but I think it possible that it may be dne to the nerve encountering the edge of the muscle plate. We now have the dorsal and ventral branches ; the latter grows on until, as shown by Paterson's observations, 9 1 . 1 , it encounters the niesotheliiun of the dorsalmost angle of splanchnoccele, whereupon the branch is forced to divide (rat embryo eight to nine days) into a , somatic and a splanchnic ' branch. Fig. 364, N.sotn, and N.spl. In this case the mechanical cause of the division seems unmistakable. The splanchnic branch, at least in the case of the dorsal and lumbar nerves of mammalia, is still further deflected to H horizontal course by the cardinal vein, and ia thus directed toward the aorta and enabled to join Paterson's fij-mpathetic cord. The somatic branch grows „ ^ ^-st

into the somatopleure, but pia ,„ T™D9ver.e section ot a »<«» Embryo of about very soon divides — cause "ST?"™? ,'," * ?I>1*wd Days hrouRh the Lumbar ReKloo. 1 ■ ■. , Md Hedulla spiuallB n donml root 01 naftUoii 1 D

unknown — mtO two superior d Tislon or nerve Et ectoderm £^ ipluicbDie

^.nn^l.^o Tk« f.'.-il braoob Son . somatic branch or iwrTs Pan puicreau SpL

branches. Ihe further ,p|B™ JL kidney « ord card nal »e n nHfTmeMnt^

history of the somatic 'K'^5thSli'^'^!;,„f'^™J^'V'"^Si;;i'!;lSS

Dotocnord tip A ^ na artcrr C , antral caoaL

nerve branches naS still After a M Paleraoa (The Bgura is compiled from several

to be ascertained. About the time these changes are going on, there is developed an increased separation of the roots of the primary dorsal and ventral rami, so that each has its discrete bundle of ganglionic and medullary nervefibres, Fig. 3G4.

Cervical Nerves. — W. His, 88.3, 360, points out that, while the medullary neuroblasts send their fibres all into the ventral roota throughout the greater part of the spinal cord, yet in the upper cervical region the neuroblasts in the zone of the future lateral horn send their fibres out in nerve bundles near the entrance of the ganglionic fibres. We have in this peculiarity a transition to the cerebrS type, in which the dorsal root is formed partly by medullary fibres.

In birds and reptiles the first and second cervical ganglia are present only during a very short early embryonic period (Chiarugi, 89. 2, 334) and then disappear entirely, as was discovered by Froriep, 82. 1, 83. 1. Froriep also observed that in mammals the ganglia continue their development, being present in the adult.

Sympathetic System

Two views have been advanced in regard to the origin of the sympathetic system. The older view, that of Bemak, was that it arose in situ from the mesoblast ; the later view, that of Balfour, was that it arose as a series of buds from the spinal nerves, the buds afterward becoming connected to form two main chains of sympathetic ganglia. Remak's view has been reestablished by A. M. Paterson, upon whose memoir, 91.1, I base the following account. It is possible that His' suggestion, 90. 1, is correct, and that the cells of the sympathetic are not mesenchjTnal, but cells which have emigrated singly from the ganglia. Good summaries of the literature on the subject are given by Onodi, 86. 1, and Paterson.

The first trace of the sympathetic may be seen in a mouse embryo of 'eight days (rat of 7 mm.) at a stage when the spinal nerve has nearly reached the mesothelium of the splanchnoccele, and the Wolffian tubules have just appeared. In the interval between the aorta and the cardinal vein the uniformity of the mesenchyma is now broken by a group of cells, which differ strikingly from their neighbors ; the cells stain deeply ; their nuclei are large and often possess a considerable number of nucleoli. This mass of specialized cells is bilaterally symmetrical and extends from the level of the cephalic border of the fore-limb to the level of the stomach. It constitutes a cord on each side, and is the anlage of the sympathetic system. The cord is comparatively large anteriorly, and gradually tapers off and becomes indistinct posteriorly. It has no connection with the spinal nerves or ganglia. Longitudinal sections show that the cells are fusiform and elongated lengthwise of the cord, and that the cord offers no tnwje of segmentation.

the next step in the development is the union of the spinal nerves with the sympathetic cord; the imion takes place only in the dorsal and lumbar region, not in the neck or in any segment of the body posterior to the bifurcation of the aorta. It is the splanchnic branch only which joins the sympathetic cord. Fig. 364, Spl. In rat embryos of 8.5 mm. (eight to nine days) the cord is slightly larger than before, but is still in close proximity to the aorta and presents no sign of constriction or segmentation ; the ventral branch of the nerve has just reached the angle of the splanclmocoele and is dividing. In mice embryos of nine days the branch has grown about half-way to the cord ; in those of ten days it has almost reached — in those of eleven days it has actually joined — the cord. The cord itself now has ventral branches and its cells mingle with the nerve-fibres, and later the cells migrate along the nerves. In the anterior thoracic region the whole of the splanchnic branch joins the cord, but in the lower thoracic and in the abdominal regions some of the fibres pass beyond. In the neck above the point of origin of the vertebral artery the splanchnic bnmches, as already stated, have no connection with the sympathetic cord. After union with the nerve, the cord loses its boundaries, and its cells acquire, Fig. 3C5, Sy, greater size and branching processes. Though the splanchnic nerve Dranch elongates considerably it continues to end in the cord. It was the observation of tbi& condition coupled with the eissumed

necessity of tracing all supposed nerve-cells to an ectodermal origin, which led Balfour to his theory of the origin of the sympathetic ^ cconl. The splanchnic nerve-fibres distribute themselves through the cord and its s pi branches, also penetrating the cervical portion of the cord which does not receive any of the cervical nerves.

" In transverse sections," says Paterson, I.e., p. 171, " of a human embryo about the end of the first month, hardened in spirit and stained with aniline blueblack, the sympathetic cord [KS.I^iif^^^^rspuS'chSicn' has very much the character just described. The cord itself is large and uniform in width, widening out anteriorly to form the inferior cervical ganglion ; beyond this it narrows, encloses the subclavian arterj-, and forms a fibrous cord; this again becomes cellular, and widens out into the "superior" eer%"ical ganglion. No splanchnic branches join the cord in front of the level of the inferior cervical ganglion. In the thorax (Plate 28, Fig. 10) the splanchnic branches are seen {spl) arising from both roots of the spinal nerve (/, Z>), and, as in the figure, terminating wholly in the sympathetic cords (»i/). Sometimes a small portion of a splanchnic branch can be traced round the ventral side of the cord, accompanied by a ceUular branch from it. In the hinder thoracic region, a small part only of the splanchnic branch joins the cord, the greater part, along with cellular outgrowths from the sympathetic, passing onward to form the solar plexus and semilunar ganglia, which are seen in process of formation on the ventral aspect of the aorta. A similar fibro-cellular bundle passes to join the supra-renal body. In the lumbar region the splanchnic branch can be seen for a considerable distance almost entirely unconnected with the sympathetic cord, and separated by an interval from it. The cord gradually narrows as it is followed l»ck:ward, and becoming attenuated disappears at the point of bifurcation of the aorta. "

The third step is the gangliation of the cord, that is to say, the formation of the series of enlargements, which constitute the adult ganglia, the thinner portions of the cord persisting as the interganglionic commissures. The commissures come gradually to consist chiefly of nerve-fibres. The ganglionic thickenings first appear (human embryo of 1&-19 mm., mouse embryo nineteen days) where

the nerves join the mesenchymal sympathetic, and presumably result from the growth locally of both the nerve-fibres and the sympathetic cells. As' the parts gradually attain their adult form, the regularity of the alternate swelling and constriction does not persist, but as the

ganglia become defined in form their position tends to become irregular ; while one may lie in the interval between two vertebrae, the next may be seen opposite the vertebra i tself . The parts derived from the sjnnpathetic cord in the neck above the inferior cervical ganglion may be regarded as belonging to the peripheral or collateral distribution of the sympathetic nerve, because they have no direct connection with the cervical nerves. A fibro-cellular bundle springs from the cord and accompanies the vertebral artery; beyond this the original cord, which is at first terminated at the level of the mouth, becomes constricted by the formation of a fibro-cellular commissure separating oflf the superior cervical ganglion. This ganglion ends head ward in a fibrous bundle, which accompanies and is lost upon the internal carotid arter}' beneath the auditory capsule. The middle cervical ganglion, when present, is to be regarded as formed of a gp'oup of cells, which have been included in the commissure. Fig. 300, MO. The connections of the sympathetic cord with the cranial nerves have 3'et to be investigated. As regards the caudal termination, the sympathetic cord is at first ill-defined behind the region of the kidney's ; it gradually extends further back, alongside the aorta and middle sacral artery, where the two cords become closely approximated. They become graduall}" more and more attenuated, and finally disappear. Near their termination they are joined together on the dorsal aspect of the middle sacral artery by cellular commissures, from which the connecting loop and ganglion impar are developed. No fusion of the two cords can be seen until they have reached their permanent posterior limit. The sympathetic cord behind the lumbar region may be regarded as belonging to the peripheral -c^ ofji a ^ *u * distribution of the cord for the same reasons as the

Ic Ganglia of One Side CCrVlCal portlOU.

onheflfth^Mon^hl^TSe The peripheral branches from the sympathetic ne™w TOnlnectedVith ^*^^^^» including the Collateral ganglia, as well as the the jfanKiionic chain, medullary portious of the supra-renal bodies, the 8.gI Miperior Kauiu- ^^P^ri^r ccrvical ganglia, etc., are formed by outon;\w.(riniddre gan- prowths from the cord, which are at first cellular.

glion; I.G, inferior fjL,

ganglion: , great Thcsc give nse to ganglia, nerves, and plexuses, ter*A!^*M/ Patereon.'^" and are accompanied by the parts of the splanchnic branches of the spinal nerves, which do not join the ganglia. In this category- are placed doubtfully the gray I'anii communicantes.

General Morphology of the Cephalic Nerves.*— It is now generally believed by embryologists that the nerves which spring from the brain form a part of the same morphological series as the spinal nerves. Unlike the spinal nerves they vary greatly among themselves both in their development and in their permanent character, and at least one of them, the optic nerve, appears to have a different morphological value from a true nerve. It is, therefore, impossible to give, as was attempted for the spinal nerves, a comprehensive history of the nerves of the head, but instead we must study each nerve separately.

The following table gives a list of the cerebral nerves and shows with which division of the brain each is connected :

Table op the Cranial Nerves.

Vesicle. Nerve.

First I. Olfactory.

II. Optic.

Second III. Oculo-motor.

IV. Trochlearis.

Third V. Trigeminus.

VI. Abducens. VII. Facial. VIII. Auditory. IX. Glosso-pharyngeal. X. Vagus.

Spinal cord XI. Spinal accessory.

XII. Hypoglossus.

I give below the separate history of each nerve, and in the following paragraphs of this section I have discussed certain general (questions of the mori)hology of the cerebral nerves.

The first point to be emphasized in regard to the cephalic nerves is that, as discovered by W. His, 88.3, there are three sets of roots, one ganglionic, the other two medullary. The ganglionic roots are part of the same series as the sensory roots of the spinal cord. The two sets of medullary roots are parts of the same series as the single set of spinal motor roots. It is, therefore, a peculiarity of the brain, that its medullary fibres have their points of exit along two longitudinal lines on each side. Both lines are situated in the ventral zone of His : one is toward the Bodenplatte and may be regarded as the prolongation of the line of the ventral roots of the spinal cord ; the other is close to the edge of the dorsal zone of His, and, therefore, immediately below the ganglionic root. It appears a justifiable hypothesis to assume that every segment in the head had originally its segmental nerve, and that every nerve had three roots, one sensory and two motor, i. e., one lateral and one ventral motor root. The lateral root is the distinguishing characteristic of a typical cephalic nerve, t but its existence has been long overlooked because

  • For an admirable r<^um<i of the profirreas up to 1888 of our knowledfce of the development of oej)halic mTves «ee W. His. 88.2. 879-409.

1 1 cannot but think that the spinal nerves also will be found to have lateral roots.

it is BO closely joined to the ganglionic or dorsal root that it has been generally mistaken for a part of a dorsal root. It is this mistake which has been the principal obstacle in the way of investigations upon the morphology of the cephalic nerves, and the correction of this mistake by His is, to my mind, the most important contribution to the morphology of the brain which has been made for a long time past. The relation of the three roots is well illustrated in Fig. 370.

As stated in Chapter IX. there are probably seventeen or eighteen segments in the vertebrate head, and perhaps seventeen or eighteen neuromeres in the brain (see above). As yet, however, only twelve nerves have been observed in any adult vertebrate. Of these nerves some are purely ganglionic, others are purely medullary, and still others are mixeil, and one of them {hypoglossiis) arises by the fusion of parts of four nerves; of the medullary nerves, some represent lateral roots, like the accessorius, others ventral roots like the abdiicens. If, therefore, the cephalic nerves were derived from seventeen or eighteen segmental nerves, they must have undergone very extensive modifications. Morphologists are endeavoring to trace out these modifications, and to establish thereby the hj^iothesis that the cranial nerves represent a series of segmental nerves. That these endeavors will be successful can hardly be doubted by competent embryologists.

The second point to be emphasized is that the gill-clefts are not segmentally arranged, and that all attempts to ascertain the segmental value of cranial nerves by determining their relations to the gill-clefts are based upon an erroneous assumption. As explained in Chapter IX., each of the three anterior gill-clefts, counting the mouth as one, corresponds to several segments. It is possible that the posterior clefts are segmentally arranged, but these clefts are without branchial nerves of their own, being innervated from the vagus. As regards the nerves connected with the clefts, to wit, the trigeminal, facial, glosso-pharyngeal, and vagus, we can conceive them as representing each several segmental nerves, either by being the product of the fusion of several primitive nerves, or by being one each of a group of nerves, the rest of which are aborted. The branchial nerves are recurred to in a paragraph below.

A third important point is the subdivision of each primary cephalic ganglion into an upper (lateral or main) ganglion, and a lower (or epibranchial) ganglion. The development of the lamprey, as worke<l out by C. Kupffer, suggests that every cephalic ganglion had primitively two direct connections with the epidermis to make the lateral and epibranchial organs, and the development in the amniota suggests that two ganglia are differentiated from the primitive one, and that in some cases a cephalic ganglion represents the primitive, in others one of the secondary, ganglia. Thus we may hj-pothetically regard the ciliary and trigeminal ganglia as primary ; the acoustic as a secondary lateral line ganglion; the facial as a secondary' epibranchial ganglion ; while in the case of the glosso-pharyngeal and vagus nerves, both secondary ganglia are preserved, Ehrenritter's and the jugular ganglia being assigned to the lateral, the petrosum and nodosum to the epibranchial series. I can, of course, only suggest this hypothesis as an obvious corollary of Kupffer's discovery, and though its justification must be left to the future, yet it seems to me now very plausible.

The nerves of the head have very different values, and are by no means morphologically equivalent one to another. It seems certain, however, that not one can be homologized with a single complete segmental nerve, that is to say, a nerve in which, aside from its commissures, there are to be found all the nerve-fibres, both ganglionic and medullary, of one segment united in one main trunk. On the contrary, no cephalic nerve is the equivalent of more than a part of a complete segmental nerve. Even those cerebral nerves which are derived from the fusion of several nerves do not include the whole of each nerve component.

We may conveniently distinguish between those nerves of the head which are derived from part of a single segmental nerve, and those derived from the fusion of parts of several segmental nerves. Unfortunately this distinction rests at present chiefly on hypothetical identifications. We have to class provisionally, as single nerves, olfactory, oculo-motor, trochlear, and abducens — and perhaps acoustic, as compound nerves, trigeminal, facial, glosso-pharyngeal (?), vagus (?), accessorius, and hypoglossal.

Concerning the roots, a few general remarks may be made. We have already insisted upon the triple division into dorsal sensory roots, lateral motor roots, and ventral motor roots. The dorsal and lateral roots are situated so closely together, the former at the ventral edge of the dorsal zone of His, the latter at the dorsal edge of the ventral zone, that they appear as one root so long as the origin of the fibres is not considered. We have, in fact, several nerves, which arise apparently from one root, but which in reality arise from tw' o roots closely united ; such are the trigeminal, facial, glossopharyngeal, and vagus nerves. If the lateral root aborts, the sensory root may remain ; such nerves are the olfactory and acoustic. In the reverse case the lateral root persists, as occurs with the oculomotor ( ?) , trochlear, and spinal accessory nerves. The ventral motor roots, like those of the spinal cord, to which they are partially equivalent, have an independent exit : they persist only in the abducens and h}TX)glossus.

A constant feature of the persistent ganglia is probably that the ganglionic fibres as soon as they enter the medulla form a longitudinal bundle, which grows tailward close to the outer surface and in the lower part of the dorsal zone of His. This bundle is homologous with the similar bundle in the spinal cord. The bundle is known as the ascending tract in the anatomy of the brain and behind the vagus as the tractus solitarius. It has been shown to receive fibres in the embryo from the trigeminal, facial, glosso-pharjTigeal, and vagus ganglia.

I will now give a synopsis of the interpretations of the twelve cerebral nerves, which appear to me indicated by our present knowledge of the development of the nerves, as reviewed in the following twelve sections, and by our knowledge of the position of the cephalic segments as described in Chapter IX.

I append a table, modified from Zimmermann, 91.1, 100, which indicates the relations of the nerves to the neuromeres so far as at present rendered probable. The assignments niade in the table are in my judgment all more or less problematical.

1« Olfactory, Probably ganglionic, though the development of its ganglion differs from that of the other ganglia; belongs to the first (and second?) segment.

2. Optic. Probably not a true nerve.

3. Oculomotor, Lateral root with sensory ganglion, which aborts very early ; belongs to the first or second segment of the midbrain.

4. Trochlear, Lateral root with sensory ganglion, which aborts very early ; belongs to third segment of mid-brain.

5. Trigeminus. Sensory and lateral roots of several segments.

6. Abdaceius, Ventral root, perhaps of a single segment, and of the same segment to which the facial nerve belongs.

7. 8. Facial is-acustic us. Sensory and lateral roots of several nerves. The acustic may include two distinct ganglia and would then represent two sensory roots. The facial intervenes between the two parts of the acustic, and may prove to be the sensory and lateral roots of one segment. 9. QlossO'pharyngeus, Sensory and lateral roots of one, possibly two segments.

10. Vagus. Sensory and lateral roots of a single segment, but secondarily connected by means of a persistent epibranchial commissure with the innervation of several gill-clefts of the hypoglossal region.

11. Accessorius, Lateral roots of four hypoglossal nerves, of which the ganglia are temporarily developed, with accessions of fibres from cervical nerves.

12. Hypoglossus. Ventral roots of four occipital nerves of which the ganglia are temporarily present and of which th6 lateral roots form the accessorius.

Fork-brain. Mid-brain.



1 2

8 4



9 10 11 12 13 14 15 IG 1

Dorsal root.

lAt<*ral root.


Motor-oculi. ' (y Motor-oculi.)


? AcuHticus. Facialis. AcuKticus. ( i lomo-pharjmgeus. Vagus.


Tripreminus. Facialis.

( r losso-pharyngeus. Vagus.





Ventral root.


nypogloflBus. HypoglosBoii. HypoglosRus. HypogloBBua.

Branchial Nerves

The relations of the nerves to the segments (myotomes and neuromeres) are primitive, the relations to the branchial arches and gill-clefts are secondary'. Indeed we must assume that the vertebrates had segmented ancestors, who acquired gill-clefts, segments being phylogenetically much older than gillclefts. The ancestral nerves were adapted to the gill-clefts, and we may some day know the history of that adaptation and the modifications consequent upon it. At present we can only say that, contrary to the assumption which has prevailed for twenty years, the gillclefts are not segmental and therefore the branchial nerves are not in segmental order.

The unquestionable branchial nerves are the facial, glosso-pharyngeus, and vagus. To the same series we must probably assign the trigeminus after subtraction of its ophthalmic branch, for it enters into the same relations to the mouth as the other nerves mentioned to the gill-clefts ; as we have seen, the mouth is probably a modified pair of gill-clefts. Coimting the mouth as a gill-cleft, we may say that each of the four nerves arises by the union of a lateral root with a ganglion to form a common nerve-trunk, which springs from or passes by the epibranchial organ of the ganglion and descends behind the cleft with which the nerve is associated, in the visceral arch between that cleft and the next following. Later there arises a branch which passes in front of the cleft; the main stem is then known as the post-trematic branch, the secondary branch as the prsB-trematic branch. In the lamprey the whole series of epibranchial organs are connected by a continuous longitudinal conmiissure. In mammalia all trace of the conmiissure is lost except behind the vagus, which thus ia permanently associated with the fourth and fifth clefts of amniota, to which it does not morphologically belong. Gegenbaur's hypothesis that the vagus represents several branchial nerves is not tenable, for reasons explained below. I regard it as probable that the hypoglossus, with which I include the accessorius, will be ultimately recognized as including the branchial nerves of the fourth and fifth clefts, if indeed these clefts ever possessed true branchial nerves.

I. Olfactory Nerve. — Van Wijhe, 82.1, 18, has sought to prove that the olfactory nerve is not really the first but the second of the cerebral nerves, and that it arises further back morphologically than the optic nerve. The development of the fore-brain, as worked out by His in the human embryo, p. 595, renders it very diflBcult to accept this notion, and the arguments presented by Chiarugi, 91.1, seem to me conclusive that the olfactory nerve is really in front of the optic.

His, 89.4, 717-723, finds in the human embryo that the nerve develops as follows : The first step is the separation of the olfactory plate, p. 575, from the wall of the brain by an ingrowth of mesenchyma. This separation has been observed by KoUiker, 90.6, in chicken embryos of the fourth day and in a cow embryo of 10 mm. The second step is the production of the olfactory ganglion; the ectodennal cells of the olfactory plate multiply, the karyokinetic figures being found next the outer or free surface of the layer ; the cells thus produced assume the appearance of medullary neuroblasts, and at four w^eeks are found migrating toward the mesenchymal surface, so that the base of the layer of the olfactory ectoderm becomes crowded with nuclei ; the protoplasm of these neuroblasts is collected on one side of the nucleus in a ]X)intetl mass ; the cells now grow forth from the ectoderm and constitute the anlage of the ganglion between the ectoderm and the brain. The third step consists in the assumption of the bi-polar form * by the cells of the ganglion, and the elongation of the poles on the one side as centripetal nerve-fibres which join the brain, on the other as centrifugal fibres which join the olfactory epithelium (embryos of five weeks). It thus appears that the development of the nerve is accomplished during the fifth week in the human embryo. Kolliker has observed that in the rabbit of thirteen days the ganglion has reached the olfactory lobe, but its centripetal fibres have not penetrated the wall of the lobe ; he also observed in the same rabbit that the nuclei of the ganglion were dividing karyokinetically, and he considers it probable that these divisions result in forming chains of cells, each chain developing into one nerve-fibre, and he thinks that in the adult the fibres are multinucleate. Chiarugi states, 91.1, that the olfactory nerve is present in the guinea-pig embryo of 4.7 mm., before there is any olfactory lobe, and that it extends from the brain wall to the olfactory plate. Miss Piatt, 91.1, 200, affirms that the olfactory ganglion is derived from the neural crest, but has published no proof of this affirmation.

Concerning the morphological interpretation of the olfactory nerve no satisfactoiy conclusions are yet possible. Marshall, 78. 1, 82. 1, advanced the theory that it is a true segmental nerve, or at least the dorsal root of one, but its development differs so much from that of the ordinary ganglionic ners^e that I hesitate to accept this theory. Marshall has sought to strengthen his theory by homologizing the nasal pits with a pair of gill-clefts, but the observations he has reported, 79. 1, do not seem to me to justify the homology, and he has failed to attribute weight to the fact that gill-clefts are primarily evaginations of the entoderm, while the nasal pits are invaginations of the ectoderm and have no connection with the pharynx in any vertebrate. J. Beard, 85.1, modified Marshall's theory, and homologizes the olfactory plate and its ganglion with an epibranchial or lateral sense organ. We know (Chap. XXVIII.) that the ganglionic sense organs arise by a union of the ganglion with the ectoderm, but the olfactory sense organs arise by a differentiation of both the sensory ectoderm and the ganglion from a common ectodennal plate. Nevertheless, it remains a tempting hypothesis, which places the nose in the series of segmental sense organs, but at present it is still merely an hypothesis with no secure bfisis. If it is verified hereafter, we may recognize in the olfactory nerv^e a true ganglionic nerve or dorsal root, or perhaps the representative of a series of roots, since it is iKDSsible that a numter of segments liave disappeared from the ])ra3-oral region, and each segment may be supi)osed to have had its nerve.

That the olfactory nerve corresponds to a spinal dorsal root is rendered probable by, 1, the formation of its fibres from bi-polar cells ; 2, the ingrowth of the fibres from the ganglion into the wall of the neuron.

II. The Optic Nerve. — The development of the optic nerve is treated together with that of the eye. Chapter XXVIII. Concerning the morphological value of the optic nerve nothing is known, nor can we hope to form any satisfactory hypothesis as to its value until the development of the optic nerve-fibres is thoroughly understood. At present we are unable to say whether it is to be regarded as a modification of a true nerve or of a cerebral commissure.

