<8 we ers eee + it * oretateraeece * . + . e+ eee + °-+ tater * ve ¢ * eree re) 7 * » . . ve * 3 . * ae o8 . <4 % . » sae Parad . +8 4-4-8 «© LPP tr . . ‘ Paras ~ane . + et ee oe coe er a ae yt e MARINE BIOLOGICAL LABORATORY. Received Accession No..... Given by Place, *,* No book or pamphlet is to be removed from the Lab- oratory without the permission of the Trustees. yeh ok if ji a ae FN, Te JOURNAL OF omparative Neurolog A OWARTERLY PERIODICAL DEVOTED TO THE Comparative Study of the Nervous System. EDITED BY C. Li: HERRICK, PRESIDENT OF THE UNIVERSITY OF NEW MEXxICco. ASSOCIATED WITH OLIVER S. STRONG, COLUMBIA UNIVERSITY, C. Jupson HERRICK, DENISON UNIVERSITY. AND WITH THE COLLABORATION OF Henry H. DoNALDsSON, Ph.D., Universety of Chicago; PROFESSOR LUDWIG EDINGER, Frankfurt a-M.; PROFESSOR A. VAN GEHUCHTEN, Université de Louvain ; C. F. HoDGE, Ph.D., Clark University; G., CARL Huser, M.D., University of Michigan ; B. F. KINGSBURY, Ph.D., Cornell University and the New York State Veterinary College; FREDERIC S. LEE, Ph.D., Col- * legeof Physicians and Surgeons, New York; ADOLF MEYER, M.D., Clark University and Worcester Insane Hospital; A. D. MorRILL, M.S., Hamilton College; G. H. PARKER, S.D, Harvard Univ. VOLUME VIII, 1808. PUBLISHED FOR THE EDIToRS BY C, JUDSON HERRICK, GRANVILLE, OHIO, U.S. A. R. Friedlander & Son, Berlin, é ‘ 3 A ‘ European Agents. DHE JournaL oF Comparative Neurocoey. ConTEeNTS OF VoL. VIII, 1898. Nos. J and 2, JULY, 1898. Pages 1—112, i—lii. Plates I—XIV. CONTRIBUTED ARTICLES, The Finer Structure of the Selachian Cerebellum (Mustelus vulgaris) as shown by Chrome-silver Preparations. By Dr. ALFRED ScHaPER, Harvard Medical School. With Plates I—IV and one Aiaurepinithevtext 22.2 520 ee ee en Oe eee Physiological Corollaries of the Equilibrium Theory of Nervous Action andu@ontrolojuby iG. Ios) EUORR TC Ke ee Ne Ce 2) The Somatic Equilibrium and Nerve Endings in the Skin. By C. L. Hurricc and G, EB '\CoGHILT, |) \Withe Plates Ve bx ao eee The Brain of the Fur Seal, Callorhinus ursinus; with a Comparative De- scription of those of Zalophus californianus, Phoca vitulina, Ursus americanus and Monachus tropicalis. By PIERRE A. FisH, D.Sc. D.V.S., N. Y. State Veterinary College, Ithaca, N. Y. With Plates ean four fores) inthe text] ase UMN Scns The Cortical Motors Centres in Lower Mammals. By C. L. HERRICK. Ditty Tek eR a a eta eee ak OU The Nerve Cell asa Unit. By PIERRE A. FisuH, D.Sc., D.V.S., Jthaca, N. Y. With 7 figures in the text EDITORIAL. @ung Collaborators ea sea er a en mtnes Re eee ee eles yaa Oe YA es Report on Neuronymy by the Association of American Anatomists -eeee LITERARY NOTICES. Morphology : Cranial Nerves of Chimera The Irigemino-facial (Complex of, Fishess=_ 222 loess The Vagus Nerve of Teleosts The Ampulle of Lorenzini 2a 32 57 92 99 Iit 112 The Brain of the Sturgeon he \(@erebellum of \Fishes2320 2 28555 De See epee ee The Development of the Retina according to Cajal Physiology : The Function of the Protoplasmic Processes of Nerve-cells Junction of Vagus with Superior Cervical Ganglion Degenerative Changes after Resection of the Vagus___________. Secondary Changes in the Primary Optic Centres in Case of Bul- busWAtropibiyy 2.02 es eee ee eae See eee 2s 08, aS On the Alleged Atrophy of the Nasal Epithelium after Section of the ‘Olectory \Werver oo ee ee eee Relations between the Nose and Sexual Organs Psychology : Absolute Sensitiveness of Various Parts of the Retina when the Eye is Accommodated for Darkness whe Psychology, of inventions: - 2.222 See oe ee eee Projection of the Retinal Image Color (Mixing in. the Byets o2 po ee ee ‘The Emotional)'Content of (Dreams a ee Technique : Experiments with the Weigert Methods Pathology : Mills”, Practical Neurclogy 2-0 > 2. Be ee eee .Chapin’s Compendium of Insanity Genesis and’ Nature of Hysteria 4-272 seo ee Neuro-Pathology and Heredity____-------- SRE ie Pee heal Fleury’s:/ Mental Medicines. Jat. so. oh ee a ee The Truth About Cigarettes Miscellaneous : The. Journal’of; Applied | Microscopy: 3... 24 eee es ee Neurologic Mlionninology 225 tose eee ee eee *etaGeue Errors and Omissions in Wilder’s Neural Terms_.____--__--_--. No. 3, NOVEMBER, 1898. Pages 113—248, liii—lxviii. Plates XV—XIX. Critical Review of the Data and General Methods and Deductions of Mod- ern Neurology. By Dr. ADOLF MEYER, Worcester Lunatic Hos- pital, Worcester, Mass. Part I, With Plates XV to XIX. --_-____ A Report of the Neurological Seminar of the Marine Biological Labora- tory, Wood’s Holl, Mass. Season of 1898. By A. D. Morri 1, Hanilian College yo) idutinsee te Ce Ae eG rn The Problem of the Vertebrate Head. By H. V. NEAL, Knox College XxX xl viii xl vii xl viii xl viii xlix li 113 149 (5) The Cranial Nerves of Bony Fishes. By C. JuDsSON HERRICK, Denison University and Pathological Institute of the New York IS EEALERPELOS PIE LES ee es ee Sey ee eee ens aa Review of Johnston on the Cranial Nerves of the Sturgeon. By OVS STRONG, . Columbia \Un2versty. =. = a ee Review of Allis’ Paper on the Cranial Nerves of Amia. By Miss CORNELIA M. Criapp, Vt. Holyoke College..__-...------------ The Giant Ganglion Cell Apparatus. By ULRric DAHLGREN, VET ARIEELOM WCLILEUEV SIL Y sents hy ee as Sees ee oe es Innervation of the Olfactory Epithelium. By A. D. Morri Lt, LEIGH LS US REP SE IE BE ee ee The Giant Ganglion Cells in the Spinal Cord of Ctenolabrus ad- spersus (Walb.-Goode). By PorTER E, SARGENT, Harvard NPAT EF SLY eet a RN a a ea On Variations in the Distribution of the Spinai Nerves Entering the Lumbar Plexus. By C. R. BARDEEN, The Johns Hopkins OPT ogy LEU TL so a oe ee Notes on the Peripheral Nervous System of Molgula manhattenis. By G. W. HuntTER, JR...--- ----------------------------- The Elements of the Central Nervous System of the Nemerteans. By THos. H. MontcoMERY, JR., PH. D., Universtty of Penn- sylvania..----------===--—-~-----_- —=-- ------~=~- ------=- On the Nerve Terminations in the Selachian Cornea. By CREss- WELL SHEARER, McGill University, Montreal....-.---------- Some Nervous Changes Accompanying Budding in Dero vaga. By T. W. GALLOWAY-______.--.-.-~----.---- ---- ---------- Epidermal Organs of Phascolosoma gouldii. By MArcaretT L. NICKERSON, University of Minnesota...-------------<----- The Histological Structure of the Eyes of Cubomeduse. By EpD- WAR Dl lga DER GER So ata en te ae eae a ie ee A Contribution to the Nervous System of the Earth-worm. By eit ICING 52 < 2 ie oo Se eee ane ea a A Comparative Study of the Functions of the Central Nervous System of Arthropods. By Albrecht Bethe. Translated from the German by W. W. NoRMAN.--.--_--------------- The Functions of the Otocyst. A Review. By E. P. Lyon-_---- An Experiment to Test Recent Theories as to Movements of Nerve Cells. By HENRY H. GODDARD, Clark University EDITORIAL. Announcement CRITICAL DIGEST. Review of Recent Text-Books of Anatomy and Pathology of the Nervous System. First Article. By Dr. ADOLF MEYER---- liii (6) \ No. 4, DECEMBER, 1898. Pages 249—336, Ixix—Ixxvi. Plates XX—XXI. Critical Review of the Data and General Methods and Deductions of Modern Neurology. By Dr. ADOLF MEYER, Worcester Insane Hospital, Worcester, Mass. Part II, With Plates XX and XXI_- Observations on the Weight and Length of the Central Nervous System and of the Legs, in Bull-frogs of different Sizes. By Henry H. Donatpson, Professor of Neurology in the University of Chicago_- LITERARY NOTICES. Physiology : Functional:Changesin\/Nerve ‘cells 23 2s ee eee Morphology: The Organ\of Jacobson-in, Mammals) 2222s eh eceeosee= == Rugeion the Facial»Netvées=+ JoJo ee eee Relation of the Chorda Tympani to the Geniculate Ganglion___--_ Cole’on' the Nerves of ‘the (Cod Wish= Sos cease a see ere 249 314 Ixxiv THE FINER STRUCTURE OF THE SELACHIAN CER- EBELLUM (MUSTELUS VULGARIS) AS SHOWN BY CHROME-SILVER PREPARATIONS. By Dr. ALFRED SCHAPER, Harvard Medical School, Boston, Mass. (WirtH PLAtTEs I TO IV AND ONE FIGURE IN THE TEXT.) During my stay at the Marine Biological Laboratory at Woods Hole, Mass., in the summer of 1897, I found opportu- nity to extend my comparative anatomical studies on the cere- bellum over the various species of Selachii to be found on that coast. I especially endeavored, among other things, to eluci- date by means of the Golez method the finer structure of the cerebellum of this group of vertebrates, so important from a comparative anatomical standpoint. Notwithstanding the zeal with which the various representatives of the vertebrate phy- lum have been investigated by means of the Golgi method dur- ing the last decade, the selachian brain has hitherto, strange to say, scarcely been included within the range of research. The first and only author so far who has given us a connected de- scription of the histological structure of the selachian brain is, as far as I know, &. Sauerbeck (4). The specimens at this author’s disposal were prepared by Professor Rudolf Burck- - hardt of Basel, under whose direction the work was prosecuted. Sauerbeck’s results appeared in an article in Band XII of the Anatomischer Anzeiger under the title ‘‘Beztrage sur Kenntntss vom feineren Bau des Selachierhins.” Although this work, as the author himself states, makes no claim to completeness, yet it is the first attempt to fill in this existing gap in our compara- tive histological knowledge of the central nervous system and thereby can serve as a worthy starting-point for further and more thorough studies in this field. 2 JOURNAL OF COMPARATIVE NEUROLOGY. In all researches on the central nervous system by means of impregnations the elements of the cerebellum, as is well known, offer the greatest difficulties. A considerable number of preparations are usually required to demonstrate all the elements of this organ. This peculiarity appears to have pre- sented itself also in the preparations at Sauerbeck’s service, as, judging from his figures, all the other parts of the brain seem to have been much more completely and certainly impregnated than the cerebellum. Thus Sauerbeck was able to demonstrate only Purkinje cells and ependyma elements in the cerebellum. He has, indeed, on the ground of stained preparations confirmed the presence of a molecular and a granular layer in agreement with earlier authors ( Viault (8), Sanders (3), and Rohon (2) ), yet he did not succeed with Golgi preparations in learning anything of the elements constituting them. Thus there remain in reserve many important points for further investigation to determine. As I have several years since directed my special attention to the morphology of the cerebellum and contemplate subjecting this part of the brain in all the vertebrate types to a compara- tive anatomical and comparative embryological investigation, the present gap in our knowledge of the selachian brain, so im- portant for my purpose, became so much the more perceptible to me. It was natural, therefore, that I should eagerly seize the opportunity, presenting itself to me in Woods Hole, of fill- ing in, as far as lay in my power, these gaps by means of suitable investigations. My effort’ toward this end were attended with some success and I take the liberty to report briefly upon them in the following. Although I propose to publish here the results obtained from silver preparations only, yet I might for general orienta- tion, preface this with a few words upon the morphology of the selachian brain. In so doing I confine myself to that of Mustelus vulgaris (Galeus canis) from which species my Golgi preparations were exclusively made. As with most of the Selachii, in Mustelus the cerebellum is greatly developed and is traversed by numerous transverse’ folds. It might from this appear that the cerebellum of the ScHaPer, Structure of Selachian Cerebellum. 3 shark already approaches that of the higher vertebrates in its morphological structure. This, however, is not the case. On a closer examination we soon learn that we have to do only with a plate thrown into transverse folds and enclosing a relatively roomy cavity, not with a solid organ lke that of birds and mammals. The essential feature in the cerebellum of the latter is the massive development of the white matter, which goes hand in hand with an extraordinarily complicated surface fold- ing of the cortex, whence the formation of the ‘‘ arbor vitae” so characteristic of birds and mammals. There zs none of this present in the selachian cerebellum. There is expressed, indeed, in the folding an evident tendency to an extension of the sur- face, yet these folds cannot be directly (also not genetically) homologized with the, ‘‘convolutions”’ and furrows of the cere- bellum of the higher vertebrates. In the first case we have, in general, at least, to do with actual folds, i. e. with structures where an infolding on one side of the lamella corresponds to an outfolding on the other, while in the second case we are con- cerned with sod ridges and protuberances separated by fur- rows. Concerning the finer structure, the presence of the three cerebellar layers typical for all vertebrates—the molecular, Pur- kinje and granular—has been already demonstrated by the ear- lier authors, as Viault (8), Sanders (3) and Rohon (2). I have to add, however, that the granular layer does not everywhere participate to the same extent in the structure of the cerebellar lamella. Indeed in definite and extensive portions of the latter it is entirely absent, in consequence of which the cortical matter in these places consists of the molecular and Purkinje layers only and the latter is separated from the ependymal layer by only a relatively thin zone of fibers. The Purkinje cells likewise are absent in certain regions of the lamella. This relation of the layers is in various respects of especial interest but I cannot enter into this here. The accompanying schema of a transverse section through the anterior portion of the cerebellum of Mus- telus may serve to illustrate this condition. 4 JouRNAL OF CoMPARATIVE NEUROLOGY. I might however add that in the Selachii during life, as in younger stages in bony fishes and in certain embryonic periods of all vertebrates the cerebellar lamella has a deep longitudinal Jurrow tn the median line (sulcus medianus) where the extremely thin lamella is composed solely of ependyma cells and com- missural fibers. ae \ Cape ewe, Diagram: Cross-section through the anterior part of the cerebellum of AM/ustelus vulgaris showing the arrangement of the different layers. I pass now to a description of my silver preparations and begin with the most characteristic cell group of the cerebel- lum, the Purkinje Cells. The presence of these cells has been already established by the older investigators (Viault, Sanders and Rohon). Sauerbeck was the first, though, to observe and describe them in Go/lgz preparations. He speaks of them as follows: ‘‘ Von der Mem- brana limitans interna ab gerechnet im zweiten Drittel der radi- alen Ausdehnung (der Kleinhirnlamelle) finden sich typische Purkinje-Zellen, die nach innen, d. h. gegen die nicht versilberte ScHAPER, Structure of Selachian Cerebellum. 5 Kornerschicht hin, einen Axencylinder senden, der sehr bald in die horizontale, resp. tangentiale Richtung umbiegt ; nach aus- sen ragen armleuchterartige Dentritenfortsatze, doch sind diese nicht so reich verzweigt wie bei hoheren Vertebraten, wie auch schon bei den Teleostiern Schaper sie abgebildet hat.”’ The figures given by Sauerbeck are, in consequence of their small scale, but little adapted to reproduce the characteristics of these cells. I have therefore again. represented two typical forms of Purkinje cells in Plate I, Figs. 1 and 2. We see from these that the size and form of the cell-body, as well as the magnitude and arrangement of the protoplasmic processes, are subject to considerable variation. This is correlated in part with the very variable thickness of the molecular layer; we meet the more massive and coarser type (Plate I, fig. 1) mostly in the more strongly developed portions of the molecular layer, the smaller and more delicate type (Plate I, fig. 2) mostly in the thinner regions of this layer. As Sauerbeck rightly remarks, the protoplasmic processes are not so richly branched as is usually the case among the higher vertebrates and one can compare them in this respect with those of the bony fishes, as some time ago (5) described and figured by me. They are distinguished, however, even from the latter by a still more sparse arborization and their less straight, irregular course to the surface. One could say they stand upon a still lower plane of development than those of the bony fishes and display throughout life an embryonic condition of the homologous cells of the higher vertebrates. Like the dendrites of all Purkinje cells, these are provided witha thick covering of very fine spines and usually tend to terminate with a slight thickening. Concerning the axzs-cylinder, this also presents certain note- worthy peculiarities. With respect to their course, they be- have differently according as they come from cells which have a granular layer beneath them or from such as, in the absence of the granular layer, lie close to the membrana limitans tn- terna. (See text-figure, p. 4). In the latter case (Plate I, fig. 1) the axis-cylinder is naturally compelled, soon after its origin from the under pole of the cell, to bend laterally and proceed, 6 JOURNAL OF COMPARATIVE NEUROLOGY. parallel with the “tans interna and immediately beneath the row of Purkinje cells, to its destination. In the preparation upon which figure 1 is based I succeeded in following such an axis-cylinder a considerable distance. In this way ‘the neuraxons of those Purkinje cells situated in the cerebellar plate where a granular layer is lacking furnish the principal contingent of the often considerable mass of parallel fibers interposed between the layer of Purkinje cells and the “mutans interna. But even where there is a granular layer, the axis- cylinder does not as a rule immediately sink into this; but first proceeds a short distance in a horizontal direction and then takes a sharp bend inward (fig. 2). It is now to be noted that I never succeeded in any of my preparations in demon- strating collaterals of the axis-cylinders of the Purkinje cells, although, as far as I know, their presence has been established in all other vertebrates hitherto investigated. Whether we have here a defective impregnation or whether these collaterals are actually absent in the Selachii, and their absence is perhaps . the expression of a lower phylogenetic stage of development, I do not venture to decide at present. Sauerbeck likewise does not mention collaterals in his article. Among his figures of Purkinje cells, however, are some with divided axis-cylinders. May one of these branches represent a collateral ? Nerve-cells of the Molecular Layer. These cells are for the most part uncommonly delicate and vary extraordinarily in the form and size of the cell-body as well as in the mode of branching of their protoplasmic processes. To properly illustrate this variety of form I have represented a larger number of cells in Plate I, figures 3 to8. They lie in all levels of the molecular layer, from the layer of Purkinje cells (even pressing in between these) to close to the surface. The rule seems to obtain that the smaller the cells and the shorter their dentrites, the closer they lie to the surface and wzce versa. The dendrites display in general the tendency to extend towards the surface, although in the deeper lying cells there is also a considerable extension of the same horizontally (Plate I, Fig. 3). ScHAPER, Structure of Selachian Cerebellum. 7 While the cells vary in form, yet the xeuraxons in all the cells observed by me have essentially the same behavior. The nervous process soon loses its individuality through the giving off of very numerous lateral branches and is resolved a short distance from the cell body into its terminal arborization. Thus all the cells of the molecular layer belong to the so-called ‘* Golgz type.”’ In course and mode of distribution, however, the ramifications of these axis-cylinders display a certain variability; we find axis-cylinders which proceed horizontally a short dis- tance and thereby give off lateral branches outward and inward (Plate I, figs. 3 and 7), others which descend and soon are lost in their terminal arborizations (Plate I, figs. 6 and 8) and others again which run outward in wide curves and then fall into more or less numerous terminal branches which mostly ex- tend toward the surface (Plate I, figs. 4 and 5). The question now arises: What relation do these nervous processes bear to the Purkinje cells? We know definitely that in the higher vertebrates, a least, a certain number of the cells of the molecular layer, the so-called ‘‘ basket cells’? enter into very close relation with the bodies of the Purkinje cells, em- bracing the latter with tassel-like terminal arborizations. Be- sides these, another group of cells of the molecular layer has long been known under the name of the ‘‘star shaped” or ** small cortical cells’? about whose axis-cylinder little was actu- ally known. Stohr (7) first succeeded about a year ago in dem- onstrating in silver-preparations of the human cerebellum well developed nervous processes on the ‘‘small cortical cells” also, whose terminal arborizations behave toward the Purkinje cells very much as the basket cells do, but without forming the typical basket of fibers. Stohr is thereby inclined 4o abandon the division of the cells of the molecular layer into basket cells and small cortical cells and to unite them all in one group with the common character that all, notwithstanding great differences of Sorm, display the tendency to enter into close contact with the bodies of Purkinje cells by means of the ramifications of their axts-cylin- ders. In this view Stohr finds himself in agreement with Dogiel (1) and Kélliker who shortly before had similarly expressed 8 JOURNAL OF COMPARATIVE NEUROLOGY. himself. Yet not long ago an article of Smirnow’s (6) appeared, also an investigation of these elements of the cerebellum of man, as well as of the dog and hare. The results brought the author to the conclusion, in opposition to Stohr’s view, chat two distinct kinds of cells are to be distinguished in the molecular layer according to the behavior of the axts-cylinder. The figures given in this work appear to me also to plainly show that there are cells in the molecular layer whose axis-cylinders do not bear compar- ison with those of the cells hitherto described as ‘‘ basket cells.” In a number of these cells (in man also) the nervous process a short distance from its origin breaks up into numerous terminal arborizations which may be distributed in all directions. The field of distribution of many of these axis-cylinders does not ex- tend down to the bodies of the Purkinje cells and if certain iso- lated branches do come into the neighborhood of the latter, yet, in view of the distribution in many directions of the remaining branches, this is to be regarded as an incidental appearance. The-tassel-like arborization of the collaterals so characteristic of the basket cells can nowhere be demonstrated, at all events, with these cells. There appears, after all, to be only one char- aateristic applicable to all the cells of the molecular layer, viz.: that their axis-cylinders soon break up into their terminal arboriza- tions and generally do not leave the region of the molecular layer. The sfeczal behavior of the axis-cylinders of the zzdividual cells, ‘however, that is the definite relations of some to the bodies of the Purkinje cells and the correlated typical adaptation of the terminal arborizations to the latter and the entirely disorderly mode of distribution of other cells, compels, or at least entitles us, in agreement with the conclusion of Smirnow previously given, to distinguish two kinds of cells in the molecular layer, 7. e. ‘‘basket cells” and the others accurately described by Smirnow which we still for the present may term ‘‘ small cortical cells.” It is not thereby rendered necessary to create a fundamental dis- tinction between these two categories of cells. It even appears probable to me that the basket cells are merely to be regarded asa specialized form of the molecular cells. Nevertheless the morphological differences in the behavior of the axis-cylinder ScHAPER, Structure of Selachian Cerebellum. zg expressed in the higher vertebrates make the above division desirable from practical grounds. I have entered into a discussion of this point more at length here to secure a basis for the elucidation of the relations which we have encountered in the above described cells of the molecular layer of the selachian cerebellum. As we have seen, the axis-cylinders of none of these cells exhibit the typical con- duct of those of the basket cells; never were tassel-like term- inal arborizations demonstrable. In certain cells lying in the deeper zone of the molecular layer, one sees here and there isolated terminal branches penetrate between the bodies of the Purkinje cells but there cannot be said to be the intimate and extensive contact-relation to the latter that there is in the case of the basket cells. Besides this, there are usually present on such axis-cylinders numerous other terminal branches which do not proceed towards the Purkinje cells. In the ma- jority of the cells of the molecular layer we saw the ramifica- tions of the axis-cylinder spread out in all directions without ever entering into any relation with the bodies of the Purkinje cells. Thus all cells observed by me in the molecular layer of the selachian cerebellum more or less resemble—apart from a slighter complexity in the ramtfication of both dendrites and axts-cylinder —those cells which Smirnow has recently described in man and the higher vertebrates and distinguished from the true basket cells. Wf we now, as mentioned above, will regard the basket cells merely as a particular specialization of the cells of the molecular layer, this specialization has not yet appeared in the cerebellum of the Selachii; these cells are here still in a more primitive or phylo- genetically younger stage of development. I might use this op- portunity to mention that I likewise have not hitherto succeeded in demonstrating true basket cells in the molecular layer of the cerebellum of Ze/eosts (5). Probably the state of affairs here is similar to that in the Selachii. Nerve-cells of the Granular Layer. The granular layer in the cerebellum of Selachii has hith- erto been established as such and homologized with the cor- 10 JouRNAL OF COMPARATIVE NEUROLOGY. responding layer of the higher vertebrates from stained prep- arations only. For such a homologization the proof was yet to be brought that this layer also contained similar e/ements and that these elements exhibited the same behavior as in the higher vertebrates. I have now succeeded in my preparations in furnishing this proof, viz.: that there are both ‘small granule cells’’ (granule cells sensu strictiort) and ‘‘large granule cells” (so-called Golgé cells of the cerebellum) in the granular layer of the Selachit. Regarding the small granule cells (Plate II, figs. 9, 10 and 11) the cell body and protoplasmic processes behave through- out like those of the higher vertebrates. The relatively small, mostly round or polygonal body sends in all directions a limited number (usually 3 or 4) of delicate protoplasmic processes of which the majority terminate in the way familiar in the other vertebrates, by means of a claw or brush-like structure. The extremely thin zeuvaxon arises in the majority of cases from a protoplasmic process and then winds zig-zag be- tween the other granules, usually proceeding directly to the molecular layer. When it reaches the molecular layer, the axts- cylinder divides in a'\ in the manner typical for all vertebrates. The two branches proceed in opposite directions parallel to the surface of the cerebellum and the long axis of its folds and finally (very probably a considerable distance from the point of bifur- cation) break up into their terminal arborizations (Plate II, fig. 12). Ihave unfortunately only once succeeded (notwithstand- ing a most careful examination of my sections), in following one individual axis-cylinder of a granule cell without interruption from its origin to its point of division in the molecular layer. But I have repeatedly come upon the pieces of these [T-shaped bifur- cations, as shown in figure 12 (Plate II). There can scarcely be any doubt as to the connection of these fragments with the axis-cylinders of the granule cells and I have no hesitation in declaring, in spite of any inadequacy of observation, that the behavior of the axis-cylinder of the small granule cells in the cere- bellum of the Selachit ts identical throughout with that in the other vertebrates. I might mention here a certain peculiarity found ScHaPer, Structure of Selachian. Cerebellum. IL in the axis-cylinder of a granule cell, which I have figured in figure 9g (Plate II). We see here the nervous process divide fork- like while still in the granular layer. J recollect having once be- fore happened upon a similar condition in the cerebellum of a mammal. We have here, doubtless, a developmental anomaly viz.: an abnormal division of the axis-cylinder. I have above called attention to the fact that most of the nerve. processes of the small granule cells appear to run in a more or less divect course to the molecular layer; frequently, how- ever, one meets with cells also where the axis-cylinder first pro- ceeds horizontally for a considerable stretch and then wends up- ward in a wide curve (Plate II, fig. 11 below). I have unfortu- nately not been able to follow such an axis-cylinder into the molecular layer. I conjecture however that these axis-cylinders are devoted to those parts of the cerebellar plate where a gran- ular layer is lacking. We also find here in the molecular layer numerous axis cylinders of granule cells ascending and are thereby forced to the conclusion that these portions of the cere- bellum are supplied: by the granule cells of other regions. It remains to be mentioned that the neuraxons of the granule cells on their entrance into the molecular layer appear to gain somewhat in caliber and are more thickly studded with varicosities than inthe granule layer (Plate II, fig. 11). Shortly before their termination in the molecular layer they usu- ally turn upwards abruptly and thus send their terminal arbori- zations to the upper portions of this layer (Plate II, fig. 12). The ‘‘ large granule cells’’ (Plate II, figs. 13, 14, Plate III, fig. 15) are, asin the other vertebrates, cells of the ‘‘Golgz-type,” i. e. those whose axis-cylinders break up in their terminal arbori- zations soon after their origin. The very voluminous cell-body is usually round and in size often exceeds that of the Purkinje cell. The extremely coarse protoplasmic processes are small in number and exhibit no great tendency to branching. The nervous process arises in the majority of cases directly from the cell body, only seldom from a dendrite (Plate II, fig. 13). Its terminalarborization is less complex than in the higher vertebrates and is distributed in an entirely irregular manner among the small granule cells. 12 JouRNAL OF COMPARATIVE NEUROLOGY. Nerve Fibers. Although the course of the nervous processes of the various types of cells has already been described individually, yet it might: be advisable to consider again here in particular the behavior of the neurites in their totality and mutual relations, especially as these relations are of a peculiar nature in the cere- bellum of the Selachii. Besides this there are still those axzs- cylinders to be considered which enter the cerebellum from other parts of the central nervous system. The presence of such nerve fibers in the cerebellum of selachians, also, is to be assumed a priort both from the standpoint of comparative anatomy and from the necessity that external impulses must be transmitted to the cerebellum. Notwithstanding this, the positive proof of these fibers has presented the greatest difficulty to me. After a painstaking search through my preparations, I have only been able to actually demonstrate one isolated fiber unquestionably of this kind. This one is shown in figure 16 (Plate III). We see this fiber ascend in an irregular course through the granular layer and break up in its terminal arborization in the molecular layer. This very meager demonstration of the existence of “ascending fibers’ in the selachian cerebellum must suffice at present.- In its caliber and morphological characteristics this axis-cylinder observed by me scarcely differs from that of the Purkinje cell. Nothing was observed of ‘‘ moss-léke outgrowths” at definite intervals, such as were described by Ramon y Cajal and others on the ascending fibers of Mammalia (fibres mous- SCUSES). The tangle of nerve fibers in all layers of the cerebellar plate is infinitely complicated and it is, in fact, very difficult to find one’s way. As already mentioned above, the neuraxons of the Purkinje cells, before they enter the granular layer, usu- ally proceed a longer or shorter stretch in a horizontal direction beneath and in the layer of the Purkinje cells. In this way they form a dense nervous plexus (Plate III, fig. 17, above). This plexus is especially strongly developed where there is no gran- ular layer under the Purkinje cells, in consequence of which a// a ScHaPER, Structure of Selachian Cerebellum. 13 the axis-cylinders are obliged to pursue their destination along the membrana limitans interna. If we further consider the granular layer, we find here in general a dense tangle of fibers proceeding in all directions (Plate III, fig. 17); in the upper part near the layer of Purkinje cells a perpendicular direction of the fibers does indeed prevail, in the deeper layers, however, they cross each other without any order. In figure 17 (Plate III), which illustrates these re- lations, the delicate axis-cylinders of the granule cells are omitted; the larger fibers present come in part from the Pur kinje cells, another part, however, undoubtedly belongs to the “‘ ascending fibers.”’ As mentioned above, there do not appear to be any characteristic marks attached to these two groups of fibers, so that they can only be distinguished by following them to their origins or to their terminal arborizations. This is, nat- urally, only practicable in a few fibers and in the cerebellum of selachians is especially difficult from reasons mentioned below. The course of the fibers within the granular layer is some- what differently arranged in those parts of the cerebellar plate where the former does not border directly upon the membrana limitans interna, but is separated from it by a compact layer of medullated fibers (the first rudiment of a central white matter) and where, besides, externally the Purkinje cells are asa rule entirely lacking so that the granular layer borders immediately upon the molecular layer, the latter being usually very thin in such places (Plate III, fig. 18). Here one sees, instead of an irregular network, well marked fbe7-bundles at definite intervals ascending in a vertical direction within the granular layer. They stand in connection below with the basal fiber layer and extend upward to the boundary between the granular and molecular layers or also somewhat into the latter. Here they end, as though cut off, usually with a small hook-like bend. The fiber-bundles are so compact that they traverse the gran- ular layer as completely closed masses. The axis-cylinders of the granule cells do not participate in their formation but proceed upwards between the bundles in the usual manner. They are not drawn in figure 18 (Plate III). I have unfortu- 14 JoURNAL OF COMPARATIVE NEUROLOGY. nately not been able to ascertain with perfect certainty the further course of those fibers. Only this much is certain, that they turn in a horizontal direction along the boundary between the granular and molecular layers and proceed a distance further here in the form of isolated bundles. The roundish cross- sections of these fiber-bundles which I have met with lying at certain intervals from each other between the molecular and granular layers demonstrate this sufficiently. I conjecture that these fiber-bundles now gradually lose their individuality and finally go over into the thick nervous plexus which we saw locally so greatly developed beneath the Purkinje cells in other regions of the cerebellar cortex. I further conjecture that these bundles contain both centrifugal and centripetal fibers, i. e. both the axis-cylinders of Purkinje cells and ‘‘ ascending fibers,’’ or, in other words, all those tracts which, by means of the crura cerebelli, furnish the functional communications of the cerebellum with the other portions of the central nervous system. Concerning the fibers of the molecular layers, here the as- cending axis-cylinders of the small granule cells and their hori- zontal branches are especially prominent. The former form locally a dense forest, especially in those parts of the cerebellar plate where the Purkinje cells are lacking (Plate II, fig. 11). The axis-cylinders, as mentioned before, here increase some- what in thickness and are quite thickly beset with varicosities. The nervous processes of the ‘‘ cortical cells’’ also participate in the tangle of fibers in the molecular layer and likewise the terminal arborizations of the ‘‘ ascending fibers,’ of which I have brought to view but little. I might mention here one peculiarity which all these fibers appear to possess which pass from the granular into the molec- ular layer and wice versa. This consists in the fact that all these fibers in their passage from one layer to the other usually undergo a double (bayonet shaped) flexure in that they bend for a shorter or longer stretch into a horizontal direction, along this intermediate zone, and then after a second bend, about at a right angle, enter the other layer. The cause of the difficulty ScHAPER, Structure of Selachian Cerebellum. 15 with which an individual fiber can be followed from one layer to the other in sectzovs is now clear from this characteristic of the neuraxons, in the transition zone the fibers usually leave the plane of the section and are thereby withdrawn from further observation. It thus comes about that we frequently see in silver preparations that the majority ofthe fibers ascending and descending through both layers end in the intermediate zone as though cut off, of which figures 11 and 18 (Plates II and III) furnish a clear illustration. These observations on the arrangement and course of nerve-fibers are still of a meager nature and require much amplification by means of further investigations. The study of these relations in the cerebellum of selachians is rendered considerably more difficult and complicated through the varia- tions in the combination of the different layers and especially through the entire absence of a granular layer in extensive re- gions of the cerebellar plate. The selachian cerebellum pre- sents conditions, owing to these peculiarities, very unlike, as far as I know, the cerebella of all other vertebrates. The Neurogha. The neuroglia of the selachian cerebellum is in various re- spects, both from the morphological and phylogenetic point of view, of especial interest. The supporting substance as a whole undoubtedly stands plylogenetically on a very low scale. Sau- erbeck has already succeeded in demonstrating true ependyma cells in the cerebellum of JZustelus, ‘‘ welche der membrana lim- itans interna ansitzen deren Fortsitze sich bis sur membrana lim- ttans externa verfolgen lassen.”’ We conjectured that they form the principal constituent of the whole supporting substance, not having succeeded in demonsirating other elements belonging to the neuroglia type. I can confirm the observation of Sauer- beck’s in so far as that the ependyma cells play a leading role wn the constitution of the supporting tussue of the selachian cere- bellum ; 1 have to add, though, that the majority of them are much modified in their morphological appearance, that they retain their connection with the membrana limitans externa ’ 16 JOURNAL OF COMPARATIVE NEUROLOGY. only in certain parts of the cerebellar plate and, furthermore, that other neuroglia elements also are present, which are prob- ably not derived from the ependyma cells. The most primitive forms of ependyma cells with an entirely embryonic habit we find along and in the immediate neighbor- hood of the median line of the cerebellar plate, where the me- dian furrow nearly reaches the surface. Cells of this kind are shown in figure Ig (plate III). We see here one or two proces- ses arise from a round-oval or triangular cell body, which pro- cesses may again divide and extend in a fairly straight course to the surface where they lie against the limitans externa with a conicalexpansion. These processes are entirely smooth. Usually a number of processes arising from different cells are united into a thick bundle and form in this way bulky column-like structures which come out very clearly even in simply stained preparations. Not infrequently there arises also from the under pole of the cell body a short frequently branched process. Purkinje cells are not present in this portion of the cerebellar plate. All other ependyma cells are to be essentially distinguished Srom those here described, above all because most of them have com- pletely lost their connection with the membrana limitans externa. The few which retain a permanent connection with the surface I have found, in my preparations, almost exclusively in the thinner regions of the cerebellar plate only and especially imme- diately cephalad of the transition of the latter into the velum medullare posterius. Such an ependyma fiber is shown in fig- ure 20 (plate IV). Wesee here the fiber arise in the typical way from a pyramidal cell body lying close against the membrana limitans interna, pursue an irregular course through granular and molecular layers and attach itself with a conical expansion to the limitans externa. On its way the fiber gives off several lateral branches which partially fall into numerous terminal twigs. Those supporting fibers of ependymal origin which have lost their connection with the surface of the cerebellum termin- ate in various planes of the granular layer or reach the Purkinje cells. These elements are impregnated with extraordinary ease Scuaper, Structure of Selachian Cerebellum. EZ, and usually in great numbers so that often a dense forest of them comes into view. Figure 21 (plate 1V) furnishes us an illustration of this. It shows us a closely packed multitude of fibers proceeding from the conical cell bodies on the sembrana limitans interna and ascending ina very irregular zig-zag course in the granular layer. They are beset with numerous richly branch- ing lateral branches which are closely interwoven with neighbor- ing twigs and thus form an extremely delicate and complicated supporting framework, in whose meshes are found lying the ner- vous elements of the granular layer. At their free ends also the fibers break up in a similar manner into delicate terminal ar- borizations. It is striking that the fibers, where they encounter a blood vessel, frequently are closely united to the wall of the same, as illustrated in the upper left corner of figure 21. It thus appears that these fibers here enter into similar intimate relations with the blood vessels as has already been often de- in other b! scribed regarding the processes of the ‘‘ astrocytes’ vertebrates. Where the granular layer is absent in the cerebellar plate, the ependyma fibers have a somewhat different appearance. They proceed usually more directly, have a smoother surface and give off only scattered branches during their passage through the zone of fibers. Such a fiber is shown in figure 22 (plate IV). Towards the layer of Purkinje cells we see it break up into sev- eral ‘slender terminal branches which are distributed between the bodies of the Purkinje cells and usually end with a knob or brush-like enlargement in the vicinity of the molecular layer. Besides the supporting fibers hitherto described of an un- doubtedly ependymal origin, we finda second very characteristic kind of neuroglia elements (fig. 23, plate IV), which are confined exclusively to the molecular layer and in my opinion are to be regarded as the homologues of Bergmann’s fibers in the mole- cular layer of the cerebellum of higher vertebrates. These fibers arise from irregular pear-shaped cell bodies lying between the Purkinje cells and extend in a tolerably direct course to the surface where they are placed against the limitans externa with conical enlargements. They are invested along their whole 18 JOURNAL OF COMPARATIVE NEUROLOGY. length with a more or less dense moss-like covering which pre- sents an uncommonly delicate aspect. They people the mole- cular layer in great numbers and form, when many are impreg- nated, an almost impenetrable thicket. These elements have in fact a great resemblance to ependyma fibers. One can imagine. that these fibers have retained their original connection with the limitans externa and that with the progressive thick- ening of the cerebellar plate the cell body, released from the limitans interna, is gradually withdrawn into the interior of the plate. I ‘am more inclined, however, to the view that these fibers are genetically dissimilar to the ependyma cells, that they owe their origin to a part of the derivates of the ‘‘ germinal cells’’ and are accordingly to be regarded as secondary support- ing elements or as gla elements in a narrower sense. The same obtains for the Bergmann’s fibers in general. However, as far as I know, a positive proof for this conception is not yet brought forward. Further investigations are necessary here. Should they actually prove to be true glia elements, the fibers described would represent, according to my observations at all events, the only ones of their kind in the cerebellum of selachians, inasmuch as I have not succeeded in demonstrating any elements comparable to ‘‘astrocytes’’ or ‘‘ mossy cells”’ in my preparations. It emerges from the foregoing considerations that the fun- damental structure of the cerebellar cortex of the selachians as a whole shows already the typical features of that of the higher vertebrates. A lower stage of development, however, can be established in regard to the individual cellular elements, which is expressed both in the less complexity of the dendrites and axis-cylinder and in the prevailing ependymal character of the supporting structure. Harvard Medical School, Boston, Mass., Nov. 17, 1897. ScHAPER, Structure of Selachian Cerebellum. Ig LITERATURE LIST. -_ . Dogiel, A. S, Die Nervenelemente im Kleinhirn der Vogel und Saugetiere. (Arch, mikr. Anat., XLVII, 1896). 2. Rohon, J. V. Das Centralorgan des Nervensystems der Selachier. (Denkschr. K. Akad. Wiss., Wien, XX XVIII, 1877). 3. Sanders, A. Contributions to the anatomy of the central nervous system in vertebrate animals. (Philos. Trans. R. Soc., London, CLXXVII, 1882). 4. Sauerbeck, E. Beitrage zir Kenntniss vom feineren Bau des Selachierhirns. (Anat. Anz., XII, 1896). 5. Schaper, A. Zur feineren Anatomie des Kleinhirns der Teleostier. (Anat. Anz., VIII., 1893). 6. Smirnow, A. E. Ueber eine besondere Art von Nervenzellen der Molecu- larschicht des Kleinhirns bei erwachsenen Saugetieren und beim Men- schen. (Anat. Anz., XIII, 1897). 7. Stéhr, Ph. Ueber die kleinen Rindenzellen des Kleinhirns des Menschen. (Anat. Anz., XII, 1896). 8. Véault, F. Recherches histologiques sur la structure des centres nerveux des Plagiostomes. (Arch. Zool. expérim. et générale, V, 1876). EXPLANATION OF FIGURES. All the figures are drawn with a magnification of about 240. B. Z.—Basal fiber-layer. #, Z.—Fiber-layer (underneath the Purkinje cells). G. Z.—Granule layer. Z. e.—Limitans externa. Z, z.—Limitans interna. P. C.—Purkinje cells. S. m.—Sulcus medianus. PLATE I. fig. z. Large Purkinje cells from a portion of the cerebellar plate where the granular layer is lacking. fig. 2. Small Purkinje cells. Figs. 3-8, Various forms of nerve cells from the molecular layer. PLATE II. Fig. 9. Small granule cells. fig. 7o. Small granule cells; their axis-cylinders ascending between the fiber-bundles of the granular layer. fig. 17. Section through a portion of the cerebellar plate, where Pur- kinje cells are lacking, Compactly arranged ascending axis-cylinders from small granule cells in the molecular layer. In the granular layer a number of cells, of which the under ones send their axis-cylinders sidewise ; besides these, numer- 20 JoURNAL OF COMPARATIVE NEUROLOGY. ous axis-cylinders which partly belong to the granule cells (very delicate), partly to the Purkinje cells and are partly derived from cells lying outside the cere- bellum. Fig. 12. Axis-cylinders of small granule cells proceeding horizontally and their terminal arborizations in the molecular layer. Two [T-shaped divisions of the ascending main fibers. Figs. 13-14. Large granule cells. PLATE III. Fig. 75. ULarge granule cell. Fig. 76. Ascending nerve fiber of extraneous origin and its terminal arbor- ization in the molecular layer. fig. 17. Irregular arrangement of the nerve fibers (neurites of the Pur- kinje cells and ‘‘ ascending fibers ’’) in the granular layer. fig. 78. Arrangement of the nerve fibers transversing the granular layer in distinct buudles which go over into a basal fiber-layer. Fig. z9. Ependyma cells from the neighborhood of the median furrow of the cerebellar plate. PLATE IV. Fig. 20. Modified ependyma fiber traversing the whole thickness of the cerebellar plate. fig. 21. Modified ependyma fibers in the granular layer. fig. 22. Modified ependyma fiber from a region of the cerebellar plate where the granular layer is missing. The slender terminal arborizations of the fiber penetrate between the Purkinje cells. fig. 23. Neuroglia cells of the molecular layer (Bergmann’s fibers). PHYSIOLOGICAL.” COROLLARIES, OF |THE \EQUL LIBRIUM THEORY OF NERVOUS ACTION AND CONTROL. By C. L. Herrick. Opportunity has afforded incidentally in connection with previous articles in this journal to point out the suggestions from anatomy in favor of a theory of nervous action based on the fundamental conception that the differentia of the various forms of nervous action consist in differences in the form of resist- ance and the reaction thereto, or, in other words, that nerve action partakes of the nature of equilibrium. It may now be permitted to offer fresh illustrations of the application of this principle. In the first place, however, we may note that in no department of physical science is it so plain as in neurology that we are dealing wholly with dynamic elements. While it is true that in the structure of the brain we have to do with morphological details of marvelous complexity and the descriptive side of our work is concerned with the varying outlines, sizés, and combin- ations of cells, fibres, etc., and the still more recondite struct- ures within the cell and their dendrites, yet it is always obvious that these morphological peculiarities are but the expressions of inner forces and their responses to others from without. Thus it may even be doubted whether such a body as a centro- some or, at any rate, a centrosphere exists as a material element. Authors have been content to interpret the ‘‘asters’”’ as the vis- ual evidence of differential attraction in the cytoplasm. It is possible to go farther and admit that all the structures with which the cytologist (and so the physiologist) has to deal are the visual interpretations of dynamic processes. This is more apparent to the neurologist than to the crystallographer 22 JOURNAL OF COMPARATIVE NEUROLOGY. because the former grows accustomed to observe the close cor- relation between structural differences and conscious experience whose dynamic nature it is impossible to doubt. There can be no more doubt that the morphological peculiarities of nervous or other tissue are the expression of the equilibrated forces of growth and other functions than that the form and polariscopic qualities ofa crystal represent molecular reactions. It is also ap- parent that the concept of matter in either case helps not at all in the explanation of these forces and that the attribute of ma- teriality is to be determined on independent grounds. Not to discuss the ontological question at this time, it may simply be said that in our use of the morphological terms it is only with the reservation that they are convenient expressions to define the constant elements in our experience of dynamic forms. There are many advantages in this more direct interpretation of vital phenomena, for by the interpolation of imaginary material elements between the objective force and the subjective experi- ence one loses sight of the constant dynamism—a dynamism which does not make necessary a fresh explanation of each new expression of force; for the existence of force may be regarded as self-evident when we recall that activity is the sole element of experience, and its varying forms are, in a sense, the alge- braic expressions for interactions. The whole question of troph- ism is robbed of most of its difficulties if we think of structure as not a thing dissimilar from function, but consider both as dif- ferent expressions of similar forces. It would seem that especially in the sphere of embryology we should be ready for the abandonment of the fruitless search for material grounds for persistence of type. The theory of pangens is one illustration out of many of the absurdities to which a materialist construction is driven. The observed con- formity to type observed in each of the thousand plants which may arise by minute subdivision of a moss, for example, shows how hopeless is the attempt to base on any specific material the capacity for heredity, no matter how eked out by the doctrine of latency. Correspondence in mode is the condition of iden- tity implied by a dynamic theory, and the heterogeneity ex- Herrick, Physzological Corollaries of Equilibrium Theory. 23 pressed in the forces of the body of a man may be expressed in the terms of the forces of a spermatozoan equally well. The assimilative power necessary when we assume that repeated nucleary division takes place without reduction of the chroma- tin is certainly dynamic and why should this dynamic deter- minant be limited to some material element? Does not the body preserve its integrity in spite of the flux of its materials ? Why should not the actual material of the nucleoplasm be ina similar flux while retaining its form, i. e., its dynamic attri- butes ? From this point of view the coordination of parts through the nervous system becomes only a special instance ofa _ coor- dination in the entire organism. ~It is true that even the unex- pected wealth of fibrous ramification in the nervous end-organs revealed by the various applications of the Golgi method is still insufficient to explain the perfect co-adjustment of part with part in nutritive and trophic equilibrium—in fact, auy con- ceivable completeness of nervous continua would leave some- thing to be explained, for, in the last analysis, the processes are intracellular or even cytoplasmic, Even if we should grant that unsuspected imperfections in our present methods deprive us of the power of detecting the anastomoses between neuro- cytes in the same circuit, yet the most perfect conceivable con- tinuity would still leave an appeal lying to protoplasmic trans- mission. A forthcoming paper will afford illustrations of what is here referred to. In the skin of many (probably all) amphi- bia and reptiles (Axolotl and Pizynosoma) there exists about the cells of Leydig a very complete and beautiful protoplasmic reticulum in such a way that each large cell is completely envel- oped, while the meshes commingle and pass from cell to cell. This reticulum arises from certain nucleated protoblasts which are devoid of cell wall and whose naked protoplasm fills in inter- stices between the larger cells. This reticulum is not an arti- fact for it is found by the use of widely different reagents and is most complete when the fixation of the protoplasmic structure is most perfect, and in some cases of applications of chrome- osmic + platinic chloride + alcohol solutions this perfec- 24 JouRNAL OF COMPARATIVE NEUROLOGY. tion leaves little to be desired. Ordinary hardening processes do not reveal the structure asa rule. It may be that the pro- toplasm is a delicate film which is thicker in certain parts than in others but the relation to the intercellular nuclei is certain. These are entirely distinct from the chromatophores. _Bethe’s methylene blue process reveals the farther fact that nerve fibers, which lose the sheaths after passing through the corium, end in knob-like tuberosities in proximity to these nuclei, though whether they penetrate the protoplasm or simply spread out upon it remains, from the nature of the method, uncertain. These nerve-fibers when stained with picro carmine or fuchsin, in contrast to haematoxylin nucleary stains, seem to blend with the protoplasm and it is difficult to decide which appear- ance is nearer the truth. Such close contiguity between a naked fiber and a naked protoblast is too vaguely different from continuity to require physiological separation, however impor- tant the distinction may appear morphologically. Here we have an illustration of a condition, which I be- lieve is more general than we now can demonstrate, in which a nervous end-organ is so connected with a meshwork of vast extent as to suggest a very extensive somatic influence of a na- ture similar to nervous reaction over vast tissue areas. We venture to suggest that there is no such sharp distinc- tion between nervous functioning and the intracellular processes of the ordinary non-nervous cell as our present terminology and usage suggest. It is certain that in the differentiation of function the cells of the body at large do not give up all of their heritage of nervous or nerve-like power. Students of histo- genesis may have been puzzled, as the writer has, to account for the fact that a very remarkable degree of coordinated tro- phic power is exhibed by the embryonic body prior to the de- velopment of nerve tracts and end-organs. The phenomena of nervous deficiency in anencephalic monsters is equally inexpli- caple from the standpoint of rigid limitation of coordinating power to the nervous system. In the sponges and Ccelenterata it is plain that the coordination necessary to individual existence and perpetuation of specific characters is possible with no cen- Herrick, Phystological Corollaries of Equilibrium Theory. 28 tral nervous system. There is a form of vital equilibrium so resident in the general system as to give rise to much the same phenomena of nervous unity as in the case of higher animals. It is not at all necessary to suppose that the cells of the body of higher animals have lost this power during the differentia- tion of the central system ; it would be more probable that the central system should be superadded. There are a number of classes of cells which seem to be, in the nature of the case, freed from all direct nervous control. The chromatophores of the Amphibia, to which the writer has devoted some study, seem in some cases not to be in a direct way associated with a definite nervous supply.’