Template talk:Ref-Kupffer1891a

From Embryology

The Development Of The Cranial Nerves Of Vertebrates

von Kupffer C. The development of the cranial nerves of vertebrates. (1891) Translated by Oliver S. Strong J Comp. Neurol. 1: 246.

Professor C. von Kupffer.

Gentlemen :

There is scarcely need of the assurance that I do not stand before you with the pretense of presenting the comprehensive subject of my report with perfect sj'mmetry on all sides. Tliat would only be possible if the development of the vertebrate head as a whole could receive thorough treatment at the same time. The works which relate to the cranial nerves are still issuing, and it is difficult to predict, up to the present, how the course of these works will influence the solution of the problem of the vertebrate head.

In consideration of the time here allotted to such a report, I limit my remarks to a condensed summary of the present trend and results of research in relation solely to the morphogeny of the cranial nerves, apart from their histogeny and from the remaining phases of the problem of the head.

So much may be said in advance: that in the formation of the vertebrate head very remarkable reductions and fusions of endodermal and mesodermal parts, and, in connection therewith, of the peripheral nervous system, extending in direction from before backwards, have occurred. And, further, I venture to assert that, besides a branchial system

I Report read at the meeting of the Anatomical Society at its fifth annual session, at Munich, May i8, i8qi. Translated for this journal, from advance sheets, by Oliver S. Strong, Fellow in Biology in Columbia University.

V. KuPFFER, Cranial Nerves of Vertebrates. 247

of cranial nerves, sharply distinguishable genetically from the spinal system and belonging to the gill apparatus, a spinal system of dorsal nerves has likewise more or less completely maintained itself.

The old theory of the spinal nature of the cranial nerves, dating its beginnings from Proschaska and Sommering, received weighty support through the Goethe-Oken vertebrate theory of the skull; it outlasted the latter doctrine, and appeared until recently consistent with Bell's law. But the embryological works of the last decade have so severely shaken the foundations of this doctrine that so decided a defender of it as C. Gegenbaur(') found himself obliged to admit that the homodynamy of the cranial and spinal nerves appeared to him no longer tenable. The first shock this theory received was through the observation of Balfour,( ') on elasmobranch embryos, that the mixed cranial nerves are decidedly of dorsal origin, and they are also to be distinguished essentially from the dorsal roots of spinal nerves owing to the motor elements in them.

While Balfour did not succeed in finding in the region of the cranial nerves any roots whatever which might be compared with the ventral roots of the spinal nerves, he, on the other hand, did not doubt the complete validity of Bell's law, and accordingly advanced the hypothesis that there had existed originally a common fundamental form of the nerves in the whole body which are only represented by the dorsal roots of a mixed nature. At this stage the diflerentiation of head and trunk took place, so that the type of the spinal nerves corresponding to Bell's law would be regarded as secondarily acquired, while the cranial nerves have preserved the original condition. This hypothesis appeared to Balfour so much the more probable in that he

1 "The Metamerism of the Head and the Vertebrate Theory of the Cranial Skeleton," Morph. Jahrb , XIII, p. 64 and 104.

2 "The Development of the Elasmobranch Fishes," Journal Anat. and Phys.,Vol. XI, 1877, and Handbook of Comp Anat , trans, by Vetter, Jena, i88r, Bd. II, p. 411.

248 Journal of Comparati\e Neurology.

decided, (') in opposition to the observations of Stieda() and Schneider, () that in Amphioxus ventral motor roots were lacking.

This hypothesis appeared untenable, since, on the one hand, ventral spinal nerves of a motor nature are found in Amphioxus, and, on the other hand, the development ot ventral nerves was observed in the region of the head, as, for example, the origin of the abducens from the ventral aspect of the medulla in the chick, as well as in Scyllium, by Marshall.

But, aside from the hypothesis, the fact remains that the undoubtedly dorsal cranial nerves do not conform to Bell's law, but are of a mixed nature.

Van Wihje's(^) treatise, appearing soon after the works of Balfour above mentioned, brougnt to light the new disclosures, known to you, concerning the ontogeny of the cranial nerves in Selachians as proof, on one side, that at least two pairs of muscle nerves in the head, the oculomotor and abducens, appear after the manner of the ventral spinal nerve roots, and, like these, innervate only muscles arising, as alleged, from somites, to which the eye-muscles should belong; while the motor elements of the cranial nerves arising dorsally are confined to the innervation of the muscles of the visceral arches, which proceed from the lateral plates. Relying upon this observation. Van Wihje impeached the validity of Bell's law, and showed the possibility that the relations discovered for the head have further application in the trunk, as in the enunciation of Bell's law the vegetative muscles in the t unk proceeding from the lateral plates were left out of consideration.

While Van Wihje sought in this way to establish the

1 " On tlie Spinal Nerves of Amphioxus," Quart. Jour. Mic. See, Jan., 1880, p. 90.

2 "Studies on Amphioxus Lanceolatus," Mem. d. I'Academ. d. Sc. de St. Petersbourg, 1873, I. 46.

3 " Contrib. to the Comp. Anat. and Devel. of Vert.," Berlin, 1878, p. 15.

4 " (Jn the Mesoderrasegments and the Development of the Nerves of the Selachian Head," Amsterdam, 1882.

^■. KuPFFER, Cranial Nerves of ]'ertebratcs. 349

agreement between cranial and spinal nerves, he himself discovered at the same time facts w^hich cause the distinction between both groups to appear more marked than was hitherto supposed. On one side, he shows that the dorsal cranial nerves in respect to their course bear relations to the somites entirely different from the spinal nerves, and it furthermore emerges from his investigations that the epidermis participates in the development of the peripheral twigs of the cranial nerves, as G6tte(') and Semper(-') had already affirmed for the N. lateralis vagi.

Concerning the relation to the mesoderm, the dorsal cranial nerves first take their course over the mesoderm and on between the somites and the epidermis, but the dorsal roots of spinal nerves pass within and below the somites. This distinction is so important that it was especially by this that C. Gegenbaur was driven to abandon the homodynamy of cranial and spinal nerves, which he had previously defended.

According to Van Wihje's discovery, the facial, glossopharyngeus and vagus enter into close connection with the epidermis in a double row of points. On the one hand, fusion of their ventral branches occurs at the upper hinder side of the gill cleft which lies cephalad; on the other hand, the rudiments (Anlage) of their dorsal branches fuse with thickened portions of the epidermis. From the dorsal places of connection are developed portions of the lateral -line system with its nerves, in such a way that Van Wihje thought the participation of elements of the epidermis in the formation of the nerves might be regarded as certain. P'rom the places of fusion of the ventral branches with the epithelium of the gill clefts the terminal twigs of those branches are developed.

1 " Embryology of the Toad."

2 "Urogenital System of Plagiostomes," Arbeiten aus d. Zool.-Zoot. Institute zu Wurzburg, Rd. II, p. 398.

250 Journal of Comparative Neitrology.

A. Froriep(') has pointed out the part played by the epidermis in the development of the cranial nerves in mammalian embryos, where, in comparison with the condition discovered by Van Wihje in Elasmobranchs, the share of the epidermis in this process seems scarcely less important. In cow embryos such processes take place in the region of the facial, glosso-pharyngeal and vagus, whose course is briefly as follows: In embrys 8-9 mm. long with three gill furrows apparent outside and four pharyngeal pouches, the spindleshaped ganglion of the facial unites with a pit-like, sunken and at the same time greatly thickened place in the epidermis, which corresponds to the dorsal extremity of the first external gill furrow. The same thing occurs with the glossopharyngeus; here also the distal end of the ganglion in the rudiment of this nerve unites with a greatly thickened and depressed portion of the epidermis dorsad to the second pharyngeal cleft. The vagus rudiment behaves similarly. Its large ganglion locates itself adjacent to a thickened epidermal surface, of the shape of a figure 8, which is situated dorsad to the third gill furrow and the vicinity of the fourth.

In cow embryos 15 mm. long the connections of the facial and glosso-pharyngeus are lost, but not of the vagus, and here the thickened mass of epidermis reaches deep into the ganglion, so that the two components cannot be definitely separated from each other.

Froriep compares these " organs of the gill clefts," though with reserve, to the connections, described by Van Wihje, of the ventral branches of the same nerves with the epidermis, dorsally to the gill clefts, and explains them as rudiments of sense organs which do not attain further development, and probably belong in the category of organs of the lateral line but with which they cannot be fully homologous, since the lateral organs of fishes are formed on the dorsal branches of the cranial nerves. As those ganglia which enter

1 " Rudiments af Sense-organs in the Facial, Glosso-pharyngeal and Vagus," Arch. f. Anat. u. Physiol., Anat. Abth., 1885.

V. KuPFFEK, Cranial Nerves of Vertebrates. 251

into the connections under discussion and accordingly originate through a union of central with epidermoid elements, Froriep indicates the gang, geniculi, gang, petrosum and gang, nodosum; they exhibit the phylogenetic remnants of former sense organs; they cannot consequently be considered homologous with the spinal ganglia, and thereby fails a chief support of the spinal hypothesis. The nerves of the visceral arches as segmental nerves can no longer be identified with spinal nerves.

In a long series of works J. Beard(') has handled the problem of the cranial nerves and their sensory end organs. Torpedo ocellata furnished the principal object of Beard's investigations, although sharks, teleosts. Amphibia and Amniota (Lacerta, chick) came under consideration. In the main. Beard follows, in facts as well as in the interpretation of the conditions, the observations and conceptions of the previously mentioned investigators, but also contributes the fact that the participation of the epidermis in the formation of ganglia and nerves occurs also in the region of the trigeminus.