• Chiurugi, 91.1, suggests that souh^ of the cells uiay Ix' more than bipolar.

Fig. 867.— Trans vprsft Section through the Posterior Part of the Mid-brain of a Human Embryo of five weeks (His' embryo Ko).

III. The Oculo-motor Nerve.— The oculo-motor nerve, according to W. His, 88.3, 366, aiises from neuroblasts of the ventral column of His in the mid-brain, Fig. 367; transverse sections of the brain of this embrj^o are represented in Figs. 368, 369, 370. W. His, 88.3, Fig. 26, has figured the nucleus of the third nerve as a broad group of pear-shaped neuroblasts, which give off the centrifugal fibres of the nerve; some of the oculo-motor neuroblasts point centralward (His, I.e., P.Martin, 90.1), and Martin states that he has obser\^ed bi-polar forms in the cat; as to the further history of these two peculiar kinds of cells we have no information. As sho^n in Fig. 363, the nerve grows in a perfectly straight line to the caudal edge of the eyeball, where it joins the anlago of the eyemuscles. Here the nerve must branch, since it is distributed in the adult to five muscles, viz. : the levator palpabrse, rectus superior, rectus intemus, rectus inferior, and oblicjuus inferior. No observations on the development of these branches in the mammalian embrvo are known to me.

The development of the motor oculi in elasmobranchs has been much studied, with conflicting results. In Scyllium and Pristiuris it appears, according to Van Wijhe, 82.1, 22, while the third gillcleft is developing, which is about the stage when the anterior roots of the spinal nerves develop according to Balfour. In Balfour's »ia^ L, the nerve after crossing the opthalmicus profundus runs to the posterior eilge of the "first myotome" of Van Wijlie; compare A. M. Marshall, 81.2. The path of the nerve passes the ciliarj' ganglion (ganglion mesocephalicum of Beard and Dohrn), but has no connection with that ganglion (Dohrn, 91.1, <>), as has been erroneously assumed by some writers. Miss Piatt, on the contrary, sa.^-s, 91.2, 90, that the nerve begins as a single cell thrown off from the ciliary ganglion. This view rests probably on erroneous interpretation of observations, for it cannot be admitted that a motor nerve is formed by ganglionic fibres. Dohni, /.c, affirms positively that medullary cells leave the wall of the brain and enter the nerve, and he traces to these cells the development of those which constitute the ganglion of the nerve; but his observations are ver>' far from convincing to me, and I still regard it as jwssible that the cells observed in the nerves are mesenchymal, and if this is the case then it is also possible that the ganglion of the nerve is of mesenchymal origin and homologous with a sympathetic ganglion.

The ganglion of the oculo-motor nerve in selachians was discovered by G. Schwalbe (Jenaische Zeitschr,^ 1879), and was identified by him with the ciliary ganglion of human anatomy. Van Wijhe found the oculo-motor ganglion in his embryos in Balfour's stage O, and pointed out that it was distinct from the true ciliary ganglion, which belongs to the ophthalmicus profundus nerve. C. K. Hoffmann, 86. 1, 302, recognized the two ganglia in reptiles, but applied the term ciliary to the ganglion of the oculo-motor, and the term ophthalmic to that of the ophthalmicus profundus. J. Beard, 87.2, put an end to confusing the two ganglia, but unfortimately proposed to restrict the term ciliary to the oculo-motor ganglion, and to introduce the name of mesocephalic for the ophthalmic or true ciliary ganglion. Beard's nomenclature is erroneous, for, as shown by His, 88.2, -1:21, the ciliary ganglion of the embryo is identical with the ciliary ganglion of the adult, and the oculo-motor ganglion is always morphologically distinct from the ciliary. Beard's proposal added to the existing confusion by misapplying the term ciliary. Antonelli, so far as one can judge from the abstract of his researches, 90. 1, has again confounded the oculo-motor and ciliary ganglion. The true oculo-motor ganglion has yet to be discovered in mammalia. For notices of the conflicting descriptions of the structure of the adult oculo-motor ganglion, see A. Dohm, 91.1, lG-28.

If the known oculo-motor ganglion is sympathetic, then it is possible that the thalamic nerve discovered by Miss Piatt and described in the following section, is really the true ganglion of the third nerve.

III. a. The Thalamic Nerve.— Julia B. Piatt, 91.2, 07, discovered a rudimentary ganglion in Acanthias embryos appended to the dorsal part of the mid-brain close to the fore-brain. In a subsequent paper, 91.1, she has added further details. The ganglion is developed from the neural crest and retains a connection ^vith the ciliary ganglion along what must be regarded as the epibranchial line commissure. The commissure is stated to give rise to the ramus ophthalmicus profundus of the adult. The ganglion proper has a transitory existence. It seems to me probable that the ganglion may prove to be, as suggested in the last section, the true primitive ganglion of the oculomotor.

IV. The Nervus Trochlearis, or Patheticus. — The origin of this nerve in the embryo long eluded investigation; thus Marshall and Spencer, 81.1, and Van Wijhe, 82. 1, 25, failed to ascertain its early history. His, in 1888, 88.3, 305, reported that in a human embryo of the fifth week the fourth nerv^e can be traced. Fig. 308, from its point of exit from near the median dorsal line of the isthmus (compare Fig. 303, IV) as a bundle of fibres running down through the mantle layer of the medullary wall to a group of neuroblasts, from which the nerve arises, and which are situated in the part of the medullary tube corresponding to the ventral zone of His. It must l>e assumed that the neuroblasts send out the fibres in a different direction from what we find in the case of all other medullary nerve-roots, but Martin's observations, noted below, indicate that the peculiar course of the fibres results from migration of the neuroblasts. It may be added that the position of the nucleus of the nerve in the adult agrees with that of the neuroblasts, as observed in the embryo by His. P. Martin, 90.1, reports that in the cat the fibres do not cross in the earliest stage, but make their exit on the same side on which their neuroblasts are situated, and that the neuroblasts themselves lie at first higher up, and later migrate to the ventral position, in which they were seen by His, as just stated. Froriep, 80.1, 57, has observed in young torpedo embryos that the nerve of either side receives fibres from both sides, and both he and Dohm, 91.1, have observed in elasmobranch embryos that the nerve forms a plexus of its own fibres on its way from the brain to the muscles it innervates.

Fig. 308.— Section of the Brain of a five Weeks embryo (His" Ko). IV Fourth nerve: A', neuroblasts of the nerve in the ventral zone of His. After W. His.

Dohm, 91.1, 9-11, has observed cells in the course of the nerve, especially at certain points where they are accumulated so as to produce a thickening of the nerve. Dohm designates these cells as nerve cells derived from the medullary canal, but neither his description nor figures justify this conclusion. It is more probable that these cells are surviving remnants of the trochlear ganglion or possibly merely immigrated mesenchymal cells.

The ganglion of the trochlearis was discovered independently by A. Froriep, 91.2, and Julia B. Piatt, 91.2, in elasmobranchs. It is a part of the neural crest, and is continuous for a time with the anlage of the trigeminal ganglion; the connected band of cells breaks down irregularly, but its scattered remnants persist for a time along the original line. At this stage the motor-fibres grow out from the medulla near the dorsal summit of the ganglion, and the permanent trochlearis is developed. Miss Piatt * speaks of the ganglion as the "primitive trochlearis, " and ^e interprets, p. 97, the ramus ophthalmicus superficialis trigemini as a survival of the original connection between the trigeminal and trochlear ganglia. As the connection here mentioned is on the level of the dorsal line of the neuron, it may be regarded as a part of a lateral line commissure. The discovery of the ganglion of the fourth nerve further demonstrates that the motor fibres represent a lateral root. In torpedo embryos of IG mm. Froriep, Z.c, 56, has found a small group of ganglion cells, which soon disappear, but at this stage are appended to the caudal side of the nerve a short distance below the ventral limit of the mid-brain. These cells are probably a remnant of the original ganglion. Miss Piatt thinks that the trochlear ganglion also contributes to the ciliary ganglion, but her proof of this appears unsatisfactory to me.

V. The Trigeminal Nerve. — This is one of the most complicated nerves of the head. It is developed from both the ganglia and the medullary tube, and has permanently both sensory and motor roots. Its ganglionic portion is double, comprising the ciliary or ophthalmic ganglion and the Gasserian, and it will be advantageous to consider these two parts as morphologically distinct. The motor root forms a single bundle ; the nerve enters into sjiecial relations with the epidermis, and finally it develops a typical system of branches. Each of these fundamental characteristics forms the subject of a separate paragraph following.

  • Miss Piatt's description Is somewhat obscured by her overlooking the fundamental difference between medullary and ganglionic nerves.

Ganglion Ciliare and Nervus Ophthalmicus Profundus. — This is the ganglion which has been long and generally known as the ciliary, and becomes the ciliary of the adult ; for mention of other names applied-to it see p. 040. The centrifugal nerve arising from the ganglion is known as the ramus ophthalmicus profundus, the centripetal nerve as the radix longa, which joins the trigeminal ganglion before the radix enters the brain. How the ciliary ganglion becomes separated from the trigeminal is unknown, so far as amniota are concerned, but in elasmobranchs Van Wijhe thinks, 82.1, 20, that a considerable middle portion of the originally continuous ganglionic mass disappears. In the human embryo at one month the ciliary ganglion is connected with the trigeminal by a bundle of fibres without cells. His, 88.3, 372. Beard, 86.1, 30, was the first to observe that the ganglion unites with an epidermal thickening of the lateral line. He says : " Cells are then proliferated off from the skin to form the ganglion, and the outer portion of the thickening begins to form the primitive branchial * sense organ. From the thickening cells are given off for some time until a large ganglionic mass is formed, which still for some time remains fused with the skin." C. Kupffer, 91.1, has found in Petromyzon embryos a large ganglion. Fig. 407, which lies in front of the trigeminal ganglion ; this ganglion is probably the ciliary and it has connection in the larva (Ammocoetes) of 4 mm. with an epibranchial organ: this suggests that there may be an epibranchial organ of the ciliary ganglion in the higher vertebrate embryos.

It is probable that the cells of the ciliary ganglion become bi-polar and produce ganglioiyc fibres, but, so far as I am aware, no observations on the origin of the nerve have been published. If the nerve arises as suggested, then the centrifugal fibres must constitute the ophthalmic nerve, the centripetal the radix longa, or as it is called in human anatomy the ophthalmic branch of the trigeminal, compare Fig. 3G3. In this figure the ciliary ganglion overlies the eye and is united with the trigeminal ganglion, (L Tx, and sends its nerve forward toward the fore-brain. Why the fibres pass to the brain by way of the trigeminal, instead of making an independent entrance, is unknown. A. M. Marshall found the nerve to run forward from the ganglion in elasmobrancli embr^'os in Balfour's stage K, past the upper border of Van Wijhe's first segment and the inner side of the eye, to end «'it a point just dorsal of the nasal pit. Some further details are given by Van Wijhe, 82.1, 20-22.

2. Ganglion G assert, or Trigeminal Proper. — After the separation of the ciliary ganglion the Gasserian (His, 88.3, 372) has in side view. Fig. 363, G.G, a somewhat triangular form in the human embryo ; ite apex points dorsalward and sends the centripetal nerve- fibres into the brain. The peripheral nerves it gives off are accompanied by some of the ganglion cells, which are thought by His to be destined to form the anlages of the gttnff I ion rhinirnm and g. oticnm. The fibres which enter the brain do so near the angle formed by the junction of the dorsal and ventral zones of His, and there take a longitudinal course as a bundle of fibreB homologous with the longitudinal bundle formed by the spinal nerves. This bundle is the tractus trigeminus or ascending trigeminal root of authors ; it lies close to the surface of the brain and is oval in section, being flattened laterally, Hie, I.e., Fig. 27. The bundle grows slowly down toward the spinal cord. In the adult it is said to extend into the ceivical cord.

  • In cuusequence of later researches we should substitute ** lateral " for **branchiaL*'

Kupffer, 91.1, 41, has observed that in Petromyzon larva of i mm. the trigeminal ganglion overlies the mouth cavity, Fig. 407; it has a strong root and the root contains fibrilhe, and its main peripheral stem branches near the ganglion to form the maxillary and mandibular branches, both of which are compact cords of fibres with nuclei among them and partially covered by a cellular sheath. The main trunk is also connected with the thickening of the epidermis, which constitutes the third of the four epibranchial organs overlying the mouth at this st^e. Froriep, 86.1, 4.1, searched carefully but unsuccessfully for an epibranchial oi^an connected with the Gasserian ganglion in mammalian embryos.

3. Motor Root or Portio MraoR. — The motor root of the trigeminus is developed from neuroblasts of the ventral zone of His in the bind-brain at the level of the Varolian bend, Fig, 363. These neuroblasts are gathered together, forming the trigeminal nucleus, which early becomes recognizable. The nucleus lies. Fig. 3G9, near the junction of the ventral and ', dorsal columns and there- \ fore close to the ascending sensory root, ov, of the trigeminus. The fibres from the neuroblasts are gathered into a single stem and make their exit, as shown in Fig. 3<)!>, near the dorso-lateral edge of the ventral zone (His 88.3, 3G5).

4. Peripheral Branches. — The trigeminus is so named because in man it was wne; mr, otaI 1>iindlB or uceodlng tract: V, fltth or trige.

observed to have three branches. One branch, as we have seen, runs to the ciliary ganglion, and must be considered as belonging morpholc^cslly rather to that ganglion, than to the Casseriau. The other two branches run r^pectively to the maxillary and mandibular regions. In the lamprey, Kupffer, 91.1, 41, the two branches arise from a common stem, but in the human embryo they arise separately from the ganglion. Whether the maxillary and mandibular nerves are to be regarded as branches of one nerve or not, must be decided by further investigations. It is possible that they are distinct and their union secondary, but the usual view is that they are primitive branches. This view has found favor chiefly from theoretical considerations : if the mouth be interpreted as representing a pair of gill-clefts, then the trigeminus may be interpreted as the nerve of that cleft, and its two branches, one in front of, the other behind, the mouth, may be compared with the branches of the branchial nerves.

No satisfactory observations on the growth of the branches are known to me. The subject would well repay a careful investigation.

VI. Abducens Nerve. — This nerve is formed exclusivelv of medullary nerve-fibres. The neuroblasts which produce these fibres have been found by His, 88.3, 365, in a human embryo of five weeks, to be situated in the ventral zone of His toward the median ventral line. Fig. 370, and the fibres pass out directly from the wall of the brain, hence the exit of the root lies in a line with that of the hypoglossal nerve and much nearer the ventral line than the exits of the main branchial nerves (trigeminus, facialis, glosso-pharyngeus, and vagus) — comj)are Fig. 3G3. Fig. 370 also shows the peculiar manner in which the abtlucens is embraced by the inner root of the facial. The fibres do not pass out in one bundle, but as firet observed by A. M. Marshall, 78.1, in several (four to seven) small bundles.

The facts that the abducens has no ganglion and arises from the ventral side of the brain, were discovered by A. M. Marshall, 78. 1, and verified by Van Wijhe, 82.1, 28. Both authors interprete<l it as a ventral root, homologous with a spinal ventral root, and correlated with a dorsal root represented by the facialis. His, 88.3, has shown that the relations are more complicated, and has rendered MarslialFs simple hypothesis untenable.

As regards the growth of the nerve, little is known. In torpedo embryos of 10 mm. (Froriep 91.2, Fig. 1) it nms straight forward to the caudal end of Van Wijhe's third segment, which is the anlage of the external rectus muscle of the eye. A. Dohm, 91.1, 11-10, states that in elasmobranchs the nerve appears in Balfour's stage L; at first only two, later more fibres could be observed. The nerve at the time it reaches the rectus anlage is very thin, later it is much thicker. Dohrn also asserts that medullary cells continue to ent^r the nerve and migrate along it during a prolonged period.

VII. -VIII. The facial and acoustic nerves are developed in all vertebrates in such intimate connection with one another, that thev are necessarily treated together. We shall take up : 1 , the development of the ganglion ; 2, the motor roots : 3, separation of the acoustic ganglion : 4, the peripheral branches.

1. Ganglion AcousTic-FACiALE. — As already stated, p. 604, this is the smaller and posterior of the two primary ganglionic masses, which may be seen in front of the otocyst in a chick of thirty to fort>' hours and in corresponding stages of other amniote embryos. His, 88.3, 372, gives the following description of the ganglion in a five weeks' human embryo: It lies close in front of the auditory vesicle. Fig. 3G3, Gv; it is somewhat triangular in form, with its apex toward the dorsal eide ; in sections ita elements present a characteristic fan-like grouping, which recurs in no other ganglion, and which is due to the twist^ course of the Bbree of the vestibular and cochlear branches of the acoustic nerve. The ganglion includes three masses of neuroblasts; the innermost or medial mass, Fig. 370, c, VIII., is the aulage of the ga ng/ioH cochfei-f, and sends its centripetal fibres as shown in the illustration dorso-laterally; the outermost or lateral mas3 is the anlage, r.VIII, of the ganglion rest ibii /are, and its fibres enter the brain with a dorso-medial inclination ; the middle mass is the anlage of the facialis or ganglion geniculi ; it lies somewliat lower down than the other two, and its centripetal fibres form a strikingly compact cord y^ within the substance of the brain.* Paul Martin, 90.3, 22!i, has observed vwr in cat embyrosof 0.«-0.'J mm. that certain fibres of -vu*^ the facialis liend over so as to form a longitudinal wDSimiriSi'I.'"'! cord which later joins the {."'ve^^iii^pig","

ascending gloSSO-pharj-n- v™tlh'ular tmnch; vn. wnin Dervr; c Mil (viwhu gealtract,which is, therefore not forme<l merely by glosso- pharyngeal fibres. The whole of the ti'ipls ganglion becomes later included in the cartilaginous mass of the »if petrosiim, but a few cells are retained on the cerebral side and form a ganglion, which is known by various names, and which His proixJSPB to call the iniracrauial. According to C. Kupffer, 81.1, the acustico-facialis ganglion of the lamprey unites in the embryo with four spots of the epidermis, two along the lateral and t\vo along the epibranchial line. Of the former one is a union with the epithelial wall of the auditorj- invf^nation, the other lies further headward, being situated between the otocyst and the trigeminal ganglion ; wliere the anterior union tabes place the epidermal cells contribute to the development of the facial ganglion. The two lower unions take place by means of ventral prolongations of the ganglion, which unite with epidermal thickenings above the first and second gill-cleftH respectively. Kupffer's statements suggest that the ganglion is really double, otherwise it is difficult to understand wliy it shoidd have two lateral line oi^ns and two epibranchial organs. Van Wijhe observed in el asmobranchs, 82.3, 20, that the facial ganglion unites along what we now regard as the lateral line and again above the first gill-cleft with the epidermis; the latter connection can be seen in Balfour ^s etage K. !Beard, 86.1, also observed the epibraiichial connection. In amniota the lateral line connection has not yet been described, but Kupffer, 91.1, 52, states that it has been found in birds. The epibranchial connection of the facial g;anglion has been very carefully studied in mammals by A. Froriep, §6.1 ; it is present in cow embryos of 0-12 mm., and is most distinct in those of from 7-9 mm., that is to say, with three gill-clefts, well developed externally ; the lower end of the ganglion is somewhat pointed and joins a small thickened area of the epidermis exactly' at the dorsal margin of the first or hyomandibular cleft (Froriep, /.r., Taf. I., Fig. I.); there is no distinct boundary between epidermis and the ganglion, and it is possible that tlio former contributes cells to the lattei-; the thickened area is slightly invaginated below the level of the surrounding epidermis; a little later the ganglion is found to have made a clean-cut separation from the skin.

it Cho Bgure of thJg ganglion In

The fate of the facial ganglion proper has yet to be traced. The embryonic facial nerve ha« in its ganglion, of course, ganglionic neuroblasts, and must be regardetl as originally a mixed nerve.

2. Motor Roots. — Our knowledge of these is derived almost exclusively from the observations of His, 88.3, 362, for His is alm«.)st the only embryologist who has studied the histological development of nerves, and it is only by such study that the history of the motor roots can be followed. In a human embryo of five weeks, the facial nerve-fibres leave the brain as a compact bundle, a little distance headward of the auditory vesicle and at a point just ventral of the root of the acusticus; this bundle may be followed, Fig. 370, for some disUmce within the brain, ascending at first, then arching over and descending near the lK)rder between the mantle layer and the inner layer toward the median ventral line, where its fibres spread out and apparently take a longitudinal course; the facialis neuroblasts are situated in the lateral part of the ventral zone of His and lie in the region of the otocyst ; the course of the fibres fn)m the neuroblasts to the actual root has not been fully traced, but His thinks they join the formatio arcuata, then enter the longitudinal bundle near the median line and form there the arching bundle of fibres just described. The circuitous course of the motor fibres is very early develo|)ed, but no reason for that course is yet known.

3. History of the Acoustic Ganglion and its Nerve Branches. — The following account is based on a jiaper by Wilh. His, jun., 89.1, in which the development in the human embryo is described, and the previous researches of others are reviewed. As stated in the previous paragraph, traces of the triple division of the ganglion are evident toward the end of the fourtli w(^k. By the middle of the fourth week the auditoiy vesicle. Fig. 371, shows the anlages of the cochlea and the semicircular (*anals, and the ganglion shows clearly its triple division ; the facial nerve has its characteristic bend, for it descends from the brain very steeply, passes through the horizontal ganglion geniculi, Cr.f/, and then descends agiiin. Theacustic ganglion lies closer to the brain-wall than the facial and is divided by the latter into the upi)er and outer ganglion vestibuli, Gt\ and the

lower and inner ganglion cochleae, Qco. The facial ganglion descends to a lower level than the acoustic, and therewith the two have finally separated. A few days later the division of the acoustic ganglion into an upper and lower part becomes still more marked, beoiuse the root of the facialis takes a more nearly horizontal course to the facial ganglion and then descends. Both parts of the acoustic ganglion lie in front of the otocyst and come in contact only with its front wall, and it is only on this wall that the maculae acusticee are developed. At five weeks the semicircular canals having formed and the twisting of the C(x:hlea having begun, the fibres of the acoustic ganglion are found united with the auditorj- vesicle. The fibres from the cochlear ganglion form a stem, the nervvs coehle.arix, and two smaller branches, corresponding to the middle branch of otologists, which run respectively to the anlage of the macula sacculi and the anlageof the macula ampullcE posterior is. The fibres from the vestibular ganglion form a single stem running to a spot which includes the anlages of three maculte, namely, of the vestibule and of the

UuiK lOVrnin) Rei-onstnicilon " i' wi ol hraln. A.ir, »ntiTior. P.w, roBtprk.. Bk.ic. cxivnul neiiiii-lrpular cankl anlaae: VIII. ■udllory nprrr; VII. [acial nerru: Ou, (cauKll'Mi Tenlbull; O.g, caiiiillon mniculi; ^c". fnuwllon cnchlen: K. r, m hull; (Y, utriculiw; Kor. Baccubiu. lea. Afier W. Hl^ Junior,

anterior and external ampullse ; in more advanced stages the macula separate and each receives a separate branch of the vestibular nerve; this is an excellent illustration of the -dependence of nerve branches uiMtn secondary changes in their peripheral connections. While the nerve branches are developing the ganglia elongate ventralward, and at the same time changes occur in the distribution of the ganglion cells. In the cochlear ganglion most of the cells remain near the cochlea, where they are ultimately converted into the spiral ganglion; otiiera ascend with the fibres to the brain, at the edge of which they acrumiilate. being stopped by the dense neuroglia (Randschleier), and give rise to His' inter-cranial ganglion, mentione<l above; still otherb remain strung out along the line where the cochlear ganglion is in contact with the vestibu^r; this line of ganglion cells is <»lled

the Zwischen-qanglion by W. His, jun.; the fibres from these cells constitute the branches running to the sacculus and the posterior ampulla. In the vestibular ganglion the cells are more evenly scattered and persist in the adult distributed along the nerve. The further development consists in little more than a series of adaptations to the advancing differentiation of the membranous labyrinth. Fig. 372 represents the parts just described, but in a somewhat more advanced stage. As to the course of the fibres within the brain, we possess no satisfactory information; see W. His. jun., /.c, p. 17-19.

4. Peripheral Branches of the Facialis. — In mammalian embryos, soon after the facial ganglion has united with the epidermis to form. the epibrauchial organ over the hyo-mandibular cleft, the nerve proper grows down into the hyoid arch, and thus develops the homologue of Van Wijhe's post-trematic branch. Somewhat later, Froriep, 86.1, 44, another brancli is formed from the oral side of the ganglion, and this branch, which is probably homologous with the rami prae-trematicus and phamygeus of selachians, extends into the mandibular arch. Froriep has observed, 87.1, in tt^rpedo embryos in Balfour's stage L, a branch of the post-trematic facial running forward below the gill-cleft into the mandibular region to there iniier\^ate a mucous canal; this brancli Froriep considers the homologue of the chorda tympani of amniota; the union of the chorda with the trigeminus is secondary. The branches in ehismobranch embryos have been carefully describe<l by Van Wijhe, 82.1, 25-21), who refers also to the earlier observations C)f Balfour and of Marshall and Spencer.