. They are, in- deed, literary migratory, though the scope and range of move- ment remains to be worked out. Two things may be quite positively stated; first, that these cells are to some extent inde- pendent of fixed nervous influence, and, second, that they are very really under indirect nervous control. Experiments tried in my laboratory many years ago showed that, in young cat fish, section of a branch of one of the cranial nerves destroyed the very marked adaptive power for the injured side. A fish, originally black, when placed in an aquarium with yellow bot- tom invariably changed to the color of the environment unless the mutilation described prevented it. The observations of G. H. Parker on photometric changes in retinal pigment cells of Palemonetes are interesting in this connection in showing that exposure to light causes actual changes in form and a segregation of the pigment of these cells. He finds that section of the nerve or severing the eye stalk from the body does not wholly prevent the reaction. This is an illus- tration of a reaction exceedingly resembling a true nervous response. The embryonic tissues of all animals possess this coordi- nating sensitiveness and trophic interaction to a high degree. In the extreme case afforded by the blood corpuscles and lymph- ocytes it seems perfectly plain that there can be no direct 1 Methylene blue seems to show connections in some instances, 26 JouRNAL OF COMPARATIVE NEUROLOGY. nervous relation and yet he would be a bold physiologist who would venture to deny that there is a most subtle and powerful coordination between the stationary tissues and the free cor- puscles. One may talk of chemotropism or vital susceptibility, but such terms express merely the fact that the corpuscles, like other cells, are coordinated with the rest of the body and bear both its specific and individual impress. The mysteries of serum therapy only increase our confidence in such an intimate re- lation. It, then, may be supposed that the circuit of nervous action in any part of the body passes through a variety of smaller somatic circuits and that the spheres of the two forms of activity overlap so that the return nerve current bears the influence of this interaction. The nervous equilibrium is only a central spec- ialized part of a vital equilibrium embracing all the activities of the body. The wandering cell, even though not in direct con- tinuity with a nerve fiber, nevertheless may be said to act ina ‘‘ nervous field”? and so is not beyond the sphere of coordina- tion, while, on the other hand, the results of changes in the ex- tra-nervous mechanism of the body all have their effect upon the central system. In the same way we may explain the effect of the sum of organic and total or somatic stimuli upon tem- perament and disposition. The processes of nutrition may be said to be common to protoplasm quite irrespective of nervous control, but the trophic influence of the latter is well authenticated and it may be as- sumed that no nervous action takes place without having its effect on growth. From the above it may be gathered that the ground of the mutual reaction may be sought in the fundamen- tal similarity of the two processes, or rather the close relation between the processes of waste and repair lying at the founda- tion of both. It is necessary to suppose, accordingly, that the central nervous system is continually affected by the vital phe- nomena of the body at large as truly as the vascular system is under the control of the nervous system. As a striking result of this effect of the somatic or extra- neural processes, one may take the phenomena connected with Herrick, Physiological Corollaries of Equilibrium Theory. 27 the restoration of mutilations. When the newt’s foot is ampu- tated, under favorable circumstances the organ is quickly repro- duced and the parts so restored differ in no obvious way from the old organ removed. What is the power which causes such a miraculous change ? Does it take place because a simulacrum of the missing limb exists in the soul and the new body devel- ops to correspond? With due allowance for use of terms, we reply, ‘‘yes, such a simulacrum does exist.’’ The form of the central equilibrium has been determined by constant reactions between the member and the central system and when the mem- ber is lost the equilibrium so established is still in force and the nervous stimuli ;which but lately served to supply tone to the limb now operate upon the stump. Intense irritation results and the tendency is to influence growth at the point of injury.’ This growth is under the directive control of the nerve just as we know the normal growth constantly to be. If the nerve of the limb be injured beyond repair monstrous growth results. It may be assumed that in case the leg were amputated and the nerve destroyed in the stump above, that the efforts at restora- tion might be abortive or result in monstrosities. It would be well to test this matter experimentally. It is believed that the application of the ideas indicated in this paper to the higher spheres of nervous activity will prove fruitful. Another application of the same principle is found in the processes connected with the regeneration of severed nerves. It is a well authenticated fact that, in the case of section of a peripheral nerve, the nuclei of the sheath of Schwann pass to the centre of the lumen and form the protoplasmic prota of the segments of the new nerve—a process wholly unintelligible if we agree with Kolliker in regarding the sheath nuclei as derived from non-nervous connective tissue corpuscles, but not so re- markable if the abundant evidence be accepted that these nuclei are but the diverted nuclei of the cells which formed the nerve 1 The assumption that irritation may produce proliferatlon is supported by the pathological karyokinesis in case of local irritation ; see also processes con- nected with development of spermatozoids, etc. 28 JOURNAL OF COMPARATIVE NEUROLOGY. originally by proliferation and moniliform concrescesce." We here have an instance where the protoplasm of the cells has be- come specialized and the nuclei switched out of the circuit and apparently related to the process of forming the cell wall. But, in spite of the specialization implied in the production of an organ for nervous conveyance alone, it appears that the early nature of the cells is dormant rather than lost, so that in the case of injury and the consequent degeneration of the myelin and axis ‘cylinder, the nuclei, with the small portion of less specialized protoplasm associated, return to the embryonic state and repeat the process of neuro-proliferation, after which the new channel is organized from the center outward and the nu- clei return to their parietal position. It is more than probable that a similar rejuvenescence is possible in the case of other tissues also. We have many instances of the same kind of differentia- tion within the cell. Take as an illustration the formation of glands in the skin of the frog, where a follicle is formed and then the several component cells are fused, the outlines being lost, and only the small nuclei which remain in the thin parietal layer of less altered protoplasm remain to indicate that the gland is really polycellular. It would be interesting in this case to institute experiments on the possibility of rejuvenescence of such cells. In the application of the neural equilibrium theory to prob- lems of heredity it would seem that there is a large and profit- able field. Without attempting details in this direction, it may be pointed out that this point of approach renders unneces- sary a vast deal of the most profitless theorizing in reference to heredity. If the neural and somatic forms of reaction are not absolutely unlike, but on the contrary are parts of a common vital type of energy (or rather force) and if it be admitted that the processes of nutrition may be and are influenced by the 1 The nuclei of the ending of the motor nerve on the muscle offer interest- ing collateral evidence. See the article by Dr. Huber in the last number of this Journal Herrick, Physiological Corollaries of Equilibrium Theory. 29 neural equilibrium, it follows that the germ is also situated in the field of these equilibrated forces and its composition, i. e. its own force formula, would be the resultant of the reaction of the existing (ontogenetic) formula as modifying the earlier (phylogenetic) force formula. Instead, then, of searching for ‘Gds,’’ ‘‘bioplasts,’’ gemmules,”’ or the like we may feel assured that, in a much more complete and integrated form, the entire life of the organism will have its effect on the germ. This con- fidence will not cause us to pay less attention to the structural appearance of the cell and, in particular, the germ cells, but will prevent loss of valuable effort in the invention of sterile theories and prepare the way for a dynamic interpretation of these phenomena. It may be noted in this connection that S. Ramon y Cajal has apparently suggested, by implicatian at least, some of the grounds for the equilibrium theory in his suggestive article in the Archiv f. Anatomie u. Physiologie, 1895. He says: ‘‘ Die Phanomen der vorerwahnten lawinenartigen Leitung, sowie die geringe Zahl der sensorischen Elemente (Zapfen der Fovea cen- tralis, akustische Zellen u. s. w.) welche alle die zahlreichen Eindricke, fir welche unsere Sinne empfanglich sind, aufneh- men miissen, zwingen zu der Annahme dass jede Sinneszelle, sowie jede subordinirte Gruppe von Pyramidenzellen des Ge- hirns successiv an der Production verschiedener Bilder sich be- theiligen. Vom anatomisch-physiologishen Standpunkt aus, wird eine Wahrnehmung von einanderen, zu derselben Empfin- dungsquantitat gehorigen, durch die Zahl und die betreffende Lage der corticalen in Erregung gesetzten Pyramidengruppen unterscheiden.”” It would seem to be evident from the above that not only the exact impression to be perceived is not pro- duced by the organ of sense (since it would then be divided into a large number of parts in being transferred to the larger number of pyramid cells) but also that inasmuch as the same cells may be participants in different percepts the physiological basis for the latter must be the particular formula of these per- mutations in a given case and thus a simple impression must be 30 JOURNAL OF COMPARATIVE NEUROLOGY. of the nature of an equilibrium constructed from the inter- actions of the cells implicated. In conclusion, it may be noticed that the ideas advocated above have a very interesting bearing on the problem of the origin of variation. The theory of the competition of parts has taken strong hold of modern biology because it is becom- ing more and more evident that the sphere of natural selection must be greatly restricted and some appeal must be made to forces residing within the organism. Even Weismann in his extreme advocacy of natural selection has been forced to yield a large place to the effects of inner coordinations. We suggest that the nature of these coordinations is rendered: much more intelligible by conceiving of all these vital-nutritive processes as equilibrated forces. If for any reason, a given part or tissue of the body is in the least exaggerated, its part in this complex coordination is increased and, accordingly, its reflex influence on the organism asa whole, or its nerve centers, will be in- creased and its quantum of the centrifugal currents will also be increased, so that the tendency will manifestly be for the newly created variation to go on increasing indefinitely until checked. The next generation will inherit this tendency and we should find that, in the absence of restraint, there would be the con- stant likelihood of the appearances of strange monstrosities with apparently unaccountable exaggerations of horn or spine. It requires very little familiarity with paleontology to discover that its records abound with cases in which no possible service- ability would account for the absurd burlesques which have been produced and only the comparative familiarity of existing types blinds us to the same fact. While not denying that there is a large element of useful adaptation in all cases (otherwise they would never have been preserved), yet it will be admitted that a very considerable proportion of the peculiarities and often the deeper seated characters have no such explanation. We should not be surprised at this, for it is apparent that the slightest variation not directly hurtful will tend to perpetuate itself. It may be said that all unnecessary parts will be elimi- nated as sapping the nutrition of the body at large. This is an Herrick, Physiological Corollaries of Equilibrium Theory. 31 abuse of a teleological principle for it is not to be assumed that the body is reasoning from present causes to distant effects. If an eye ceases to be used it is atrophied, not because it is no longer useful and is therefore a cumberer of the ground, but be- cause, the function having ceased, it is actually participating less in the equilibrium than formerly and also less than other organs. Buta newly formly wart on the skin may be abso- lutely useless, yet, like a corn, it may be the seat of irritative processes which stimulate nutrition. It is then not the ideal utility but the degree of participation in the vital equilibrium which is the primary determinant. It is necessary to seek no farther for the source of variation and it is not surprising, when we consider the infinite possibilities for the increased vital ac- tivity of one group of cells over another that natural mimicry has found at hand all the necessary variations upon which it is to work, though we must not hope to find in their number and variety the complete explanations of the imitations produced. University of New Mexico, Feb. 20th, 1898. THE SOMATIC EQUILIBRIUM AND THE NERVE ENDINGS IN THE SKIN. By C. L. Herrick and G. E. CoGHILt. PART ONE. WITH PLATES V—IX: Few problems have proven more attractive or more illusory than the general question as to the nature of the nerve termini in the membranes, for it would seem that our concepts of the histogenesis and so of the real nature of the sense organs de- pend very largely upon the conclusion at which we arrive as to the relation between the various types of sensory epithelium. The senior writer suggested, in a series of papers on the brain of the lower vertebrates, reasons for believing that the first sense to come into the field of consciousness was that of smell, and a little later Edinger emphasized the same idea by his in- vestigations of the olfactory tracts of the reptile brain. It may now be taken as fairly proven that, if the seat of consciousness is in the cerebrum, smell was the first of the special senses to find its way to recognition by it. It would then be natural that we should expect the peripheral organs of olfaction to retain a primitive character and so to afford us a clue to the early state of such organs. Then too the development of the accessory or non-nervous organs of sense has here hardly made any pro- gress even in those most highly differentiated cases in which Jacobson’s organ has assumed great proportions. From studies of the development of the olfactory organs in reptiles, as reported briefly in earlier numbers of this Journal, the writer has been abundantly convinced of the truth of Beard’s statement that the olfactory prota arise from the skin and, by a HERRICK-CoGHILL, Merve Endings in the Skin. 33 proliferation, extend to the brain, there to enter into communi- cation in the glomerules with the processes of the mitral cells of the tuber. As studied in the embryos of snakes the process is as fol- lows: The first indication of the change of the ordinary to the sensory epithelium is seen in the thickening of a portion of the superficial layer from the morphological front of the head (the region of the future infundibular recess) in relatively broad bands, one on either side of the head. As the head flexures increase, these areas are carried ventrad and come to occupy the roof of the mouth and adjacent parts of the buccal cavity. The development of the taste buds from this epithelium we have not traced in these subjects, though there is no reason to doubt that they are formed from this proton, as it is easy to see that the mucous part of the hypophysis is. At the time the first olfactory rudiments appear, the curvature is such that the hem- ispheres are protuberant in front and so come nearly in contact with the prota of the olfactory in the two bands of germinative epithelium above mentioned. Still there is no difficulty in see- ing that the original proliferations take place in the skin and that the constant proliferation by division of the earlier cells spins the nerve fiber from the original source to the point where the tuber subsequently arises. In fact, the tuber, which has frequently been compared to the ganglion of origin of a cranial nerve, does not seem to afford origin for any centrifugal fibers whatever. In preparations by the silver method it is easy to see that the neurite of the moniliform chain of the olfactory nerve comes into relations in the glomerules with dendrites of the mitral cells. Thougha considerable wealth of detail has been secured by study of Golgi preparations during the last few years, nothing has been brought to light to invalidate our original view. For a long time during the development of the brain an obvious ganglionic mass lies below the skin at the base of the point of origin of the olfactory. The gradual elaboration of the cavities of the nares only serves to redistribute the prota without materially disturbing the simplicity of the arrangement. 