The complete representation of the development of a dorsal cranial nerve would, according to him, be as follows: The rudiments of the ganglia in the head, agreeing throughout with those of the spinal nerves, arise as differentiations of the inner layer of the ectoderm, just outside the limits of the neural plate. They separate from the ectoderm, become displaced upwards in the closing of the neural tube, and come to lie between its lips, but are always distinguishable from it. After the closure of the neural tube, the portion of

I On the Segmental Sense Organs of the Lateral Line and on the Morphology of the \ertebrate Auditory Organ," Zool. Anz., 1884, p. 123; "The System of Branchial Sense Organs and their Associated Ganglia in Ichthyopsida," Quart. Jour. Mic:osc Sc, 1886. Vol. XXVI, new series, p. 95 ; '• The Ciliary or Motor-Oculi Gangl. and the Gangl. of the Ophthal. Profund. in Sharks," Anat. Anz., 1887, p. 585; " The Old Mouth and the New," Anat, Anz., 1888, p. 15 ; "A Contribution to the Morphology and Development of the Nervous Syst. of Verteb.," Anat. Anz., 1888, pp. 874, 899; "The Development of the Peripheral Nerv. Syst. of Vertebr ," Quart. Jour. Micros. Sc, Vol. XXIX, new series, 1889, p' 153; "Prof. Rabl on the Mode of Development of the \'ertebrale Peripheral Nervous System," Anat. Anz., 1890, p. 125.

252 JouRNAi> OF Comparative Neurolocjy.

these rudiments corresponding to each cranial nerve grows ventrad, unites with the central organ by means of root fibres, probably growing out centripetally, and comes to lie on the outer side of the mesoderm, between this and the epidermis; while the rudiments of the spinal ganglia (of the trunk) proceed ventrally on the inner side of the mesoderm, thus between this and the central organ. These cranial ganglia rudiments, designated by Beard ^'■neuralganglia^ fuse at the level of the chorda with a thickened place in the epidermis dorsad of the neighboring gill cleft, which thickening represents externally the rudiment of a " branchial sense organ," internally that of a ganglion belonging to the latter, the '•'•lateral ganglion.' Here the elements of the neural and lateral ganglia blend indistinguishably with each other; there thus arises the definitive and apparently single ganglion of the cranial nerve in question. This separates from the epidermis — that is, from the rudiment of the branchial sense organ — but remains connected with the latter by a nerve strand, the suprabranchial nerve (dorsal branch of Van Wihje). From the ganglion are developed distally three nerves, namely, the JV. prcebrajtchialis , JV. postbranchialis , and \. pharyngezis. As to the N. suprabranchialis, it is, according to Beard, clear that it arises from the epidermis; for the N. praebranchialis he accepts the same mode of formation. With regard to the two other nerves, he remains undecided whether the epidermis takes part in their formation. (^) Beard regards it as certain in the case of the postbranchial twigs of the vagus(-) and in the case of the trigeminus (mandibularis).(')

For the trigeminus the same law of formation is said to obtain, inasmuch as the mouth represents a pair of gill clefts. But where gill clefts, with the musculature belonging, have entirely disappeared, yet sense organs persist, there, accord ing to this author, the structure is simplified; the N. postand

1 Quart. Jour. Mic. Sc, Vol. XXVI, 18S6, p. 102.

2 Ibidem, p. no.

3 Ibidem, p. 113.

\'. KuPFKEK, Cranial Nerves of \'ertebrates. 21^3

prcebranchialis disappear, and the sole ones remaining are the N. suprabranchiales.

As such may be regarded: (i) The N. olfactorius (I segment), more clearly so after it was shown, through the olfactory buds discovered by Blaue, that the organ of smell belongs to the system of the lateral line, /.c, the branchial sense organs. 2. The ophthalmicus profundus (II segment), whose meso-cephalic ganglion (Gn. ciliare, Van Wihje, His) fuses with the epidermis close above and behind the eyes; later this root unites with that of the trigeminus, and the Gn. meso-cephalicum with the Gn. Gasseri.(') 3. The acusticus (VI segment), which is regarded as a remnant of a segmental nerve. The development of the nerve, of its ganglion, and of the ear, corresponds essentially to that of the homodynamous parts of complete segments. The auditory vesicle is the persistent, functionally modified branchial sense organ; the pertaining segment remains in the hyoid arch. Whether the aborted gill cleft is to be found, as Van Wihje thinks, behind the facial, or, as Dohrn takes it, is to be sought before the hyoid cleft, remains undecided.

Regarding the relation of cranial to spinal nerves. Beard at first declared himself in complete agreement with Froriep in so far as the dorsal roots and ganglia of the one could not be homologized with those of the other. Possibly Balfour was right, that the cranial nerves showed a more primitive condition than the spinal nerves, but it may be doubtful whether the spinal nerves ever had the same primitive character-(^) Later he changes this opinion, and, in regard to the supposed similar method of formation of the first ganglion rudiment in head and trunk, declares that there is a partial homology between the neural ganglia of the head and the spinal ganglia, but the first might possibly be only homologous with the sympathetic portions of the latter(*)

1 Anat. Anz., 1887, p. 565.

2 Quart. Jour. Mic. Sc, Vol. XXVI, 1886, p. 142-143.

3 Quart. Jour. Mic, Sc, Vol. XXIX, 1889, p. 153.

2^4 JoiKNAi. OK Com I'AKA ri\K Nki K01,0(;V.

Judging from this, the entirely different position of the neural cranial ganglia on the one hand, of the spinal ganglia on the other, relative to the dorsal mesoderm, was not taken into account by Beard in the comparison.

In a lecently published article(') on the development of Petromyzon Planeri, I have likewise treated of the method of formation of the peripheral nervous system, in the course of which the previous works of Scott and Shipley on the same object received thorough consideration.

On the one hand, my observations not only yielded a confirmation of the participation, discovered in Gnathostomata, of the peripheral regions of the epidermis in the development of the cranial nerves, but caused this participation to appear still more considerable than could be admitted from what was previously known. But, on the other hand, facts emerged which do not accord with the results of the above mentioned authors, and display the complicated nature of the ci"anial nerves in a new light.

I was then led to essentially the following conception of the composition of the dorsal cranial nerves: Each one is composed of two parts, a spinal and a lateral., which latter, comprising all its components, can also be designated as branchial. The first behaves, with respect to its origin, its course and its relation to the dorsal mesoderm, entirely like a dorsal spinal nerve of the trunk. The dorsal cranial nerves thus originally contain parts homodynamous with the spinal nerves, but thereto is added the second variously formed component, which, arising with the spinal, proceeds over the dorsal border of the mesoderm and is situated on the outer side, between mesoderm and epidermis. It is these lateral components of the cranial nerves into whose composition growths of the epidermis enter, and that occurs in two series lying the one above the other. I distinguished them as lateral and epibranchial ganglia. The first lie in the hori-

X "The Development of I'etromy/on I'laneri," Arch. f. Mik. Anat., Kd. 35, 1890.

\. KifpFFKU, Cratiial Nerves of ^ crtehrates. zz^t^

zontal plane of the auditory vesicle, and arise at three separate places, i.e., in the region of the trigeminus, acusticofacialis and vagus. The epibranchial ganglia likew^ise emerge descretely, and there is always one close above each gill pouch. The part of the corresponding cranial nerve proceeding over and outside the mesoderm secondarily unites with these structures, and, indeed, the union with the lateral ganglia is of such a nature that the rudiment of the nerve itself swells into a new ganglion, the medial (neural ganglion. Beard). From this union of the medial and lateral ganglia proceeds the definitive ganglion (Hauptganglion) of the cranial nerve concerned. The epibranchial ganglia take part in the development of the terminal twigs of the cranial nerves. In the separation of the lateral, as well as the epibranchial ganglia from the epidermis, there is nowhere shown the rudiment of a sense organ. With the sole exception of the auditory vesicle, which is formed in the closest proximity to the lateral ganglion of the acustico-facialis region and is homodynamous with this lateral ganglion, all these ganglia of both series are entirely independent structures, standing in no connection with principal sense organs.

These were the results which I had reached. I had to break ofi' these investigations, on account of lack of material, at a stage in which the formation of the peripheral nervous system is far from concluded. It was the moment of the escape of the larvae (from the egg). I have since, with new and more complete material, carried the work on further, and extended it to the stage of larvae 4 mm. in length. At the same time I discovered that my older material, upon which the work cited had been based, contained gaps which prevented me from arriving at a complete understanding of the earliest beginnings of the cranial nerves. After interpolation of the then lacking stages of development, it is necessary to complete in many respects and partially correct my earlier representation.

The first rudiment I have sufficiently described and


drawn. Towards the conclusion of the folding process taking place in the ectoderm, by means of which the massive central organ is formed, but before its separation from the ectoderm, one sees in the fore half of the embryo (head region) three cords, the median neural and the paired lateral cords, which latter, on either side, correspond entirely to the intermediate fascicles (Zwischenstrang) of His (Fig. i z).(') But I have not ascertained the succeeding phase since, as I now see, I have lacked the connecting links. I assumed that in the separation from the ectoderm the paired — /.e. , the intermediate — fascicles came at the same time to lie on both sides of the median neural cord, that is to say, of the massive brain. That is not the case; the paired rudiments move farther mesad together, so that they unite dorsad of the brain cord, into a plate lying between this and the epidermis (Fig. 2 d P), and which is distinguished from the regularly biserially arranged elongated epithelium-like cells of the brain by reason of the irregular and, subsequently, loose disposition of their elements. This plate exhibits the same structure which His had already sketched, in 1879, in a Scyllium embryo and Beard more recently has drawn in various elasmobranchs and of the chick.