Rabl, 87.1, ascribes the peculiar distribution of the facialis in the adult mammal to the fact that it innervates the myotheliimi of the hyoid arch ; this myothelium develops into the embryonic platysma, and the platysma spreads out and is ultimately differentiated into the superficial facial muscles. The nerve follows the muscle, and as the latter subdivides the former branches correspondingly.

IX. The glosso-pharyngeal nerve has been taken by many embryologists as the most typical nerve of the head, because it has two distinct r(X)ts and its relations to the second gill-cleft are very clear, and it has l>ec»n assumed that the cranial nerves typically all have two roots and are similarlv related to gill-clofts; compare His, 88.2, 423. It is to be remembered that the assumption that the glosso-pharyngeus is par excellence the typical cerebral nerve is the outcome of the necessities of a certain school of speculative morphologists. The assumption is by hypothesis, and is by no means sufficiently upheld by observation. We will consider: 1, the ganglion and its sense organs; 2, the motor roots; .'J, the peripheral branches.

1. The Ganglion and its Sense Organs. — The ganglion is the third of the four primary ganglionic masses of the head, and is situated immediatelv Miind the otocvst. It forms at first a continnous anlage with the vagus ganglion. In a chick of thirty to forty hours, seen from above, it ap|>ears as a rounded mass about equal to the auditor^' vesicle in size (His, 88. 1, 41 7) . It has lK?en commonly stated since Marshall's pajKT, 78. 1, in 1878, that there is first formed a common ganglionic mass behind the ear, and that this ma.ss divides

into two ganglia, the glosso-pharyngeal and vagus. Chiarugi, however, 90.1, 336, believes that the ganglion of the eighth nerve arises in Sauropsida as an independent outgrowth of the ganglionic cord (neural crest), and appears before the vagus. He finds, p. 426, that in the rabbit the two ganglia are distinct though they appear at nearly the same time (embryos of 4.5 mm.). In the human embryo the cells become bi-polar and produce nerve-fibres during the fifth week. The primitive mass, according to His; 88.3, 374, early divides into an upper or dorsal smaller spindle-shaped part, Ehrenritter\s qanglion^ and a lower or ventral larger oval part, the ganglioii petrosum proper. Fig. 303, Gp, The former lies close behind the auditor}' vesicle and later is covered by the cochlea ; the latter moves away from the otocyst to take a place on a level with the pharynx. the centripetal fibres form a single bundle, which enters the brain near the lower e<lge of the dorsal zones of His, and there taking a longitudinal course descends toward the spinal cord;* Nvithin the medullar^' wall the fibres constitute the ascending glossopharyngeal tract. In an embryo of 6 1> mm. NL (His' Br') His, .('., found the tract not to have reacheil the vagus region, but later it is longer and the fibres mingling with those of the vagus form a very characteristic cord, the tracfns soUtarivs^ which can be followed into the .spinal cord. It is probable that both the trigeminal and facial ganglia send fibres to this tractus.

The nerve was erroneously supi)osed by Balfour (" Works," I., 425), Marshall and Van Wijhe, 82.1, i), to arise exclusively from the ganglion, as owing to their neglect to consider the origin of the nerve-fibres they failed to see the true motor roots. Proceeding upon this false assumption they have endeavored to interpret the nerve as the morphological equivalent of a dorsal spinal root. His' observations oblige us to discard this interpretation.

C. Kupffer, 01.1, 44, found in the lamprey that the glosso-pharj'ngeal ganglion is diflferentiated later than the other cephalic ganglia, and is at first intimately associated with the anlages of the auditory vesicle and facial ganglion. Like the other ganglia it is soldered in the embryo to the ej)idermis of the lateral line, and after widening out at its ventral end it unites (ammocoetes of 4 mm.) broadly with the epidermis a second time to form the epibranchial organ above and in front of the third gill-deft, Fig. 407.

In Petromyzon, as just stated, the ganglion has the lateral and epibranchial organs, and ft is probiible that both exist in other vertei)rates; but as yet only the mammalian epibranchial organs have IxK'u accurately studied l)y Froriep, 86.1, although the lateral line organ was seen by Van Wijhe, 8».l, 20, in shark embryos in Balfour's stage K. Froriep, /.c, p. 12, observed in cow embryos of 8.5 nnn. that at the dorsal border of the second gill-cleft there is a slightly depressed area of thickened epidermis, which is united with the lower part of the ganglion; in embryos of 16 mm. the organ has disappeared, but its final history is somewhat uncertain (p. 46).

2. Motor Root. — The origin of the motor-roots in the embryo has, so far as I am aware, been studied only l)y W. His, 88.3, 361.

  • the arrangeineut is flKureii by W. His, 88.3, Fig. 'il\ it is Himilar to that of the va^us. See Fif?.

The neuroblasts are gathered in the upper part of the ventral zone of His, as a group which is quite clearly separated from the neuroblasts of the facial and vagus nerves. The fibres from these neuroblasts are gathered into a single bundle, which leaves the medullary wall near the dorsal end of the ganglion, so that it seems to form, if we disregard the origin of the fibres, a part of the true dorsal or ganglionic root; compare p. G48.

3. Peripheral Branches. — The glosso-pharyngeus enters into close relations with the second gill-cleft. As long known through comparative anatomy, the nerve typically forms two branches when it reaches the gill-cleft, and the general history of these branches has been followed in elasmobranch embryos by Balfour, Van Wijhe, and Beard. One branch rims in front of the gill-cleft — in other words, in the posterior part of the hyoid arch ; this branch is the prcetrematic of Van Wijhe, 82. 1 (prae-branchial of Beard). The other branch runs behind the gill-cleft — in other words, in the anterior part of the first branchial arch ; this branch is the post-trematic of Van Wijhe (post-branchial of J. Beard). These branches are both developed after the epi branchial organ, and in fishes are nearly equal in size.

In mammals, according to Froriep, 86.1, 13, 20, 44, the posttrematic becomes the main stem, which is found in cow embryos of 8.8 mm. running through the first branchial arch and curving forward below the gill-cleft, while the prse-trematic branch is a very small bundle of fibres at this stage, and apparently persists a.s the vermis tytnpanictts of the adult. The post-trematic is the ramus lingnalis of the adult, the ramus pharyngeus being added in later developmental stages. It may be noted that the so-called phar^Tigeus of elasmobranch s belongs to the prae- trematic.

In the human embryo the nerve grows straight down from the medulla at first (His' Br'), but in an embryo of four and one-half weeks (His' Ko) it is already bent at its end owing to the dislocation of the parts of the pharynx. His, 88.3, 3Ti>. Noteworthy is the early union of the ganglion |)etrosum with the ganglion nodosum by an oblique anastamosing branch, Fig. 3G3, the development of which has not yet been followed.

X. The Vagus Nerve. — A few words on the general morphology of this nerve may be prefixed to the history of its development. Gegenbjuir, 71.1, 72.1, directed es})ecial attention to the fact that, unlike any other nerve of the head, the vagus supj)lies several gillclefts; all the clefts, whatever their number, behind the glossopharyngeal cleft being innervated by the tenth nerve, which in fishes shows its relations clearly , since it sends off a pne-trematic and post-trematic branch for each gill-cleft of the vagus series. The numl)er of the branches in any form, of course, depends upon the number of clefts preserved in that form. As Gegenbaur had formed the theory that the cephalic nerves correspond with the gill-clefts, there being a nerve for each cleft, he necessarily concluded that the vagus was the morphological equivalent of several branchial nerves. This conception of the vagus has been generally adoptetl, and has been so generally taught, that many of the younger morphologists seem to have forgotten that it has remained a bold hypothesis, and that there is no evidence whatever of an actual fusion of several nerves into one vagus nerve to be obtained from vertebrate embryology. Nevertheless, Gegenbaur's theory has dominated all investigations of the last twenty years.

We now know — compare p. 200 — that the gill-pouches only imitate the segmental arrangement, and are in reality much less numerous than the true segments of the branchial region, and that the nerves do not correspond to the number of segments. In view of the great irregularities of the nerves as compared with the myotomes of the head, we are no longer justified in interpreting the vagus so as to make it conform to a theoretical order, which is definitely ascertained not to agree with the real order — in other words, it is not necessary to suppose that each gill-cleft had a separate nerve and just one nerve. Further we must conceive that there was primitively a chain of epibranchial organs, which were connected longitudinally with one another, and transversely with several hypoglossal nerves, but we have at present no reason for assuming that the series of cephalic nerves extended as far as the epibranchial organs. On the contrary the series of epibranchial organs (like those of the lateral line) may have extendeil tailward, by the growth of a branch consisting of nei've fibres derived from probably several hypoglossal nerv^es ; both the lateral and epibranchial branches while they grow are united with the epidermis.

I conceive the primitive condition to have been one in which there were, presumably, four cephalic nerves behind the vagus, and that these nerves had each its epibranchial organ ; the four nerves are now repre^sented by the hypoglossus and accessorius. The epibranchial organs were connected with one another by a longitudinal commissure, which persisted while the four hypoglossal ganglia disappeared, and thus the epibranchial organs and the nerve branches running from them to the gill-clefts became, apparently, branches of the vagus. While one thus recognizes the relation of the vagus to several gill-clefts, that relation is not primar}', but secondary and acquired, and does not in my judgment lend support to Gegenbaur's hypothesis. Another consequence of the abortion of the hyi)oglo«sal ganglia has been to leave their lateral medullary roots to be modified into a separate nerve-stem, the accessorius, and to join the ganglion of the vagus.

The considerations advanced above lead me to the conviction that Gegenbaur's conception of the vagus as morphologically equivalent to several nerves can no longer be maintainoil, and instead we must return to the older view and again look upon the relation of the vagus to the posterior gill-clefts as the result of the distribution of a branch, which may l)e named the nerims epibranchialis, and which, so far i\s its coimections with the epidermis are concerned, may he comi^iircd with the lateral nerve. That 6egenl)aur*s theory is untenable is shown by the development of the hypoglossal nerve, which includes the nerves of the segments immecliately behind the vagus nerve and above the posterior branchial clefts, so that, as a matter of fact, the segmental ner\"es of the posterior branchial region are incorporated not in the tenth, but in the twelfth nerve.

1. Ganglion and Ganglionic Organs. — The ganglionic crest behind the otocyst develops its two large ganglia somewhat later than does the mass in front of the otocyst; thus in a torpedo emhrj'o of G mm. Froriep, 98.2, GO, found the two anterior ganglia divided, but the two posterior were undivided. The ganglion is in amniota at first a rounded mass, which may be seen in a chick of thirty to forty hours lying immediately behind the glosso-pharyngeal ganglion, which it about equals in size (W. His, 88.2, 417, I'ig. •^). The exact history of the ganglion has never been followed. Chianigi, 90. 1, observ^ed that the ganglion arises in reptiles as a conical bud, which grows dowii from the neural crest ; later (Lacerta embryos of 5.5 mm.) it arises by three bimdles of fibres, of which the first and last represent the persistent neural crest and unite the ganglion respectively to the glosso-pharjTigeus and first cervical ganglia, while the middle bundle is the root proper, connecting the vagus ganglion with the neuron. This stage has been describetl ])y Beraneck and was the earliest seen by him. In mammals (Chiarugi, 90.1, 4"i-4<i) the ganglion also arises independently, and as it grows ventralward, passes outside of the jugular vein and aorta, unlike the glossopharj^ngeal ganglion, which passes inside these vessels. In rabbit embryos of <).o mm. Chiarugi found the ganglion attached by slender commissures to the glosso-phaiyngeal ganglion in front and the first cervical behind. the medullary rcxit next lengthens considerably, and in embryos of 1 1 mm. the ganglion is subdivided into the dorsal ganglion JHffvIxt re and the ventral ganglian nodosum. In a cow embryo of 8-9 mm. the ganglion is mu(*h larger than the glossopharyngeal or facial, and extends over the third, fourth, and fifth clefts (Froriep, 86.1). It elongates in an obli(|ue dorso- ventral direction. In a human embryo of four and one-half weeks (W. His, 88.3, 375) it isvery long and divided, as just mentioned for rabbits, into an upper ganglion jugularo and a lower ganglion nodosum. Fig. 303, (?/ and Gn^ connected by a narrower fibrous tract, along which are scattered a few ganglion cells. The jugular ganglion is spindle -shai^ed, and has on its inner side a bundle of fibres which enters the lower edge of the dorsal zone of His, Fig. 370, there takes a longitudinal course toward the spinal cord a« a well-marked ascending vagus tract: the tract is at first very short; it is soon joined by the fibres of the ti'igeminal tract, and the two sets of fibres uniting constitute the so-calleil tractus solitarins, as mentioned above in describing the trigeminal ganglion, p. r»P2. The tractus solitarius, Jis sho\>ni in Fig. 371), has at first a suixu-ficial position, but later it losos this place in appearance, being coveriMl over by the Randlippe of His, compare p. GiWj and Fig. 3S1.

The vagus ganglion probably has lK)th lateral and epibranchial organs in the embryos of all vertebrates. In elasmobranchs Ijoth were scnmi by Van Wijhe, 82.1, and have lx»en more accurately descril>ed by Froriep, 91.2. In Petromyzon th(^y have Ixhmi described by Kupffer, 91.1. In teleosts the lateral line organs are greatly dev(»lo|>ed, and there are a good many observations on the organs themselves, ])ut 1 recall none on the do|x*n(l«'nce of the organs upon the ganglion propcT. Kupffer, I.e., states that in birds l)oth the lateral and e])ibran<'hial fusion of the ganglion with the e])idermis can Ix? seen. The epibranchial organ in mannnalian embryos has l)een

carefully studied by Froriep, 85.1. Fig. 373 represents a transverse section of a torpedo embryo of 12 mm. in which the vagus ganglion shows its two connections with the epidermis, first at the lateral line, secondly over the fourth gill-cleft, where the thickening of ectoderm is very considerable. In this embryo (Froriep, 91.2, 61) the vagus ganglion is connected with the epibranchial organs over the second to sixth cleft, and with six smaller lateral organs, which all lie in the region of the fourth and fifth clefts, and in the headw^ard prolongation of the lateral line proper. The section figured passes through the fourth epibranchial organ. In a lamprey larva of 4 mm. the vagus ganglion. Fig. 407, as seen from the side, is triangular, the apex pointing tailward and being prolonged as the lateral line ; the upper angle forms the dorsal root ; the lower angle is prolonged and joins the epibranchial organ in front of and above the fourth gill-cleft ; this organ is the seventh at this stage, and is connected with the epibranchial organ in Fig. sra. -Torpedo Embr>'o of la

- Cross section. Md.ob, Me front and the chain of nve organs over the uuiia obionmtta; nch, notoc^hord: fourth to eighth gill-cleft. As regards ■*:^-^^^-J^\^!^^^rJ-o^;

mammals, Fronep, 86.1, states that the ehr, epibranchial orj^an; r, vein. ' 1 . ^. ^ 1 . , After Froriep.

vagus ganglion is found m cow s embryos of 8.7 mm. to be the largest of the cephalic ganglia, and to overlie the third cleft and the region of the still undeveloped fourth and fifth clefts ; above the third cleft and from these down beyond the level of the fifth aortic arch, it is found united with the epidermis over an area about 0.75 mm. long and from 0.19 to 0.23 mm. wide. In an embryo of 12 mm. the epidermis of the area of epibranchial contact has become invaginated and lies at the bottom of a narrow fissure, but is mlich reduced in size. The fissure and contact can be still found in embryos of 15 mm. ,

2. Motor Roots. — The neuroblasts, which form the motor roots of the vagus, are situated according to His, 88.3, 360-362, in the ventral zone of His, but toward its dorsal side, and the fibres make their exit from near the dorsal limit of the ventral zone and close to the entrance of the ganglionic root. The vagus neuroblasts are situated along the same line as those of the spinal accessory nerve and are not marked off from them in any way; compare Fig. 379.

3. Peripheral Branches. — The vagus ganglion in young elasmobranchs sends off four branches to the gill-clefts; each branch runs behind the gill-cleft to which it belongs, and is associated with the corresponding epibranchial organ ; the first to third branches are nearly alike in size, but are smaller than the large fourth branch, which is further distinguished by continuing on beyond the pharyngeal region and by becoming the ramus intestinalis of the adult (Van Wi jhe, 82. 1 , 32, Froriep, 9 1 . 2, 61) . Later this fourth branch is also connected with the epibranchial organ of the sixth cleft. The four branches behind the gill-clefts are to be regarded as post-trematio nerves and the fourth is presumably equivalent to two post-trematic nerves. The pr»-trematic branches arise later as outgrowths of the ganglion from the region of the epibranchial oreans, and also the pharyngeal branches arise similarly. Van Wijhe, I.e., p. 31, found that in Balfour's stage K the ganglion has one dorsal branch, and gives off the so-called lateral nerve ; the dorsal branch Van Wijhe identifies as the ramus supra-temporalis ; it is connected with the epidermis of the lateral line. It is probable that both the dorsal and the lateral nerves are derivatives of the connection of the ganglion with the lateral line. As Van Wijhe neglected the difference between the ganglion and the nerve his investigations must be extended before we can decide whether the four branches to the gill-cleft arise from the ganglion proper or from a nerve-tnmk which was mistaken for a prolongation of the ganglion. The question raised is important, since upon the answer must depend, to a large extent, our notion of the origin of the nerve, whether it represents one nerve much branched or several nerves which have been fused. Balfour's account of the development of the vagus in sharks differs somewhat from Van Wijhe's — see Balfour's '*Comp. Embryolog}%" II., 457.

In mammals the early condition of the vagus branches has been partially described by A. Froriep, 85. 1. In cow embryos of S.T-JS.d mm. the vagus surj^assos all other nerves in size; in those of 12 nmi. the ganglion jugulare is well differentiated from the ganglion ncxiosum, and from the former the main trunk extends for about 0.4 mm. as the anlage of the ramus intestinalis ; the tiimk at this stage consists entirely of nerve-fibres and contains no cells; the fibres pjiss through the medial half of the ganglion. As to the branches to the gill-arches and the lateral line no published observations are known to me.

Lateral Xerce. — This branch of the vagus is one of the best known nerves of the Ichthyopsida, and is connected with the sensory organs of the lateral line. The homologues in amniota of the lateral nerve have never been satisfactorily determined. The nerve itself is perhaps a imrtial surv^ival of a connection of the epidermis with the ganglia, wliich originally extended along the head as well as along the body, and which was ass(x;iatetl with the series of lateral sense organs; compare C. Julin, 87.3. In amphibia (A. Qoette, 76.1, 672) and in elasmobranchs (C. Semper, 76.3, 3!»8, Van Wijhe, 82. 1, 33) the growing end of the lateral nerve has lx?on seen to merge in the epidermis, and these observers suggest that the ner\'e may grow at the exjiense of the epidermis ; but this notion is scarcely compatible with our present knowledge of the genesis of nerve-fibres.

XL The spinal accessory nerve (accessorius Willissi) is characteristic of the amniota and is not found in the anamniota. It must, therefore, be regarded as a nerve which has been evolved within the vortc»brate series, and its development indicates that it arose by a collective modification of the motor fibres of the dorsal root« of the hypoglossus. It comprises no gjuiglionic fibres. Chiarugi, 90.1, found in reptiles, birds, and mammals that the neural crest persists, as it does in elasmobranchs according to Van Wijhe, 82.1, 32, betwet»n the vagus and first cervical ganglion, and continues as a cellular cord, both while the hypoglosHal ganglia growout from it and after these ganglia abort. He r^emls it aa the anlage of the accesBorius, and thiu is probably correct, but not in the sen^e that its cells produce the nerve, for the nerve contains no ganglionic fibres, but in the sense that it prescribes the path for the motor-fibres and conducts them to the vagus ganglion. I venture the liypotheaisj that if the hypoglossal ganglia were preserved the fibres of the accessoriua would not run to the vagus, but chiefly if not wholly to the twelfth nerve. His, 88.3, 3(iO-;t(i2, found the neuroblasts which give rise to the accessorius fibres to be distributed, as shown in Fig. 374, along the dorsal part of the ventral zone, throughout the vagus and hypoglossal regions, i.e., roughly the lower third of the medulla oblongata; the fibres, unlike those of the hyp<^los- ~

sus, make their exit near the dorsal zone; the fibres leave the medullar' wall an a series of little bundles, which unite into a nerve which runs forward nearly parallel with the medulla, being probably guided by the ganglionic cord, and joins first the vagus ganglion, then the main vagus-tnmk. Fig. 303, XI. The longitudinal trunk of the accessorius is regarded by Chiarugi, 90. 1, 317, as a modification of the original neural crest transformed in the occipital r^ion (j into a commissural cord. Some further details are given by Froriep, 86.1, as to this nerve in ruminant embryos. As regards the branches of the nerve. His, aWr koj" K^E^dym™ 88.3, ;180, finds that in the human embryo of four and one-half weeks the adult relations are already established. Fig. 363, in that the fibres all join the vagus and run for the greater part with its descending stem, but a part of them pass off as the independent ramua extermts N. accexsorii; compare also Froriep, 86.1, i:i-U.

XII. The bsrpogloBsal nerve of mammals has been shown by Froriep, 8B„ 1, to be the result of the fusion of several nerves, probably four, closely similar to the true spinal nerves in character. Froriep's results have had confirmation by P. Martin's observations, 80.3 on the cat, Cbiarugi's, 89.2, on several mammals, and Van Bemmelen's, 89.1, on reptiles. As the homologies of the hypoglossus among Ichthyopsiaa are not clearly understood, I shall confine myself to the development of the nerve in the higher forms. We shall consider in order, 1, the ganglia; 2, the motor roots; .3, the branches.

The development of the hypogloesns suggests that it arose by modification and fusion of at least four segmental nerves situated between the vagus and the first ce^^•ical nerve. The modifications consist in the disappearance of the ganglia and the conversion of the motor-fibres of the dorsal roots into the accessorius nerve, and in the disapi)earance of at least the anterior of the ventral roots. The nerve retains its primitive relations, since the lingual muscles it innervates are developed from the occipital myotomes.

1 . The Ganglia. — There are foundjn the occipital region of yoimg mammalian embryos three ganglia, which abort before they are fully dflferentiated. These ganglia have a marked resemblance to the true spinal ganglia. They are connected with a part of the neuron which belongs presumably not to the medulla oblongata, but to the spinal cord. If this is really the case the ganglia are true spinal ganglia, not cephalic. Chiarugi, 89.2, found that the ganglia are preceded by a continuous stretch of the neural crest, which appears as if a commissm-al link between the vagus and first cervical ganglia, e,g,y in Lacerta embryos of 2.7 mm. From this pseudocommissure there grow out in Lacerta at first two ganglia, which overlie and extend in front of respectively the third and fourth occipital myotomes, and there is perhaps a third ganglion, that is to say, one for the second occipital myotome; the three ganglia have only a fugitiv^e existence, and are no longer present in embryos of 5.5 mm. It may be well to recall that the first cervical ganglion also aborts in Sauropsida during early embrj-onic life, compare p. 030. Chiarugi, /.c, 331>, foimd the three rudimentary occipital ganglia in the chick embryo of the third day, corresponding to the second, third, and fourth occipital myotomes. In the rabbit only two ganglia are known in the occipital region ; these have been observed by Chiarugi, /.c, 430, in embr^^os of 0.5 mm. associated with the third and fourth occipital myotomes ; the posterior of the ganglia is the larger. In cow embryos of 8.7 mm., Froriep, 82. 1, 86. 1, 16, found one cx3ci pi tal ganglion in association with the last occipital myotome, there being three myotomes. We may assmne that there are earlier two ganglia and four segments in the cow embr\'o as in the rabbit, and that by the stage studied by Froriep the foremost segment and foremost ganglion have disapi)eared. In cow embryos of 12 mm., Froriep, 86.1, 24, found the ganglion of the fourth segment still present and its ventral end united with the hypoglossal motor roots of the same segment, but in embryos of 15 mm. the ganglion shows indications of abortion, L c, p. 33. In the human embryo. Fig. 300, the ganglion of the fourth occipital segment lias been observed by His (" Anat. menschl. embryonen," Heft III., sii, also 88.1, 401) in embryos of 13-14 mm. ; later it is found to 'have disappeared. His proposes to name the ganglion after its discoverer, Froriep^ s ganglion, Kazzander, 91.1, has directed attention to various cases in which a hypoglossal (Froriep's?) ganglion has been observed in man and other mammals in the adult stage, and reiK)rts a new case of its presence in a human adult.

The facts presenteil in the preceding paragraph render it probable that in all amniota there are at least three * ganglia present during ver}' early st^iges in the occipital region; that these ganglia belong to the second, third, and fourth segments of the region, and to the hypoglossal nerve, and that they successively disapj)ear, the last persisting for some time longer in mammalian than in sauropsidau embryos. I think that we may expect to obtain evidence that there is still another hypoglossal ganglion, namely, for the first segment.