34 JoURNAL OF COMPARATIVE NEUROLOGY. In a wide range of types it has been possible to make out the adult conditions which have often been correctly described. Merkel in his classical work gives a figure of sensory endings from a cirrus of Amphioxus that compares in every detail with the specific cells of the olfactory epithelium of a reptile or am: phibian. (Plate III, figure 10.) Few if any of those who have studied the development of the olfactory will venture to deny that the ‘‘Stiftzelle’”’ at the peripheral end of the olfactory nerve is a member of the nervous series having the same origin, though it is doubless conceivable that, through some strange fatality, every observer has failed to notice the intrusion of a foreign element at some stage of the process. (Fig. 31.) If, however, we take for granted that the fiber is continuous, we claim that there is an equal necessity for admitting the same for other clusters of nerve endings on the surface of the body. Although there was for a long time considerable disagree- ment as to the actual connections of the olfactory nerve fibers, and the classical studies of Kolliker, Klein and Piana left the matter open, it seems as though the later studies of Ehrlich, Arnstein, Cajal, Gehucten, Retzius, Brunn and Lenhossek, who employed the silver and methylene blue methods, were sufficient to prove conclusively that the olfactory epithelium possesses rod cells whose proximal end is an actual continuity with the fiber of an olfactory nerve filament. The writer has frequently verified this in specimens of Amphibia double stained with hematoxylin and picrocarmine in which very unambiguous views can be secured. A few figures from these preparations were published by Mr. Bawden, then a student in the writer's laboratory (Jour. Comp. Neurol. IV). Our studies in the development of the olfactory nerve show that the proton of the nerve is formed in or under the epithelium ofthe nasal area and that the nerve grows by moniliform concrescence of cells which arise by mitosis from this proton. From this stand-point, then, it would be expected that the neurocytes of origin would be found in the epithelium. In all essential respects the relations in Jacobson’s organ are the same as in the true nasal olfactory epithelium. The accompanying figure (Plate V, Fig. 10) from an article by Lenhossék (Anatom. Anzeiger, VII, 19-20.) illus- tuates these conditions and also the fact that other nerve fibers, HERRICK-COGHILL, Nerve Endings in the Skin. 35 apparently from the, trigeminus, terminate in free arborizations between the epithelium cells. A very large following of the new school are prepared to claim that the conditions in the ol- factory epithelium are peculiar to it alone and it is even at- tempted to correlate this with a supposed fundamental differ- ence in origin and structure of the olfactory from all other nerves of the body. But we are able to show that in the epi- dermal sense buds of the tree frog and other amphibia the same continuity of nerve fiber and cell can be determined. It has not been an altogether unnatural result of the re- markable complications of nervous structure revealed by the so-called specific methods that the results obtained by the old histological methods have been discredited and it has required some year’s experience to teach us the danger of too explicit reliance on the former. Perhaps the greatest of these sources of ambiguity arises simply from the fact that has been regarded as the chief excellence of these methods, namely that the selection is so perfect that other tissues than those selected not shown at all or, even if the after-staining of sections suc- ceeds, the conditions of impregnation are so unlike that the tracing of connections or definite relations is difficult or impos- sible. The absolutely contrary results of Dogiel and Cajal in the matter of the anastomoses in the retina illustrate the diffi- culty that exists even where the methods used are similar. The results of our own studies are rather to confirm many of the old observations and to show that there are two distinct classes of dermal endings. Of these the olfactory illustrates one and the most primitive one. In this case we have to deal with the remnants of nervous aggregates which were originally formed in or near the outer layer and in the phylogenetic development have not been diverted to a deeper level as is true in so many other instances. In our laboratory in 1891 we made out the fact that in the oral region of the earth-worm there are cells in the skin which have a nervous nature and whose processes pass entad to the central system. Owing toa delay in the other aspects of the research the observation was not made public till 36 JouRNAL OF ComPaRATIVE NEUROLOGY. the brilliant work of Gehuchten had afforded proof of the same thing, but the suggestion was of course inevitable that we have in the lower forms a permanent retention of cells in the skin which in higher types have tended to become concentrated in the central organs. What more natural, however, than that this concentration should be incomplete, especially where these cells have have acquired a specific sensory function. When the application of the Golgi and methylene blue methods revealed the fact that there is a most complicated set of free endings in the skin and that in many cases where a nervous continuity had been described there is simply a secondary apposition of a den- drite to preexisting non-nervous cells it was inevitable that the existence of cellular nerve endings should be discredited en- tirely. It is true that the greater part of the sensory prota are collected in the spinal and cranial ganglia and seem to prolifer- ate thence to the periphery; but in various regions, particularly of the head, these ganglia never concentrate in a neural ridge but retain their original place in the neighborhood of pharyn- geal clefts and the like and the possibility must be allowed that other cell-clusters elsewhere may have done the same. How- ever, there is another possibility to be considered; namely, that the terminal portion of the peripherally proliferating nerve fiber may under certain circumstances develop a specialized terminal dendrite. When the nerve is in process of developing the sub- division of the distal member is repeated progressively until the definite terminus is reached and then the extreme element is charged with the function of adapting itself to the conditions there prevailing. Inthe case of the motor ending, even the careful researches of Huber and De Witt do not finally dispose of the question as to the origin of the end-structures. We may interpret them as follows: when the fiber reaches the muscle its terminal element, together with the nucleus, applies itself to the surface of the latter and prior to the formation of the mus- cle-sheath, proliferation goes on in a less regular way than dur- ing the development of the nerve itself, in this way is formed the ‘‘sole,’’ which would, accordingly, be of a nervous nature. On the other hand, it is possible that the nerve on entering the HERRICK-COGHILL, Werve Endings in the Skin. 37 muscle comes in contact with a nucleus of the muscle which, under the stimulus afforded, begins to proliferate and the pro- toplasm of the cells so formed assumes an intermediary char- acter and spreads out upon the surface of the muscular band as a means of applying the stimulation. To us the first is in the absence of direct evidence the more probable solution. Observations are at hand which tend to show that exten- sive nervous proliferation takes place below the corium of the skin at an early stage. In section of the skin of Amphibia these proliferating cells can be seen and this is probably the origin of the ganglion plexus of the skin. (Figs. 3, 5 and 6, Plate V.) To pass then to the nerve endings in the skin, we may first note the isolated sensory cells. These may be seen in suitably prepared sections of the head in the tree frog and other Anura and also in the neighborhood of the eye in the axolotl and other tailed Amphibia. In the tree frog, where they most numerous, these cells are grouped in threes and fours in close clustres ly- ing in a special cavity passing through the entire thickness of the epithelial layer. The terminal segment is a slender nucle- ated cell, the nucleus being very narrow. The peripheral part of the cell is a narrow rod which at the periphery bears a few rigid bristles. Entally from the nucleus the cell walls are very delicate but obvious and the nerve fiber within is easily disting- uishable in the doubly stained specimens. The fiber is easily followed to the corium layer and in many cases through it. It seems too that more than one nucleus can be seen in the course of the fiber before the passage through the corium. The skin is at this point very thick and the presence of large glands serves to separate the corium from the epithelial layer, so that the course of these fibers is readily followed for a long distance. In the case of certain teased preparations it was possible to iso- late these fibers and study them with oil immersions and there can be no doubt as to the relations here described. So far as could be told, these fibers do not connect with the subepithelial plexus as do the fibers of the free arborizations to be described Jater. (Figs. 2, 12, 13, 14.) The terminal segment seems to 38 JouRNAL OF ComPaARATIVE NEUROLOGY. be entirely homologous with the segments of the nerve and its peripheral portion is perhaps simply a modified dendrite. The endings above described must not be confused with the sense buds found elsewhere in the skin. In the latter there is a well-developed accessory apparatus in the form of the well- known beaker or ‘‘Stutz’’ cells, here there is simply a cavity or tube in the midst of unmodified epithelium cells. Yet it is not to be assumed without better evidence than is now at command that these two classes are of entirely distinct nature and origin. In the first place it is scarcely to be credited that two sets of sensory organs derived from the same proton and so similar in function as are the organs of smell and taste should be of an ab- solutely different type, and what may be said of the taste buds applies mutatis mutandis to the sensory buds of the skin. The contrast between the results of different methods is no- where better illustrated than in the different conclusions reached by Fusari and Panasci on the one hand (Arch. italiennes de Biol. XIV, p. 240) and those of Arnstein (Archiv f. mikro- skop. Anat. XXXXI, 2). The former authors worked with the chrome-silver method and describe a direct communication of the nerve fiber with the axial (rod) cells of the taste buds. (This we are able to substantiate from personal observation.) Arnstein, on the other hand, denies such connection most em- phatically and claims that teased preparations with methylene blue show with all possible clearness that there is no such con- nection, but instead that the varicose nerve fibers form a felt- ing of fibers around the axial and outer cells of the bud and end free in the pore. Arnstein finds quite similar nerve endings in the filiform papilla. He does not find forked cells, but inclines to the view that such cells result from the separation of the true nerve fiber from the peripheral end of the cell to which it is at- tached. The appearance of continuity between the cell and the nerve fiber is said to be illusory and is explained as due to the blackening of the cell as well as the fiber. Ehrlich (Deutsch. med. Wochenschrift, 1886, 4) described intensely colored cells in the mucous membrane of the olfactory region which pass without interruption into anerve fiber, but these cases Arnstein also dismisses as illusory. Dr. Niemack has also reached sim- ilar conclusions by the use of different material (Anat. Heften, Merkel und Bonnet, Anat. Anzeiger, VIII, p. 20.) HERRICK-COGHILL, Werve Endings in the Skin. 39 Inasmuch as the epithelial layers of the mouth and tongue are morphologically only portions of the skin, it is necessary to ex- amine these regions for light on the nerve endings as they may be modified under the special conditions here existing. In the frog, which has been the subject of the most elaborate invstiga- tion, the sense of taste cannot be at all highly developed, for the animal is accustomed to swallow its food, chiefly horny coated insects, without mastication; and experiments (Bethe) prove a very sluggish response to chemical irritants. In the tongue of the frog, as well as in the palate, there are num- erous scattered specific sense organs, those of the tongue being flat end-plates, while those of the palate are protuberant sensory papillae. Athough these organs were described by Leydig in 1858 they have frequently been the objects of special study since then and even now authors are not wholly in agreement as to the details of the structure. The cellular elements in these sense organs consist of the cylinder of flask cells forming the protection for the sensory rod cells, a subordinate variety of which has been termed forked cells by reason of the divided peripheral projection. Alate, or winged cells, around the cup -or flask have also been noticed by some authors. Bethe, who has recently studied these buds by means of the modification of the methylene blue method which bears his name, finds two sorts of nervous termini in them: first, free termini lying be- tween the cylinder cells and reaching the surface, second ter- mini with bulb-like expansions on various cells. (Fig. 8.) One type of such endings is three-lobed and such endings are affixed to the sides of the cylinder cells; the other variety has simple circular end-plates and these endings are found on the rod cells, fork-cells and possibly also on cylinder cells. Inno case did Bethe succeed in finding actual continuity between the rod-cells and the nerve. Hein fact seems to find greater intimacy of connection between the cylinder cells, which are not supposed to have a nervous function, than with the rod-cells and in no case is there more than a contact with the cell wall. He explains the continuity detected by Arnstein and others as the result of faulty observation and imperfect methods. In the ordinary pave- ment epithelium of the palate Bethe finds termini on gland cells and ciliated cells, as well as deeper elements. It should be noted that the finding of the three-lobed end-plates on the cy- linder cells was not a uniform occurrence but rather exceptional and the suggestion is near that this is the result of an accidental state of the fibers and not a natural or permanent organ. 40 JOURNAL OF OMPARATIVE NEUROLOGY. Our own studies of the gustatory epithelium of the axo- lotle are in accord with the results of Bethe upon the frog so far as the diffuse endings are concerned, though the methylene blue does not give adequate insight into the connections between fibers and cells. The taste buds, on the other hand, afford sim- ilar results to those obtained from the sensory buds of the skin. The source of many of\the erroneous conclusions reached is, as mentioned beyond, the fact that in successful methylene blue preparations it often happens that fibrous elements stain when the cells of origin for the same fibers do not. Diffuse Peripheral Connections. —Various early writers have reported the existence of a dense net-work or felting of nervous material among the epithelial and even the corneum cells of the skin. This structure was first made out by the use of gold chloride and there was always left open the possibility that the appearance was due to the disposition of metallic salts in the interstices between the cells. Dogiel in his paper on the nerve endings of the genitalia figures a very extensive mesh-work of this kind with here and there a free knob-like termination and he traces the lower part of the reticulum to a direct communi- cation with a set of nerve fibers passing perpendicular to the skin. (Fig. 1.) Strong in his paper on the cranial nerves of the frog figures a similarly minute meshwork which is revealed in this case by the use of the Golgi method. In all of the above cases there is the element of uncertainty growing out of the fact that the methods are impregnation rather than staining pro- cesses and are histologically uncertain. It would then be emi- nently desirable to supplement the evidence from these sources by other means. In the study of the skin of the Amphibia it is easily noted that there exists at the base or ental aspect of the layer of Malpighi a layer or stratum which is in a peculiarly nascent state. These cells are devoid of the thick and rigid walls chatacteristic of the superficial cells and are protoblasts rather than complete cells. In this layer we may find, at all stages, the evidences of mitotic division. In fact there is a per- manent proliferating zone in this region. Comparison of this stratum with that of higher vertebrates shows that the latter form HERRICK-CoGHILL, Nerve Endings in the Skin. 41 no exception, though it is not always easy to detect the proto- blastic elements. A single theoretical consideration is sufficient to convince one that this is what should be expected, for it is of course recognized that every type of vertebrate has some pro- vision for the constant or occasional removal of the skin. In some cases the process of removal of the corneum is intermit- tent, while in others it is gradual. In either case it is obvious that there must bea proton of undifferentiated material—of cells that have not passed beyond the plastic stage. In those parts of the skin where there is little differentiation between the various layers the difference between the corneum and deeper cells is not readily detected in preparations by the usual pro- cesses, but in the thicker portions where the so-called Leydig cells appear the basal protoblasts are crowded into the inter- spaces and pried apart. One effect of this process has been to stretch the connecting protoplasm into an excessively thin layer or film enveloping the Leydig cell either completely or asa coarse mesh-work of naked protoplasm. In all the preparations we have seen, even those in which the preservation has been as perfect as possible, without the least evidence of shrinkage, the appearance is that of a broad reticulum arising in the intercal- lary or basal protoblasts and enveloping the cell in such a way as to wrap it completely in the products of the adjacent proto- blasts. The most perfect process of preservation for such struc- tures is a combination of chrom-acetic and platinic chloride di- luted in alcohol. The use of Merkel’s solution also gave very good results, while the various osmic acid solutions invariably produce too great shrinkage of some parts, especially of the ' reticulum. In the first mentioned solution it appears that the natural tendencies of the alcohol and the chromic acid counter- act each other while the fixing action of the platinic chloride is in no way interfered with. The avidity to all the usual stains after this treatment is also very great, while in the osmic pre- parations there is not only general diminution of the receptivity, but, what is worse, the effect is not uniform even in the same class of tissue in the same preparation. In properly prepared sections the reticular structure of the protoplasm of the Leydig 42 JOURNAL OF COMPARATIVE NEUROLOGY. cells is most beautful, but when osmic solutions are used the contents of the vesicles is blackened and the result is a granular appearance instead. The pericellular mesh-work is stained red by picrocarmine, as is all protoplasmic matter, while the nuclet are all selected by'the hematoxylin. Nerve fibers stain red but their nuclei are purple. The nerve supply is abundant and the fibers can be traced without difficulty through the corium layer in all preparations. The sheaths seem to cease after passing the corium and the subsequent course is less easy to make out. In a considerale number of cases it has been possible to trace such fibers with all desirable clearness to actual connection with the bases of the lower protoblasts above mentioned. The fiber is red, as is the protoplasm, so that it remains possible that the exact nature of the union is not obvious, yet from the fact that two masses of naked protoplasm thus come in contact, the range for possible modes of union cannot be extensive. In any case the most careful examination under immersion lenses of well-stained specimens does not reveal any form of intermedation between the fiber and the protoplasm of the cell. Nor is this relation limited to the lowest layer of protoblasts alone, for it is possi- ble to trace fibers to some of the higher members as well. The attempt has repeatedly been been made to count the number of fibers entering the given area and then to compare this number with the number of protoblasts in the same area, with the re- sult that the fibers proved more numerous than the cells in the lower series, thus offering independent evidence to the effect that these fibers are destined to more than the single basal row of protoblasts. The pericellular net-work has been described by a number of the earlier observers, but in each case the real nature of the structure has not been detected. Paulicki and Pfitzner both re- garded it as a mesh-like thickening of the cell wall. The latter thinks these ‘‘ribs’’ serve for the point of attachment of the ‘intercellular bridges.”” Part of Paulicki’s description is given in full. ‘An einigen Leydig’schen Zellen wurde ich auf kleine kreisformige, lanzende, dunkelconturirte Figuren aufmerksam, die in ziemlich regelmassigen Abstanden von einander entfernt HERRICK-CoOGHILL, Nerve Endings in the Skin. 43 der ausseren Flache der zellmembran aufsassen. Es stellte sich nun alsbald heraus, das dieser Befund bei allen Leidig’schen Zellen ein ganz constanter ist. Ueber die Deutung dieser Ge- bilde erhielt ich durch Zellen, wie deren mehrere abgebildet sind, Aufschuss. Hier fand sich ein doupltconturirtes Gitter- werk, welches uber die Protoplasmakorner hinwegging. Die Balken des Gitterwerks theilten sich ofters gabelformig und wa- ren bald dinner, bald dicker. Es ist nun anzunehmen, dass das Gitterwerk hervorgebracht wird durch rippenartige, partielle Verdickungen der Zellenmembran, und dass bei solchen Zellen, wo ein derartiges Gitterwerk zu sehen ist, der Schnitt die Zelle tangential getroffen hat, wahrend bei den Zellen, die dieses Gitter- werk nicht zeigen, die dagegen in der Zellmembran von Strcek zu Streck kleine, glanzenden Ringe besitzen, der Schnitt mitten durch die Zeile gegangen ist. Die kleinen Kreise, die der Zel- lenmenbran aufsitzen, stellen die Querschnitte der rippenartigen Verdickungen der Membran dar. Die rippenartigen Verdick- ungen der Zellenmembran Zeigen sich durchs ammtliche Farbe- mittel ebenso gefarbt, wie das Protoplasma, wesshalb sie leicht ubersehen werden konen.’”’ The author also notices that these bands are sometimes sharply stained by fuchsin, a fact that, in connection with the above, might well have suggested that these supposed ridges on the cell wall have a nature more in common with that of protoplasm. Still more suggestive was the addi- tional observation that these ridges are not limited to any single cell, but often pass to neighboring cells without interruption. He says ‘‘Ich sah, dass die Balken von einer Leydig’scher Zelle continuirlich zusammenhingen mit den Balken benachbar- ter Leydig’scher Zellen, dass ein zusammenhangendes Balken- werk sich uber mehere Leydig’sche Zellen ausdehnte. Ausser- dam sah ich aber auch, dass ganz ahnlich gestaltete Balken sich auf die benachbarten Epithelzellen fortsetzen.”’ Our observations leave no doubt that this meshwork is not only of a protoplasmic nature but that the meshes are connected with the nuclei of the basal and intercallary series. (Figs. 17-20). It is easy to trace the meshes into communication with the pro- plasm surrounding these protoblasts. It is more difficult, ex- 44 JourNAL OF CoMPARATIVE NEUROLOGY. cept in the case of perfectly preserved material, to follow the nerve fibers to the bases of the cells of the higher series, i. e., those about the sides and ectad of the Leydig cells. In good methylene blue specimens stained zxva vitam (Figs. 21-23), the fibers can be traced for a considerable distance into the epithe- lial layer among the intercallary nuclei, but it is only in speci- mens stained with picrocarmine and hematoxylin that the ac- tual connection with the cells can be made out. Even here the question (always left wholly undecided by the methylene blue method) as to the nature of the association is not entirely de- prived of its ambiguity. When a fiber of naked nerve-plasm unites with a protoblast of naked cytoplasm, who shall say whether the connection is primary or secondary in the absence of the most intimate embryological evidence or regeneration ex- periments ? An important question in this connection is that as to the source of the nerve fibers. Do they arise in the prota of the skin or do they enter the skin from out-growths of the spinal ganglia? It would seem natural to conclude that the latter is the case, and yet it is not a little puzzling to see that nearly every cell in this series has its fiber. Then, too, the fact has been repeatedly observed that the protoblasts are continually dividing, even in rather large specimens of axolotl. (Fig. 20). It must be left to careful embryological studies to decide wheth- er there are cells of origin in the skin for centripetal nerves or not. Another question must await either an embryological or pathological solution, and that is the detection of centrifugal fibers among those entering the skin. Such non-medullated fibers doubtless occur and we may think of the plexus immedi- ately below the epithelium is the probable site. We have sought to verify the results above described by the application of the methylene blue zztva vitam method as well as the tissue methods used by Dogiel, Bethe and Huber. Making all due allowance for the ambiguity of these methods, it seems that the results are in harmony with those above men- tioned. It is not difficult to secure impregnations in which every fiber is stained throughout its course through the corium, HERRICK-CoGHILL, Werve Endings in the Skin. 45 but to our surprise they seemed to stop short in the vast major- ity of cases in the zone at the base of the layer of protoblasts, while only in comparatively few cases did we trace connections like those described by Bethe with cells of higher layers. In the chromatophore zone just ectad of the corium in many parts of the skin it was possible to trace fibers horizontally long dis- tances and in some cases supposed communications with the chromatophores or similar bodies were noted. (Fig. 21). In most cases these cells were nearly destitute of pigment and pass by all gradations into undoubted ganglion cells. In this connection mention should be made of the remark- able results reported by Dr. W. Pfitzner.' This writer claims to trace the fibers after their passage through the corium into the substance of the cells and to follow them to small knob-like endings free in the protoplasm of the cells. More than this, he traces to each cell, not only of the deeper layers but also of the stratum corneum, two independent fibers from quite distinct sources and founds upon this observation an elaborate hypothe- sis, which unfortunately is deprived of all standing-room by the evidence now at hand. Mr. Massie has pointed out that there is a stage in the young amphibian skin when a curious skein of a material staining deeply with some reagents is found in the cells. The senior writer, who made the preparations used by Mr. Massie, can vouch for the accuracy of this observation. It is not unlikely that the suggestion is waranted that this skein is an embryonic and transitory element in the development of gland cells, as it is not found in all the cells but in a certain class dispersed among narrower cells having a different reaction. This skein (Fig. 4) is as certainly intracellular as the nerve fibers are extracellular in their course. Figures almost identical with those published by Pfitzner as the results of his observation can be secured by his methods, especially if the sections are taken a little oblique (Fig. 24.) The process serves to stain very dis- tinctly the part of the nerve that is medullated, i. e. that part extending through the corium, but not that part which extends 1 Nervenengungen im Epithel. Morphol. Jahrbuch, 1882, p. 726. 46 JourRNAL oF ComPaRATIVE NEUROLOGY. above the corium among the cells. Such fibers can be seen, it is true, but they are so different in appearance from the medul- lated part of these fibers that we are forced to conclude that what Dr. Pfitzner really saw is the intracellular skein of which mention has been made. It is a most natural mistake in the absence of more reliable methods and especially as the methyl- ene blue process was not at his disposal. The finding of two nerve termini in each cell is apparently to be explained as a re- sult of the fact that the base of the skein is hidden, as we found it to be in oblique or thick sections, so that the appearance fig- ured by Pfitzner frequently recurs and if one had a preconcep- tion in favor of the the nervous structure of the element one might easily construe it as he has done. After the above we may be released from the obligation to consider the extensive and interesting theories based upon the supposed intracellular endings. Transitional Cells. In certain regions of the skin the epi- thelium layer is greatly thickened and the Leydig cells are re- reduced in number or carried to a higher (ectal) level. In such portions of the skin, as on the dorsal region, an interesting mod- ification of the structure above described is found. Here the lower series of cells is elongated in a direction perpendicular to the surface forming a sort of palisade type of cells. A definite wall is often apparent in the lower portion proximad of the nucleus, while the peripheral part seems to fray out into a rep- resentative of the pericellular mesh-work. Where the Leydig cells are present there is every reason to believe that these cells participate in the formation of such of a pericellular network as has been described above but somewhat modified by the changed conditions. In a large number of cases we have ob- served a nerve fiber after passing through the corium seeking the base of these cells and making an intimate connection with one of them. Here the opportunity to observe the union is much better than the other case and the connection is perfect. In a certain sense these cells are intermediate between the rod cells and those that supply the pericellular meshwork. (Fig. 25.) HERRICK-CoGHILL, Werve Endings in the Skin. 47 Dogiel* has shown that in the eyelids of man, for example, where the number and complexity of the sense organs is ex- treme, the terminal bodies consist of a covering of several con- nective tissue layers separated by zones of flat epithelial cells enclosing the nerve net. The nerve net is described as lying free in the interior of the bulb, though a faintly stained material was noticed and regarded as coagulated lymph which may rep- resent cellular elements not competent to be revealed by the methylene blue method. (Fig. 9.) The nerve fiber loses its sheath before it penetrates the bulb and at once divides into spirals or coils forming a loose mesh-work. Aside from these specific cells, there are extensive arborizations and nets of fibers diffusely scattered in the epithelium at large. In some respects the fullest description of the highly differ- entiated sense organs of the skin of the genitalia has been given by Dogiel and his results are pertinent to our purpose, inasmuch as he finds that all the end-organs reduce to one type—a _term- inal recticulum. The so-called genital sense organs and the Krause’s and Meissner’s bodies all prove to consist of a capsule containing a reticulum of varicose fibers and, especially in the case of the genital corpuscles, those of the same order are fre- quently connected by lateral anastomoses. In addition to these special organs, Dogiel traces medullated fibers into an inter-cel- lular reticulum within the epithelium so fine and dense as to come apparently into relations with all the cells of the deeper parts of this layer. Occasionally a branch turns peripherally and ends in a knob at some distance. below the surface. We seem, then, to have evidence that the typical form of nerve ending is a close pericellular network, though Dogiel’s method is not such as to allow of determining the relation of the fibers te te ‘cells. (Fig. 1.) The most remarkable suggestion respecting the homologies of the sense organs of the skin in amphibians is that of Maurer who thinks that the hair of vertebrates can be traced back phy- 14.S. DocieL. Die Nervenendigungen i. Lidrande, etc. Archiv f. Mik, Anatomie, XLIV, 1, 1894. 48 JOURNAL OF COMPARATIVE NEUROLOGY. logenetically to these sense organs. Leydig in Biolog. Central- blatt, XIII, scouts this idea and derives the hair from the so- called ‘‘ Perlorgan”’ of certain fishes. The resemblance and affinity of the sense organs is rather with the auditory apparatus, as shown by Ayers and others. The Sense Buds. It is interesting to observe the wide dif- ferences of opinion of competent observers as to the endings in the end buds. Lenhoss&k (Anat. Anzeiger, VIII, 4) denies absolutely Fusari and Panisci’s statement that the proximal ex- tremity of the sensory cells in the taste bud passes directly into a nerve fiber and states that the nerves always end free in the bud, or rather form a meshwork surrounding it, thus constitut- ing a peri-gemmal reticulum. Nerve fibers pass in a horizontal course below the epithelium and give off collaterals from time to time which form a felting of free fibers among the general epithelium cells. Essentially similar conditions prevail in the sense buds of the mouth of fishes and the author concludes that the rod cells are to be considered as short apolar nerve-cells and that the class of nerve endings found in the earth-worm is found in vertebrates only in the olfactory organ. (Figs. 15 and 16.) Retzius takes the same view, but finds that the nerve fibers are not perigemmal but intragemmal, thus illustrating the difficul- ties growing out of a reliance on the Golgi and methylene blue methods alone. A. Geberg in a brief article in the Anat. Anzeiger, VIII, I, claims to be able to demonstrate the endings of the auditory nerve in the cochlea by the methylene method, but, inasmuch as the tissues were not stained, it seems that his conclusion, that the fibers attach themselves to the hair cells without communi- cating with the latter, must be considered as non-conclusive. Having reinvestigated the nerve endings in the sensory buds of the skin of the axolotl with material leaving little to be desired as to the fixation and hardening, and which had ‘been double stained successfully, we are able to assert with great con- fidence that, in this case, there is a special cellular nerve termi- nus having a direct basal connection with a nerve fiber. The nucleus of these cells (which cannot be termed appropriately HERRICK-COGHILL, Nerve Endings in the Skin. 49 rod cells or ‘‘ Stiftzelle’’) is narrower and more deeply stained than the supporting cells and occupies the entire width of the cell. The peripheral part of these cells has not been correctly described as yet. In reality it consists of a projection of the cell walls to form a narrow tube. These walls are delicate and very thin but easily seen because of the contrast with the proto- plasmic fiber contained in it. The latter structure is delicate but stains a deep red with the picrocarmine, while the walls are not stained by that reagent. (Figs. 26-30.) This axial fiber dif- fers not at all from that seen in the clusters found in the scat- tered sense organs on the head of the tree frog and the frog. (Fig. 32.) The proximal portion of the cell is not as easy to trace, for the corium and often the chromatophores obscure the connections to a degree. Yet it now and then happens that the direct communication with a nerve fiber rising through the corium can be made out. Of course it may be insisted that this connection is only a secondary one, but nothing but evidence from embryology or degeneration experiments will substantiate or refute the claim. So far as the evidence now goes, the scat- tered cells above mentioned and those in the buds stand or fall together, and for the former the evidence of direct continuity between cell and nerve is unimpeachable. The Plexus Beneath the Corium.—tIn portions of the skin stained zztra vitam by the methylene blue method and examined at once in glycerine very perfect views of the marvelously elab- orate plexus beneath the corium can be gained. The fibers are of two sorts, the larger being connected with the fibers from the nerve bundles from the central system, while a part at least of the fibers of smaller calibre have a local origin in certain gan- glion cells of this region. These cells were first detected in preparations double-stained with haematoxylin and picrocarmine and were seen in section in a plane parallel to the surface. In the methylene blue preparations they are very conspicuous and surprisingly numerous. The nuclei are large, while the proto- plasm of the cell does not stain or only slightly with the blue. It is an interesting and most instructive fact that the cell body remains transparent, while its own neurite or axis cylinder pro- 50 JouRNAL oF COMPARATIVE NEUROLOGY. cess is mostly intensely stained through its entire length. The hiatus between the fiber and its cell is slight but sufficient to cast a doubt on the fact of communication were the conditions not absolutely favorable. With a high power it is possible to see the sheath and the faintly tinged protoplasm so that no doubt is in this case possible. It may be noted also that other methods seem to show that it is entirely possible for the protoplasm of a cell to react differently from that of the axis cylinder derived from it. Thus may be explained many of the ambiguous and conflicting re- sults of the applications of the methylene blue process. Fig. 33 illustrates the appearance of a section stained with hema- toxylin and picrocarmine, while Fig. 23 is froma methylene blue preparation. Figs. 33-37 are from surface views of the plexus, showing the ganglion cells. Figs. 38 and 39 are from the same region, showing connections with vessels and chromatophores (Fig. 3.) It will be seen that the fibers of this plexus below the cor- ium are of two sorts. The fine fibers arise, in part at least, in the local ganglion cells and can be traced to the nerve bundles, which they enter and then mingle with the fibers of the larger sort. In the perpendicular sections it is easy to see that a cer- tain number of fibers from the general ‘‘mixed’’ nerves pass without interruption into the skin and so do not participate in the formation of the plexus. Others, on the other hand, divide dichotomously in the level of the plexus and the branches give off ‘‘collaterals’’ that pass through the corium and so reach the epithelial layer. It is not possible to state positively that fibers from the ganglion cells of the plexus give off fibers to the skin, though such certainly is the appearance. After passing through the corium, the fibers do not all at once seek out their definite termini in the cells of the epithelial layer, but they often turn sharply at right angles at the ectal surface of the corium and pass long distances parallel to the surface. This tendency is more marked in some regions than in others, This fact greatly com- plicates the study of the endings. In the case of taste buds and the organs of the lateral line this is one of the most serious dif- HERRICK-CoGHILL, Nerve Endings in the Skin. 51 ficulties in the way of a correct interpretation of the appearances presented by sections. A discussion of the theoretical bearings of these facts and further details must be deferred to the second part of this paper. Since writing the above we have been able to settle several points previously in doubt. None of our preparations of the skin of amphibians gave unambiguous results for the glands of the skin. We have at last succeeded in securing excellent zx- tra vitam impregnations in the toad ( Bufo sp.) in which it is easy to trace the non-medulated fibers from the plexus ectad of the corium, and also from that entad of it, into the most inti- mate connection with the superficial walls of the glands, which in this species are very large and highly functional. The fibers are of small caliber but are excessively numerous and envelop the whole gland in what at first looks like a closely woven re- ticulum, but a close study shows that the appearance of a retic- ulum is due to the repeated dichotomous branching of a large number of distinct nerve fibers. These fibers cross at slightly different levels and there is no doubt in most cases of the com- plete distinctness of the fibers as they cross. | Upon these fibers are frequent varicosities which may be due to imperfec- tions of the process or may be the points of attachment of the fibers upon the cells of the gland. Of course this method does not admit of determining the exact relation of the nodosities to the several cells, but there can be no doubt of the existence of a very intimate and necessary connection. One is forcibly struck by the close resemblance of this periglandular felting to the perigemmular reticulum described by many authors in the case of the sense buds. The latter is, as we have before