Not to anticipate the difficult question as to which parts of the peripheral nervous system are derived from this plate, in our terminology I will refer to it neither as ganglionnor nerve-plate, but call it the dorsal brain-plate., since, although divided into three segments, it extends over the whole length of the brain. Furthermore, in the spinal cord I find the plate not so distinct and represented by a double row of cells, which my pupil. Dr. Victor Rohon, has first described, (■') albeit from later stages, in the trout.

Next the cells of the dorsal brain-plate advance laterad

1 Compare my drawings in the Arch. f. Mikr. Anat., 1890, Bd, 35, Taf. XXVIII, Fig. 25, and His' figure in the Arch. f. Anat. u. Phys., 1879, Anat. Abt., p. 465.

2 " On the Histiog. of the Spinal Cord of the Trout," Sitzgsber. d. math, physik. Kl. d. K. Bayer. Acad. d. W., Munchen, 1884, Heft I, p. 39.

\'. KuPFFEK, Cranial Nerves of \ ertcbratcs . 257

between brain and epidermis, forming the well-known ganglionor nerve-" border" (Leiste) of writers, and now it becomes necessary in considering the development of the cranial nerves to keep distinctly in mind the different regions.

I distinguish accordingly: ( i ) The region of the fore head to the eye, inclusive; (2) the fore gill region, comprising the mouth and the three fore gill pouches; (3) the hind gill region to the eighth gill pouch, inclusive; (4) the trunk region.

In the region of \k\.e. fore head (Fig. 3) begins the formation of the borders (Leiste), which I designate as rootborders ( Wurzelleiste), from the dorsal brain-plate, before the formation of the eye has begun; and the cells soon extend along the whole lateral surface of the fore-brain, whereby the connection with the brain here becomes severed. These cells are at first rounded, then become spindle-shaped and touch each other, so that they appear in connected layers. They are ranged mostly in two rows, but so that in one place a larger collection always shows itself. In my treatise I gave a figure(') of this, indicated this accumulation as a ganglion, and regarded it as the first rudiment of the first trigeminus ganglion. That it behaves like the rudiment of a ganglion I still hold, but I have come to doubt whether it goes into the first trigeminus ganglion or forms the first spinal ganglion, to be mentioned later. The determination is difficult, and requires more extended and especially more comparative investigations. But the whole mass of cells does not in every case collect itself into a compact ganglion, but it arranges itself in lines, which, after the appearance of the sense organs, extend both towards the nose and eye and also towards the hypophysial pouch. Later they unite, in the main, with the first trigeminus ganglion, whose branches they appear.

t Arch. f. Mikrosk. Anat., B<1. 35, 1890. Taf. XXFX, Fig. 39 g).


In the fore gill region^ which extends from the eye to the third primitive gill pouch, the process takes place otherwise. Here the mesoderm occurs in addition, which, until the time of the appearance of the dorsal root-borders at the brain, still shows an entirely epithelial arrangement. There are nowhere present detached cells which enter into the formation of connective tissue.

The root-border grows on rapidly to the dorsal border of mesoderm, and spreads out so that it lies cap-like upon the border cells of the mesoderm (Fig. 4). Then two tracts (Ziige) of cells separate, of which the inner proceeds between mesoderm and brain, while the other grows on laterad from the mesoderm between it and the epidermis. This latter course is characteristic of the fore gill region. It is wanting in the hind gill region and in the trunk.

As these tracts (Ziige) display separate rudiments, I choose for them exact designations. I call the inner tract, which remains in connection with the dorsal border of mesoderm and proceeds ventrad between brain and mesoderm, the dorsal spinal nerve ^ since tJiis rndiinent extends in the same way in all parts of the body; the outer tract I designate the branchial nerve (Figs. 5 and 6, nb).

The border (Leiste), and what proceeds from it, is not continuous in the fore gill region, but at the beginning is divided into three successive segments, which correspond to the regions of the trigeminus, acustico-facialis and vagus, and consists exclusively of elongated, serially arranged cells.

While the separation of the border into the tracts mentioned, the spinal and the branchial rudiments, takes place over the dorsal border of mesoderm — it takes place about the time of the appearance of the e3'e-rudiment — growths of the epidermis begin in three places; they are the trigeminus, the acustico-facialis and the vagus swellings. They lie laterad of the dorsal mesoderm, and arise, not simultaneously but successively, from before backwards. The method of formation of these swellings I have minutely

\. KrPKFKR, Cranial Xcrvvs of ^'ertehratcs. 2c;9

described and drawn in my article. But I must here, on the the basis of later investigations, make some additions and corrections.

The trigeminus swelling, appearing close behind the eyerudiment, forms a projection (Fig. 4,^'/) directed entad and dorsad, and consisting of closely crowded cells, upon which, although still continuous, may be noticed two prominences lying the one behind and under the other. The root of the branchial nerve grows toward this swelling, unites with it by means of a tract of cells, and passes along, with another tract close by, growing in a ventral direction. The growth of these nerverudiments results partly from division of these cells, partly through the invasion of new elements from the dorsal brain-plate. At the point of junction of the root of the branchial nerve with the swelling of epidermis, there results a multiplication of the cells belonging to the nerve whereby two heaps of cells, corresponding to the two prominences on the swelling, arise, the elements of which are inserted between the epidermis cells of the swelling, so that later on they could not be distinguished according to their origin.

From this complex two ganglia are isolated successively with simultaneous separation from the epidermis, each of which consists of two parts, one epidermal and one arising from the dorsal brain-plate, of which the first part has been designated by me the lateral, the other as the medial ganglion (Figs. 5 and 7,^^/, ^"w). But the two parts are gradually merged into each other, so that finally no clear boundary between them can be pointed out. It is thus necessary to designate the whole body as one, and 1 propose the term '•'■ principaV ganglion (Hauptganglion) for it. There belong to the trigeminus two principal ganglia, formed in the same way, the Jirst and the second. The two are honiodynanious. I correct herewith my earlier view that the first trigeminus ganglion is exclusively of central origin, and has the same value as the medial part of the second trigeminus

26o Journal of Comparative Neurolo(;y.

ganglion.(') A more complete series of stages has subsequently convinced me of the incorrectness of this view.

The first trigeminus ganglion lies before and over the second, can always be distinguished from this and is connected with the nerve-rudiments of the fore head.

The epidermal swelling next behind gives rise to the labyrinth vesicle and the lateral part of the yacialis ganglion, as likewise homodynamous parts, a portion of its cells also entering into the rudiment of the glossopharyngeus. Vesicle and ganglion arise in close connection, so that in the beginning of the invagination of the pit of the labyrinth, the ganglion is situated within, like a large knob, on the wall of the pit. It becomes further enlarged through proliferation of the cells of the wall. The root of the pertaining branchial nerve itself enters by way of addition, as well with the ganglion as with labyrinth vesicle, and stretches to one other part in this structure, lengthening in a ventral part. The ganglion acquires a median portion from the latter.

The third epidermal swelling, which belongs to the vagus, is more stmple than the two preceding. There proceeds from it a rounded, simple lateral ganglion, which, however, by means of the root of the pertaining branchial nerve connecting with it, acquires a considerable medial part and thereby becomes a principal ganglion.

During the formation of this principal ganglion, and before it has yet completely separated from the epidermis, there appear new growths of the epidermis. They lead, in like manner, to the formation of ganglia, without the connection with them of rudiments of sense organs, at least directly. These are the epibranchial ganglia (Figs. 4, 6, 7 ge). They arise singly, close above the three primitive gill pouches, and lie thus in a second series removed ventrally from that of the principal ganglia. The epibranchial ganglion appearing first is not the formost, but

t Arch. f. Mikr. Anat., Bd. 35, 1890, S. 538.

\'. KuPFFEK, Cranial Nerves of Vertebrates. 261

the one over the second gill pouch. At the end of the embryonic period, and after the release of the larvae, this process advances cephalad into the region of the trigeminus, and also caudad in proportion as the hind gill pouches successively arise. These ganglia thus exhibit regular branchiomerism, and those arising in the region of the trigeminus point to aborted anterior gill pouches.

How the branchial nerves, passing along by the principal ganglia and proceeding to the visceral arches, connect with the epibranchial ganglia, I shall show later.

In the kind gill cegion, from the fourth primary gill pouch to the eighth, the development of the peripheral nervous system shows a less complicated course. The border (leiste) growing out from the dorsal brain-plate appears first as a continuous one, connects with the dorsal border cells of the mesoderm segments and sends a ventral extension between brain and mesoderm: there thus arises from it only the dorsal spinal nerve, the rudiment of branchial nerves remains apart from the border. There nevertheless arise, as already mentioned, in regular order from the epidermis, the epibranchial ganglia corresponding to the gill pouches of this region.

So long as the rudiments of the nerves consist of continuous chains of cells, it is not difficult to follow their course and establish the connections. But there comes a stage where, connected with the cells, fibrillfe appear and separate the cells, and with this difficulty ensues for the investigation, so long as the fibrill^e are not united into thicker cords. The latter has occurred after the release of the embryo, so the investigation is prosecuted on a firmer basis. One can then enumerate the roots in the different regions, and follow the ramifications of the distal nerves. The stage intervening between the first rudiment, consisting of continuous rows, and the later first evident compact nerve cords renders uncommonly difficult the determination of the question whether the rudiments and ganglia proceeding

262 Joi KNAL OK Comparative Neurology.

from the dorsal brain-plate and border respectively at any time give up their connection with the dorsal region of the brain. I assume this separation only for the region of the fore brain., where the rudiments also possess secondarily no connection with this region of the brain, but are united with the trigeminus and the mid-brain respectively. Behind this it does not occur, at least not up to the point of time when the formations of ganglia from epidermis begin and come forth in connection with the roots of the branchial nerves. With this question is bound up the other question of the final fate of the dorsal brain-plate. Do its cells advance in a body laterad and ventrad, or a middle portion of the cells of the plate remain in loco, while these cells intercalate themselves in the epithelial covering of the neural tube? According to my observations the latter takes place. A great part of the plate is applied to the formation of the peripheral nerves, but the remainder appears wedged in between the cells of the roof of the brain without the interruption of the connection wnth the peripheral parts.