  • P. Martin, UO.3, 230, affirms that he fluds in the cat five rudirnnitary hypoKlossal ganglia.

Although the occipital ganglia entirely disappear, the ganglionic cord, from which they arise, persists and serv^es as the anlage of the accessorius as stated in the preceding section.

No ganglionic sense organs connected with the hypoglossus have yet b^n recognized, but it is to me probable that the part of the lateral line near the vagus represents hypoglossal lateral organs. Suitable investigations on Ichthyopsida might result in confirming this suggestion.

Historical Note. — The last hypoglossal ganglion was discovered by Froriep in 1882, in ruminant embryos, and its history has since been further studied by him. His, in 1888, recognized its presence in the human embryo. Chiarugi, 89.2, 90.1, has studied the ganglia in reptiles, birds, and mammals, and our present knowledge rests to a large extent solely upon his observations. P. Martin, 90.3, has observed the ganglia in cat embryos.

2. Motor Roots. — The neuroblasts which give rise to the hypoglossus lie, in the human embryo, in the ventral part of the ventral zone of His, Fig. 374, and their fibres make their exit from the medulla not far from the Bodenplatte (His, 88.3, 3G1). The fibres are gathered into bundles. According to His, these bundles are quite numerous and are found even below the vagus ganglion. I consider it probable that His is mistaken in regard to this, and that the fibres leave the medulla in man only in the region behind the vagus — in other words, in the region of the four occipital segments, and in four segmentally arranged bundles. That there are three, and probably four, segmental motor roots in cow embryos has been shown by Froriep, 85.1, 10, but P. Martin records, 90.3, 230, that in cat embrj^os, representing younger stages than Froriep studied, there are five distinct roots (? of which one cervical). Chiarugi, 89.2, 90. 1, has observed five segmentally arranged roots in Lacerta, the first root lying in front of, the remaining four corresponding to, the four occipital segments ; four roots in Tropidonotus ; three roots in chicks toward the end of the third day, the first occipital segment having no root ; and finally two roots in rabbit embryos. Van Bemmelen, 89. 1, 243, describes in Lacerta five well -developed hypoglossal ventral roots, and has noticed fibres further forward toward the vagus, which suggest to him the possibility of yet more roots ; he further records that motor fibres are added from the first and a little later also from the second cervical nerve.

3. Branches. — It will be remembered that the posterior branchial arches are invaginated, the invagination constituting the simis certncalis. The hypoglossal nerve in a himaan embryo of the fifth week. Fig. 363, was observed by W. His, 88.3, 380, to pass around this sinus, going behind and below it and there curving forward into the tongue; as shown in the figure, the nerve crosses the vagus below the ganglion nodosum, and after crossing gives oflE a branch, ramus descetidens, which nms along the lateral side of the jugular vein parallel to the vagus trunk. The mechanical cause of the formation of this branch, I do not know. Chiarugi, 90.1, 432, has observed that the distribution of the nerve is essentially the same in rabbit embryos as in human.

In Lacerta, Van Bemmelen, 89.1, finds that the course of the nerve, as it curves around to enter the tongue, is closely parallel to the united prolongation of the five myotomes (four occipital and one cervical) which grow like a single cord (Froriep's Schulterzungenstrang) into the tongue to produce the lingual muscles. Chiarugi, 90. 1, 321, states that in lizard embryos the nerve-trunk runs outside the jugular vein, from which it is separated by the intervention of the vagus and of the carotid artery, and accompanies a branch of the jugular, which runs to the mandible and is probably the sub-maxillary vein.

Spinal Cord. — The diflEerentiation of the cord and brain is effected by the development of the cerebral vesicles. The histogenesis of the

cord has been described in the sections on ><^7J^ /^l^xT^N ^py the neuroglia and the nerve-fibres. The /!^-:-^k-^-!^'^^^'^ following paragraphs refer chiefly to the

cord without regard to the peculiarities offered by the lower end of cord, Fig. 375, which we find the t}T)ical developmental features very imperfectly foUowecl. This presumably, to the partially abortive history of the caudal end of the neuron in mammalia. The following descripFI0.87.V -Lower emi of the Spinal tions are based in large part on His'

Conl of a Human Embryo of three Months. Epy, Ependymal layer; memoir, OD.«6.

General GROWTH.- The following account is based upon that of Kolliker ("Qrundriss," 2te Aufl., 200). The medullary groove is found completely closet! in the region of the spinal cord in a chick embryo with thirteen primitive segments, and in slightly more advant^ad human embryos. But the posterior end remains for a while aa a solid mass, which terminates by fusion with the ectoderm. When the primitive segments are all formed, the end of the cord separates from the ectoderm. At this stage the cord extends as far as the segments. In human embryos the cord equals the vertebral column in length up to the end of the third month. After the fourth month the vertebral column outgrows the spinal cord, which, although it absolutely lengthens, becomes relatively shorter, so that the distance from its end to the end of the spinal canal increases. This apparent ascent of the cord (ascensus medullar spinalis) might be more properly described as a descent of the vertebrae. A secondary result of the changed position is that the nerves running out from the lower end of the cord, since their exits l)etween the vertebrae are below the end of the cord, are forced to take a more and more longitudinal course within the spinal canal. There results a series of nerveroots, which after the fourth month elongate as the vertebrae descend, and thus gradually produce the so-called cauda equina. The filum tenninalis is developed, according to Kolliker, from the pia mater, and is therefore, i)roperry speaking, not a nervous structure. The upper ])art of the filum, however, even in the adult contains a prolongation of the s))inal cord with its central canal; compare Toumeux et Hermann, 87.3.

The cervical and lumlnir enlargement of the spinal cord are indicated in the human embryo at two months and are well developed at three months.

2. Central Canal. — The central canal has at first the form in sections which is shown in Figs. 159, 160, and 161, being flattened from side and elongated dorsal-ventrally, but is often more or less irregular in shape. I have observed that in birds and mammals there is a tendency for the walls of the canal to come temporarily into close contact along two longitudinal lines, so that the canal appears at the first glance to be divided into three channels. This condition may be well seen in the rabbit, and it is probably of wide, possibly of constant, occurrence. As to its significance, I have no clew. The contact is soon lost, and the canal becomes freely open again throughout its extent.

There now occurs a change of shape, of, p. 007, by which the canal cuts into the thick medullary wall on each side, dividing it into the upper and small dorsal zone of His, and the lower and larger ventral zone of His, see W. His, 86.2, p. 497, Fig. 6. This change occurs in the human embryo toward the end of the fourth week and attains its maximum about the beginning of the sixth week. It is precisely during this period that the medullary nerves grow out from, and the ganglionic nerves grow into, the spinal cord; the former arising from the ventral zone, the latter entering the dorsal zone. The dorsal plate is curved inward, making a median ridge internally and a median groove externally ; on either side of the latter there is a projecting fold, where the deck-plate curves over (the fold is the anlage of Goirs cord.

About the eighth week the canal begins to contract (compare His, 86.2, Figs. 0-9) between the dorsal zones until the walls first meet and then unite. Thus in a foetus of the tenth week, Fig. 376, the union has already taken place except at the very dorsalmost part of the canal, where a small remnant of the original cavity persists.; whether this is always the case, I do not know. In older stages all traces of the canal (both its cavity and its epithelium) have disappeared, not only between the dorsal zones of His, but also between the upper part of the ventral zones. In Fig. 376, the boundary between the dorsal and ventral zones is marked by the insertion of the dorsal nerve-root. The lower part of the central canal remains open, and presents in section certain definite curves of outline, which deserve closer study. The open part of the canal is elongated dorsalventrally, but toward the close of foetal life it becomes more rounded in form, and in the adult is elongated transversely. In the caudal end of the human cord the cavity is large, Fig. 375, and does not go through the same changes of shape as in the rest of the cord.

It seems to me that the dorsal part of the central canal is obliterated by the union of its walls and the subsequent atrophy of its so-called epithelium, although the exact steps of the atrophy are unknown. In the adult the line of the central canal on the dorsal side is represented by the posterior fissure^ which is merely a thin partition of vascularized tissue and not a true fissure. It seems probable that the tissue as claimed by Barnes, 84. 1, is really derived from the cells lining the central canal, and is, therefore, to be classed morphologically with neuroglia. Waldeyer, 76. 1, and others speak of a contraction of the canal, and of its being pushed in by the ingrowth of the posterior columns. This view is incorrect, for, as shown in Fig. 37fi, the central canal exists between the posterior eohmiiiti, even after the columns of (Joll iin<l Biirdaoh ciin Iw? recognize*!, in what are essentially their [lermanent positions. A, Kobinsun, 91.1, yo, calls attention to tho facrt that the cord widens at an early stage in rodents, so that in section it appears nearly round insteatl t>f oval ; this change causes a slight diminution in the dorsal-vontral diameter of the central canal. I cannot ri^jurd this diminution as a step toward the obliteration of the canal.

The niiferiitr ti.is'ire iKjgins to develop during the early part of the eighth week, and arises by the growth of the cord, which takt^ place, an indicated in Fig. :tT7, bo as to prodiK-e two bulging ridges on the ventral side of the conl. The median s|>dce between the ridgtfs is the future anterior fissure; it is occujned by fibr;>us connective tissue enveloping the <'ord; it is, therefore, a true fissure, for aoroa* It there is no (Connection l>etwot>n the ner^'ous tissue of the two sides. Iudee<l, part of the original surface of the cord bounds the fissure on either side, and therefore we may correctly describe the ti.-wue in the fissure as |mrt of the enveloi»e (pia mater) of the conl. As the embryo lulvances the ridges grow and the fissure deepens; the growth nt the ridges is hirgi^ly due to the expansion of the gray matter to form the anterior horns.

There is no atro])hy of the ventr.d iM>rtion of the canal as Liiwe, 80.1, 114, asserted, but the central canal of the adult rejiresents the ventral portion of the primitive canal.

3. Growth op the Mantle Layer. — The mantle layer in man (HIb) and rodents (A. Bobinaon, 91.1) first appears in the region of the ventral zones of His, forming in sections a triangular inaas on each side between the inner layer and the Bandschleier ; it gradually tliickens, and at the same time its development progresses dorsalward and encroaches also upon the inner layer. There is thus a stage (in rats when the cartilaginous bodies of the vertebrae arise) in which the inner layer is very much reduced in the ventral col umns. and gradually increases in thickness dorsalward becoming in the dorsal zone thicker than the mantle layer which however soon grows at the expense of the inner layer which is ultimately reduced to the lining or so-called epithelium of the central canaj The mantle layer is easily recognized by the large size of its elon gated nuclei, and by the fact that some of the nuclei are elongated dorsal-ventrally and others radially; in the inner lajer the nuclei are smaller and all point radially.

4. Dorsal Zone of His. — The origin of this dn ision of the cord has already been described, p. 60" In a human embryo of 12.5 mm,, W. His, 86.2,497, found the dorsal zone to begin with a broad arch from the deck-plate, to-form a marked projection into the central canal, and to have upon its outer sur face a rounded projection, ov.b, which His calls the Ofal bundle (ovales Bundel); the projection is a product of the Randschleier and contains the ganglionic fibres, which have entered the medullary wall and run lengthwise within it; the oval bundle at this stage extends about half the distance from the sensory ro<rt, which enters the ventral boi-der of the bundle, to the median dorsiil line; tho oval bundle is the anlage of the greater part of the posterior column of the adult. The oval bundle now steadily eu larges and creeps dorsalward until it reaches the arch formed by the passage of the dorsal zone into the deck-plate, d.pl. The arch gives rise to G oil's cords. The two cords of &t 11

become closely united with one another bj fo t--t™iisv n«s«:icm the obliteration of the central canal between ^ ^^ DoiwtiR^r^'oi't huthem. Fig. llTd. The oval bundle meanwhile mau Emhiro of nix wlwIu creeps still further and makes its way between the cords of GoU and the Kray matter until it I?'"^■ ^A- f"""™' ™t: ""^ meets its fellow from the opposite side below b,floor-i]iiite: r.R, wnirairoot, the ronls of Goll; thus arise the cords of «cii»m8. Btinhuh (Biirdaclische Keilstriinge), Fig. 378, b. At this stage — embryo of the tenth week — the dorsal zone of His is no longer distinctly marked off from the ventral zone except by the position of the sensory root. The inner and mantle layers have become the gray matter and they are completely covered by the expansion of the oval luindle, that is to say, by a layer produced from the primitive Randsi'hleier of His, p. <UG, and containing chiefly longitudinal ganglionic tibi'es. The layer developed from the oval bundle may be subdivided into two parts : the medial Burdach's cords and the lateral portion of the posterior columns. Outside of and above Burdach's cords are Qoll's cords, which are developed from the arch by which the deckplate originally passed into the dorsal zone of His in the embryo. The fibres in Qoll's cords are developed rather late.

The neuroblasts of the mantle layer of the dorsal zone send their nerve-fibres ventralward ; the fibres constitute the formatio arcuata. As indicated in Fig. 378, only part of the posterior horn of the adult probably is developed from the dorsal zone. The inner and mantle layer give rise to the gray matter, which increases rapidly after the middle of the second month, owing partlj^ to the multiplication of its cells, partly to the penetration of blood-vessels, and the accompanying loosening of the tissue; this loosening {Auflockerung) progresses from the head backward. At three months the posterior horn is still broad and short in section, but it gp-adually becomes long and narrow.

Sabstantia Oelatiyiosa Rolandi. — This tissue is probably the neuroglia plus numerous nerve-cells of the tip of the anteriorlhom, develoi)ed in the mantle layer. As the cells of the embryonic mantle layer are apparently all neuroblasts. His, 86.2, 508, assumed that there are cells, which migrate into the laj^er to form the gelatinous substance. The origin of these cells he did not observe. Gierke, 86.1, 144, pointed out that most of the elements are very small nerv^e-cells. H. K. Coming, 88.1, found that in the dorsal part of the inner layer of the dorsal zone of His, the development is greatly retarded, and he interprets the substance of Roland as a tissue persisting in a somewhat embryonic condition, and not having the same differentiation of its cellular elements that we find in other parts of the cord.

6. Ventral Zones op His. — The ventral zones are larger and more complex than the dorsal zones. At six weeks. Fig. 377, they comprise at least three- fourths of the cord; each zone consists of an upper connecting piece. His' Schalfstiick, and a wider lower segment; the width of the latter is due to the great thickening of the inner and mantle layers ; the Randschleier or anlage of the white substance extends completely over the outer surface of the zone as a layer or envelope, which varies but little in thickness. Owing to the projection of the oval bundle and of the lower segment, the Schaltstiick is marked externally as a wide gr<H)ve ; His designates the angle of this groove next the oval bundle as the Randfurche^ the angle next the lower segment as the Cylinder fnrche (His, 86.2, 498). As development progresses, the Schaltstiick relatively diminishes, while the lower segment increases, so that the groove just described is gradually obliterated ; nevertheless it can long be recognized. The gray matter of the Schaltstiick is to be considered as the anlage of the cervix cornu. For a considerable period the Randschleier or anlage of the white substance of the connecting piece remains thin, compare Fig. 377, but toward the end of the second month it begins to thicken until the groove is obliterated, but the thickened portion still retains, according to His, a certain individuality, and may be identified as the anlage of the lateral pyramidal cord (Hinterseitenstrang) .

The lower segment of the ventral zone is the anlage of the anterior horn, the anterior column, and a large part of the lateral column. It is characterized by its early and rapid growth, at first chiefly of the gray matter, later of the white matter (Randschleier) also, compare Figs. 377 and 37(i. The exit of the ventral root divides the white substance into the anlage of the lateral column and the aul^e of the anterior column or cord.

The growth of the gray matter depends chiefly on the multiplication of the germinating cells and the growth of the neuroblasts in the mantle layer. As the neuroblasts are most numerous in the ventral part, there results the precocious enlargement of the lower segment as compared with that of the rest of the cord. The neuroblasts of the lower segment send out their nerve-fibres mostly in small bundles. The nerve-fibres of the formatio arcuata coming from the neuroblasts of the dorsal zone also enter the lower s^ment, and as some of these fibres are developed later their paths cross those of earlier fibres, owing to the changed relative positions. Not all the neuroblasts send their fibres directly int« the ventral roots ; on the contrary, some of them are found placed longitudinally in the lower segment. Thus the gray matter of the anterior horn becomes verj- complicated at an early stage. The growth of the nerve-cells of the ventral column has already been described, p. C2+.

5. Ghav and White Mattes of the Fctiti's. — Concemiug the development of the cord during the foetal period (middle of the third month until birth) we know very little.

As reganls the outline of the gray matter we find that the anterior and posterior boms at three months are of about the same size and shape, and have a very bn)ad connection with one another, compare W. His, 86.2, &U5. At five months the cord has grown very much, Fig. :t78, and the central canal having remained stationary is relatively much smaller. In sections frtmi an embryo of this age, ), I observe a perijiheral den- / ser hiyer surrounding a cen- i tral l(KWier area, which is i ^ divided, Fig. 37f*, into t^?o \ parts by the anterior fissure and Burdach's cords; if this lighter area correspoods to the gniy matter, then at this stage the anterior and posterior horns are fuse<l, and the horns are not finally < _ _ _ _ shaped out until later. ^i^^^^autJi^Tom^^' '"'*°' *■

As regards the white matter: some scattered observations are recorded by KoUiker in the second edition of his "Entwickelungsgeschichte," and there are a good manj^ observations by various authors on the appearance of the medullary sheaths of the nerve-fibres, which are at first naked. Flechsig, 76.1, drew attention to the fact that the sheaths appeared at diflEerent periods for different tracts, and he sought by extended observations to trace the course of the fibres within the cord of the embryo by following the course of the tracts with sheaths as distinguished from those without. Flechsig's observations have been extended by Bechterew, M. von Lenhossek, 89. 1, and several others. References to the various authorities are given by von Lenhossek. A proper collation of the results obtained has yet to be made. Lenhossek finds that the medullary sheaths appear on the fibres of the posterior roots and on the fibres of Burdach's column about the same time, but that the fibres of Goll's column are not medulbited until a few days later. He has discovered, further, that at a certain i)eriod the fibres of the lower part of Gk)irs column are medullated, while in the lower cervical region only those fibres which form the ventral part of the column have received white sheaths, and that in the upper cervical region none of the fibres of this column are medullateil. He concludes, therefore, that the medullation of GoU's column is centripetal in direction, and that the fibres which form it have a long course, but he thinks that there is no anatomical proof that any of the fibres of the p<isterior r(K)ts i>ass directly into the posterio-intemal columns. It is now generally allowed that the deposition of white matter upon the axis cylintlers takes place first in the neighlx)rho(xl of the cells from which they spring, and proceeds thence toward the termination of the axial process. This Ixnng the case, it follows that as the columns of Burdach consist principally of posterior nx)t fibres which have just entered the cord, thev will become medullated verv shortlv after the fibres of the posterior nK)ts themselves, while the column of Goll, which is formeil of fibres of the iK)sterior ro<^is which have enteretl the cord at a considerably lower level, will become medullated at a later period.

Cajal, 90.1, discovered that the fibres of the white substance of the spinal cord give off fine branches nearly at right angles, which penetrate the gray matter ; these branches he names the collaterals, I.e., p. 88, and they have since been found in the adult by Kolliker, 90.2. They appear very early in the embryo, and after the medullary sheaths api)ear they are seen to go off from the main fibre at the nodes of Ranvier.

0. Blood- Vessels. — The first appearance of the blood-vessels in the cord has l)een studicnl by W. His, 05.1, 15, 86.2, 41>3. The spinal cord lies in a canal, the walls of which are formed by embryonic connective tissue and repesent the anlage of the pia mater. During the embryonic ix»ri(xl of the human embryo the conl is in conta(*t with the wall of the spinal canal only along the mediaii dorsal line. The walls of the canal contain capillaries which are developed during the third week in the region of the head from the aorta, in the rump from the intervertebral arteries.

These capillaries form anastomoses which produce four longitudinal vess<»ls, two near the ventral minlian line, one close below each sen.s<>ry root. From these four vessels vascular buds penetrate the spinal cord, the branches from the ventral arteries preceding the others in their development; c/. His, 86.2, Taf. I., Fig. 2. The two ventral arteries become included in the anterior fissure ; during the sixth or seventh week they imite into a single median vessel at the bottom of the fissure, i. e., near the central canal; this vessel is the arteria sulciy and is the principal source of supply for the gray matter. From the two vessels next the sensory roots branches enter to the region of the future posterior horn.

The vascular buds consist of elongated vasoformative cells, which force their way through the neuroglia network ; by the time the buds have become vessels, there are considerable perivascular spaces, as if the neuroglia had contracted away from the blood-vessel.

After the vessels have penetrated it, the cord develops more rapidly, as if better nourished (His, 86.2, 496).

Medulla Oblongata. — The term is now restricted to the portion of the brain extending from the spinal cord to the Varolian bend. Our knowledge of its development is derived mainly from the superb researches of His, whose predecessors had given us little more than generalized descriptions of the external form. This section is, therefore, based on His' pai)er, 90.2, which, however, deals with the development of the region of the calamus scriptorius only, to which region accordingly the following account mainly refers. The presence of the zones of His and the appearance of the Rautenlippe have already been described, p. 608. The division of the medullary walls into four zones (p. 606) dominates the structure of the medulla oblongata throughout life, and the division of the ventral and dorsal zones of His can be traced in the floor of the fourth ventricle of the adult. The secondary complications of the medulla are largely owing to the modifications due to the transformation of the Rautenlippe, and in lesser degree to the fact that the anterior fissure of the spinal cord is replaced by a thickening of the Bodenplatte, which allows the nerve-fibres to cross from one side to the other directly. The following more detailed history may be more easily understood if these general characteristics of the medulla are born in mind, than would be otherwise possible.

As His points out, 90.2, 6ij^ the ailult medulla contains in every transverse section parts which have been present from the start and others which have been added later; the former as a rule lie nearer the ventricle; the added parts lie nearer the outside, but a portion of them mingle with the older parts, it being especially the fibres which traverse the medulla as they develop in manifold directions. Nevertheless, in a general way, we may affirm that the further from the ventricle in the adult, the later was the development in the embryo. The first cells to be differentiated are the spongioblasts, which constitute the ependyma in the adult. Next arise the neuroblasts which migrate into the mantle layer ; the earliest nerve-fibres alone give rise directly to iierve-roijts ; the later ones take their paths within the medulla. Third arise the fibres of the formatio arcuata, which lies in the outer part of the mantle layer and sends its fibres from side tu side, and the homologue (tractus solitarius) of the oval bundle of spinal cord sensory fibres. Fourth, the i)arts already formed are covered in by the Rautenlippe and the stream of neuroblasts which it sends toward the median line. Outside of all these finally ensues a development of transverse and lon^tudinal fibres, the latter iiioliiding the f uniciil;is restifonnis and the tractus intermedins of His. Zones op His in the Adult. — As will be shown below, the tractus solitarius is a bundle of fibres running longitudinally and homologous with the " oval bundle" of sensorj' fibres in the spinal cord, and it indicates pennanentlj' the lower boundary of the dorsal zone of His. In the embryo the two columns primitively meet at a decided angle, and this an{j*le is marked by a gnxive in the wall ot the central canal, or, as we should say in referring to the adult, in , the floor of the fourth ventricle. There is always a median groove, which extends from the opening of the central canal to the aijiieductus Sylviffi, and marks the limit between the two ventral zones i)f His, although they partially concresce during embryonic life; on each side of the groove is the ventral zone, the surface of which projects slightly and is known in descriptive anatomy as the eminentia teres. The groove between the dorsal and ventral zone is very shallow and partially obliterated in the adult; it persists, however, in three depressions, namely, the fovea posterior of the ala cinerea, the fovea anterior, and the sharp depression between the eminentia teres and the peduncles of the cerebellum; opposite Schwalbe's tuberculum acusticum the groove is almost obliterateil. By this division we see that the alie cinerefe and corpora restiformia are parts of the dorsal zone of His.

Dorsal Zone op His. — This part of the medulla (FUigeleisie of W. His) undergoes fundamental modifications owing to the development of the I^uteiilippe, p. COS. It also changes its position with relation to the ventral zone in consequence of the long continued expansion of the deck-plate, or, in other wonls, in consequence of the so-called opening of the medulla. When first fully differentiated the ventral zones, as seen in cross sections, Figs. 348 and ai9, ascend obliijuely from the median line, but the dc>r8al zones aj>|>ear nearly [tarallel with the median t^ane. In the next stage. Fig. 37i), the ventral zones diverge so much from one another that they both lie nearly in the same horizontal plane; itt the Siime time (beginning of the fifth week) the dorsal zone liends over throughout its length to form the Rautenlippe, RL; the lower limit '^ oi the dorsal zone is marked if"KM»gi^^';«fJr"' wd^iiw"' ^' '■>' *^ position of the tractus solitarius, T.H. Within a few days the Rautenlippe unites with the main fol<l of the zime and continues to grow toward the median ventral line passing outside of the tractus solitarius, which thus l)ec(>meR buried, and, instead of lying su|)erficially, is thereafter deep I>elow the outer surface. The modified dorsal zone formed by the intion of the two folds is termed by Hie, 90.8, 33, Fliigelwulst, With the beginning of the sixth week the Fliigelvralst bends over outward so that its inner surface faces dorsalward and its outer surface ventralward. The dorsal and ventral zones are now nearly in the same plane, and the groove on the inner surface between the zones is nearly obhterated, Fig, 380. There next arises the secondary Rautenlippe of His, Fig. 380, RL, which is apparently not a nervous structure, but merely a transition from the dorsal zone to the ependyma or expanded deck-plate; it must not be confounded with the true or primary Rautenlippe. If the size of the parts developed from the dorsal zone be compared with that of the ventral column in Fig. 380— the tractus solitarius, T, marks the boimdarj- — it will be evident that scarcely a fifth of the adult medulla is developed from the dorsal zone.