in- sisted, entirely distinct from and totally unlike the intragem- mular endings in distinct cells which may be demonstiated by a wide range of independent methods. The same preparations used in the earlier parts of this paper have also afforded to a more extended study a number of sat- isfactory views of the connection of the ganglion cells of the §2 JourNaL oF CoMPARATIVE NEUROLOGY. reticulum below the corium with fibers—not only with such as pass directly into the nerve bundles but, as we now find, with with non-medullated fibers which pass through the corium and end in relation with the cells of the epithelium layer. We also find that these and other fibers, after passing through the cor- ium, turn and pass for long distances parallel to the surface to their final destination in the upper layer. This seems to be par- - ticularly true of the fibers of the perigemmular series of the sense buds. In some cases well defined bundles of nerves in a common sheath pass through the corium, while in those cases where the nerve sheath is present it is soon lost after passing the corium. It seems natural to conclude that the non-medul- lated fibers of the epithelium are essentially similar to the fibers of the same structure that supply the glands. If so, we may add that these are in both cases centrifugal and we have a sug- gestisn at least toward the solution of the puzzle as to the re- spective functions of the several classes of fibers. That the general cells of the skin have more or less power of absorption and excretion, as well as secretion, can hardly be doubted and, if so, why may not these fibers from the disperse ganglia of the peripheral sympathetic system be the neural sponsors for these functions ? The methylene blue method reveals the same sens- ory endings in the skin that we have described fully from histo- logical preparations, but curiously enough they appear as fibers simply because the nuclei are not stained and this fact explains the discrepancy in the two methods. It is interesting to compare the intercellular net-work de- scribed above with the similar so-called connective tissue net- work described by Bruyne (Arch. de Biol., XII, 1892) sur- rounding the muscle fibers. The figure given in the article by the same author in Anat. Anzeiger, X, 18, is so remarkably similar to the apearance we have called attention to that one may be pardoned for suspecting similarity of nature. It may be that more than one instance of intercellular bridges rests on the mis- interpretation of similar structures. The relation of the space so kept open between the cells to the circulatory fluid is a ques- tion of greater interest than seems to have been suspected. HERRICK-CoGHILL, Nerve Endings in the Skin. 53 Nore ON THE METHYLENE BLuE Process. It appears that we have had in one respect the usual experience with the meth- ylene blue zxztra vitam impregnation process. It is not difficult to secure excellent impregnations of the nerves of the skin of the Amphibia in which the nerve fibers are deeply and quite selectively stained, yet it appears that there is a strong ten- dency for the stain to be extracted or rendered diffuse by the process of imbedding so that tissues which were very perfectly stained in the glycerine are quite unsatisfactory in thin section. It appears that the difficulty is in the action of the alcohol, which is required in both the paraffin and the celloidin methods of im- bedding. To obviate this difficulty we have resorted with good results to the use of a mixture of gum arabic and glycerine. The fragment is placed in glycerine or may be placed at once in the gum-glycerine. After an impregnation of a day or two in a closed bottle the specimen is mounted in a paper tray with the mixture and the latter is allowed to evaporate till a consist- ency is reachcd which will permit it being placed in the micro- tome and sectioned. In this way sections are secured thin enough to serve the pupose desired and these may be mounted in gum-glycerine or may then be dissolved out of the gum and treated in any way desired. 54 JourNAL OF CoMPARATIVE NEUROLOGY. EXPLANATION OF FIGURES. PLATE V. Fig. 7. Diagram of the skin of the sexual organs, after Dogiel. Fig. 2. End-organs in the skin of the tree frog, original. eased prep- aration. . Fig. 3. Sense bud of young salamander. Original. Fig. 4. Skin of tadpole with nerve endings and the transitory skeins inter- preted as nerve endings by Pfitzner. Fig. 5. Skin of very young tadpole. Original. Fig. 6. Skin of tadpole, near angle of mouth, Original. Fig. 7. Sense bud of Amblystoma. Original. Fig. 8. Nerve endings in the epithelium of the frog, according to Bethe. A.—* Gabelzelle,’”’ from sensory papillz of tongue. B.—Cylinder cells. C.—Isolated rod cell. D.—Upper part of papilla. £.—Ciliate cell of palate. Fig. 9. Nerve ending in the human conjunctiva. Dogiel. Fig. ro. Nerve endings in Jacobson’s organ, Lenhossék. Fig. 11. Nerve endings in the taste buds. Arnstein. PLATE VI. Fig. 72. Section from the skin of the head of a tree-toad. a, nerve bun- dle and endings; 4, gland ; ¢, corium; d@, small gland; ¢, chromatophore. Fig. 73. Skin of head of leopard frog showing cellular nerve endings in groups penetrating the skin. Fig. 14. Similar endings from the tree frog. Figs. 15, 76. See Plate VIII. Fig. 717. Part of the skin of the axolotl showing the nerve bundle on its way to the skin and the pericellular net-work. Fig. 18. Skin of axolotl showing pericellular net-work and the nerve-fibers entering from below. Fig. 79. Similar section fixed in Flemming’s solution. Fig. 20. A section of portion of axolot] skin where the Leydig cells (Z. c.) are two-layered. Proliferating cells (4) in lower series of protoblasts; ¢c, corium; B. V., capillary ; nerve fibers entering from below. Figs, 21-23. See Plate VIII. Fig. 24. Skin of tadpole as figured by Pfitzner. HERRICK-COGHILL, Nerve Endings in the Skin. 55 PLATE VII. Fig. 25. “Section from a different part of the skin with cellular nerve ter- mini. This is probably to be explained as the result of the elongation of the basal series of the epithelial cells. fig. 26. Sensory bud from skin of axolotl, showing the tubular peripheral ending of sensory cells with fine thread of protoplasm extending to periphery and the basal connective with nerves. fig. 27. Sensory bud from another part of skin. Fig. 28. Similar bud in which the peripheral portion of the sensory ele- ment seems divided. Explained as due to the shrinkage and “ fraying out” of the wall. fig. 29. See Plate VIII. fig. 30. Isolated supporting cells from specimens similar to Fig. 28, stained with haematoxylin, picro-carmine and methylene blue. Are the blue fibers nerves, or are they lines of precipitation in folds of the cell wall due to shrinkage ? Compare Fig. 11. fig. 37. Cells from nasal cavity of leopard frog. Fig. 32. Nerve endings from skin of same to illustrate similarity to the last. Fig. 33. Skin of gills of axolotl to show ganglion cells beneath the corium. PLATE VIII. Fig. 75. Pericellular nerve fibers from sensory bud of conger eel. Fig. 76. Intrabulbar endings in Bardus. (Both 15 and 16 from Lenhossék.) Fig. 21. Skin of the axolotl showing nerve endings in or near the chroma tophores and in the skin of the axolotl. Methylene blue. Fig. 22. Similar to Fig. 21, showing endings in layer of protoblasts. Fig. 23. Perpendicular section through skin of axolotl stained zntra vitam with methylene blue and cleared in glycerine. The plexus beneath the corium. is clearly visible. Fig. 29. Cells similar to to Fig. 28, stained with methylene blue. Fig. 34. Surface view of methylene blue preparation, similar to Fig. 33, showing connection of ganglion cells with nerve bundles. Fig. 35. Same as Fig. 34. Figs. 36,37. Ganglion cells of large ramose form from same layer as above. Fig. 38. Relation of nervous reticulum below the corium to the capillaries. Fig. 39. Chromatophore-like ganglion cells. 56 JoURNAL OF COMPARATIVE NEUROLOGY. PLATE IX. Fig. 4o. Section of the skin of the head of a toad (Bufo) after txtra vitam injection with methylene blue and fixation with Bethe’s solution of molybdate of ammonia. Examined in glycerine. The section is somewhat oblique so that the duct and part of the body of the gland is removed. The delicate non- medulated fibers are seen generously distributed over the uncut surface of the gland, Coarser fibers are also seen in the lower and upper plexuses, also a bun- dle of sensory rods at the left. Fig. 41. Intra vitam methylene blue preparation of skin of axolotl, show- ing connection of cells of the ganglionic meshwork beneath the corium with the epidermis. a, fiber passing to cells of the intracellular reticulum; 4, non-med- ullated fibers from a nerve piercing the corium; ¢, ¢,! c and ¢’, ganglion cells of the plexus beneath the corium. tHE BRAIN OF THE FUR SEAL,. CALLORHINUS URSINUS:) WITH A COMPARATIVE DESCRIP- TION OF THOSE OF ZALOPHUS CALIFORNIAN- US, PHOCA VITULINA, URSUS AMERICANUS AMO OMONACIUS TROPICALIS: = By EYERRE: A. Fisa, 1:)Sc.,..D..V./S; NV. Y. State Veterinary College, Ithaca, N. Y. WITH PLATES X To XIII. TABLE OF CONTENTS. INTRODUCTION, Fi A Removal of dura, Terminology, CALLORHINUS URSINUS, Lateral aspect, Mesal aspect, . PHOCA VITULINA, Lateral aspect, Mesal aspect, URSUS AMERICANUS, Lateral aspect, WMesal aspect, ZALOPHUS CALIFORNIANUS, Lateral aspect, Mesal aspect, FISSURAL INTERPRETATIONS OF OTHER WRITERS, THE LATERAL VENTRICLE (paracoele), MONACHUS TROPICALIS, GENERAL CONSIDERATIONS, DESCRIPTION OF PLATES, *This article was written at the request of a member of the Bering Sea Commission and will appear in their Report of the Bering Sea Fur Seal Investi- gations, 58 JOURNAL OF COMPARATIVE NEUROLOGY. INTRODUCTION. The specimen was from a young male pup twenty five inches in length, weighing about twelve pounds. The brain was still incased in the dura and on the basal surface portions of the cranial bones were left adherent to this membrane. An occasional cut through the dura caused a protrusion or hernia of the cerebral substance. The weight of the brain in the fresh condition, as reported by Mr. Lucas, was ten ounces and two hundred and forty grains. This included the dura with the attached cranial fragments. The specimen was preserved in a ‘‘rather strong solution of formalin” and except for some swelling of the tissue and soft- ening of the interior was in a very good condition. The _ bloat- ing was indicated by the increased weight which, immediatly after the receipt of the specimen, Dec. 12, ’96 was found to be 13 ounces, a gain of nearly three ounces, by the closure of the fissures and by the cerebral hernias. The weight without dura and attached fragments of cranial bones after preservation from Sep. 1 to Dec. 12 was 9% ounces and 80 grains (avoir.). The lat- eral girth was 26 centimeters, the longitudinal girth with the oblongata cut off at an even level with the caudal surface of the cerebellum was 24 centimeters, being slightly less than the former. This may, perhaps, be accounted for, to some extent, by the tape resting slightly in the inter-cerebral cleft, and to the bloating, as this would affect the lateral rather than the longi- tudinal circumference. The brain as indicated by the girth measurements was of a subglobular form slightly tapering at the ends and its outer sub- stance though firm was not unyielding. Twenty four hours immersion in 95% alchol served to contract the nervous tissue sufficiently to open the fissures and yet to retain enough flexi- bility of their walls to permit of an easy examination of their Fisu, Brain of the Fur Seal. 59 depths. In order to obtain the desired results, after photo- graphing the dorsal and ventral surfaces of the entire brain, it was cut across and the crura cerebri or mesencephal, and the cerebellum and oblongata separated. The cerebrum was then divided by a section along the median line, separating it as nearly as possible into two equal halves. Removal of dura. The falx showed an interesting devel- opment, its frontal portion, especially in the region of the olfactory bulbs, being of considerable depth, then becoming very shallow along the middle of the length of the cerebrum and be- coming very deep again in the intercerebral cleft in the caudal region of the cerebrum. A distinct longitudinal venous sinus as in the human brain is not present; but in place of it is a vein of some size lying to the right of the (intercerebral) cleft and receiving the contents of the dorsal cerebral veins. In connec- tion with the weak development of the falx along the middle of its length, there was noticed an interdigitation of the gyres of the mesal surface of the hemicerebrums in this region. This intimate overlapping of the gyres on the mesal surfaces of the two hemicerebrums is possibly correlated with the deficiency of growth of the falx here and may serve, in a measure, to increase the firmness of the union of this region and prevent any undue strain upon the callosum which lies some little distance from the dorsal surface of the cerebrum. This interdigitation of the mesal gyres is also present in the sheep where the falx is also deficiently developed. If the hem- icerebrums be divided with a sharp knife without first separating the pial adhesion of the gyres, the gyres will be cut. An artifact of this nature has, indeed, been mistaken by one writer in an article on Phoca, for the cut surface of a bundle of fibers dorsal to and larger than the callosum and designated by him as the commissura suprema. The tentorium in Cal/lor/inus is very strongly developed, apparently extending the whole depth of the transverse arch- like cleft between the cerebrum and cerebellum. The tough fibrous tissue of the tentorium is, moreover, very noticeably re- inforced by the presence of osseous tissue. Where the falx 60 JouRNAL OF CoMPARATIVE NEUROLOGY. joins the tentorium there is an extension of this osseous tissue in a vertical direction into the falx, a circumstance which cer- tainly is not common in the majority of other animals but has been noted by Turner in Macrorhinus. Terminology. WWith the existing uncertainties relating to the homology of the fissures of the brains of the carnivora and that of the human species, much confusion has resulted in the pre- sent nomenclature. Some have made a direct homology, others have proposed a fissural type solely and only for the lower forms, while still others have blended the two and some have utilized a system of names devised by themselves. On the lateral surface of the various fissured brain types there is at least one fissure—the Sylvian—which is quite constantly present,and on the mesal surface, the hippocampal fissure. In the matter of nomenclature no attempt has been made to follow the law of priority, but those fissural names, whether of old or recent date which seemed most appropriate concern- ing position and relation, have been adopted, and, with perhaps but one or two exceptions, no new names have been introduced. It has been the purpose to use an intrinsic terminology and to substitute for the sometimes indefinite terms, anterior, poster- ior, superior and inferior, terms of more universal applicability, cephalic, caudal, dorsal and ventral. For cephalic and caudal Professor Wilder has recently suggested praeal and postal as equivalents, and for cephalad and caudad, praead and postad. Where certain of the fissures or gyres have been submerged for a portion or the whole of their course, they have been des- ignated as such, or the equivalent terms, subfissure or subgyre proposed by Wilder, have been used. In the study of fissures mere surface appearances are not accepted as final. A fissural entity is not always easy to define. The best apparent guide is the relative depth throughout the course of the fissure. We may commonly assume that the greatest depth is at about the middle of its length and that it becomes gradually shallow toward each end until it reaches the surface. Such a simple condition, however, does not usually exist. One fissure may join the end of another, giving the ap- Fisu, Brain of the Fur Seal. 61 pearance at the surface of a long continuous fissure. By separating its walls or ‘‘sounding”’ its depth the true state of affairs is easily perceived. The presence of a shallow whether it be near or at a distance from the end of a fissure would seem to indicate that at some time during development this shallow has been or will be represented at the surface and separate two independent fissures. CALLORHINUS URSINUS. Cranial Nerves. The cranial nerve roots of Callorhinus are well developed and need no special comment. In the case of the optic nerves we do not find the X-shaped chiasma as in Phoca, but the nerves run parallel to each other for a short distance from the chiasma before diverging toward the eyes. The third pair or oculomotor nerves have a straight lateral direction from their apparent origins, but at the lateral border ‘of the hypophysis they bend abruptly upon themselves and proceed cephalad forming a very distinct right angle. The olfactory lobes are fairly well developed. Fissures. No special mention will be made of the gyres (convolutions). These are naturally formed by the fissural de- pressions and it is believed that a careful description of these furrows will by implication include that of the gyres sufficiently for our present purpose. The olfactory fissure is completely hidden by the olfactory crus and bulb; when these are removed a shallow fissure is ap- parent which becomes deeper toward the base of the lobe. Forming the lateral boundry of the olfactory lobe is the rhinal fissure which passes in a caudo-lateral direction to the Sylvian. An apparent continuation of the rhinal from the Syl- vian is known as the post-rhinal fissure. It extends in a meso- caudal direction for a centimeter and a half, stopping just short of the cleft between the cerebrum and the cerebellum. A care- ful examination of the postrhinal shows that it has no connec- tion whatever with the rhinal but is continuous, superficially at least, with a subfissure (postica ?) lying in the caudal wall of the Sylvian. 62 JOURNAL OF COMPARATIVE NEUROLOGY. Lateral Aspect. The Sylvian is a convenient fissure to be- gin with. There is usually some evidence of it if the brain is at all fissured, and in the lower animals, at least, it forms a cen- ter around which other fissures are more or less regularly ar- ranged. In Callorfinus the Sylvian extends in a dorso-caudal direction, inclining somewhat toward the vertical. Apparently it terminates in a fork, but when the walls of the fissure are di- varicated it is seen that the cephalic or anterior branch is really another fissure, which, after its superficial union with the Syl- vian, becomes a submerged fissure lying just beneath the surface of its cephalic wall and running parallel with it to the base of the brain, but not actually connecting either with the Sylvian or with the rhinal. The Sylvian on account of the subfissural complication appears to be a larger fissure than it really is. In a former paper’ attention was called to the fact that this vertical fissure (superficial vertical branch of the Sylvian) had been mistaken for the true Sylvian. Both fissures are well marked and cannot be ignored, but it is an unusual circumstance for the Sylvian to assume a strictly vertical position in the adult and there would, moreover, remain a fissure in the usual situa- tion of the Sylvian unaccounted for. In my former paper I des- ignated this vertical fissure as the Anterior of the Felidae, and found at a later date, while consulting Krueg’s article? that he questioningly represents a similar fissure by the same name in Calocephalus (Phoca) vituhinus. Callorhinus, while showing this fissure similarly situated, instead of elucidating the compli- cations, seems rather to add to them and to suggest a probable doubt as to the correctness of the homology with the anterior fissure. Indeed, the conditions are strongly suggestive of its being nothing more than the detached frontal portion of the super-sylvian fissure. An examination of the brains of certain bears tends to illuminate this view. In the family Uyrs¢dae as 1°96, P. A, Fish. A note on the Cerebral Fissuration of the Seal (Phoca Vitulina). Jour. Comp. Neurol. VI, 15-19. 