I do not propose to discuss the histogeny of the nerves in this place, since the details are still not clearly enough established to admit of generalizations; but I may devote space for the view to which the embryo and the youngest larval stages of Petromyzon have led me. None of my observations contradicts the view, but rather everything indicates that the fibrill^e arise as processes of cells, but not merely from cells of the ganglia and central organ but also from those cells which, ranged in chains, form the first rudiments of peripheral nerves. This being accepted, it appears to me, further, most probable that the growth of the fibrillas in the dorsal nerves extends in both directions, centripetal as well as centrifugal. Thus, when the rudiments have attained the stage of formation at which they display fibrillge along with the cells, the cells appear moved apart from each other, and at both ends, the central as well as the peripheral, proceeding forth in fine filaments.

\'. KiPFFKR, Cranial Nerves of }'i'rtcl}rates. 363

All other questions bearing further upon the peculiarities in this process appear to me at the present not ripe for dis^cussion. Yet I believe I may say one thing definitely, that the rudiments of the dorsal nerves, as w^ell in the earliest phase of the cell chains as also later, when fibrillae have already appeared, alw^ays show the connection with the central organ.

According to what has been hitherto communicated, there enter into the composition of the cranial nerves two systems, the spinal and the hrafichial. The first is common to the head and trunk; the latter appears exclusively in the head, is most developed in the fore gill region, and one part is produced into the hind gill region.

Concerning the spinal system of the head, I thus confine myself here to the dorsal rudiments, reserving the ventral spinal nerves for later mention. The complete interruption of the roots between the regions of the trigeminus, acusticofacialis and vagus self-evidently also affects the spinal system. It also divides, therefore, into three regions, of which the hindermost adjoins the spinal system of the trunk. Proceed-, ing from the root-border, the rudiment of a spinal nerve divides into two tracts. The lateral tract (zug) remains in connection with the dorsal border of mesoderm — I denote it as the dorsal branch of the dorsal spinal nerve — the median tract, or ventral branch of the dorsal spinal nerve, passes between mesoderm and brain towards the chorda, passes around this and arrives at the outer side of the aorta. At first consisting of serially arranged cells, this tract includes in itself the rudiments of the spi)ial gafiglia and of the sympathetic ganglia (Fig. 6).

So far as I have been yet able to discover, the spinal spinal system of the head does not keep pace in further development with the branchial system, but instead undergoes remarkable reductions.

The nerves of the branchial system are, in respect to their mode of origin, more composite. The rudiments proceeding

264 Journal of Comparative Neurology.

from the root-border grow, with a participation in the principal ganglion then originating, over into the ventral region (Fig. ^, nd). It there connects secondarily with a portion proceeding from the distal end of the associated principal ganglion, of which I cannot definitely say whether it is of central or of epidermal derivation — that is to say, whether it is derived from the medial or lateral part of the ganglion, or from both. Then, later, when the epibranchial ganglia arise, further complications ensue. In conjunction with the formation of these ganglia, there appears a peculiar subepidermal layer of cells, which gradually spreads from the eye to the hindermost gill pouch, dzif remains conjined to thc ventral side.

If it were not preoccupied, I would suggest the name "hypodermis" for this layer, but, as it is, propose the term " neurodermis" (Figs. 6, 7, 10, nb). The point of origin of this structure is the vicinity of earliest epibranchial ganglion, dorsad and cephalad of the second gill pouch. At the lateral line, i.e., the outer boundary between the dorsal and ventral region, the neurodermis is not strongly developed; here it consists of several layers of cells which form an inwardly projecting ridge, but in general it remains one-layered and consists of closely arranged but disconnected epithelium-like cells, with a prevailingly cylindrical form.

Note. — The figures which are referred to in this paper will be collected in plates at the end of the article.

[to be CONTINCE]).]


( Coiitimied. )

C. H. Turner.

Additional Remark upon the External Morphology of the Epencephalon. — As has been remarked above, the most remarkable characteristic of the avian brain is a tendency towards great compactness. This tendency has left its stamp upon the epencephalon. In addition to being wedged into the caudal V of the hemispheres, (-) and to being transversely convoluted, (') the epencephalon suffers a most noteworthy modification. Its cephaloventral extremity projects mesad into the cavity of the fourth ventricle, thus producing the almost unique phenomenon of a portion of the epencephalon being embraced by the metencephalon (Plate VII, Fig. 3; Plate XIV, Fig. 10; Plate XV, Fig. 7; Plate XVIII, Figs. 9, 10, 21).


Although in its external form the avian epencephalon resembles one of the embryonic stages of the higher verte-

1 A Correction.— Dr. R. W. Shufeldt has called my attention to an error that occurs on page 56, lines 16 to 18. Those lines should read : " My notes upon these two birds are based upon a study of sketches given in the United States Geological Survey, J. W. Powell, Director; third aiinual report (1881-82), p. 56, Fig. 8, and p. 70, Fig 20 "'

2 Supra, p. 41.

3 Supra, p. 51. . .

266 Journal of Compahatuk Neurology.

brates, yet in its internal structure it differs but little from the adult mammalian cerebellum. As in the mammalia, so here, the epencephalon consists of three major histological regions; the epithelium, the cerebellar cortex, and the body. Around the periphery, lining the convolutions, we find the narrow epithelial layer; entad to this we find the cortex, while further entad, forming the core, we find the body.

Cerebellar Cortex (Plate XVIII, Fig. 6). — The transversely and unequally convoluted cerebellar cortex is composed of three laminae. The most ectal lamina is the widest of all. It is composed of neuroglia through which are scattered a few of Deiter's corpuscles. The second lamina is very narrow. It consists of a single layer of the well-known Purkinje's cells. These are large and gibbous flask cells. In hsematoxylin and aluminium-sulphate cochineal preparations, these cells are densely stained and present large, clear, spherical nuclei and large, dense nucleoli. In different brains the size of these cells varies greatly, but in all cases they rank with the largest cells in the brain. The width of the next layer is a variable quantity. This dimension varies not only in different brains, but also in different parts of the same epencephalon. But, in every epencephalon, the major local thickenings of this lamina take place in a definite and constant manner. In all cases this layer is narrowest at the proximal extremity of each canvolution. Thence the width increases gradually, although irregularly, until the distal extremity of the convolution is reached. There the layer is widest. Although at the proximal extremity of each convolution this layer is often but little wider than the layer of Purkinje's cells, yet at the distal extremity of each convolution it will be almost as wide as the external neuroglia layer (Plate VII, Fig. 4; Plate XIV, Fig. 8). This fact seems to warrant the following suppositions:

I. This layer of the avian epencephalon was deposited before that bbdy became convoluted.

Turner, Morphology of the Avimt Brain. 267

2. Originally this was a uniform layer parallel to the surface of the epencephalon.

3. Its present shape is due to a tendency to remain parallel to the surface which has now become greatly increased by convolutions.

Although the cerebellar cortex is here described as being composed of three distinct layers, it must be borne in mind that neither membranes nor ventricles separate these lamina". Nor is there either membrane or ventricle between the cortex and the remainder of the epencephalon.

White Substance of the Epencephalon . — The body or core of the epencephalon is an irregular solid polyhedron, from the periphery of which projections extend into all the irregularities of the ental surface of the cerebellar cortex, and completely fill them. The centre of this core is occupied by a small ventricle, which is connected with the fourth ventricle by a narrow median isthmus. Histologically, this region is composed almost exclusively of fibres. It contains, however, two niduli, one of which is near the ventricle, while the other is in the peduncle.

Dentate Niduhis. — The nidulus near the ventricle is large and conspicuous, and is known as the dentate nidulus. It lies, for the most part, cephalad of the ventricle, and is very irregular in outline. In many cases the nerve cells constituting this nidulus are large pyramidal cells, in other cases they are multipolar, while in still others both pyramidal and multipolar cells are found in the same nidulus. In some brains these cells are densely compacted, while in others they are only loosely aggregated. In all cases they lie in a bed of nerve fibres. In my preparations these cells are so densely stained that their nuclei and their nucleoli are invisible. This nidulus is abundantly supplied with Deiter's corpuscles.

268 Journal of Comparative Neurology.


Nidiili of tlic Fiftli Nerve. — The mesencephalic nidulus of the trigeminal nerve has been described in connection w^ith the niduli of the mesencephalon. (') One tract of the fifth nerve arises in the epencephalon, but the niduli of that region also have been described above. There remains then to describe in this connection those niduli only of the trigeminal nerve which are located in the metencephalon.

Lateral Motor Nidulus of the Fifth Nerve (Plate XVIII, Figs. 8, 15). — This is a well-defined nidulus which is situated near the lateral surface of the medulla, adjacent to the cephalad root of the trigeminal nerve. Although situated quite near to the lateral surface of the metencephalon, yet it lies entad to that tract of the fifth nerve which passes to the myelon. It also lies cephalad to the tract which passes to the deep motor nidulus of the trigeminus. In difierent birds the shape of this nidulus undergoes considerable variation. As far as my observations go, the two extremes are represented by the young dove ( Columba livio) and by the adult Swainson's thrush [Mylocichla Sivai7iso?ii). In the first case it is a compact spherical nidulus; in the other it is an elongated, straggling nidulus, which lies parallel to the deep motor root of the fifth nerve.