The dorsal zone becomes the corpus rentiforme of the adult, including the tracts of longitudinal fibres associated with it; these B-rGth^iractus solitarius. Fig. 380, 7", the fvniculufi reatiformis, F.r, Ritd probably the ascending trigeminal tract, a. Tr. This last probably only, because at the time and place it appears the exact boundary between the two zones cannot be determined. Further toward the spinal cord the restiform body merges into the clava, which passes into the fasciculus gracilis, which in its turn is prolonged into the columns of GoU in the spinal cord. During the fifth month the clava occupies nearly a transverse direction (Kolliker, " Entwickelungsges. , " 2te Aufl., 54!i). The detailed history of the restiform bod\' has still to be traced. The trac-tus solitarius arises very early, owing to the penetration of fibres from the cerebral ganglia into the medulla ; these fibres, like those of the spinal nerves, take a longitudinal course and appear in sections as a compact bundle situate(l in the Ptandschleier of the dorsal zone of His, as has been already described in detail in the account of the cephalic nerves. As stated above, the Rautenlippe during the fifth week buries the solitary tract. Its development shows that it is homok^^ous with the columns of Burdach of the spinal cord, although in the medulla it loses its original superficial position, which it retains in the cord. After the Rautenlippe has united with tlio inner fold of the dorsal zone a layer of neuroglia is developed over the new external surface of the zone; this layer is continuoufl with the Randschleier of the ventral zone, compare Fig. 300; in it appear two bundles of longitudinal fibres, Fr and a.Tvy also transverse or so-called arcuate fibres. The most lateral of these bundles, Fi\ is the funiculus rostiformis; it is scarcely noticeable until the secondary Rautenlippe is formed ; the fibrt»s are coarse ; the ventro-medial portion is penetrated b} arcuate fibres ; the fibres of the bundle first appear near the cord, later higher up; most of its fibres are arcuate ones, which bend and tiike a longitudinal course; these arcuate fibres of the funiculus probably arise from the cells of the olivary body of the ventral zone of the opposite side (His, 90.2, 57). The ventro-medial bundle, Fig. 38(», a.Tr, in the outer neuroglia layer, is the tract iia intermediuH of His, a term which he employs l)ecause the bundle includes not only ascending sensory fibres, but probably alst) fibres running from the cerebellmn to the spinal cord ; in descriptive anatomy the bundle is usually known as the iiscending trigeminal tract or root; the bundle is oval in section and consists of coarse longitudiufd fibres, and is crossed by arcuate or transverse fibres ; its development begins anteriorly and progresses tailward (His, 90.2, 50).

The neuroblasts of the dorsal zone have a remarkable history, according to W. His, 88.3, 90.2, 35-44. They arise early and rapidly become abundant (see p. Oil), and their production continues until the end of the stjcond month, when it cejises altogether. His, 90.2, 47. The neuroblasts develop during the fourth week, that is to say, before the formation of the Rautenlippe begins, and produce the arcuate fibres and the primitive cerebral motor roots, as above described for the single nerves. These neuroblasts, therefore, resemble in their development those of the spinal cord. The neuroblasts which arise later have in large part a different history, accomplishing a peculiar migration, which has no parallel in the spinal cord. In the medulla, as in the cord, the pnxluction of neuroblasts begins on the ventral side and progresses for a week or more dorsal ward, and consequently, when the germinating cells or parent cells of the neuroblasts have disapi^ared in the ventral zone, they are still present in the dorsal zone and continue to change into young nerve-cells while the R^iutenlipjie is bending over. The concrescence of the Lippe witli the main fold of the dorsal zone oj^ens the way for the neuroblasts of the Rautenlipi)e to migrate* from their site of origin past the outside of the tractus solitarius toward the ventral zone of His, which they enter, and accumulating in its lower part, Fig. 3SI, there contribute, together with other neuroblasts which come from the dorsal zone by migrating in ]Miths inside the tractus solitarius, to the development of the olivary- bodies. The cause of the migration of the neurblasts is entirely unkninm, but their wandering from the RautenlipiK? is one of the most distinctive characteristics of the medulla obl<^ngata.

Ventral Zone of His. — This zone is at first alx)ut the same as the dorsad in size, Fig. 37*J, but it rapidly outgnnvs the dorsal zone and constitutes more than tliret^-fonrths of the adult medulla. Its development has an obvious resemblance to that of the ventral zone in the spinal conl, for there is a similar nipid ex|>Jinsion and consequent bulging inward and outward, and the exi)ansiou

is due chiefly to the mantle laj'er, the Kandechleier remaining thin. There are ttiree chief factors which cauBe the development to differ from that in the spinal cord. These are, 1, the bending of the zones outward and downward until they come to lie in nearly the same horizontal plane, compare Fig. ^7'J with Fig. 380 ; 'i, the absence of the anterior fissure, which is obliterated by the growth of the Bodenplatto to constitute the raphe ; '6, the peculiar arrangement wliich is gradually assumed by the gray matter, developed out of the mantle laj'er; 4, the ner^-e-fibres in the Kandschleier also take different courses from that which they take in the white matter of the Bpinal cord. These four sets of features are considered in the four following paragraphs.

1. The bending of the ventral zones, UVe that of the dorsal, is part of the process of the so-called opening of the medulla correlated with the expansion of the deck-plate. The general character of the movement has been already described, p. {iO!t. We have merely to add that, while it is going on, the inner surface of the zone, which constitutes the lai^r part of the floor of the fourth ventricle, becomes protul)erant and bulges inward, forming a wide, rounded, longitudinal ridge, Fig. 381 ; the two ridges are separated from one another by a narrow, deep median fiasure or groove, which in later stages opens somewhat, so as to appear V-shaped in cross section, and persists throughout life in that form. As the groove deepens but little, if at all, after the second month, while the medulla continues to enlarge, it follows that the groove becomes not absolutely, aa sometimes stated, but relatively smaller. His speaks of its opposite walls uniting and the groove thus diminishing, but he gives no direct evidence of such concrescence, and his figures show no diminution of size in the gnK)ve during later stages. In Fig. 381, another effect of the interior bulging is sbo^vn, namely, that that part of the surface of the ventral zone is brought into nearly the same plane as the inner surface of the dorsal zone, and as the groove Ijetween the two zones is nearly obliterated, the floor of the medullary cavity (fourth ventricle) is rendered comparatively even.

2. The raphe arises by a thickening of the Bodenplatte and is primitively a partition of neurt^lia, which is sub8e<|uently penetrated by flbres crossing from side to side. In the spinal cord the Bodenplatte remains thin though it gives rise to neuroglia, and by the passage through it of nerve-fibres forms the anterior white commissure. We must, therefore, homologize the raphe with this commissure.

As the ventral zones thicken during the second month and project more and more ventralward, the growth of the Bodenplatte in the medulla oblongata obliterates the fissure almost, but not quite, completely, which would otherwise be formed between them, as in the cord. The growth of the Platte depends on the elongation of its cells (spongioblasts, for it contains few or no neuroblasts), which is accompanied by a movement outward of some of its nucleated cellfibres, which are at first all situated close to the central canal. By the end of the first month fibres cross the septum, and thereafter the number of fibres crossing it steadily increases. It allows no neuroblasts to pass (His, 90.2, 27, 55).

3. The gray matter or mantle layer increases very rapidh and is the principal factor in the enlargement of the ventral zone. Its development involves, as elsewhere in the neuron, the gradual reduction of the inner layer until only the ependjTna remains. The grajmatter is, of course, homologous w^itli the tmterior horn of the spinal cord; but whereiis in the spinal cord the nerve-cells and nervefibres are irregularly arranged, in the medulla they produce a highly characteristic pattern by their distribution. The greater part of the gray matter in the ventral zone of the medulla is converted into the formatio reticularis. His, 90.2, 51. The formatio reticularis has from the very start a more or less distinctly four-sided outline, as seen in cross sections ; it is marked out by the bundles of nervefibres crossing one another at right angles. One side faces the fourth ventricle. Fig. 381; one faces the raphe; the third faces the outer wall of the medulla, and the fourth, which is irregular and somewhat undefine<l, faces the dorsal zone. The reticulate appearance of this area is due to the crossing of the fibres at right angles to one another. The fibres are first radial, second arcuate or transverse nmning toward or from the raphe, and third longitudinal; the last set of fibres are develo|)ed later than the first two. The fibres are united in bundles, which grow in size b\' the addition of fibres which join them as development progresses; the fibres are accompanied by a limited number of neurc>blasts migrating along the bundles. The formatio reticularis is clearly mappeil out by the end of the fourth week, ajid its development commences as soon as the nerve-fibres begin to grow out from the neuroblasts, for the fibres at once follow their definite courses, one set taking radial paths, another set taking transverse courses. A similar arrangement is found in the mantle layer of the spinal cord, but is obscured by the further development, instead of being preserved and emphasized as in the medulla.

In embryos of six weeks and older the formatio reticularis is entirelv surrounded by a crowd of neuroblasts. Of these the accumulation on the inner side, or toward the fourth ventricle, is the oldest and consists of neuroblasts developed in toco; it is very distinct at the l)eginning of the fifth week. The neuroblasts on the lateral aide are, of course, those which belong to the dorsal zones of His. Those on the medial and outer side, on the contrary, are immigrant cells, which have travelled to their location after the imion of the Rautenlippe with the main wall. The stream of cells passes, as we have seen, on both sides of the tract us solitarius. Fig. 381 ; that outside the tractus comes from the Rautenlippe, and is at first (fifth week) small, but later increases very much. The stream progresses arouudthe edge, or, better expressed, over the surface of the formatio until its outer and medial surfaces are covered by scattered neuroblasts, forming a continuoiis sheet (Ordnzplatte of His, 90.2, 42), which the subsequent development transforms into the cellular layer of the olivary body, compare Fig. 381. The olivary cell band has at first no very definite boundary ; the cells are here and there more crowded than elsewhere (His, 90.2, 62); the fibres which spring from them gather into bundles and run toward the raphe ; by the end of the third month the olivary band has become folded and appears to contain all the cells it is to receive. The band gives rise ultimately to both the upper and lower olivary bodies — in the region of the hypoglossus to the accessory olivary body (Neubenolwe) and in the region of the pons to the zackiger Briickenkern of His. The layer of neuroblasts between the formatio reticularis and the ependyma is the anlage of the sub-ependjTnal motor nuclei, His, /.c, p. 50. It may prove an assistance in following the description of the medullary structure to point out that in a rough way there are four layers distinguishable: 1, externally is the layer of white matter developed from the Randschleier, and which may be followed into the dorsal zone, see Fig. 381; 2, intemall}'- is the sub-ependymal laj^er of neuroblasts, which is continued laterally into the gray matter of the dorsal zone (corpus restiforme) ; 3, inside the external fibrous layer is the sheet of olivary neuroblasts, which merge into the lateral gray matter of the dorsal zone ; 4, the layer of the formatio reticularis between the sub-ependymal layer and the olivary bodj' ; this layer may be considered as continued laterally by the tractus solitarius, but topographically only, for the formatio reticularis arises from the gray matter, the tractus, as we have seen, from the primitive Randschleier, so that one cannot be the morphological continuation of the other.

4. The Randschleier includes the homologues of the anterior and lateral cords of white substance in the spinal cord, being divided by the exit of the ventral roots (hypoglossus and abducens) into two regions. Fig. 381, one medial region adjoining the raphe, the other ventral, situated at the exposed outer ventral surface; the former corresponds to the anterior, the latter to the lateral columns, and the latter spreads, as we have seen, over the dorsal zone after the concrescence of the Rautenlippe. The two regions meet, of course, along the line of the ventrm roots, forming a rounded angle with one another (His, 90.2, 54). The medial region, as soon as the nerve-fibres begin to develop, acquires both longitudinal fibres and transverse fibres, the latter running to the raphe or thickened Bodenplatte ; as development progresses the number of fibres increases and they group themselves into bundles : the primitive longitudinal fibres, like those of the spinal cord, are derived from the gray matter of the opposite side ; this primitive longitudinal bundle persists throughout life; it is the h interer Langsbiindel o{ FlechQig. During the second month the medial region grows rapidly, expanding at the same rate as the raphe, but the primitive longitudinal bundle is kept confined near the ventricle so that below it is a layer of neuroglia between the f ormatio reticularis and the raphe ; this layer is crossed by bundles of arcuate fibres, which enter through the raphe from the opposite side and most of which join the formatio reticularis; during the fourth month longitudinal fibres constituting the so-called pyramids are developed in this ventral part of the medial region. The very late development of the pyramids was discovered by Flechsig, 76.1, 132, 142.

The ventral region of the Randschleier, extending from the exit of the ventral roots, Fig. 381, XII, to the dorsal zone, is, of course, homologous with the region of the lateral columns of the spinal cord; it is identictal with the medial part of the ireisse Randzone of Flechsig, 76.1. When the olivarj^ band of neuroblasts becomes folded, some of the folds cut so deep into the white layer that it is almost obliterated at those points. About the middle of the second month fibres from the raphe enter the layer and ultimately pass on to form the funiculus restiformis; the nimiber of these fibres, though small at first, is large by the end of the second month. There appear during the second month fine longitudinal fibres in the layer.

Pons Varolii. — The pons is developed out of the floor of the third primitive vesicle of the braiu, in front of the Varolian bend. Concerning its history we possess no detailed information. Kolliker (Grundriss, 2te Aufl., 250) states that the characteristic transverse fibres appear during the third month as a narrow, thin band, and that the pons grows as the lobes of the cerebellum become larger and more distinct. He notes further as characteristic of the foetal brain that the corpus restiforme seems to merge in part with the lateral part of the pons, and apparently some of the fibres of the corpus bend toward the median ventral line and enter the pons. The growth of the pons is rapid. In embryos of the fourth month and older, the pons can be at once recognized as a commissure between the two sides of the cerebellum.

As to the fate of neuroblasts present in the pons, and as to the origin of the nerve-fibres of the pons, nothing is, as yet, known. It will probably be found that the development of the pons is similar to that of the ventral zones of His in the medulla oblongata.


The morphologically primitive relations and position of the cerebellum are well shown in the frog's brain. Fig. 303. It is a thickening of the brain walls extending across the median dorsiil line; its formation, therefore, involves the thickening of the deckplate ; the cerebellum is situated between the medulla oblongata and the isthmus or constricted portion of the medullary tube connecting the hind- and mid-brains. His, 90.2, 24, states incidentally that it is developed in man at least from the dorsal zone of His (Fliigelplntte) ; unfortunately his investigations on the cerebellum are still unpublished.

The following account of the development of the external form of the cerebellum is based on Mihalkovics, 77.1, 53-57. In its first stage the cerebellum is merely a lamella across the dorsal side of the hind-brain; its posterior limit is marked by the point where the expansion of the deck-plate begins ; toward the mid-brain the lamella merges into the isthmus without any demarcation (human embryo of the fourth week), compare Fig. 343. At this stage the cerebeUum rises as a transverse plate inclined at a wide angle to the axis of the medulla oblongata, and bears an obvious resemblance to the amphibian cerebellum, compare Figs. ^83 and 303. While the Varolian bend (Briickenbeuge) is developing the lamella thickens and widens; its posterior border passes gradually into the thin and exjianded deck-plate of the medulla; the transitional part is the secondary Rautenlippe, and does not participate in the formation of the cerebellum, but is the anlage of the velum niedullare imaficum {hinteres Markaeget) p. C". Thefurther development of the lamella in the chick has been investigated by Lahousse, 88.1; it continues growing in all dimensions ; by the third day ' it b^ns to arch forward and upward \ until it encloses a space which is a ]™^„ „,^ ,^, „„. diverticulum of the fourth ventricle " "'" ""' ""

Fig. 383; the convolutions are distinctly marked on the ninth day and are merely superficial transverse ridges not folds of the wall.

It is probable that in the mammalian embryo a similar bending of the lamella takes place, but that the diverticulum is obliterated by the growth of the cerebellar walls, but observations are wanting to verify this supposition. There is never present any large open diverticulum in the mammalian embryo (Kolliker, ** Entwickelungsges." 2te Auti., 537). The lamellar anlage of the mammalian cerebellum grows rapidly into a rounded protuberance, the transverse diameter of which exceeds the longitudinal. As seen from above, the cerebellum now appears somewhat pointed laterally. The lateral ends of the lamella expand and form the anlage of the cerebellar hemispheres, leaving the median portion as the anlage of the vermis. There now soon apjx^ar (beginning of the fourth month in man, cow embryo of 80 mm.) a series of four transverse grooves, by w^hich the surface of the vermis is divided into five primary ridges (g}'ri), which persist as five primary lobes throughout life ; two of the transverse lobes belong to the upper surface ; three to the lower surface ; they are respectively the quadrate or anteroH superior, the jH)Sten>superior, the posteroinferior, pyramidal, and the uvula. During the fourth month the hemispheres grow rajv idly, so that at five months they etjual and thereafter surpass the central vermis more and more in size. The primary transverse lobes spread onto the hemispheres during

FiQ. 3S4.— Section through the Cerebellum and Medulla Oh- fViri f/mrfli TnnnfVi nn^l

lon^atu i.f a Human Embr>'o of one hundred and sixty Days. *'"*' luurni muuiu, rtiiu

Minot(^ill. No. (irt. r. Vermis; /f, hemispheres; Fl, flocculus; they pcrSlst there as in

Jirf. m^lulla oblongata; C, Central canal. X 4 diams. ^^^ ^^^j^ throughout

life. In descriptive anatomy an astounding variety of names are applied to the various parts of each lobe; it would be an essential gain if at least three-fourths of these names could be discardetl. Each of the five primary lobes becomes subdivided by additional grooves, most of which are approximately parallel to the primary grooves ; the subdivision continues until the full munber of folia are j)roduced, which is probably accomplished before birth. The fifth or most posterior lote forms an independent expansion on each side, l)eginning in the fourth month to form the nocculus, Fig. 384, FL A numlH?r of additional details as to the human cerebellum at various stages are given by Kolliker (^'Entwickelungsges.," 2te Aufl., 54r*<J548).

The histogenesis of the cerebellum has been studied in the chick by Lahousse, 88.1, and in man by Bellonci et Stefani, 89.1, and Vignal, 88.1. In the chick at six days (Lahousse, Fig. 28) both the mantle and inner layers are crowded with nuclei and form alx>ut thrt»e-fourths of the wall in section, the remaining fourth being constituted by the Randschleier in which there are a few nuclei ; between the spongioblasts are seen the dividing germinating cells (p. 613) close to the central canal. The Bandechleier of the cerebellum is _ the graue moleculare Decklamelle of Lciwe, 80. S, or envellope ' moleculaire grise of Lahousse, I.e., p. 63. At the sixth day there begins to appear in the Randschleier a stream of cells, which probably come from the Kautenlippe of the cerebellum, but as the Xippe was not known to Lahousse he gives no information on this point.* These cells are elongated parallel to the surface of the cerebellum, close to which they appear. Their immigration results finally in the conversion of the Bandschleier into the outer layer of the adult. During the eighth day the nerve-fibres appear, the differentiation of the mantle and inner laj'ers is easily recognized, and the Randschleier now comprises three layers, a thin outer or superficial neuroglia layer, a middle gray richly imoleated layer, in which the immigrant cells are situated, and a third layer next the mantle layer, having scattered nuclei. The ninth day we can make out the following layers beginning within: I, the ependyma: 3, the inner layer; 3, the mantle or gray molecular layer, some of the cells along the outer edge of which are changing into Purkinje's cells, making another layer, 4; 5, the neuroglia layer or inner part of the Randschleier; and, (i, the outermost layer containing immigrant cells {Lowe's Zellstrief), cells which are probably neuroblasts. The six layers just enumerated can be recognized in the mammalian embryo, and have been described by W. Vignal, 88.1, ;t2r-;J34, PI. XIL, w^ho, however, failed entirely to rec<^nizo the early differentiation of the neuroblasts and neuri^lia. There is a thin outermost layer without nuclei, next follows a broader layer crowdecl with nuclei ; these belong to the cells which have migrated into the embtj-onic Randschleier, and they form a well-marked layer throughout foetal life; this layer, so far as I know, was first obeer\-ed by Obersteiner (Sitzber. Wiener Akad.Wiss., 1870), in the cerebellum of new-bom children; and it may be conveniently designated as the outer nuclear layer; it disappears as a distinct layer during childhood. Belionci et Ste fani, 89.1, 53, state that two zones itaj^~~wuMtoiu"No,BS~~uiidwtW™r. may be distinguished in Obersteiner's '?!ti!dIKy?do*'iS'^™lS'i^i?SrK^ layer, an outer zone with numerous "[■J.hfch'S^'^HoDwwed™**' ""'ei' •™"*' karjokinetic figures and crowded

round nucleolated nuclei, and an inner zone with the nuclei elongated and less crowded. In pigeons of twelve days' incubation, some of

• Lfiwe camp vpry wwr iII«covm-Iiib Ih^ BftUCi^iilippe. lor he olwrred Ui»t the epeDdJUU wu n-ilKtcd ou to tbe ouUide of the wrFbellum.


Rumui Enihr70 of one huTulnd utd ■ Itays, Miuot Ci

iho cells of the outer nuclei have develop)eil dendritic processes, wliich extend even into the inner nui*le«ir layer. Immediately below the outer nuclear layer is one with few nuclei, and then we come to the broad band of crowded nuclei belonging to the mantle layer proper ; everything out^side the niiuitle layer is derived from the primitive Kandscbleier. The ivlls of Purkinje art> recognizable in the human f (vtus of five months and their external branches at six months ; at seven niontlis their inner ends are rounded, but at birth pointed and apparently prolongeil as a process running into the mantle layer (axis-cylinder priK^ess?).

The Foiirtn Ventricle and its Roof.— The fourth ventricle lias long betni known to einbryologists as the expandetl central canal of the hind-brain, and as enclosed completely by the meilullary wall. The exiwuision of the deck-plate and consequent thinness of the dorsal wall of the ventricle was known to Von Biier, but he supposed that this wall was lost in the adult, 28.2, 74, 37.1, 108. Kemak, 60.1, 33, maintnineil this opinion for the chick; R^ithke, 39.1, 37,38, 20.1, Th. IV, 14, for reptiles and anamniota. We owe to KoUiker ('* Entwickelungsges. , " 1 ste Autl. , '^43) the discovery of its i>ersi8tence ; to Hensen (Arch.f. J/'/Ato.s-A*. Attat,^ II. 4'24) the demonstration that it forms the epithelial Ci)vering of the choroid plexus.* Several writers have thought that the membrane was broken through at cert^iin points, but it prolmbly is rt»ally continuous throughout life. The fourth ventricle is to he regardeil, then, as an exi)ansi<)n of the central canal jn^rmanently bounded by the original medullarj' walls.

The fourth ventrii*le has, as seen from alK)ve, a rhomlwid 8ha|)e, Figs. 34*2, 343; it tai)ers down anteriorly to the central canal (acjueductus Sylvia^) of the mid-brain, iK)steriorly to the central camd of the spinal cortl. It is widest at the level of the Varolian bend and in the adult the lateral angles of the embryo persist as the recessus httevales. The so-called H(H>r of the ventricle is constituted by the inner surface of the doi*sal and ventral zones of His, already descrilxMl.

The nH>f of the ventricle behind the cerebellum is derived from the <Uvk-phite, nmipare p. (iOH; it bec^omes subdivided into three parts, the iU)rsid eiK»ndyma, the seix)ndary Rautenlippe, and the epithelial covering of the i'hon>id plexus of the fourth ventricle. The de(*k-plate is a layer of epithelium and preserves its simple epithelial character through most oi its extent and throughout life. In the human embryo at four wi*eks (His, 90.2, 20) it is a single laj'er of cells, S/t higii by !()/* wide, but toward the edges of the plate the cells l>econie a little higher and narrower; the number of cells increases (whether by their own division or not, is uncertain) so that the Ci»lls lH»come higher (1 1-1 3/«) during the second month, although the area of the membrane greatly enlarges.

Where the deck-plate joins the lateral wall of the medulla it lH^<M>ines thi(*kened, fonning the secondary Rautenlip^ie, p.G7. When the main deck-plat(» and the choroid ])lexu8 are removed from an embryo of two months or older, the Rautenlippe api)ears as a

•Thf»«» refi'ivnct's an* taktMi without verifli'atltm from Mihalkovios. T7. 1, TiO. The reference to HeDHeu lias been verifled, beiu^ to an incideutal obserration in a paper on Um) eyes of snaila.

narrow whitish band along the edges of the fovea rhotuboidalis, or, in other words, of the medulla oblongata and cerebellum. The band persists throughout life and is known in descriptive anatomy by three different names ; the part attache<l to the ccrel)ellum is termed the velum medullare posticum (hinteres Marksegel) ; the part along the edge of the medulla oblongata is tenned the Teen in fosHW, rhomboidalis or ligula (Riemchen) ; the part at the posterior ajKJX of the rhomboid opening is termed the obex (liieyel).

The choroid plexus is an ingrowth of the deck-plate accompanied by vascular mesenchyma and projecting into the cavity of the fourth ventricle. In the amphibia. Fig. 303, the half of the deck-i)lalo nearest the cerebellum forms a series of irregular rounded projections into the cavity of the fourth ventricle, and each of th(»He proj(»ctions contains mesenchyma (/. e. connective tissue and blood-vessels). In mammals we find the same choroid area, but it is j)ushed in, as a whole, into the cavit}\ In the placentalia at least, the invagination of the whole area precedes in the embr>'o the; fonnation of the irregularities of the surface. The invagination, cf. Fig. 3K(;, may Ikj seen in the human embryo of five or six w(»eks as a transverse fold <>f the deck-plate extending quite deep down, and resulting, a[)par<»ntly, from the excessive development of the Varolian Ix^nd. The fold is the anlage of the choroid plexus. By its further development the* aniago assumes a more and more complex and irrc»gular fcjrm, l>ut it remains always a fold of mesenchyma richly vascular and cover(?d by the epithelial deck-plate. In the human embryo at four months (KrJliker, " Entwickelungsges.," 2te Aufl., 540) thcj position of the fold can be seen, when the medulla oblongata is view(Ml from alK)V(», jis a narrow transverse line, along which the mew^nchyma (cormc^crtive tissue of the pia mater) enters the fold, and wliicli is situatf?d close behind the cerebellum; in front of and Ix'hind this line th(» deckplate forms a transverse ridge ((Jlfrus choroidtmH anterior and ;>o^terior) ; the two ridges might, at first sight, Vx? mistaken for fx^rticnm of the cerebellum.

The Mid-brain. — Concerning the sec^ond cerebral vesicle our information is ver>' imperfect, and amounts to little more than a knowledge of its general form at successive stagcjs; it is dffriveil chiefly from Mihalkovics, 77.1, «3-f;H, and Kolliker, "Entwickelungsges." 2te Aufl., 535. The mid-brain is remarkable for its precocious expansion. Fig. 341, and fr>r the fact that in young embr>'OH it occupies — owing to the cephalic flexures — the highest part or summit of the head. Fig. 338. In b<jth the figures just referr(!il to, the mid-brain appears as a vesicle with a large cavity and thin walls constricted in fnmt as it joins the fore-brain — Ix/hind, as it jf^ins the hind-brain. We have no knowledge of the sej/arate histories of the six longitudinal zones (dfck-plate, the four zfines of His, and the Bixlenplatte). The floor of the mid-V>rain ver>' early l)egins to thicken, and the thickening includes the Bfxlenplatte, for it extends a^-ross the methan line. On the dorsal side the median line has, in young human embrA'os at lea.st, an external ridge with a c^>rrffSfx/nding internal «^roov«\ lx>th resulting frr>m a meilian fold of the m<filullar>' wall. The whole dorsal side of the mid-brain expfinds considerably Oiiiman embryos of four weeks) ; esfx^cially is this the case in Saurrjiisida, as may be well seen in a cbick embrro of the fourth day, Fig. 3S2. The mid-brain now grows steadily, though much less than the foreand hind-brains, so that the cerebrum and cerebellum outstrip it. Its growth is principally a thickening of its walls and an increase of its length, but with little enlargement of its ca\-ity ; hence the cavity becomes relatively smaller, though it persists throughout life as the part of the central canal known as the aqiiednctus Sylvire, and intervening between fore-brain (third ventricle) and hind-brain (fourth ventricle) .

The ventral part (? ventral zones of His) of the mid-brain develops into the peduncles nf the cerebrum; the projecting of the peduncles as rounded longitudinal ridges on either side of the median ventral line becomes noticeable during the third month; they remain small until the fifth month, when the fibres from the pyramids of the medulla oblongata begin to penetrate them, and thereupon they enlarge and at the same time the longitudinal concavity of the ventral side is obliterated. It is probable that the Bodenplatte thickens, somewhat as in the medulla, and persists as a median raphe.

Fig. 9ee.-M«IiaD Seotinn nf Embri-o iif 3a uim. i. SkdIi cerebri: /.m. foranieD of Sui

The corpora qttadrigemina arise from the dorsal side of the midbrain, and will, perhaps, be found to represent the dorsal zones of His. The dorsal wall of the mid-brain is at first evenly arched and smooth ; at five weeks there is a median ridge, as already noted ; during the third month the ridge is replaced by a groove; during the fifth month there appear two oblique grooves which run inward and backward, one on each side, Fig. 3*i7, and complete the subdivision of the surface into the four corpora quadrigemina. Concerning the development of the posterior commissure, which is a bundle of fibres crossing the dorsal wall of the brain just in front of the corpora, see p. C8G,

Owing to the fact that the mid-brain grows much leas than the foreand hind-brains, it is gradually covei-ed over, principally by the expansion of the hemispheres. At the beginning of tho third month the hemispheres have expanded to the edge of the mid-brain ; at three months they half cover it; at four monms they coyer all but a small piece; at five months the whole of the mid-brain.

Median Portion of the Fore-Brain. — The manner in which the primitive fore-brain is divided into two lateral parts or hemispheres and one median ^lart {Thalamencephalon, Zwischenhirn), aft«r the outgrowth of the optic vesicles, has been described. The cavity (enlarged central canal) of the median part is the third ventricle of descriptive anatomy ; therefore, the median part is sometimes called the region of flie third ventricle. For convenience the hemispheres are treated in a separate section. It has alreaily been i>ointeil out that it is misleading to describe the primitive fore-brain (first vesicle) as dividing , into two secondarj' vesicles. To divide the median portion of the fore-brain into two parts, as is traditionally done, is arbitrHiy. We shall, therefore, in this section consider not only the thalamencephalon as usually delmed, but alf^o the lamina terminahs and the commissures, etc.

1. General Shape. — By the fifth week the median portion of the forebrain has assumed nearly its definite form. Across the anterior median line extends that portion of the medullar}' wall connecting tho two hemispheres known as the lamina terminalis. Fig. 340, between /.HiandiJ.o. Aboveand around the dorsiil side of the foramen of Monro the medullarj' wall is continued, in the median line, Fig. 340, but is modified first to form the corpus callosum, second the choroid plexus, two structures of which the history is < presented below. The corpus callosum "\wl';iS"oV"'Ki'rb.Sn"z'^^'i: la a thickening produced by fibres, E^^^ft^w^HiS**^"*'"' ■" '"'^' forming a transverse commissure between the two hemispheres. The choroid plexus is a fold of the medullari- wall which projects into the cavity of tho brain. Fig. 388, Fix. The cerebral hemispheres are outgrowths from the anterior part of the forebrain. Fig. 3;iO; the passage from the cavity of the hemispheres to the median cavity is the foramen of Monro, Fig. 300, m : the part of the fore-brain between the foramen of Monro and the mid-brain corre^>onds t-o the thalamencephalon. or Zwischenhim as ordinarily defined. The thalamencephalon as viewed in dorsal a-spect in a human embryo of five weeks, Fig, 388, Z, has somewhat of a caak-shape. The anterior end adjoining the hemispheres is narrower than the posterior end adjoining the mid-brain ; the anterior half of the thalamencephalon slopes inward. Along the median dorsal line is a ridge, a, which is developed as a fold of the deck-plate during the fifth week ; toward the hemispheres the ridge widens out and disappears ; the continuation of the deck-plate between the hemispheres corresponds to the tela choroidea; toward the mid-brain the ridge merges into a median evagination of the brain-wall ; this evagination is the anlage of the pineal gland, p. 088 ; there are soon developed the two ridges which diverge V-like from the pineal anlage to run forward along the median ridge, and which are destined to form the pars habenularis (ganglia habenulse, laminae medullares, and pineal stalk) of the pineal lobe.

Viewed in median section. Fig. 340, the median fore-brain is seen to have a great downward prolongation which begins to form during the fourth week, develops rapidly during the fifth week, and persists throughout life. The enlargement may be designateil as the subthalamic or infundibular; subthalamic because it lies below the region in which the optic thalami arise, infundibular l)ecau8e its apex is the recessus infundibuli. As seen in section the enlargement has, 1, a i)osterior wall, Jf, which descends at nearly a right angle to the axis of the mid-brain ; the posterior wall is convex, and it is the anlage of the mammillary tubercles; 2, a lower wall which includes the anlage of the tuber cinereum, /.c. of the infundibulum, and of the optic chiasma; 3, an anterior wall constituted by the lamina terminalis. At the angle where the anterior and lower walls meet, the recessus opticus, R.o, leads off laterally into the hollow stalk (anlage of the optic nerve) of the optic vesicle. Higher up lies the foramen of Monro, fm, leading into the cavity of the hemispheres, //. In the figure there is seen a groove which runs from the recessus opticus, JK.o, to the mid-brain; this groove marks the division line between the dorsal and ventral zones of His ; it persists in part throughout life. The persistent part was named the sulcus Monroi by Reichert because it runs later from the lower e<lge of the foramen of Monro, the foramen extending as it devel()])s much closer to the recessus opticus than it does in the early stage of Fig. 34( ».

In older stages the median fore-brain shows many minor mcxlifications, but its fundamental shape and division, as found at five weeks, are permanently retained. The most important alterations are due^ first, to thickening of the walls, which is especially great in the region of the optic thalami ; second, to the fact that the foramen of Monro does not enlarge with the growth of the brain, and therefore becomes relativelij small, comjmre Figs. 340 and 38(3.

Appearance in Cross Sections. — Fig. 389 is a section of the thalamencephalon of a five wrecks' embryo nearly at right angles to its axis. In the median line is the deck-plate, d.pl, with its three folds already described; the division between the dorsal, Th, and ventral, s.Th, zones of His is well marked by the sulcus Monroi. The Bodenplatte forms the mammillary groove, J/a, v/hich is b<:)rdered by two eminences internally; the eminences are the cross sections of two ridges, whic'h l)order the groove and unite with one another in the median line beween the tuber cinereum and the

Duunmillary region proper; the ridges are the aulages of the mammillary tubercles.

Fig. 390 represents a much older stage and serres to show the thickening of the walls and the origin of the choroid plexus ; the section passes through the foramen of Monro, m, and the optic chiasma, cii; and the piane of the section may be approximately rec<^nized from Fig. 388. Very striking is the great thickening of the brain-walls to form the anlage of the corpus striatum, st, in the hemispheres, and of the optic thalamus, tb, and the pars subthalamica in the middle part. The deck-plate, s, closes the third ventricle t, above. The medial wall of each hemisphere is bent in, n, owing to the Bo(jenfvrche, p. ti96. The wall of the hemisphere does not join the deck-plate, s, directly below the Bogenfurche, but changes into an epithelial membrane, which forms an irregular fold, pi, projecting far into the cavity, /, of the hemisphere, or lateral ventricle ; this fold is the choroid plexus, see below.

The Deck-Plate. — The entire deck-plate except the pineal (and paraphysal) parts assumes an epithelial character. It produces the pineal gland, see p. 688, the paraphysis, see p. 6D0, anti the choroid plexus, and persists in part as the tela choroidea. The pineal gland and paraphysis are so far independent organs that they are treated in separate sections. We are, therefore, here chiefly concerned with the choroid plexus.

The Choroid Plexus. — When the hemispheres begin to grow out, the deck-plate between tfiem and above the foramen of Monro is convex, but it soon becomes concave and during the fifth week the deck-plate forms a fold on each side projecting into the lateral ventricle. The space between the two hemispheres is occupied by mesenchyma, which grows into the lateral fold carrying blood-vessels with it; the fold is the anlage of the choroid plexus; its relations are well shown in Fig. 391. At first the choroid fold contains no connective tissue, the ingrowth of mesenchyma following after the fold is formed ; the fold, therefore, owes its origin to the growth of the deck-plate. Examined in a side view the fold is seen to be thin, but long ; it ends abruptly in front, but disappears posteriorily more gradually (His, 89.4, 605). The deck-plate becomes a layer of cuboidal epithelium covering the choroidal fold, and merging on the one hand into the wall of the hemisphere and on the other, Fig. 391, into the median part, tela choroidea, of the deck-plate. The tela is itself an epithelial layer, which is continuous in front with the lamina terminalis, behind with the pineal anlage. During its further development (c/. Mihalkovics, 77.1, 114-117), the folcl increases in length and diameter, and its surface is thrown up into rounded protuberances, which grow into irregular processes. The fold takes its place in the lower part of the lateral ventricle, lying close against the basal surface (ganglia) of the hemispheres (Mihalkovics, 77.1, Taf. 1, Fig. 10). The size and complexity of the choroid plexus are correlated with the degree of development of the hemispheres, and the plexus is, therefore, largest and most specialized in the mammalia. The plexus in the human embryo enlarges more rapidly than the lateral ventricle so that by the fourth or fifth month it quite fills the lateral ventricle, but after that period the plexus lags somewhat, and there is gradually produced the space around it as found in the adult.

The connection of the choroid fold with the medullary walls of the hemispheres extends during embryonic life for some distance backward from the foramen of Monro. The exact history of this modification has never been traced.

Lamina Terminalis. — The embryonic history of the lamina terminalis was long imperfectly understood, but it has been cleared up by F. Marchand's investigations, 91.1, on human embryos. It may be regarded either as a prolongation of the deck-plate, or, as suggested by His, 88.3, as the result of the union of the dorsal zones of His {Fliigelplatten) in front. It is the median portion of the medullary wall. Figs. 340, 341, in front of the recessus opticus and foramen of Monro; it unites the two hemispheres, being, of course, continuous with their walls, and it closes the third ventricle anteriorly; it is continuous above with the tela choroidea. Fig. 395, below with the optic chiasma (or anlage thereof). At five weeks it is a thin plate, Fig. 340, of about the same thickness as the deckplate, and with cells but little if at all differentiated.

The upper part of the lamina terminalis becomes wQvy much thickened, and forms (Mihalkovics, 77. 1, 12*2) a broad band of triangular section after the fourth week, uniting the two hemispheres. This band is the anlage of the septum lucidum, the corpus callosum, the fornix, and the anterior commissure. Fig. 391 ; the lower apex of the triangle is the anlage of the anterior commissure, ca; the posterior vertical bi^rder of the fornix; the upper horizontal border of the corpus callosum, c.r, and the remainder of the area is the anlage of the septum pellucidum. It is usually described as resulting from the concrescence of the two hemispheres, but I consider it simpler and more natural to regard it as a thickening of the lamina terminalia, which it ia morpholc^cally. The anla^ may be well seen in a median longitudinal section of the brain of a cow embryo of 8 cm. (Mi- "

halkovics, I.e., Fig. 17) or of a human embryo of the third to fifth month, Fig. 391. The anterior commissure acquires its fibres before they appear in any other part of the lamina terminaiis, and early become separated from the fornix and septum lucidumashort distance. The thickening hes below the Bc^nfurche, bf, and in front of the foramen of Monro. In it the fibres to form the anterior commissure and the fornix have been observed to appear in rabbit embryos of

25-30 mm., and those to form the corpus callosum in rabbit embryos of 35- m«»i v. BoRenfureiie; cc. conn» osiio..ik Bum: 6'p. aeptum lucldum; c a. uilerlur

40 mm. (jllhalkOVICB, I.e., 123, Vii) commlHurei CU, olfactory lotw: CAi,op in pig embo'os of 8 mm. (Blumenau, {^^.V^il *S(' «J?bfK™'6 mid:

81.1, ti). hU\o: pin.Diaeni gUail. After R Mar The fornix, corpus callosum, and septum lucidum tc^ther form a triangle, which after its formation expands throughout foetal life. The anterior apex, where the fornix and callosum meet, grows forward, and the posterior apex, corresponding to the end or splenium of the callosum, grows backward ; the corpus callosum is thus not only lengthened but carried backward. Fig. 302, cc, above the foramen of Monro and the optic thalamus, Th. The development of the corpus callosum also extends beyond the anterior apex ; the part below the apex i.s short, ro, and corresponds to the rostrum i of descriptive anato. my. From a morpho\ logical point of view, Giacomini's statement {Giomale d.r. Accad. Med. Torino, Nov.-Dec, 1883), that the corpus callosum is covered by a thin but constant layer of gray matter, is verj' significant. The statement has been verified by Blumenau, 01. 1.

Tlin septum pell ucidiim (or luciduin)\& developed from of the thickened lamina terminalis between the corpus calloiium and the fornix. The area is at first very small, but rapidly enlarges. At four months a small cavity appears in it. Fig. ^91, Sp, which enlat^es as the septum grows and becomes the ventricle of the septum {ventriculus quintus, pseudoc(fle),iiC€iF. Marcband, 91.1, SI. The origin of the cavity is uncertain ; it has no connection with any of the brain cavities proper; Prof. B. G. Wilder writes me that in man and anthropoids it is wholly circumscribed by brain tissue; it is much narrower in other mammals, but the pia does not extend into it. Marchand thinks it probably arises as a cleft in the tissue.

Co.MMissuKE» AND FoRSix.— A commissure is a tract of transverse fibres connecting the two sides of the nervous system. In the mammalian brain three such tracts are known to arise in the territory of the first vesicle; they are: 1, the anterior commissure; 2, the corpus cuUosiim; '.\, the posterior commissure. The anterior commissure and corpus callosum are developed, one might also say, as jiarts of the septum pelluciduni and belong morphologically to the lamina terminalis. The fornix may be defined as a longitudinal commissure. For the general relations of the commissures and septum to the lamina, see alwve. Osbom, with great ability, has traced the homologies of the three commissures throughout nearly the entire vertebrate series, and has shown, 86.1, 87.1, that, contrary to previous belief, the cor|>us callosum is not confined to the miimmalia, but is present in birds, reptiles, and amphibia, and probably in fishes, and further that in amniota and amphibia the anterior commistjure comprises always two divisions — an olfactory and a temporal. Mammals, therefore, are distinguished from other vertebrates, not by the possession of the corpus callosum, but by its great size, which we may safely correlate with the great size of the mammalian hemispheres. The typical position of the commissures is shown in Fig. 3it;f; the posterior commissure, P, lies behind the pineal gland, pi, close to the corpora bigemina (mid-bram); the corpus callosum, c, lies close to the foramen of Monro, fm, and the anterior commissure is situated lower doivn, a, in the lamina terminalis, and it consists of two bundles of fibres, an upper lai^er pars olfactoria and a lower smaller pars temporalis: the fibres of the tem]>oral bundle are distributed to the temporal portion of the so-called mantle, Fig. 394 ; the fibres of the olfactory portion run in part to the olfactory lobes, but also ^ive off a frontal branch bundle to the frontal region of the mantle. The mammalian corpus callosum consists of an anterior or frontal division supplying the dorso-mcdial portions of the mantle, and a posterior division, the commissura comu Ammonis, supplying the mantle area above the Ammon's hom (H. F. Osborn, 87.1, 540).

Flo. SM.-MiMiftn

The development of the commissures in marsupials (Osbom, 87. 1, 531)) shows that the homolt^es established by Osbom are correct. But in sheep the development is so far modified that these homologies are less clearly brought out. The development in sheep is thus de- ' scribe<i by Osbom, 87.1, SUS: "In the .10 mm. stage the hemispheres have already partially united in frout of the primitive lamina terminalis forming the terminal plate. The anterior commissure now appears as a delicate thread of fibres in the lateral region of the brain stem. The hippocampal sulcus is 1 well markeil. At 35 mm. the an

. „ .(, Oltaetory lobe; imlBSure, dlvlJIag Into tbe ol, and the pan tempont terior commissure extends slightly ir?, 'opiletiial'amus;' V.'thlrd veotrlcle.'

nearer the median line. In an embryo of '^7 mm. the terminal plate has extended considerably forward. The anterior commissure shows a division into the pars olfactoria and temporalis, while in the median line its fibres begin to nnite with those of the opposite hemisphere. This unioa doee not take place in the terminal plate, as stated by Mihalkovics, but in front of it, t\e., the plate does not form the ground substance to be traversed by these fibres. On the other hand the fibres bridge the fissure which is gradually closing in front of the terminal plate. Immediately above the anterior commissure, on either side, are descending fibres which represent the first stage of the fornix. These appear fofore the anterior commissure crosses the median line. This stage corresponds closely to that figured by Mihalkovics, 77.1, Taf. VII. Fig. 60. In the next stage the terminal plate has extended in front of the anterior commissure, the fornix fibres are more numerous, and at their upper limit a few fibres are observed extending toward the median line; these are the earliest callosal elements. At 4il mm,, which follows a considerable interval of development, the hippocampal sulcus is very deep and the terminal plate is much more extensive. In its lower portion the anterior commissure, now a compact bundle, extends laterally above the cerebral peduncles. The columns of the fornix are well defined, and between them in the upper portion of the plate pass the fibres of the corpus callosum. A careful study of these fibres shows that, like those of the anterior commissure, they unite with each other in front of the terminal plate. The callosal fibres disappear as they pass around the hippocampal sulcus, Above this sulcus is an interval in the inner wall of the ventricle in which no fibres can be observed, but in the root of the ventricle are the fibres of the corona radiata. This leads me to doubt whether the fibres extend at an early stage from the corona radiata into the corpus callosum, as stated by Mihalkovics. It seems rather that this is a subsequent union. This stage differs considerably from that figured by Mihalkovics as the initial stage of the corpus callosum. "

The posterior commissure has been but little 8tuthe<l embryologically. ItspoRitionmay berecoguized (K611iker,"EntwickeIungsges.," 2te Aufl., 525) in a sheep embryo of 2!t mm. as a slight thickening of the dorsal wall of the fore-brain close to the mid-brain. The fibres of this commissure appear in the chick the latter part of the fourth day, according to Mihalkovics, 77.1, 73.

Dorsal Zone of His {Optic Tlialami). — The dorsal zone of His in the fore-brain forms the hemispheres and in the median portion produces the optic thalami. The thataini may be defined as thickenings of the dorsal zones continuous with the thickenings which produce the corpora striata of the hemispheres. It will be remembered that the lower limit of the dorsal zone is marked by the sulcus Monroi. The development of the thalamus has been outlined by KoUiker in both his text-books ; the early stages in man {fourth to twelfth week) have been investigatetl by W, His, 89.4, 701, 731. The sulcus Monroi becomes evident during the fourth week and very distinct dtiring the fifth; later it becomes shallower, but persists.

At the beginning of the fourth week the thalamic region is concave toward the ventricle. During that week the thickening of the _ _ , walls in both the thala mic and sub-thalaniic regions begins, and bv the end of the fifth week the wall projects in both regions convexly into the cavity of the third ventricle. The thalamic thickening does not extend throughout the dorsal zone of the thalamencepbalon, but only in a circumscribed reEmbryo of gion. It accordingly " w'tT^^ produces a lai^ tulier, ■(er w. Hi«; Fig. 395, 77t, the longcontinued growth of which converts the third ventricle into a narrow fissure. The tuliers meet toward the end of the second month and actually unite over a email area across the median line, their union constituting the comminsura mollis, cm. Mihalkovics, 77.1, 71, assigned the formation of the commiasura mollis to the fifth month, and this date is confirmed by F. Marchand, 91.1, 310. His thinks that the commissure is formed earlier. Above the tuber thalaniicum is a groove named the sulcus habenulce by His. It corresponds to the external ridge described, p. 680, as running obliquely along the upper surface of the thalamencephalon to the pineal anlage. Below the tuber is, of course, the sulcus Monroi ; the two grooves are united behind the tuber, where they are also joined by a transverse groove, the sulcus pineale; where all these grooves meet there is a slight lateral enlargement or recess, recessus geniculi, of the ventricle. As the tuber enlarges the recessus deepens and is narrowed so that at two months and a half there is only a small fissure visible. Later even this fissure disappears ; its walls probably give origin to the " centre median " of Xtuys {mediane Sehhugelcentrum),

The part of the dorsal zone between the thalamic tuber and the mid-brain is known as thepars retrothalamica; it includes the pulvinar, the brachia of the corpora quadrigemina, and the corpus geniculatum.

Ventral Zone op His. — This comprises, as already stated, the region of the thalamencephalon below the sulcus Monroi; for this part Forel (" Untersuch. lib. d. Haubenregion," Arch, /. Psychiatries Bd. VII.) has proposed the convenient name of pars subthalamica. Concerning its embryonic history almost nothing is known. It becomes very thick and is usually described as part of the optic thalamus.