This nidulus is well supplied with large pyramidal nerve cells. The apex of each cell is prolonged into a long process, which process is often curved. In hsematoxylin and in aluminium-sulphate cochineal preparations, these cells are densely stained, and each one presents a densely stained nucleus and nucleolus. This nidulus is well supplied with Deiters' corpuscles, and nerve fibres traverse it in several directions.

Ventrad to the above nidulus there exists, in some of my sections, an ill-defined cell cluster, which probably deserves

I Supra, p. 121 ; Plate XV, Fig. 3.

Turner, MorpJiology of the Aviaii Brain. zGc^

to be considered a distinct nidulus. In the medulla of the young dove ( Cohimba livia) and in a few other cases this cell cluster appears to be undoubtedly distinct from the lateral motor nidulus of the trigeminus, while in other cases (some of the Turdidcr), it is evidently amalgamated with that nidulus. Histologically, it could hardly be considered a distinct nidulus. For, although the cells are not typical pyramidal cells, and although they appear to differ from the typical cells of the lateral motor nidulus of the trigeminal nerve, neither are they typical fusiform or flask cells. Although tending towards the flask cells in outline, yet they agree with the pyramidal cells in the structure of their nuclei. In haematoxylin and in aluminium-sulphate cochineal preparations, these nuclei are densely stained. After due consideration it has been thought best to consider this cell cluster as a slightly modified portion of the lateral motor nidulus of the trigeminus, which portion occasionally becomes distinct.

Lateral Sensory Nidulus of the Fifth Nerve (Plate XVIII, Fig. 8). — This nidulus lies near the lateral surface of the metencephalon, and on the same level as the lateral motor nidulus of the trigeminus. It lies about as near to the lateral surface of the medulla as does the latter nidulus, but it lies upon the opposite side of the deep inotor fasciculus of the trigeminal nerve. The form of this nidulus is not constant throughout the class Aves. In some cases [Hylocichla Swainsoni) the outline of this nidulus is sub-spherical, while in others ( Coluniba livia) it is somewhat irregular. This nidulus is smaller than the lateral motor nidulus of the same nerve.

The cells of this nidulus are flask shaped. In haematoxylin and in aluminium-sulphate cochineal preparations, these cells are densely stained, and each one presents a faintly stained nucleus and a densely stained nucleolus. The niduli of these cells are relatively smaller than the niduli of sensory cells usually are. These cells are loosely and irregularly

370 JOURNAL OK Comparative Neurology.

aggregated, and among them is dirtributed a large number of Deiter's corpuscles,

Deep Motor Xidulus of the Fifth Nerve (Plate XVIII, Fig. 16). — Near the meson, on a level with the niduli of the glossopharyngeal and pneumogastric nerves and extending as far cephalad as the roots of the trigeminal nerve, there exists an elongated sub-ellipsoidal nidulus. This small cell cluster lies immediately cephalad of the mesal extremity of the deep motor root of the fifth nerve, and has its major axis approximately parallel to the longitudinal axis of the metencephalon.

The nidulus is composed of a loose aggregate of large pyramidal cells, among vv^hich numerous Deiter's corpuscles are distributed. The apex of each cell is prolonged into a long process, while the base is supplied with several shorter processes. The apical process is often curved. In haematoxylon and in aluminium-sulphate cochineal preparations, these cells are densely stained, and each one presents a densely stained nucleus and a densely stained nucleolus. Several small fasciculi of nerve fibres pass from this nidulus to the raphe.

In the avian brain there does not appear to be any homologue of what in the human medulla(') is known as the inferior sensory root of the trigeminus.

Gasserian Ganglion (Plate XVIII, Figs. S, 15, 19). — This is a large ganglion which is situated upon the root of the trigeminal nerve. This ganglion consists of large typical bipolar cells, which are arranged with their longitudinal axes perpendicular to the metencephalon. In addition to the ordinary cell wall, each of these cells is surrounded by an additional sheath. In this sheath several nuclei are visible. In haematoxylin and in aluminium -sulphate cochineal preparations, these cells are densely stained, and each one presents a faintly stained spherical nucleus, within which is a

I Ambrose L. Rannev, " The Applied Anatomy of the Nervous System,^' second edition, p. 254, Figs. 55, 5.

Ti'RNEK, MorpJiology of tJic Avian Brain. z'^ i

densely stained nucleolus. From the proximal extremity of each cell a nerve fibre passes into the metencephalon, while from the distal extremity of the same cell a nerve fibre passes into the nerve.

Nidiilus of the Abducens Nerve (Plate XVIII, Fig. lo, 13). — This is an ill-defined nidulus which lies hear the meson and which extends as far cephalad as the external root of the sixth nerve. In the alligator brain(') and in the human brain(^) this cell cluster is situated on the floor of the fourth ventricle. In the avian brain, however, this nidulus is not adjacent to the ventricle, but is separated from it by a large faciculus of nerve fibres.

This nidulus is composed chiefly of rather small, irregular, pyramidal cells, which resemble those of the niduli of the oculo-motor and pathetic nerves. True, these cells are more irregular than those figured in Plate XVI, Figs. 13, 14; but the cells of the niduli of the third and fourth nerves also are usually more irregular than the cells there delineated. In hasmatoxylin and in alumininium-sulphate cochineal preparations, these cells are usually densely stained, and each cell presents a small densely stained nucleus, within which is a densely stained nucleolus. This nidulus is well supplied with Deiter's corpuscles.

Niduli of the Facial and of the Auditory Nerves (Plate XVIII, Figs. 3, 4, II, 13, 18). — Since in the avian brain the seventh and eighth nerves have a common root, it has been thought wise to describe the niduli pertaining to that root as though they were the niduli of a single mixed nerve. In the avian metencephalon at least three distinct niduli are related to the fibres of facial and auditory nerves. For convenience these are here designated as " nidulus L," " nidulus B," " nidulus Y."

Nidulus L (Plate XVIII, Fig. 18).— Far laterad,near the the root of the auditory nerve, there is a small sub-spherical

1 Prof. C. L. Herrick, op. cit., p. 153.

2 Prof. Ambrose L. Rannev, op. cit., p, 339.

272 Journal of Comparative Neurology.

nidulus. The cells of this nidulus are very irregular in outline. They are neither typical pyramidal cells nor typical flask cells. Probably they resemble the former more than than they do the latter. In haematoxylin and in aluminiumsulphate cochineal preparations, these cells are densely and obscurely stained, and each one presents a densely stained nucleus, within which is a densely stained nucleolus.

The cmit/entia acustica is a slight projection into the cavity of the fourth ventricle. This projection is situated upon the floor of the fourth ventricle, about half way between the external root of the eighth nerve and the meson (Plate XVIII, Fig. 11). This is a sub-elliptical body, and is divided by a band of fibres into two unequal portions. That division which lies nearest the ventricle is much larger than the other portion. This eminentia contains nidulus B.

Nidulus B (Plate XVIII, Fig. 18). — As has been mentioned above, this cell cluster constitutes the nervous portion of the eminentia acustica. The band of fibres which divides the eminentia into two unequal portions divides this nidulus in a similar manner. In that section of this nidulus which is nearest the fourth ventricle the nerve cells are crowded into the mesal half, while the remainder of that portion of the nidulus is filled with Deiter's corpuscles. In that section of the nidulus which is on the other side of the band of fibres the nerve cells are uniformly distributed. The cells of this nidulus are small, gibbous, flask cells, which, in hcematoxylin and in aluminium-sulphate cochineal preparations, are faintly and obscurely stained, and each of which presents a faintly stained nucleus, within which is a densely stained nucleolus. This nidulus is surrounded on all sides by nerve fibres.

It is quite probable that the two divisions of this nidulus correspond to the two small and distinct but adjacent niduli which occupy a similar position in the alligator brain.(')

Nidulus T (Plate XVIII, Figs. 4, 11). — Mesad to the

X Prof. C. L. Herrick, " Notes on the Brain of the Alligator," 1. cit., p. 153.

Turner, Alorpholog'y of the Avian Brain. 373

above nidulus and between the surface and that tract of the auditory nerve w^hich passes to the raphe, there is an inconspicuous nidulus. This nidulus contains a fevv^ scattered fusiform cells and numerous Deiter's corpuscles. In hcematoxylin and in aluminium-sulphate cochineale preparations, these cells are obscurely stained, and each one presents a faintly stained nucleus, within which is a densely stained nucleolus.

Root Ganglion of the Auditory Nerve (Plate XVIII, Figs. I, 4). — Within the skull cavity the avian eighth nerve bears a small root ganglion. This ganglion is closely appressed upon the metencephalon. Indeed, in some cases it is so intimately connected with the brain that it resembles a superficial nidulus. This ganglion contains large spindle-shaped cells, which apparently are of the same type as those in the Gasserion genglion. In hsematoxylin and in aluminium-sulphate cochineal preparations, these cells are densely stained, and each cell presents a faintly stained spherical nucleus, within which is a densely stained nucleolus. Unfortunately, in all my sections that show this ganglion these cells are cut approximately at right angles, thus rendering it impossible to demonstrate whether or not each extremity of the cell is prolonged into a nerve fibre. HoweveR, each cell is surrounded by a nuclei-bearing sheath, which resembles the sheaths described and figured for the cells of Gasser's ganglion (Plate XVIII, Fig. i).