Floor of the Third Ventricle. — Along the floor of the ventricle on or near the median line are developed the following structures: a, substantia perforata posterior; 6, mammillary tubercles; c, tuber cinereum ; d, infundibulum ; e, optic chiasma ; /, lamina terminalis, but this last does not properly belong to the floor. What relation the Bodenplatte bears to the production of the first five structures is still uncertain. In regard to this development little is known.

o. Substantia perforata posterior perhaps really all belongs to the mid-brain. It becomes distinct during the fourth month.

&. Mammillary tubercles (corpora albicantia, candicantia, Markkugelchen) begins. Fig. 340, m, as a single relatively large convex projection of the medullary wall. As the brain enlarges the mammillary region grows very slowly and hence becomes relatively small (W. His, 89.4). According to Mihalkovics, 77.1, 72, a median groove arises early in the fourth month dividing the region into two tubercles, which later become white (owing to the development of meduUated nerve-fibres ?).

c. Tuber Cinereum. — This is part of the floor between the mammillary tubercles and the 'infundibulum proper, see Fig. 340, <.c, Concerning its development no details are known.

d. Infundibulum. — In rabbit embryos of 12-16 mm. and inhuman embryos of five weeks there is found developing a small cylindrical outgrowth of the brain, which is known as the processus infundibuli. The outgrowth takes place in the median line immediately in front of the tuber cinereum and behind the optic chiasma. Figs. 301 and 401, Inf. It very soon comes in contact with the hypophysal outgrowth of the mouth, and is ultimately transformed into the posterior lobe of the pituitary body as already described, p. 674. His' observations, 89.4, 706, on the human embryo confirm in most respcrls Milhalkovirs' account of the development in the rabbit.

e. Optic dhlnsma and Recessus. — The optic chiasma and tracts together constitute a transverse ridge-like thickening of the wall of the brain to allow the passage of the nerve-fibres of the optic nerves. Laterally the ridgc?s merge into the optic nerve ; the recessus opticus is bounded liehind by the ridges, in front by the lamina terminalis. The optic ri<lges {Sehstreif of Mihalkovics, 77.1, 78) accordingly are first indicated when the groove of the recessus opticus develops, and they bec^>me strongly marked, as the optic nerve-fibres appear. In the chick the fibres have been oliservea in the latter part of the fourth day passing from one side of the brain through the optic ridge to the optic nerve of the opjKjsite side, Mihalkovics, l.c. The growth of the fibres is centrifugal.

The rereH.sn.H (tpfinis, which was first described by J. Michel in 187"2, leads to the optic nerve, being a transverse groove, Fig. 301), R.op. It is more marked at birth than in the adult, but may be traced throughout life. For notices of the scanty literature previous to 1877, ui>on the chiasma and recessus, see Mihalkovics, 77.1, 8r>-S2.

Pineal Gland {Epiphysis, comirinm^ pineal or parietal eye^ Zirlxd^ZirbeUlrilse) . — The pineal gland or epiphysis is develo[>eti as a median dorsal evagination of the medullary wall of the fore-brain a short distance in front of the mid-brain; between it and the midbrain is situatf^l the jx>sterior cr)mmissure, p. 684. Its site is said by A. Got^tte, 76.1, t<) lx> identical in Bombinator with that of the anterior neuroporus, or i^oint where the meilullary groove closes last

in the head ; if this coincidence is true of vertebrates generally, it must have some, perhaps important, significance. The development of the epiphysis in reptiles and lower invertebrates indicates that it was primitively a median eye, which survives as a rudiment, compare lx?low.

The pineal evagination appears after the development of the brain is quite advanced (chick, end of fourth day, in the rabbit the fourteenth dav, in white mice of l>.r> mm., in sheep embrvos of

3.5 mm., in man at about the" sixth /.; h-iiK. Hurroiiniiwi by ^veck) ; it therefore cannot — as Mi th«' optic v«»«icU'; of, , n • mm i r^c • j.i i

oN)oyHt : 3/</, hind-brain, halkovics, 77.1, 05, justly obscrves Aft.'r Duval. agaiust A. CWtte, 76.1, 315-3ir,—

be interpreted as resulting from the connection at the neuropore of the medullary wall with the epidermis, for the two layers are 8e|)arat(Ml by intervening mesenchyma in amniote embryos long before the evagination appears. In birds the evagination points forward, in mammals backward; this difference is probably due to the greater d(^velopment of the corpus callosum forcing the pineal gland back in mammalia. Our knowledge of its development in mammals and birds we owe chietiy to Mihalkovics, 77.1, 94, whose results have been confirmed by Kraushaar's observations on white mice, 86.1. The evagination lengthens out until it nearly reaches the epidermis; it is a tube or sac communicating with the fourth ventricle, ending blindly, and having walls composed of cylinder cells. The sac next enlarges at its upper end, and the wall of the enlargement after thickening forms buds (chick fifth day, rabbit embryos of 20-25 mm. ) , which are hollow and retain in the chick their primitive form but in mammals the hollow buds become filled with proliferated epithelial cells, which take on roimded and polygonal forms and are presumably degenerated elements ; the cells have processes and lie more or less separated from one another.

Vm. .'>.>4i. Hralii of a Chick Einbryt), fourth Day. I, KirKt, II, wn-ond c«*n'i)ral vcKicIe: K}>, «*pi

in reptiles the epiphysis assumes a more complicated structure, but in many forms its differentiation is more or less imperfect. When carried to its highest known development the pineal sac is differentiated into three parts, a distal eye-like enlargement close to the epidermis, a middle, narrow part like an optic nerve, and a proximal larger part, as shown by W .B. Spencer, 86.1, whose results have since been verified and extended by Beraneck, 87.1, Beard, 88.2, Francotte, 87.1, 88.1, McKay, 89.1, Owsjannikow, 88.1, Ritter, 91.1, Strahl and Martin, 88.1, and Wiedersheim, 86.1. A synopsis of the development of the reptilian epiphvsis is given by C. K. Hoffmann in Bronn's " Thierreich," VI., Abth. III., 1981-11)93. The distal end of the evagination lies near the epidermis ; it early enlarges into a hollow globe, which soon flattens out somewhat; the wall on the side next the epidermis thickens and assumes a lens-like character; the wall on the opposite side is, of course, united with the stalk and assumes a retinal character. Strahl and Martin, /.c, observed in the retinal region the differentiation of the Randschleier and of the nuclear layer, and the presence of karyokinetic figures next the cavitj% so that the primary stratification is the same as in the wall of the brain proper; later pigment granules are deposited in the part of the retinal layer toward the lens, and nerve-fibres can be observed in the stalk. There can be little question that the structure in question is a true, though rudimentary eye. It has also been observd in lampreys and amphibians.

The morphological significance of the pineal bod}^ is still under debate. The fact that it forms an eye in Petromyzon indicates that the optic character was primitive, but it appears to have lost that character along the lines of descent leading to the teleosts and elasmobranchs, while it has retained it along the lines leading to the amphibians and reptiles, becoming in them more or less rudimentary and disappearing altogether in the birds. As the pineal eye is the distal part of the epiphysis only, and is wanting in mammals (compare, however, H. F. Osbom, Science, Jan., 1880), the suggestion is inevitable that the pineal gland of mammalian anatomy is homologous with the proximal part only of the reptilian epiphysis.

Historical Note, — The first suggestion that the epiphysis might represent a visual organ was, so far as known to me, made by RablRiickhard, 82.1; it was renewed by Ahll)orn, 84.1, but was first definitely verified by De Qraaf, 86.1, whose article, together with Baldwin Spencer's admirable memoir, 86.1, forms the basis of our present knowledge of the pineal eye. Le3'dig, 88.4, 90. 1 , attempted, but unsuccessfully, to prove that the pineal eye could not be a sense organ. As regards the development, the principal authorities are Mihalkovics, 77.1, for manunals, Beraneck, 87.1, and Strahl and Martin, 88.1, for reptiles.

Paraphysis, — Under the name of paraphysis, or Stinwrgan^"^ Selenka, 90.1, has described a second evagination from the median dorsal wall of the fore-brain, which is similar to the epiphysis. It is further for\\*urd, being between the orig'n of the hemispheres. Selenka very doubtfully compares it with the median auditory organ of ascidians, as the epiphysis has l>een compared to the median eye of ascidians. Selenka has observed the organ in sharks, reptiles, and marsupials. In reptiles, just after the pineal evagination has begim in the embryo, there appears another evagination some distance in front of it and also in the median dorsal line, to develop the paraphysis. The evagination grows backward until it reaches the epiphysis ; after the pineal eye is cut oflF, it shoves itself imder the pineal eye, but without uniting with it ; the end of the paraphysis is enlarged and forms a number of fine hollow buds ; its proximal part or stalk is round or oval in cross sections ; throughout the embryonic period the cavity remains in (communication with the third ventricle: the fate of the organ after birth is unknown. Unfortunately Selenka gives no figures.

The paraphysis has been observed by Charles Hill, 91.1, in the embryo of 7 mm. of Corregonus (a teleost) to grow out asymmetrically from the wall of the brain just in front of the epiphysis; it is about half the size of the epiphysis.

Cerebral Hemispheres. — A previous section is devoted to the development of the median portion of the fore-brain, and accordingly in this section wo confine ourselves to the lateral outgrowths or hemispheres of the fore-brain. The hemispheres arise, as has been descril>ed, as diverticula of the dorsal zone of His in the anterior half of the fore-brain, and therefore they can never develop any structures homologous with parts arising from the deck-plate, the ventral zones of His, or the Bodenplatte. The choroid plexus might be t^iken as an exception to this law, but, as its development teaches us, it is not morphologically part of the hemispheres. For convenience the cerebral convolutions are considered in a separate section, p. 095.

General Growth. — The hemispheres of the human embryo of four weeks have been described, p. 500. They continue to enlarge throughout the entire foetal period, but their connection with the middle portion of the fore-brain does not enlarge correspondingly. There is but little (? if any) enlargement of the iforamen of Monro after the fifth week, but there is a considerable growth of the walls of the foramen, so that the actual size of the structures connecting the hemispheres with the median fore-brain increases very considerably, but the enlargement of the hemispheres is still more rapid, so that they become and remain large, pedunculate, vesicular lat<?ral apjiendages, and project beyond the median fore-brain forward, upward, and in later stages backward so as to cover the mid-brain also. The enonnons expansion of the hemispheres is one of the most characteristic features of the amniote embryo, but the expansion is greater in mammals than ia reptiles, in man than in any other mammal. The size of the hemispheres in the adult is closely correlated with the degree of mental development of the species. The fundamental relations of the hemispheres to the fore-brain are clear from Fig, 388 ; while retaining their strictly limited connection with the anterior part of the median fore-brain by means of the medullary walls bomiding the foramen of Monro, /.3f, the hemispheres are expanding in every direction. In front and above the two hemispheres have walls which face one another and are separated by a narrow constantly deepening fissure, Fig. 390, /, which is fiUed with mesenchymal tissue constituting the anlage of the falx cerebri {Hirnsivhel). Posteriorly a groove separates tie hemisphere from the Zwischenhim. In a lateral view the hemisphere shows a wide, shallow depression at five weeks, which gradually becomes more marked. Fig. 3!>7, and is ultimately transformed into the fissure of Sylvius. Corresponding to the external depression there is an internal projection of the wall of the hemisphere into the cavity of the lateral ventricle; this projection is the first indication of the corpus striatum, which arises as a thickening of Q_)i the wall extending not only over the region of the fossii of Sylvius, but also past the foramen ^j^, of Monro, to be continued as the thickening of the thalamencephalic wall, which produces the tha- fio lamus opticus, p. 080. We thus tJ™^ have a hemisphere the floor wall N^Va. of which is thickened to form the anlage of the so-called basal ganglia, while the rest of the wall is thin and is designated as the mantle ( pallium) . While the Sylvian foeea is appearing the anlage of the olfactory lobe is differentiated. Fig, 341, R, by the bulging forth of the lower anterior wall of the hemisphere,* and is soon marked oflf from the hemisphere proper by a distinct groove, the rhinal fissure of comparative anatomy. We find that the hemispheral vesicle is now divisible into three primary regions, which all persist throughout life; these are:

1. Tfie mantle (for detailed history, spe p, GflJ).

2. The basal ganglia (for detailed history, see p. 604-5).

3. Olfactory lobe (for detailed history, see p, 703).

The mantle outgrows the other parts and forms nearly the whole of the convoluted surface of the adult brain. During the fifth week the choroid plexus grows into the lateral ventricle, compare p. 681, and thereafter forms a conspicuous stnicture, but it is not part of the hemisphere in a strictly morphological sense.

d Includes tlu: lobus hippo

obi, medulla obloDgBts.

As stated above, the upward expansion of the hemispheres causes them each to have a medial wall, bounding the fissure in which the falx cerebri is develope<l. Along this wall is developed a fold, which is marked by an external groove and an internal ridge ; the groove is the Bogenfurche of German embrj^ologists, Fig. 395, a, 401, a. Fig. 391, bf; and is in part equivalent to the callosal fissure or groove of the adult, while the internal ridge is the anlage of the comu Ammonis. The Bogenfurche begins at the olfactory lobe, which it crosses, and divides it into an anterior and posterior part (see p. 703) ; the Bogenfurche then curves around, Fig. 401, a, parallel with the edge of the foramen of Monro. It begins to develop anteriorly and gradually extends further and further backward, until it is a long arching groove, terminating in the temporal lobe (lobus h ippocampi) . It must be remembered that the posterior end of the groove arises in reality by itself as the hippocampal groove (Ammonsfurche) but the two ends soon join, making one long fissure as described (W. His, 89.4, ()07). At its posterior end the groove forms two branches, each corresponding to a fold of the brain wall ; one branch is the anlage of tlio parioto-occipital, the other of the calcarine fissure. These three fissures (Bogenfurche and its two branches) and the S.v Ivian fissure are the only fissures which arise as folds of the brain. Mihalkovics, 77.1, has proposed for them the distinctive name of Total/ urchen (total grooves). All other fissures (sulci) are mei-ely depressions of the cortical surface, not folds of the brain-wall. When the corpus callosum is developed, p. G83, it gradually occupies by its enormous expansion most of the space under the Bogenfurche, so that the fissure (sulcus corp, callosi), is almost hidden above the corpus callosum in the adult. The internal ridge corresponding to the Bogenfurche, has, of course, the same arched course; it begins at the olf actor}' lobe, curves upward and backward around the foramen of Monro, and bending downward terminates behind the corpus striatum in the temporal region. Its course may be understocxl from Fig. 388 and Fig. 395. As to the fate of the frontal end of the ridge, we have no satisfactory knowledge ; the posterior end is the anlage of the liippoctunpus, the ridge corresponding to the main groove developing into the hippocampus major (comu Ammonis), and the ridge corresponding to the branch (sulcus calcarinus) developing into the hippocampus minor (calcar avis).

The three lobes (frontal, temporal, and occipital) of the adult are very gradually evolved. The first step in their differentiation is the development of the fossa Sylvise. The fossa may be recognized in a human embryo of five weeks. It seems to owe its origin to the fact that the brain-wall forming the fossa grows principally in thickness to produce the corpus striatum, while the mantle grows very rapidly in superficies; it follows that the mantle region expands and project^ beyond the thick- walled fossa. Fig. 397; the mantle at this stage forms a vesicular /row/a/ lobe^ -F, and a vesicular post-Sylvian lobe, each with thin walls and each including a portion of the wide lateral ventricle. The post-Sylvian lobe ]:)ecomes in part the temporal lof)e^ T, but it also expands toward the cerebellum, and its expansion forms the occipital lobe, Fig. 397, Oc, The frontal and temporal lobes may therefore be regarded as primary, the occipital lobe as a secondary or later acquisition.

Each lobe includes a portion of the lateral ventricle ; the portion in the frontal lobe becomes the anterior comu ; the portion in the temporal lobe the descending comu; the portion {recessus occipitalis) in the occipital lobe the posterior cornu. Now the Bogenfurche extends flown behind the fossa of Sylvius, therefore along the medial wall of the temporal lobe ; hence the inner ridge corresponding to the Bogenfurche projects into the ventricular cavity of that lobe; now the ridge is the anlage of the hippocampus major (comu Ammonis), which remains throughout life a ridge projecting into the descending comu. It will be recalled further that the Bogenfurche has a branch, the calcarine sulcus, Fig. 392, cat, which runs on to the medial wall of the occipital lobe, and has corresponding to it a ridge projecting into the ventricular cavity of that lobe ; this ridge likewise persists throughout life and is the hippocampus minor (calcar avis) of descriptive anatomy. The exact history of the modifications in the shape of the lateral ventricle during the foetal period has still to be worked out.

The fossa of Sylvius undergoes important modifications (compare Mihalkovics, 77. 1, 149). At the end of the second month the hemisphere in side view has a bean-like shape, the hilus facing downward ; the fossa is situated at the hilus. At three months the fossa is about as high as broad ; during the fourth and fifth months it becomes more sharply defined and has a marked inclination toward the occiput. The floor of the fossa corresponds to the corpus striatum and island of Reil; the brain- wall constituting the floor is very much thickened ; the external surface of the floor, which is seen when the brain is viewed from the side, is the so-called island of Reil. •Morphologically the island and the corpus striatum are parts of the same structure. During the sixth month the edges of the fossa begin to spread over the island and cover it in, so that by the ninth month it is entirely buried, and can be seen only by opening the Sylvian fissure.

The thickness of the walls of the hemisphere apparently increases throughout the second to ninth month. In the region of the basal ganglia the thickening takes place very early and becomes very great. The mantle thickens more slowly and never equals the basal ganglia in thickness.

The size of the hemispheres, as a whole, increases very rapidly for a long period, so that at birth they more than equal all the rest of the brain in volume. They cover first the thalamencephalon, later the mid-brain also, still later the cerebellum also. Owing to the growth of the cerebellum after the fifth month it is less completely covered by the hemispheres at the end than during the middle period of foetal life.

Foramen of Monro. — The foramen of Monro is at first, Fig. 337, a rounded opening, which soon becomes pointed at its lower side. As to its actual size in successive stages we have no measurements; it is converted into a fissure-like opening, and is commonly said to diminish in size, but I think it probable that the diminution is relative only, not absolute. A knowlftdge of the f<Bta1 history of the foramen would he a desirable addition to Embryology.

Mantle or Pallium. — The mantle comprises all that part of the hemispheres which enters into the formation of neither the olfactory \ohcs (rhinencephalon) nor basal ganglia {Bodentheil, Stamyiiiheil). Its general history we have already reviewed ; the ^development of its convolutions is treated in the next section ; wo have, therefore, to present only what little is known of the histogenesis of the cortex cerebri, the cortex being the superficial stratum of the mantle.

Histogenesis. — ^For the development of the nerve-cells, see p. 624. We have to add here what little is known concerning the development of the laj'ers of the cortex, following Vigiial, 88. 1, 24'J, who alBO gives, p. :J32, a summary of previous work. In a rabbit embryo of fourteen days the Randachleier is still thin, while the mantle layer with rounded nuclei and the inner layer with elongated nuclei have both grown very much. In later stages, Fig, ;Ui8, I find six main layers, the , * homologies of which with the layers, both of earlier and of adult stages, have still to be determined. The outermost layer is thin, 0, and con* tains \ory few nuclei; below is a broa<ler layer with the nuclei grou|>ed chieH> in radial lines, 5; this layer IS the anlage of the cortex cerebri. sentii f,frict)i, and is seen to consist of three strata; it is along the inner edge of this layer that the great pyramidal cells arise to form the third lajer of Meynert, while the rest of the la^Lr produces the second layer of M.e\ nert. Now if, na I hold to b© proballe, the large pyramidal cells ^^f are h )mo!ogous with the Purkinje ^H ctlls cf the cerebellum, then layers 5 "Id and ( (f Fig. IV.>S are derive*! from ' •" the original Randachleier. But in the present state of our knowledge another interpretation is equally possible — namely, that layers 1-4 are derived from the inner layer of the embryo, layer 5 from the mantle layer, layer (5 from the Randschleier.

For a comparison of the layers of the cortex in various air-breathing vertebrates see Nakagawa, 90. 1.

The raedullated nerve-fibres of the mantle do not appear until after birth, S. Fuchs, 84.1, 181.

Basal Ganglia. — The corpus striatum and the various parts associated with it arise from the thickened wall of the fossa of Sylvius. This thickening is continuous past the posterior side of the foramen of Monro with that thickening of the dorsal zone uf His, which produces the thalamus opticiiM of tlic median fore-brain, p. (iS6. The part of the thickening, which connects the corpus striatum proper with the median foro-brain, develops into part of the bocalled peduncles of the hemispheres of the adult; constituting what is termed by W. His, 89.4, 7C)U, the Streifenhiigelstiel. The commencement of the thickening may be observed in rabbit embryos of 12-i;( mm. (Mihalkovics, 77.1, 110), in the human embryo at four weeks; it necessarily coincides with the first formation of the fissure of Sylvius, The tbickeoing soon becomes (His, 89.4, 699) a considerable prominence; it is broad, forming what may be called the floor of the hemisphere; it stretches from the lamina terminalis and the anlage of the olfactory lobe across the fossa of Sylvius and behind the foramen of Monro, where it joins the anlage of the optic thalamus. Even at the beginning of the fifth week traces of the division of the corpus into three limbs can be detect<xi ; a lower limb {hinterer Scheiikel of His) runs to the lamina terminalis; on the upper limb {vorderer Schenkel of His) to the anterior olfactory lobe, and a middle limb to the posterior olfactory lobe. The middle and lower limbs together form th© cms Mediate, the upper limb the cnts laterale of descriptive anatomy — compare Fig. 39!), which well illu8trat<« the primitive form and subdivision of the arching corpus striatum. Later, in the same measure as the hemisphere expands toward the cerebellum, the posterior part of the corpus striatum grows. A groove, which persists into late fcetal periods, marks the division between the S("h™hir"Moni" ^("su" of-'iiS corpus and the optic thalamus; the u'*^'jofn^'to"h"pa™*u'btJ«raS™^ groove ultimately becomes obliterated, em,cniB mwiiaie of ihrconiuBatriaand the tis.sue which fills it up is the ^taiut"iu»:° R^^^r^^'un optkiLlt anhure of the stria cornea or fermina- £/■ 'T^^t^M"i!?i"' '""■ *"*^ ^' lis {tCEuia semictrcularts); His proposes, therefore, to designate the groove aa the .sulcus stria coniece. The origin of the stria cornea (Hornstrcif) wan discovered by Mihalkovics 77.1. 133.

Cerebral ConTOlutions. — We must divide the so-called fissures which produce the convolutions (gyri) of the brain into two classes, the primary folds and the secondary- fissures. The former are literally folds of the entire brain-wall, and were, therefore, appropriately termed " Tofatfurcfien" by Mihalkovics, 77. 1, H'i, who first clearly recongized them as a distinct class of fissures. The latter are merely narrow groove-like depressions of the surface of the hemispheres.

1. Primary Folds, — These are the fossa Sylvi and the Bt^enfurche; the latter has at its posterior end two branches, known as the calcarine and parieto-occipital fissures respectively. To these we ought possibly to add thejissura collateralis, p. 701, which is situated on the lower surface of the temporal lobe.

The fossa of Silrius, as already stated, jt. I't'J'i, is at first a wide, shallow depression, which gradually deepens. The wall of the depression is verj' much thickened to make the corpus stria^fim ; the outer part of the wall makes, of course, the floor of the fossa, and

this floor becomes the island of Rail in the adult. The growth and expansion of the mantle cauisea it to project farther and farther, thereby deepening the fossa. At four months, Fig. 397, S, it ia very wide and aacends backward between the frontal, P, and temporal, T, lobes. At the b^in' ning of the fifth mouth _ ,, f / r I \ / \i 1/ (Minalkovics, 7 7.1,

f\ /C / (. J L )L X / 150) the fosea ^^^ be 3 4 5 6 7 s 9-IO come deeper, longer, and

Fig. -III"!.— Oullintnof theFiMure of SflTiim of Human Em- moreobllque, and its anbryo, « Bu««.l™ Lu«ar Hontha Afwr G, M.h.lkovica. ^^.;^^ ^^^^ -^ ^^^^^^

by an angle, Fig. 4(m*, 5. the two margins gradtmlly approach one another, concealing the floor of the fossa or island of Reil, and meanwhile the angular notch of the anterior margin becomes more marked. The changes continue until, as shown in the figure, the opening of the fossa is a narrow Y-8hai>e«.l cleft, 0, leading down into the fossa proper ami the island of Reil. The walls of the fossa of Sylvius, including the island, accjuire secondary furrows during the niutb month. The part of the margin of the fissure between the two forks of the Y is sometimes termed the operculum . The Bogeiifurche, or fissura prima, arises very early. Its anterior part appears first, beginning as 8tate<l above, p. 092, at the olfactory lobe, thenc« passing along the medial wall of the hemisphere in a cnr\ed line which maj be roughlv described as parallel with the lamma terminalis and t«la choroidea," Fig. 401, a. The pts terior part of the Bogenfiirche ap pears later; it corresponds to the Ammon.ifurche of Mihalkovici 77.1, 145 (sulcus hippocaiiijii of Huxley) ; it begins on the medial wall of the temporal lobe, and gradually extends upwaM and for ward until t^jward the latter part of the second month it joins the anterior part, and the union of the two produces the great Bogen- .fi'S^^^^,^^^,'S.MZ,^'S^^Si furche, which begins at the olfac- fSI^,!*'j."*'optt ,hJ,uSlu"jSriiu?bu8^"; tory lobe, runs widely arching toiriuB:'/„/;infundibuiuiii. 'Afwrw. His. x along the medial wall, and temii- ' "'" '*"^ nat«s at the lobus hippocampi, p. 091 . Corresponding to the external groove is an internal ridge, the ridge persists in the posterior part as the hippocampus, but its fate in the region of the frontal l()be is obscure. Below the ridge is a strip of the hemispheral wall, the Randbogen {(/yrus aicHatns) of F. Schmidt, 63.1. In the adult a large part of the Randbogen is occupied by the verj' large corpus callosum, above which persists the Bogenfurche as the callosal groove. The portion of the Randbogen immediately behind the callosum develops during the first half of the fifth month little transverse ridges upon its surface, and thereby becomes the recognizable anlage of the gyrus deniatus (Mihalkovics, 77.1, 147). The extreme posterior end of the Randbc^^n is bent upon itself, hook -like, and is easily identified as the anlage of the </ynis iniciiiatHS.

From the posterior part of the Bogenf urche run out, both from nearly the same point, its two branches, the fissura calcarina and the fissura parieto-occipitaliSj Fig. 402, a, 6, which both appear while the occipital lobe is growing out, the calcarine fissure, a, usually, but not always (His, 74.1, 114), arising before the parietooccipital, &, which last first develops at the beginning of the fourth month (Mihalkovics, 77.1, 146). Both fissures run upward and backward on the medial wall of the hemisphere, and as they diverge they enclose a space between them, which corresponds to the socalled cuneate lobe of the adult. To these two branches of the Bogenfurche correspond internal ridges (c/. His, 74. 1, Fig. 113), but the ridg ? corresponding to the parieto-occipital fissure is subsequently obliterated as the brain wall thickens, while that corresponding to the calcarine fissure persists, and, as indicated by its name, becomes the calcar avis, or hippocampus minor, p. 693.

During the seventh month the parieto-occipital fissure extends beyond the medial wall on to the external wall of the hemispheres, and by its extension establishes the life-long boundary between the

{)arietal and occipital lobes. The anterior boundary of the parietal obe is the fissure of Rolando, see below.

2. Secondary Furrows. — These, as defined above, are merely grooves upon the surface, not folds of the walls, and they have, therefore, no corresponding internal ridges on the ventricular side of the brain- wall. We may conveniently divide them into main or essential fissures and accessory or unessential fissures.

The MAIN FISSURES may be enumerated as follows :

1. Calloso-marginal or splenial.

2. Fissure of Rolando.

3. Fissures of the frontal lobe.

a. prsBcentral.

b. superior frontal.

c. inferior frontal.

d. olfactory or rectus.

e. tri-radiate.

f. internal frontal.

4. Fissures of the parietal lobe.

a. Intraparietal.

b. retrocentral.

5. Fissures of the temporal lobe.

a. superior temporal.

b. inferior temporal.

c. occi pi to-temporal or collateral (compare 6, d). G. Fissures of the occipital lobe.

a. ascending perpendicular.

b. superior occipital.

c. inferior occipital (sagittal).

d. occipito-temporal (compare 5, c).

7. Fissures of the island of Reil. a. central, fo. praBcentral. c. postcentral.

The primitive type of the fissures and of the convolutions between them is marked in the adult by the accessory fissures, which join the primary fissures or arise from them, and also by secondary bridges by which two adjacent convolutions are connected with one another across a fissure.

The calloso-marginal or splenial fissure. Fig. 402, e, arises about the middle of the fifth month, in front of and above the corpus callosum, cc, by the fusion of two or three shorter fissures ; the area of the hemispheral mantle between the calloso-marginal fissure and the corpus calloeum is the gyrus fornicatus. Behind the main tiasure, , are several subsidiary fissures which vary considerably in different bmins in both number and arrangement; they appear usually to unite with the calloso-marginal fissure, which is thus prolonged further back above the corpus callosum, cc, and usually the added secondary fissures caase the calloso- marginal to terminate posteriorly with an upwanl turn, a short distance beliind the upper end of the fissure of Kolimdo.

Flo. «>3.— Bri

The development of the fissure of Rolando has been carefully studied by D, J. Cunningham, 90.1, whose account is as follows: There is some variability in the time at which the fissure makes itet appearauce. The more u»ual time is the last week or ten days of the fifth month, but it is not uncommon to meet with hemisphores well on in the sixth month of development with no sign of the fissure. As a general rule, the fissure of Rolando is developed in two separate and distinct pieces, Fig. 403, Bo'. Ho". The lower portion appears in the form of a shallow oblique groove, which represents the lower twothirds of the fully-formed sulcus. It always makes its appearance before the upper piece. Its lower end is place<l close to the coronal suture — perha[>s, indeed, it may lie immediately subjacent to the suture — while the upper end lies further back, and reaches n point midway between the upper margin of the hemisphere and the Sylvian fossa. The upper piece of the fissure makes its appearance in the form of a deep pit or depression between the upper end of the lower portion and the margin of the hemisphere. An eminence separates the two portions of the fissure from each other. Soon, however, a faint furrow runs over the siunrait of this elevated intervening piece of the cortex, and the two primitive portions of the sulcus are i>artially unite<l to each other. As development (?oe8 on the more complete dix-s the union become, and t!ie more fully is the intervening eminence borne down into the bottom of

the fissure. As a rule, the confluence takes place rapidly, but in many cases the process appears to be retarded. Among my specimens I have several hemispheres which, although close upon the seventh month, show still a complete severance of the two constituent elements of the furrow. But the portion of cerebral cortex which intervenes between the two parts of the fissure is not entirely obliterated. It disappears from the surface, it is true, but it is still to be discerned, even in the adult brain, in the bottom of the fissure, in that shallowing or deep annectant gyrus which we have described at the junction of the upper and middle thirds of the sulcus. In some rare cases, as stated by Cunningham, the two original portions of the fissure of Rolando remain quite distinct throughout life. In these the intervening bridge of cortex remains on the surface, and is not pressed down by the fusion of the upper and lower divisions of the fissure. We have noted that the same deep annectant gyrus may be observed in the fissure of Rolando of the chimpanzee and

Fig. 408. — Rif^ht HemisphtTo, Natural Size, of n Fnptus of nearly soven Months, ip, Interparietal fissure partly formed; Ro', upper, Ro\ lower piece of flsRure of Kolando; Pre. 8^ superior prsocentral fissure; sup./, superior frontal : Prc.i, inferior prsBcentral; .V, Sylvian fissure; Temp. ft, temporalis superior; Temp.i, tem]K>ralis inferior; Ext, external i)erpendicular fissure of Bischoff. After D. J. Cunningham.

orang. We may assume, therefore, that the interrupted form of development of this sulcus holds good among the anthropoid apes as well as in man. With regard to the lower apes, we have no evidence one way or the other. The development of the fissures in the brain of the ape is still virtually unkno^vn ; and if we examine the bottom of the fissure of Rolando and the other primary furrows in a low ape, we find a uniform depth througliout, and an absolute absence of deep annectant gyri. It is dangerous to argue from ihe adult condition alone, but still the appearances are such as would lead us to infer that the continuous and not the disrupted form of development of the primary fissures holds good among the lower apes. The lower end of the fissure of Rolando is sometimes lengthened out by union with a small accessory fissure (fissure of Ober

Btaller, " Daa Stirnhim," 1890) so as to be prolonged to the fissure ot Sylvius. The inferior genu of the fissure of Rolando appears usually about the seventh month and always before the superior genu, in the lower piece of the lissure ; the superior genu is developed at the junction of the upper and lower piece. From the seventh month onward the convolution (posterior central), behind the fissure grows more rapidly than the convolution (ascending frontal) in front of it.

The fissure of Rolando was first so named by Leuret in 1 8;t9 (" Anat. comp. du Hysteme nerveux") ; in Germany it is usually termed the central fissure. It is the now accepted division between the frontal and parietal lobes. Next to the central or interhemi spheral fissure and the Sylvian fissure it is the most important landmark in the topography of the human cerebrum. It is, however, not a primary or essential fissure throughout tliose mammalia having ccmvolutions. The Fissures of the Frontal Lobe. — The prie-central arises generally toward the end of the sixtli month, and, therefore, some time after the fissure of Rolando, but not invariably, for it has been obser\*ed to precede the fissure of R4>lando. see D.J.Cunningbam, 90.1,P1. 1.,Fig. 1 ; it can 1)0 identified by its ijosition, it lying in front of the parietal boue, which covers the fissure of Rolando. It runs nearly parallel with the fissure of Rolando, and arises from two pieces (Cunningham, /.c, p. 7, tx)mpare Fig. 4I):{, Prc.i, and Prc.s), which usually remain distinct but are sometimes united. The development of the superior and inferior frontal JissHren is obscure. If the brain be viewed from below. Fig. 404, the lower surface of the frontal lobe offers at five to six months three depressions, which I have found to mil Month, cbi. be remarkably Constant. One of these nen"?*^foi(ncl '^ tlio sulcus vpctiis ov olfactoriits, (fympaw Fig. ill which the olfactory hulhus, 01, is lodged. The other two are small; they unite later with one another, and forming branches give origin to the inappropriately named fri-radiate Jis.iure. I have to add here what may be called a new fissure, which ajjpears not to have been hitherto generally recognized, although so far as my uncompleted observations go, it seems very constant Ixitli in embryos and adults ;* I name it the internal frontal fissure; it is situated on the medial wall of the frontal lobe, Fig. 402, /, and runs approximately parallel with the calloso-marginal fissure; it divides the marginal or si> called first frontal convolution into two parts; if the conclusion that the internal frontal fissure is primary and constant be verified, it will 1)6 necessary to subdivide the marginal convolution as now definetl.

pis. vi. and vli,, unJ iu «.-Vrnil m-ll-Liiuiiii teJU-buoks It is clearly flKurwl.

Human Eniln-vn ol i)i (Vrvbellum ^ .f/'I. nmli iDrundlhulum: <>ii. oji

The Fissures of the Parietal Lobe. — The intra-parietal fissure arises (Mihalkovics, 77.1, 154) as two limbs during the sixth month ; one limb is parallel with the fissure of Rolando, and not far behind it ; the other limb has a more longitudinal course and lies not far from the median plane; during the eighth month the two limbs unite. During the seventh month, according to Mihalkovics, Z.C., a retrO'Central fissure appears between the ascending limbs of the intraparietal and the fissure of Rolando. Along the median wall of the parietal lobe extends only the calloso-marginal fissure, Fig. 402, e, and its branches, c.

The Fissures of the Temporal Lobe. — During the sixth month there appears the superior temporal fissure on the external surftice of the lobe and parallel with the adjacent margin of the great Sylvian fissure, compare Fig. 403, Temp. s. Usually somewhat later appears another fissure, the inferior temporal^ immediately below and parallel with the last mentioned ; the second fissure. Fig. 403, Temp, i., is often discontinuous. On the lower surface of the lobe is developed during the sixth month also the great occipito-temporal fissure, the fissura collateralis of Huxley; this fissure varies greatly in length; it normally extends far into the occipital lobe, hence its name, and sometimes runs so far forward as to border the gyrus hippocampi. The collateral fissure is very deep, and there is a projection on the inner side of the brain corresponding to it, and which is known as the eminen tia collateralis of Meckel ; this fissure ought, perhaps, to be classed with the primary folds, p. 605. The collateral fissure, according to D. J. Cunningham, 91.2, 344, is continued forward in the middle foetal life by the incisura temporalis and the limiting fissure of the insula Reilii ; these three grooves may be taken as making the limits of the temporal lobe, but in later stages tho originally evident relations of the three grooves to one another become obscured. During the ninth month of foetal life an accessory transverse fissure on the under side of the temporal lobe unites with the limiting fissure of the insula, and therefore in the adult the fissure appears to have changed its primitive course.

The Fissures of the Occipital Lobe. — This lobe has three surfaces, an inner or medial, an external, and a lower or cerebellar. On the medial surface the lobe is bounded anteriorly b}' the parieto-occipital fissure, Fig. 402, 6, and shows the calcarine fissure, a, the origin of both which is described p. 697. The area between these two fissures is the cuneate lobule (Zwickel). On the external surface the first fissure to appear is a small, short one. Fig. 403, Ext^ the ascending perpendicular of Bischoff, 68. 1 , 447 ; the horizontal limb of the intraparietal fissure extends on to the occipital lobe, and probably joins the fissure of Bischoff; the prolonged intraparietal is known as the superior occipital fissure. Later (eighth month) arises lower the longitudinal inferior occipital (sulcus sagittalis). On the lower surface during the sixth month appears the great occipital temporal fissure, which, as stated above, also belongs to the temporal lobe. The data of this paragraph are chiefly from Mihalkovics, 77. 1, 155.

The Fissures of the Island of Reil. — The best account of the adult fissures of the insula is probably that of Oberstaller [An at. Anzeiger, 1887, p. 739). He finds four vertical fissures: the first and second prsBcentral, the central, and the post-central, as they may be called; as the second prae-central is small and insignificant, it may be regarded as accessory. Their development has been studied by D. J. Cmmingham, 91.2. In the latter half of the fifth month the central fissure (sulcus centralis insulce) becomes evident as a faint linear furrow which runs upward and backward from the lower part of the Sylvian fossa ; from the very first it lies accurately in the line of the fissure of Rolando, and it appears at the same date ; it is situated much nearer the hinder end of the insula than at later stages, owing to the growth of the posterior part of the insula. The first prcv-central fissure is developetl a little later, but as a general rule before the end of the fifth month, and lies accurately in line with the sulcus prae-centralis inferior of the frontal lobe ; during the last month of foetal life its upper end generally moves forward to a slight extent so that its relation to the frontal prae-central is marked ; it is remarkable that for a certain period the prsB-central fissure is l)etter marked than the central, but during the eighth month it loses this pre-eminence. Guldberg {Anat, Anzeiger^ Oct., 1887) mistook the prsB-central for the central fissure. The post -central fissure is much later in making its appearance. As a rule, it does not show until the middle of the sixth month or even later ; its development coincides with that of the intraparietal fissure, the line of which it prolongs.

The remarkable coincidence of three main fissures of the island of Reil with the lines prolonging respectively the prae-central inferior, the fissure of Rolando, and the intraparietal necessarily suggests that the insular fissures are parts of the same fissures as those of the mantle enumerated.

The Accessory Fissures. — Beside the main fissures the human brain has a large number of short fissures of an irregular and variable character, and which modify and mask the primary fissures to a variable extent. These accessory fissures appear during the last month of foetal life, for the most part as branches of earlier fissures, but in small part as independent grooves. Whether or not other fissures are developed after birth I do not know. The laws governing the appearance of the accessor}'^ sulci have not yet been ascertained.

3. Transitory Fissures. — The question is still under debate as to whether there are in early stages of the foetus temporary folds or not. Bischoflf, His, and others, with whom I am strongly inclined to agree, consider the irregular folds, which are often to be observed on the surface of the cerebrum from the first to the fourth month, as artificial and accidental. On the other hand, KoUiker, Ecker, Mihalkovics, 77.1, 144, and others, consider that the folds are normally present.

4. Evolution of the Fissures. — It is well known that there are several types of convolutions, and that different fissures are typical of different orders of mammalia. It is probable that all the fissures (/. e., secondary furrows) of the human brain were evolved \vithin the series of primates, and it is doubtful whether they are any of them homologous with the fissures in other mammalian orders ; compare Sir Wm. Turner, 90.2.

5. Historical Note. — Our knowledge of the fcetal fissures and convolutions was very slight until Reichert ('* Der Bau des menschlichen Grehims," 1859, 76-90). More thorough were the valuable memoirs of Bischoflf, 68. 1, and Ecker, 83. 1. The mechanical factors concerned in the production of the convolutions have been discussed by His, 74. 1, 110-117. D. J. Cunningham has made, 90. 1, 91.2, important additions to our knowledge of the development. Of the anatomical papers on the convolutions man}" are of morphological value; among them Sir Wm. Turner's address, 90.2, is of the first value to the embryologist. Of the knowledge of the subject up to 1877, Mihalkovics gives an admirable summary, 77.1, upon which I have dra\vn freely.

Olfactory Lobes (Riechlappen).— The following account is based upon the researches of His, 89.4. The olfactory lobe arises by differentiation of an area of the wall of the primitive hemisphere. The differentiation begins in the human embryo during the fourth week as the hemispheres begin to enlarge, and affects the aren adjoining the median lamina terminalis, compare Fig. 339, 01. The olfactory area, as it may be called, expands with the hemispheres, and thus soon extends well forward in front of the lamina terminalis ; it then constitutes ja slight longitudinal ridge with a corresponding internal groove, along tha under side of the cerebral hemisphere. Fig. 341. The area now appears as a fold of the hemispheral wall. There now develops the primary groove {primdre Bogenfurche) p. 692, and this extends aot only in an arch along the medial surface of the hemispheral wafi but also curves on to the olfactory ridge, and by crossing it transversely divides the ridge into an anterior and a posterior segment, Fig. 399. The ridge next separates from the hemisphere, so as to be converted into a blind tubular diverticulum, which remains connected at its posterior en<l with the hemisphere, and we now have an olfactory lobe, which has a central cavity in direct communication with the lateral ventricle ; the lobe has two segments, one posterior connected with the brain, the other anterior and comprising a narrower part or stalk, and an enlarged end ; the stalk is the anlage of the tract us olfactorius and trigonum; the enlarged end is the anlage of the bulbus olfactorius. The posterior segment becomes the posterior olfactory lobe, a part of the brain which has been long imperfectly recognized ; it comprises the pedunculus corporis callosi {or gyrus subcallosiis of. Zuckerkandl), the outer and inner roots of the olfactory nerve, and the substantia perforata anterior.

The olfactory ganglion of the embryo unites with the bulbus olfactorius. The union takes place during the latter part of the fifth week, Fig. 405. The origin of the olfactory ganglion is described p. G37. It grows upward, and as during the fifth week the end or bulbus of the olfactory lobe bends toward the median line, Fig. 405, the ganglion lies close behind the bulbus, in the groove which may be regarded as the prolongation above described of the Bogenfurche or primary fissure across the lobe. The ganglion now spreads around the bulbus, and unites with it, as may be seen on the right side in Fig. 405, and forms a superficial layer over the surface of the bulbus, which thus has three layers — the outer ganglionic layer, the neuroglia layer corresponding to the Randschleier, and the inner nucleated layer corresponding to the inner layer and mantle lay^r of the spinal cord. The transformation of thene into the adult layers has still to be worked out. Owing to the fusion of the ganglion with the bulbus there can be no nerve-trunk running from the ganghon to the brain; the centrifugal fibres from the gangliolic layer run off in bundles, which fonn I plexus-like network on their way to be distribut«d to the olfactory epithelium covering the upper turbioal fold (obere Muschcri. The fusion of the ganglion with the bulbus explains why the olfactory fibres appear in the adult to arise from the wall of the lobe, although not medullary nerve-fibres.

During the second month the hemispheres expand so rapidly that they carry the base or posterior part of the olfactory lobe forward, while the bulbus remains attached to the ganglion so that oi^^ nfni^'^n'o ^^t" ^^iy"iZ ^^^ bulbus at the end of the secood weekH (His flch) 01 anwrior oc. month is bent bsck and lies under the Kf^VSlto?y "ITokIi™ ^"^ ™™1 posterior segment, but during the third tuber c^rcuiii'^lfiBr w' H™* '^' mouth the bulbus bends forward again and assumes its permanent position. During the third month also, the anterior half of the lobe lengthens out and becomes clearly differentiated into bulbus, tractus, and trigonum. The cavity of the olfactory lobe becomes obliterated in great part before adult life, but exactly how or when is not known.

Evolution of the Head. — We are now in a position to review briefly the factors which have determined the differentiation of the head. The conception that the head is composed of a number of segments has now been current for nearly a centurj-. For a long time the attempts to determine the number of cephalic segments were confined to the study of the skull, following Oken's idea that the skull ia composed of a number of vertebrie. We have already seen, p. 4fi9, that all such attempts were necessarily fruitless. A great advance was made when Gegenbaur, in I8T2, sought to determine the segmental value of the cranial nerves, compare p, 4fi9, but the correct and only way was pointed out by Balfour, who sought to detennine the number of actual segments in the embryonic head, compare p. 100. Van Wijhe found of the true myotomes at least nine in shark embryos, and Dohm has found in a very young stage of the torpedo about twice that number, compare p. '.JOO and Fig, 118. It has thus been proved that the head is a segmented region in which the majority of ihe segments abort very early in the embryo. The next step must be to ascertain what causes have restjt^ in, and what effects have resulted from, the disappearance of myotomes in the head. The first thing to indicate the formation of the bead is, in the embryo of all classes of vertebrates, the dilatation of the medullary tube to form this brain, a dilatation which crowds the mesoderm down on the ventral side of the neural tube. I think also that the enlargement of the brain is the direct cause of the fonnation of the head-bend, and that probably the proamniotic area has the role to play of preventing the directly forward growth of brain, because there being no mesoderm in the proamnion the entoderm and ectoderm are united and the head cannot develop across the area, and consequently bends to allow the elongation of the cerebral vesicles. The head-bend still further crowds the myotomes, compare Fig. 1 1 8, and it is to this crowding that the abortion of the myotomes is to be attributed, according to my hypothesis.

The effects of the abortion of the cephalic segments have been to prevent the development of the primordial skeleton into separate vertebral masses, and to prevent the development of the cranial nerves on the type of those of the spinal cord. It must, however, be admitted that the correlation between the arrangement of the nerves and the development or abortion of segments is very obscure.

Another factor in the evolution of the head, which enters into action much later, is the development of gill pouches with their resultant modifications of the gill-arches, formation of the branchial skeleton, etc. Noteworthy is the fact that the coelom of each arch connects a myotome with the splanchnocoele (pericardial ciivity) and is apparently homologous with the nephrotome of a rump segment. If this homology is correct we must describe it as a further peculiarity of the head that its nephrotomes give rise, not to excretory tubules, but to branchial striated muscles, see p. 478.

A third factor which comes into play still later is the annexation to the occipital region of at least four true cervical (hypoglossal) segments with their vertebraB and nerves, compare p. 429 and Ofjri.

The skull plays a subsidiary part and is an accessory structure added after all the essential morphological cliaracrteri sties of the head are present. The erron(?ous notion that the skeleton is the framework upon which the body is built has been discardwl by embryology. That the organs of special sense have had a profound influence on the head during its evolution cannot Vxj doubted, but, while we put down the possession of the ol factors, visual, and auditory organs as essential chanicteri sties of the head, we cannot say, so far as we can recognize at present, that they have influenced the constitution of the head nearlv as much as the other factors.

We must for the present define the head as the anterior region of the body, in which the medullar}' tube is enlarged, the segments consequently aborted, and the skeleton therefore not divide<l into vertebrae, nor the nerves \vith dorsal and ventral roots united; which possesses the three organs of special sense; in which the gillclefts are developed ; and which has increased its original territory by the annexation (at least in amniota) of several cervical segments.

Historic Disclaimer - information about historic embryology pages 
Mark Hill.jpg
Pages where the terms "Historic Textbook" and "Historic Embryology" 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 and interpretations may not reflect our current scientific understanding.     (More? Embryology History | Historic Embryology Papers)

Human Embryology: Introduction | The Uterus | General Outline of Human Development | The Genital Products | History of the Genoblasts and the Theory of Sex | The Germ-Layers | Segmentation | Primitive Streak | Mesoderm and the Coelom | Germ-Layers General Remarks | The Embryo | The Medullary Groove, Notochord and Neurenteric Canals | Coelom Divisions; Mesenchyma Origin | Blood, Blood-Vessels and Heart Origin | Urogenital System Origin | The Archenteron and the Gill Clefts | Germinal Area, the Embryo and its Appendages | The Foetal Appendages | Chorion | Amnion and Proamnion | The Yolk Sack, Allantois and Umbilical Cord | Placenta | The Foetus | Growth and External Development Embryo and Foetus | Mesenchymal Tissues | Skeleton and Limbs | Muscular System | Splanchnocoele and Diaphragm | Urogenital System | Transformations of the Heart and Blood-Vessels | The Epidermal System | Mouth Cavity and Face | The Nervous System | Sense Organs | Entodermal Canal | Figures | References | Embryology History

Cite this page: Hill, M.A. 2017 Embryology 1897 Human Embryology 27. Retrieved October 22, 2017, from https://embryology.med.unsw.edu.au/embryology/index.php/1897_Human_Embryology_27

What Links Here?
© Dr Mark Hill 2017, UNSW Embryology ISBN: 978 0 7334 2609 4 - UNSW CRICOS Provider Code No. 00098G