Nidulus of the Glosso-pharyngeal Nerve (Plate XVIII, Fig. 16). — At the surface it is almost impossible to distinguish beteen the roots of the glosso-pharyngeal and pneumogastric nerve, but when we pass entad a knowledge of the physiological functions of each nerve renders the separation of the glosso-pharyngeal and pneumogastric niduli an easy matter. Both niduli lie at the meson and in the floor of the fourth ventricle. But the nidulus of the glosso-pharyngeal nerve lies further cephalad and extends forther dorsad than the nidulus of the pneumogastric nerve.

274 loi'RNAI. OF CoMPAKA'ri\"K N KUK()I,<)<;^ .

The glosso-pharyngeal nidulus is composed of a close aggregate of small pyramidal cells, among which a few fusiform cells are scattered. The fusiform cells are mostly confined to the caudal portion. The pyramidal cells are quite irregular in outline. In hasmatoxylin and in aluminium sulphate cochineal preparations, these cells are densely stained, and each one presents a densely stained nucleus, within which is a densely stained nucleolus. This nidulus is abundantly supplied with Deiter's corpuscles.

Nidulus of the Pneumogastric Nerve (Plate XVIII, Fig. i6). — Anatomically, the nidulus of the tenth nerve is sometimes distinct from the nidulus of the ninth nerve and sometimes not. Histologically, it is always distinct. It lies caudad to the nidulus of the glosso-pharyngeal nerve, and, usually, does not extend so far dorsad as that nidulus does. Like the above nidulus, the pneumogastric nidulus is elongated, with its major axis parallel to the longitudinal axis of the metencephalon. This nidulus is not quite so wide as the nidulus of the ninth nerve.

The pneumogastric nidulus is composed of lax-ge, typical, irregular gibbous cells, which present a strong contrast to the small pyramidal cells of the glosso-pharyngeal nidulus. In hfematoxylin and in aluminium -sulphate cochineal preparations, these cells are densely stained, and each one presents a large, clearly stained nucleus, within which is a densely stained nucleolus. This nidulus is well supplied with Deiter's corpuscles.

Nidulus of the Spinal Accessory Nerve (Plate XVIII, Fig. i6). — Immediately ventrad of the nidulus of the pneumogastric nerve, and extending from the nidulus of the glosso-pharyngeal nerve caudad into the myelon, there is a narrow, elongated cluster of cells. This cell cluster is supposed to be the nidulus of the spinal accessory nerve. This nidulus is composed of large, irregularly arranged, pyramidal cells, which are larger than those of the glosso-pharyngeal nidulus. In haematoxylin and in aluminium-sulphate cochi-

Turner, Morphology of tJie Avian Brain. 275

neal preparations, these cells are densely stained, and each one presents a densely stained nucleus, within which is a densely stained nucleolus. This nidulus is well supplied with Deiter's corpuscles.

Xidulus of the Hypoglossal Nerve (Plate XVIII, Fig. 17). — Further ventrad, and separated from the nidulus of the spinal accessory by a cell-less region, lies the large nidulus of the hypoglossal nerve. This nidulus is sub-pyramidal in shape. Its base is about at the junction of the myelon with the metencephalon, while its apex is as far cephalad as the caudad extremity of the nidulus of the glosso-pharyngeal nerve.

This nidulus is composed of large pyramidal cells, which resemble those of the nidulus of the spinal accessory nerve. In haematoxylin and in aluminiumsulphate cochineal preparations, these cells are densely stained, and each one presents a densely stained nucleus, within which is a densely stained nucleolus. This nidulus is vs^ell supplied with Deiter's corpuscles.

Olives. — Near the ventral surface of the metencephalon, and immediately laterad of the root of the hypoglossal nerve, there is a small, ill-defined cluster of cells. This nidulus is probably the homologue of the olivary body. However, this nidulus does not exhibit the slightest trace of the complex structure of the human olivary body.(') The cells of this nidulus are small. In ha^matoxylin and in aluminivim-sulphate cochineal preparations, the nuclei of these cells are much more densely stained than is usually the case in flask cells.

Accessory Olives (Plate XVIII, Fig. 13). — Near the ventral surface of the metencephalon of a young dove ( Coluniba livia) a small, ill-defined cell cluster has been observed. It is composed of fusiform cells, and is probably a homologue of the accessory olives. In haematoxylin preparations these cells are faintly stained, and each one presents a large, clear

I See Rannev, op. cit., p. 268 and p. 262, Fig, 59.

376 T^i'KNAi. OF Comparative Neurology.

nucleus, within which is a large, densely stained granular nucleolus.


Root of the Trigeminal Nerve. — In the amphiban brain, according to Professor Osborn,(') the fifth nerve root is composed of the following tracts: i. Ascending tract of the cervical region, reinforced by 2, fibres from the deep motor nidulus, representing two tracts. 3. Fibres from the sensory nidulus. 4. Descending tract from the mesencephalic nidulus. 4. Direct encephalic tract.

In the avian brain the composition of the root of the trigeminal nerve tallies even more closely with the composition of the corresponding nerve of the human metencephalon.(-) As in the human brain, so here this nerve is composed of two distinct roots, each of which is composed of several distinct fasciculi. In the lower types of birds ( Columba livia, etc.) this appearance is quite distinct, but in the higher types {^Hylocichla sxvainsoni , etc.) it is often somewhat obscured. One of these roots lies caudad to the other. The cephalad root is composed of two fasciculi, the lateral motor fasciculus and the ascending cervical fasciculus. The caudad root is also composed of two tracts, the lateral sensory fasciculus and the deep motor fasciculus.

Lateral Motor Fasciculus of the Trigeminal Nerve (Plate XVIII, Fig. 8). This is a short bundle of loosely aggregated fibres which passes from the lateral motor nidulus of the trigeminal nerve laterad into that nerve. This tract is feebly convex, the convexity projecting caudad.

Ascending Cervical Fasciculus of the Trigeminal Nerve (Plate XVIII, Fig. 16). — This is a narrow fasciculus, which, after passing entad for a short distance, turns abruptly and passes caudad into the myelon.

Lateral Sensory Fasciculus of the Trigeminal Nerve

I " Amphibian Brain Studies," Jour, of Morphology, Vol. II, p. 69. a See Rannev's " Applied Anatomy of the Nervous System," p. 338.

Turner, Morfliology of the Avian Brain. 377

(Plate XVIII, Fig. 8). — This is a small tract which passes from the lateral sensory nidulus of the trigeminal nerve laterad into the root of that nerve.

Deep Motor I-^'asciculus of the Trigeminal Nerve (Plate XVIII, Figs. 8, 15). — This is a broad bundle of fibres v^^hich passes from the root of the trigeminal nerve directly mesad to the raphe. It intersects the raphe immediately dorsad of the deep motor nidulus of the fifth nerve. This faciculus is composed of several narrow, isolated bundles.

In addition to the these tracts, there is a tract which descends from the epencephalon and enters one of the roots of the fifth nerve. There appears to be no trace of what Professor Osborn has called the direct encephalic tract, but there appears to be a tract passing into the mesencephalon. Probably that tract communicates with the mesencephalic nidulus of the fifth.

Tract of the Abducens Nerve (Plate XV, Fig. 12; Plate XVIII, Figs. 10, 13). — As in the amphibian(') and in the reptilian(-) and in the mammalian( ') brain, so here, the internal course of the abducens nerve consists of a single narrow fasciculus. This bundle passes from the nidulus of the abducens nerve ventro-laterad to the external root of that nerve. Although small, this tract is relatively larger than the corresponding tract of the mammalian brain.

Fasciculi of the Facial and Auditory Nerves (Plate XVIII, Fig. 18). — In birds the root-fibres of the facial and auditory nerves are so intimately associated that it is not now possible to say which fibres belong to the eighth and which to the seventh nerve. In this connection we find four fasciculi, (^) one of which passes to the epencephalon. For convenience,

1 H. F. Osborn, "Amphibian Brain Studies," p. 70.

2 C. L. Herrick, " Notes on the Alligator Brain," p. 153.

3 Ambrose L Rankev, op. cit., p. 339.

4 Among my notes I have one, dated several months ago, in which is described for the avian brain a tract resembling the genu of the seventh nerve as it appears in the mammalian brain. Since that time the series that furnished the note has faded so much that verification has been impossible. Other series have been carefully studied, but no such tract has been since encountered. — C. H. T.

278 JoiTRNAL OK Comparative Neurology.

the three tracts that are confined to the medulla are designated " tractus L," " tractus B,"' " tractus D."

Tracius I^. — This is a short fasciculus which passes from nidulus L directly laterad to the common root of the auditory and facial nerves.

Tractus B. — This fasciculus passes direct from the root of the auditory nerve meso-dorsad to the eminentia acustica. There it envelops nidulus B, after which it passes mesocaudad to the raphe. This latter portion of the tract is convex, with the convexity projecting caudad. A similar tract is found in almost all vertebrate brains.

Tractus D (Plate XVIII, Fig. 18). — This is a narrow bundle which passes from the common origin of the auditory and facial nerves mesad to the raphe. There it probably decussates. This fasciculus appears to be the homologue of a similar tract discovered by Professor C L. Herrick in the alligator brain,(') a tract which he considers to be a fasciculus of the facial nerve.

There does not appear to be any very great resemblance between the tracts of the auditory nerve of the birds and the tracts of the corresponding nerve of the amphibia. According to Professor ()sborn,(') the eighth nerve of the amphibia is composed of the following tracts: i . A tract from the posterior longitudinal fasciculus, connecting with the myelon. 3. A tract from a large nucleus situated directly above the exit of the ninth nerve and above the motor nidulus of the trigeminal nerve. (This cell cluster is probablv Deiter's nidulus.) 3. A tract from a small group of cells in the lower angle of the metencephalon. 4. A tract from the fasciculus communis. v ^ tract from the epencephalon. Certainly the reseinblance is not very striking. Better preparations, however, than those, at my disposal might strengthen the homologies between these two groups, for occasionally I have observed a faint indication of what might be a connec-

1 Op. cit., p. 153.

2 "Amphibian Brain Studies," p. 66.

Turner, Morphology of the Avian Brain. 379

tion between the root of the auditory nerve and the longitudinal fasciculus.

Fasciculi of the Glosso-pharyngeal and Pncuniogastric Nerves. — The fibres of these nerves pass directly entad to their respective niduli.

Fasciculi of the Spinal Accessory Nerve. — The internal root of this nerve consists of a longitudinal series of fasciculi which pass entad from the root to the nidulus of the spinal accessory nerve.

Fasciculus of the Hypoglossal Nerve. — This fasciculus passes from its nidulus directly ventro-laterad to its external root.

Raphe. — The only commissure in this region is the commissure which connects the two halves of the metencephalon. This commissure extends along almost the entire length of the metencephalon, and is familiarly known as the raphe.

Crossed Pyramidal Tract. — The course of this bundle is the same in the avian brain as it is in the human.

Direct Pyramidal Tract. — The course of this tract in the avian brain is similar to the corresponding tract of the human. But it is composed of a small fraction only of the median longitudinal fibres that are found near the ventral surface of the medulla. The remainder go to form the dorso-median fasciculus.

Direct Cerebellar Tract. — The homologue of this tract is present and has a course similar to its course in the human brain.

Posterior I^ongitudinal P^asciculus (dorso-median fasciculus). — Among anatomists the cephalad terminus of this bundle has long been a open problem. During the past two years two distinguished anatomists, Edinger(') and Honegger,(') have each oflered a solution of this problem. Each

1 Dk. Ludwig Euingek, " Twelve Lectures on the Structure of the Central Nervous System." Trans, by Willis Hall Vittum, M.D., p. 120.

2 Jacob Honegger, " Vergleichende Anatomische Uutersuchung iiber den Fornix und die zu ihm Beziehung Gebrachten Gebilde in Cehirns des jMenschen und der Saiigethiere."

28o Journal of Comparative Neurology.

is describing it as it appears in the human subject. Edinger, after remarking that certain of the fibres of the posterior commissure curve caudad, continues: " These fibres, together with others which arise in the depth of the inter-brain, are met with as a fine fasciculus ventrad of the anterior oculomotor nidulus. As we pass back this fasciculus progressively increases. There are added to it numerous fibres from the nucleus of the oculo-motor. We shall, from :iow on, meet with the triangular cross-section of this bundle, which is composed of fibres from such various regions on every transverse section of the brain, from the corpora quadrigemina down to the beginning of the spinal dord. This bundle has been called the fasciculus longitudinalis posterior. Inasmuch as fibres are given off" along the whole course of this bundle to the nerve-nuclei, as can be plainly seen in embryos of the sixth to seventh month, when few other fibres are medullated, and, as this bundle projects further back than the nucleus of the abducens, it is probable that the fasciculus longitudinalis posterior not only contains the fibres of communication between the nuclei of the ocular muscles, but that it also contains fibres to other cranial nerves. Flechsig is also of this opinion," etc.

Honegger's solution, written contemporaneously, certainly is not in accord with the above. This author considers that this tract does not give oft' fibres to the posterior commissure. He has also demonstrated that a portion of the fibres of the posterior longitudinal fasciculus pass to the mammillary body, and that other fibres of this bundle decussate in the ventral portion of the diencephalon.

In the avian brain that portion of this fasciculus which lies cephalad of the oculo-motor nidulus consists of a few scattered fibres. Cephalad, these fibres terminate abruptly, near the posterior commissure. Passing caudad, between the oculomotor and the trochlear niduli, these few fibres become a large bundle. From this point to the myelon there is a progressive increase in the size of this bundle. A short dis-

TuKNYLK^ M^orphology of the Aviafz Brain. z8i

tance cephalad of the trochlear niduhis the major portion of this fasciculus turns and passes obliquely to the ventral surface of the medulla, thence into the myelon. A few fibres, however, continue caudad and remain near the dorsal surface of the metencephalon. Throughout its course this tract continues near the meson. Although it has not been possible to trace fibres from this fasciculus into any of the niduli of the avian medulla, yet it is evident that fibres are given off' all along the line. Although the evidence at hand is not quite conclusive, yet I am inclined to believe the relationship of the avian posterior longitudinal fasciculus is the same as that described above by Edinger for the human posterior longitudinal fasciculus.

Fibres Arcuatce (Plate XVIII, Fig. 14). — On a level with the auditory nidulus a convex tract, with its convexity projecting laterad, passes from a nidulus in the peduncle of the epencephalon to a nidulus in the ventral portion of the metencephalon.


I. Economy of space is evidenced in all parts of the avian brain, and the phrase " higher type of birds" carries with it the significance " greater compactness of the brain." Indeed, progressive compactness has played so important a part in the evolution of birds that there is a vast difterence between the lowest avian brains, with their large projecting olfactory lobes and exposed optic lobes, and the highest avian brains, with their small, inconspicuous olfactory lobes and covered optic lobes. The difference between these two extremes is almost as great as that between the brain of the lizard and the brain of the lower types of birds. Yet there is no impassable gulf between the brains of the lowest and the brains of the highest types of birds, for all the intervening stages are supplied by the brains of the various avian groups. In reviewing this remarkable sequence, we are almost forced to believe that this tendency towards compactness of the brain

282 Journal of CoMi'AJiATi\E Neurology.

existed long before the first bii'd was evolved. If this be true, then this tendency towards a progressive compactness of the brain, combined with a tendency to develop all parts pertaining to vision and to atrophy all parts pertaining to smell, will account for all the major differences between the avian and reptilian brain.

Furthermore, within this class of animals, this progressive compactness of the bi-ain is a factor of taxonomic importance. So far, at least, as the major groups are concerned, a classification based upon this alone is in harmony with those classifications that are based upon a study of the structural elements of birds.

2. Neurologically considered, birds are preeminently seeing animals, and all parts that appertain to vision are developed to a great degree. The optic nerve is the largest of all the cranial nerves, and the optic lobes are complexly difl^erentiated bodies. Even the third, fourth and sixth nerves, although quite small, are relatively larger than the corresponding nerves of the mammalian brain.

An extraordinary development of one set of organs is never accomplished but at the expense of some other set. In this case the organs of smell have been the martyrs. Although in the lower avian types the olfactory lobes are paired and conspicuous, yet in the highest type of birds the rhinencephalon is a small unpaired body, which is partly embedded in the base of the cerebrum.

3. Histologically, the avian brain is composed of nerve fibres, nerve cells and neuroglia. In this connection we are concerned with nerve cells only. Although these cells present a great diversity of forms, yet they may all be grouped in the following classes: ganglionic cells, Deiter's corpuscles, fusiform or flask cells, pyramidal cells, and multipolar cells. The ganglionic cells are large bipolar cells, which are never found outside of the root ganglia. Each extremity of each of these cells is prolonged into a nerve fibre. In addition to the ordinary cell wall, each cell is surrounded by a special,

TifRNEu, Morphology of the Avian Bra'n/. 283

nuclei-bearing sheath. Deiter's corpuscles are small cells, which contain so small an amount of protoplasm that ordinary preparations reveal only their nuclei. These minute cells are universally distributed. The remaining three types are encountered throughout the brain; but in anyone nidulus some one type always predominates, often to the exclusion of the other two. The flask cells resemble a flask in shape, and when stained each cell presents a faintly stained nucleus, within which is a densely stained nucleolus. Such cells are supposed to be sensory in function. The pyramidal cells are sub-pyramidal in outline. These cells stain densely, when each one presents a densely stained nucleus, within which is a densely stained nucleolus. Such cells are probably motor in function. The multipolar cells resemble distorted, manybranched pyramidal cells. Such cells probably act as switch stations for nervous energy.


Fig. 1. Transverse section of the brain oi Hylocickla Szuainsona, taken through the root ganglion ot" the auditory nerve; a., tractus B of the seventh and eighth nerves; B, nidulus B of the seventh and nerves; d, nidulus d of the seventh and eighth nerves; Sg, root ganglion of the auditory nerve.

Fig..', ■', 4, <5Transverse section of the metencephalon of Sialia sia/is; \', root of fifth nerve; I'l//, root of seventh and eighth nerve.

Fig. I). Section through the cortex of the epencephalon.

Fig. 8. Horizontal -longitudinal section through the fifth nerve root of Hylocichla Sxvainsoni ; IV, root of trochlear nerve; V^, lateral sensory root of the trigeminal nerve; -5^, lateral motor nidulus of the trigeminal nerve; .5*, lateral sensory nidulus of the trigeminal nerve.

Fig. .9. Horizontal longitudinal section through the metencephalon of the domestic turkey brain, taken through the tenth nerve root; Jf, tenth nerve root; !>, nidulus of the ninth nerve; 10, nidulus of tenth nerve.

Figs, JO-lJ. Successive transverse sections of the metencephalon of Hylocichla Stvainsoni; a. tractus B of the auditory nerve; B, nidulus B of the seventh and eighth nerves; d, nidulus Y of the seventh and eighth nerves; V, root of the trigeminal nerve; VI, root of the abducens nerve; 7^ nidulus B of the seventh and eighth nerves; '.), nidulus of the ninth nerve.

Fi.gs\ 1S'14Transverse section through the metencephalon of


Col II III ha //-.■III (nestling); A. O., accesory olives; /-'. u., Fibra; arcuatic; I'/, root of abducens nerve; fV//, root of seventh and eighth nerves.

Ft'o-: l')-17. Horizontal longitudinal sections of the metencephalon of Coliniilni Uvia: Va, lateral motor fasciculus of the trigeminal nerve; \'h, deep motor fasciculus of do; IV, cervical fasciculus of do; TX, root of ninth nerve; A', root of tenth nerve; '1, deep motor nidulus of the trigeminal nerve; ii, nidulus of the eleventh nerve.

/■Vi,'. /<s'. Transverse section through the metencephalon of Agckeiis phicniceus, taken through the root of the auditory nerve; }'TI^, tractus d of the seventh and eighth nerves.

Fio-. UK Cells from the Gasserian ganglion of Hylocichla Sxvaiii-


Fig. ..'IK Diagram illustrating the course of a few of the tracts of metencephalon. No attempt has been made to represent the relative thickness of the tracts; A. C, anterior commissure; A. F., anterior peduncle of the cerebellum; C. C, corpus callosum; C. S., commissura sylvii; I). C, direct cerebellar tract; /-. F., posterior longitudinal fasciculus; F. C, posterior commissure. F. F., pons fibres (these are very few and scattered); /*. A'., peduncular nidulus; F. 7'., pyramidal tracts; 7\ F., tract from the epencephalon (see p. 125); T. T., taenfa thalami. Roman numerals indicate the corresponding nerve niduli.

Fig. Jl. Diagram illustrating the course of a few tracts of the metencephalon. No attempt is made to represent the relative size of the tracts; C. g., corpus geniculatum externum; c. //., crescent-shape nidulus; c. /., crossed pyramidal tracts; F. F., direct pyramidal tracts; /. C, inferior commissure; M. T'., mesencephalic ventricle; X. p.. nucleus posterious; F. g., prosencephalic tract from the corpus geniculatum; F. S\\, nidulus pyriformis and nidulus lenticularis; F. f'., prosencephalic ventricle; 7\ B., Tractus Bummi. Roman numerals indicate the corresponding nerve tracts. Arabic numerals indicate the corresponding nerve niduli.

Fig. ..'J. Transverse section of the brain of Sia/in sia/is.

Fig. ..'.1. Cells from lateral sensory nidulus of the trigeminal nerve of Colli mini Uvia (nestling).

Fig. 24. Cells from lateral motor nidulus of the trigeminal nerve of Columha Uvia (nestling).


The Relative Weight of the Brain to the Body in Birds. The American Naturalist., 1888, p. 537.

Bechterew, W. Nactrag zu der Arbeit, Ueber die verschiedenen Lagen und Dimensionen der Pyramidenbahn beim Menschen und den Thieren, und ueber das Vorkommenden von Fasern in denselben, welche sich durch eine friihere Entwickelung anzeichnen. Nenrologisclics Ccntr alblntt, ]a.\\rga.r\g 10, 1891, No. 4.

Ti'RNER, MorpJiology of the Avian Brain. 385

Bellonci, Josef. Ueber die Centrale Endigung des Nervus Opticus bei den Verbraten. ZcifscJirift f. IVt'ss. Zoo/o<rie^ Bd. XL VII, 1888.

BuMM, A. Das Groshirn der Vogel. Zritschrift f. Wiss. Zoologie, Bd. XXXVIII, 1883.

CiCACcio, G. N. Intorno alle piastre nervosa tinali ne tendine ne Vertebrati, 6 plates. Memorie della R. Accademia delle Scieme delV Institute di Bologna ., Series 4, Tomo X, fasc. 2, 1890.

CouES, E. Coues' Key to North American Birds, second edition, 1884, p. 176.

Darkschewitsche, L. Ueber die Kreuzung von Sehnervenfasern. ArcJiiv f. Opffinliiiolo<ric, Bd. 37, 1891, pp. 1-27.

DesmonliN!^. Anatomie des systemes nerveux des Animaux a Vertebres, two vol. with atlas, Paris, 1825.

Edinger, Luuwig. I. — Untersuchungen ueber die Vergleichende Anatomie des Gehirns. I. — Das Vorderhirn, Frankfurt, 1888.

II. — Zwolf Verlesungen ueber den Bau der Nervosen Central organe, 1889.

III. — Twelve Lectures on the Structure of the Central Nervous System, for Physicians and Students. Second revised edition, with 133 illustrations. Translated by Willis Hall Vittum, M.D. Edited by C. Eugene Riggs, A.M., M.D. Philadelphia and London, i8go.

Gagenbaur, C. Grundniss der Vergleichende Anatomie, 2 Aufl., Leipzig, 1878.

GoLGi. Sulla Fina Anatomia degli Organi Centrali del Sistema Nervoso, Milana, 1886.

GoLOWiN, E. P. On the Development of the Ganglia in the Chick. Proc. of the St. Petersburg Society of Naturalists, Section of Zoology and Physiology, Vol. XX I, book 2, (in Russian).

Herrick, C. L. I. — Notes on the Brain of the Alligator. Jour, of the Ciucinnati Nat . Hist. Societv, Vol. XII.

II. — Topography and Histology of the Brain of Certain Reptiles. yourual of Comparative Neurology, Vol. I, pp. 14-37, piIX, X, 1891.

His, W. Die Neuroblasten and deren Entstehung in Embryonalen Mark. Abhandl.d. Math, und Phys. Klassen d. h., Sachs. Gesellsch. de Wisscuschafteu, Bd. XV, IV, 1889.

Honegger, Jacob. Vergleichende Anatomische Untersuchungen ueber den Fornix und die zu ihm in Beziehung Gebrachten Gebilde in Gehirn des Menschen und der Sangethiere, mit zehn Lichtdrucktafeln, Genf, 1890.

HuYLEY, Thomas. On the Classification of Birds and on the Taxonomic Value of the Modifications of Certain of the Cranial Bones. Proc. of the Zoological Society, London, 1867.

Jergerow, S. Ueber das Verhiiltniss des Sympatheticus zur Kopfverzierung einiger Vogel. Archiv f. Anatomie uud Phvsiologie, /'hys. Ahth., Suppl. 1890.

286 Journal of Comparatue Neurology.

Leiret et Gratiolet. Anatomic Comparee du Sjst^me Nerveux, Considere dans ses Rapports avec 1' Intelligence, deux vol. et atlas, Paris, 1839-57, Tome premier par Leuret.

Marsh, E. O. United States Geological Survey. J. U. Powell, director. Third annual report, 18S1-92, p. 56.

Meckel, A. Anatomic des Gehirns der Vogel. Dciitsches Archiv f. Physiologic, von J. F. Meckel, Bd. 11, Halle und Berlin, 1816.

MoNDiNO, C. Richerche Macroe Microscopische sui Centri Nervosi. Ref. in Jahrsbericht f. Anatomic und Physiologic, von Hermann und Schwalbe, Bd. XVH, i Abth.

Nakagawa, Isaac. The Origin of the Cerebral Cortex and the Homologies of the Optic Lobes in the Lower Vertebrates. Journal of Morphology. Vol. IV, No. i, i plate.

Osborn, Henry Fairfield. I. — The Origin of the Corpus Callosum, a Contribution Upon the Cerebral Commissures of the Vertebrates. Morphologischc Jahrbuchcr, Bd. XII, 1887.

II. — A Contribution to the Internal Structure of the Amphibian Brain. Journal of Morphology, Vol. II, 1888.

Owen. Comparative Anatomy of the Vertebrates, 1866, Vol. II, p. 117.

Pansini, S. Delle Terminazion dei Nervi sui Tendini nei Vertebrati. Progresso Medico, Tomo III, 1889, pp. 66-69.

Parker, W. K. Article on Birds. Encyclopcpdia Brittaunica. ninth edition. Vol. Ill, 1876.

Pemberton, Henry Russell. Recent Investigations on the Structure and Relations of the Optic Thalmi. Jour, of Comparative Neurology, Vol. I, pp. 135-148.

Penso, Rudolpho. Sui ganglio geniculato e sui Tendini nei Verbrati. Progresso Alcdico, Napoli, Tome III, 18S9, pp. 66-69.

Perlia, Dr. Ueber ein Neues Opticus Centrum beim Huhne. Albrecht von Gracf's Archiv f. Ophthalmologic, Bd. XXXV, 1889.

Ranney, Ambrose L. The Applied Anatomy of the Nervous System. Second edition. New York, 1888.

Rabl-Ruckhard. Das Central nervensystem des Alligators. Zeitsclirift f. Wiss. Zoologic, Bd. XXX.

Serres, E. R. a. Anatomic Comparee du Cerveau dans les Animaux Vertebres, Appliquee a la physiologic et A la Pathologic du Systeme Nerveux, deux vol. et atlas, Paris, 1824-26.

ScHUFELDr, R. W. North American Passeres. Jour, of Morphology, Vol. HI, 1S89.

Stieda, Ludwig. Studien Ueber das Centrale Nervensystem der Vogel und Sangethiere. Leipzig, 1868.

Stilling. Ueber die Centrale Endigungen des Nervus Opticus. Schutze's Archiv f. Mik. Anatomie, Bd. XVIII, 1880, p. 468.

Stricker, S. A Manual of Histology. American translation, e<lited by Albert H. Buck, New York, 1872.

Journal of Comparative Neurology, Vol. I.


Journal of Comparative Neurology, V'ol. I.

/ tt


^>^<v. -


k I






» .. "[^ 16







h 22 /A -^

Journal of Comparative Neurology, Vol. I.


Journal of Comparative Neurology, Vol. I.


Journal of Comparative Neurology, Vol. I.