2°80. J. Krueg. Ueber die Furchen auf der Grosshirnrinde der zonopla- centalen Siugethiere. Zeit. f. wiss. Zoologie, XX XIII, 595-672, 5 plates. Fisu, Brain of the Fur: Seal. 63 e) in the Cazzdae the super-sylvian forms a complete arch, the cau- dal portion being known as the posterior supersylvian (Krueg), or postsylvian (Owen). The frontal portion of this arch varies in its distance from the Sylvian. Occasionally the frontal and caudal portions are about equally distant, but when there is any difference in this distance, it appears that the frontal portion ap- proaches more closely to the Sylvian than does the caudal. In Ursus arctos, or the brown bear, Krueg figures the frontal por- tion of the supersylvian as approximating very closely to the Sylvian. The condition in Callorhinus might be considered as a stage just beyond this. In the brown bear the frontal portion of the supersylvian is still visible upon the lateral surface close to the Sylvian. In the case of the seal it has passed over the brink, so to speak, and is no longer visible its entire length on the lateral surface. The following diagrams will illustrate the conditions more clearly. _Prss. : Suh puss gio Fig. 1. Wig. 2: Figs. 1 and 2. A diagrammatic representation of the relation of the Sylvian and supersylvian fissures in the bear and seal, as if seenin section, rss, pre- supersylvian. ss, postsupersylvian. Sy/, Sylvian fissure. At the bottom of the Sylvian fissure lies the insula, pre- senting but a slight degree of development. There is a sugges- tion of a circuminsular fissure but in other respects the surface is entirely smooth. In the caudal wall of the Sylvian is a well marked subfissure. It separates a portion of the concealed cor- tex, forming a subgyre, which from its size and position might be easily mistaken for the insula. The appearances would sug- gest that the subfissure is the postica and the subgyre a rem- nant of the Sylvian gyre. The supersylvian fissure shows some variation on the two sides. It presents the usual arrangement on the right hemice- brum, forming, superficially at least, a complete arch around 64 JouRNAL OF COMPARATIVE NEUROLOGY. the Sylvian. The presence ofa shallow and a slight bifurcation near the level of the free end of the Sylvian indicates the sep- aration of a postsupersylvian fissure, postsylvian of other writers. Plate I, Fig. 4. The supersylvian curves around the free end of the Sylvian at a rather sharp angle and soon appar- ently enters the Sylvian, but in reality is submerged in its cephalic wall. A very short cephalic branch is given off toward the ansate fissure before the supersylvian enters the Sylvian. On the left hemicerebrum there are three distinct portions; the postsupersylvian has a slightly more oblique dorso-caudal course, the supersylvian proper is quite branching and more inclined to a vertical than a horizontal course. One of its branches appears to enter the Sylvian from behind but a shallow shuts off any deep connection. The frontal portion appears as a surface fissure for only one third ofits course, then, as on the other side, it becomes submerged in the Sylvian. As this portion bears much the same relation to the supersylvian as the postsupersylvian whether they be disconnected or not, the frontal portion will be designated as the presupersylvian fissure. In the second speci- men of the brain of an adult Callorhinus, kindly loaned to me by Mr. True, the executive curator of the U. S. National Mu- seum, both hemicerebrums showed a distinct separation of the postsupersylvian, more pronounced than on the right hemice- rebrum of the pup; but there was no separation nor distinct ap- pearance of a shallow indicating an independent presupersylvian as in the left hemicerebrum ofthe pup. In the adult, as in the pup, each supersylvian gave off a short cephalic branch before entering the Sylvian. The Lateral fissure, on account of the breadth of the brain, does not show in its entirety upon the lateral aspect. It is twelve centimeters long, by far the longest fissure, and is seen for a short portion of its course upon the ventral aspect extend- ing, on the left hemicerebrum, to within five millimeters of the ventral portion of the postsupersylvian. It lies in this region just in advance of the margin of the cleft between the cerebrum and the cerebellum. It then arches caudo-dorsally approxi- mately parallel with the hemicerebral margin but receding from Fisu, Brain of the Fur Seal. 65 it until it fully reaches the dorsal surface, then approaching to within eight or nine millimeters of the intercerebral cleft, it con- tinues its arched course in a cephalo-ventral direction approach- ing to within five millimeters of the presupersylvian fissure at about the level where the latter becomes submerged in the Syl- Vian. The lateral is a deep fissure and no distinct evidence of shallows could be detected along its course although in certain places the presence of submerged buttresses interfered to some extent with the soundings, the average depth being from ten to thirteen millimeters. The cephalic extremity of the fissure ter- minates in a fork, more marked on the left hemicerebrum than on the right. Does this widely forked termination represent the ansate fissure? It has the same appearance and relation to the lateral as seen in the cat, and provisionally, it is here so designated. The gyre, bounded by the lateral and supersylvian fissures and its parts, is indented by numerous branches originating from the above named fissures. There are also occasionally in- dependent minor fissures present in this gyre. The Ectolateral fissure. The ectolateral on the right hemicerebrum is a distinct fissure. It begins on the ventral surface near the termination of the postrhinal ; it then proceeds dorso-caudally, parallel with the postsupersylvian and for about the same distance. On the left side it is a shorter fissure and superficially is continuous with the dorsal portion of the postsu- persylvian but a shallow separates a deeper connection. On the left side side of the adult Cal/orhinus, a somewhat similar condition exists except that the superficial union of the ectolat- eral is with the ventral portion of the postsupersylvian. The Coronal fissure is about three centimeters in length and extends except for a slight caudal convexity in an almost verti- cal (dorso-caudal) direction. Its greatest depth is eight milli- meters. On the right hemicerebrum it gives off a slight spur pointing toward the Sylvian. In Callorhinus it represents, per- haps, the least complicated fissure in the brain. The Cruciate fissure is not at all represented upon the me- 66 JOURNAL OF COMPARATIVE NEUROLOGY. sal surface of the brain. It is seen best from a dorsal view. It arises at the margin of the intercerebral cleft. It arches in an obliquely cephalo-lateral direction. From the cephalic extrem- ity of the cruciate at a depth of fifteen millimeters there passes off another fissure, which Krueg has represented as the precru- ciate in certain carnivora, nearly to the mesal margin just dor- sal to the olfactory bulb. The depth of these fissures at their junction is from 12-15 millimeters. Between these fissures and the intercerebral cleft there is a triangular shaped area to which Mivart has applied the name of ‘‘ursine lozenge’ (Turner), thought by Mivart to be of considerable significance. Just caudal to the cruciate fissure is a small fissure corresponding to the postcruciate of Krueg. On the left hemicerebrum it is tri- radiate, on the right it is straight. The Superorbital fissure has no connection with the rhinal. Its length is 25 millimeters and its depth 8-10 millimeters. It has a slight lateral convexity but has no branches. The Medilateral fissure. The name of this fissure is par- ticularly appropriate in Cal/orhinus ; not only is it on the mesal side of the lateral fissure, but for a portion of its course is actually on the mesal aspect of the brain. It curves around the caudal margin of the hemicerebrum just on the verge of the cerebro- cerebellar cleft. Between the lateral and medilateral fissures there is a gyre averaging about 15 millimeters in width in which there are two or three secondary fissures, which would seem to indicate an attempt at the division of this gyre into two. Mesal Aspect. The callosal fissure presents no marked pe- culiarity except upon the left hemicerebrum where, instead of continuing around the genu of the callosum, it proceeds toward the dorsal margin, or is continuous with a fissure coming from this margin. On neither hemicerebrum is there any appear- ance of a fissure immediately surrounding the genu. The hip- pocampal fissure occupies its usual position, arching from the splenium around the optic thalamus to the tip of the pyriform or temporal lobe. The Splenial fissure. On the right hemicerebrum, this fis- sure, if prolonged on the dorsal aspect, would be continuous Fisu, Brain of the Fur Seal. 67 with the cruciate. It is separated by a gyre 4 millimeters in width. The fissure passes ventro-caudally and a little beyond the splenium on the ventral aspect and it apparently terminates in a wide fork, or else enters a fissure passing at right angles to its own course. Sounding the fissure at this point gives some indication of a shallow separating the caudal branch of the fork. Following the appearance designated by Krueg in his diagrams of the conditions found in some of the carnivora, the splenial proper includes the ventral branch of the fork, while the dorsal branch may represent what he calls the postsplenial. On the left hemicerebrum the splenial fissure penetrates the hemice- rebral margin and appears for a short distance on the dorsal sur- face. A smaller but well defined fissure lies in front of the splenial. On the left side it cuts the dorsal margin. For the present we may designate it as the presplenial fissure. It cor- responds very well with the fissure which Kiikenthal has called fissura sublimica anterior. The Marginal or supersplenial just passes the meso-ventral margin of the hemicerebrum about ten millimeters caudad of the splenial. It extends approximately parallel with it to the dor- sal margin which it cuts and on the right hemicerebrum extends on the dorsal surface for about 15 millimeters. On the left hemicerebrum the fissure branches just at the margin. The main portion however continues obliquely latero-cephalad for about 20 millimeters. In the gyre between the splenial and su- persplenial fissures a well represented secondary fissure is seen. A well defined but unnamed fissure lies on the meso-ven- tral surface. It arises at the caudal margin and proceeds in an angular course toward the ventral end of the splenial, it then swerves cephalo-laterad and terminates not far from the post- rhinal. Its position corresponds approximately to the collateral fissure in the human brain. This tentorial surface of the cere- brum has numerous secondary fissures and branchings some of which seem large enough to merit special mention. One such fissure lying parallel with the postsplenial suggests a similarity to the occipital. It cuts the hemicerebral margin slightly and the relation of the lateral fissure at this point suggests in a way 68 JouRNAL OF CoMPARATIVE NEUROLOGY. the paroccipital of man. This occurs on the left hemicerebrum. On the right the postsplenial has much the same appearance. At the cephalic end of the mesal surface beyond the genu of the callosum, there are two pretty well marked fissures. The one nearest the callosum corresponds to the genualis of Krueg, part of falcial—Owen, or falcial—Wilder. On each hemicere- brum this fissure cuts the dorsal margin slightly. The other and more slightly developed fissure lies nearer to the olfactory bulb. It does not reach the dorsal margin but extends farther in the ventral direction. This fissure corresponds to the rost- ralis of Krueg, part of falcial—Owen, subfalcial—Wilder. PHOCA VITULINA. The frontal portion of the cerebrum is more foreshortened than in Callorhinus and there is therefore a slightly different ar- rangement of corresponding fissures in that region. One of the most striking differences is the olfactory portion of the brain. In Callorhinus it is the larger, the olfactory bulb is of considerable size, the crus is correspondingly wide and lies flush with the mesal surface. In Phoca the bulb is relatively smaller and the crus has atrophied to scarcely more than a pedicle, it lies deep- ly imbedded in the olfactory fissure and it is removed 6-8 milli- meters from the mesal surface by a portion of the cortex which projects fully 5 millimeters beyond the crus. The precribrum (anterior perforated space) is well devel- oped and shows with greater distinctness than in Cadlorhinus. The rhinal fissure is apparently continuous with the Sylvian, but upon raising the overlapping portion of the frontal lobe, it is seen to maintain its continuity and to appear again caudal to the Sylvian as the postrhinal, differentiating a larger pyriform lobe than in the case of Callorfinus. There is no connection between the postrhinal and the subfissure in the caudal wall of the Sylvian as in Callorhinus. Lateral Aspect. The Sylvian fissure pursues a much more obliquely dorso-caudal course than in Ca/lorhinus and presents the same amount of complexity with relation to the surround- ing fissures. In its caudal wall lies a subfissure (postica ?) and Fisu, Brain of the Fur Seal. 69 the intervening Sylvian gyre. Both are relatively better de- veloped than in Callorhinus. The supersylvian has much the same relation to the Sylvian as in Callorhinus. It is not distinct- ly separated from the postsupersylvian although the interlocking of some of the subgyral buttresses suggests the possibility of an attempt at separation. On each hemicerebrum there is a continuation of the postsupersylvian dorso-caudad beyond the supersylvian. Fig. 3. Fig. 4. Fig. 3. Cross section of a fissure, showing the obliquity of the walls. fig. 4. A diagram to show the difference in the course of a fissure at its surface and depth. The heavy lines represent the fissure walls at the surface. The dotted lines and arrows represent the buttresses (b) formed by the bulging of the deeper portion of the wall of the fissure. The frontal end of the supersylvian apparently forks, one branch bending toward the Sylvian, the other continuing cephal- ad. The ventral branch has a superficial union with the verti- cal fissure which has been mistaken for the Sylvian. In my former paper (I. c.) I designated this fissure as the anterior. Krueg also had taken the same view. From the conditions al- ready described in Cadlorhinus, it seems to me that this fissure is after all a disconnected portion of the supersylvian and that presupersylvian would in some ways be a suitable name for it. It is submerged in the cephalic wall of the Sylvian for the ventral third of its course. In Cadllorhinus the ventral two-thirds of the corresponding fissure becomes submerged. The lateral fissure, as in the case of Callorhinus, is the longest fissure in the brain. In Phoca, however, it is confined entirely to the dorsal aspect of the cerebrum, and at its caudal end it appears to terminate in a widely diverging fork or per- haps a small transverse fissure, possibly corresponding to the lunate (Wilder) of the cat. Its course is approximately paral- lel with the intercerebral cleft and is somewhat tortuous. At 70 JouRNAL OF COMPARATIVE NEUROLOGY. its cephalic end it appears to communicate with the cephalic branch of the supersylvian. This appearance will be discussed more fully under the description of the ansate fissure. The ectolateral fissure occupies a relatively higher or more dorsal and caudal position than in Callor/inus. It is of a more secondary character and courses approximately parallel with the postsupersylvian. The cruciate, unlike that of Ca/loriinus, is represented upon both the mesal and dorsal aspects. On the left hemicerebrum a shallow is present in the dorsal portion not far from the mar- gin. No distinct ‘‘ursine lozenge” is present here as in Cal- lordinus. The foreshortened condition of this region may have something to do with its absence. ’ A well defined postcruciate fissure is present on the left side. It presents a zygal (Wilder) or quadriradiate form. A slight secondary fissure near the olfactory bulb may represent a rudimentary precruciate fissure. The superorbital fissure shows a better development than in Callorfinus and similarly has no connection with the rhinal. But the opposite end, dissimilarly, extends farther and is over- lapped by the olfactory bulb. The medilateral is not present in Poca asa distinct fissure. Its location is occupied by a series of short disconnected fis- sures. The coronal fissure is a relatively longer fissure than in Cal lorhinus but is not so entirely disconnected from adjacent fis- sures. Its dorsal end lies caudal to the cruciate. On the left hemicerebrum it is separated by a shallow from an apparent connection with a continuation of the cephalic branch of the supersylvian. On the right hemicerebrum the shallow is sug- gested by the interlocking at this point of two submerged but- tresses. The ansate fissure, while not distinctly represented as an independent fissure, would, it seems to me, be indicated by the fissure extending from the coronal to the cephalic branch of the supersylvian, where, on each hemicerebrum, the interlocking of submerged buttresses would again suggest a shallow shutting it Fisu, Bram of the Fur Seal. 71 off from the branch of the supersylvian, and then continuing to the lateral fissure where a slight spur pointing toward the inter- cerebral cleft might indicate its separation from the lateral. Owen in his figure of the hemicerebrum of Poca represents a corresponding fissure as the coronal. Mesal Aspect. There is a slight appearance of the callosal fissure in the splenial half of the callosum, but none at all for the remaining half. The hippocampal fissure is well developed and needs no special comment. The splenial fissure is well developed and in general is as described for Callorhinus, except that its position is farther re- moved from the frontal portion of the cerebrum and that its cephalic end cuts the margin and is shown upon the dorsal sur- face. The postsplenial has about the same relations as in Cadlor- hinus. The fissura sublimica of Kikenthal' is poorly represented in my specimen of Poca and is somewhat confused with smaller branches and secondary fissures. It lies between the splenial and the callosum. Kiukenthal finds this fissure also present in Phoca grocnlandica, Phoca barbata, Macrorhinus leoninus and Otaria jubata. In Callorfinus there was no appearance of this fissure whatever. The fissura sublimica anterior of the same author is more clearly represented. In my former paper, on account of its position dorsal to the callosum, I designated it questioningly as the supercallosal. On the left hemicerebrum it is well developed and connects with the cruciate. On the right side, however, the fissure is independent and much smaller. In addition to this fissure, on each hemicerebrum, there is an- other dorsal to it and in front of the splenial. In Cadllorhinus I have called it the presplenial. The marginal or supersplenial is well shown in Phoca as in Callorhinus but lies nearer to the dorso-caudal margin, approxi- mately parallel with the splenial. Inthe intervening gyre there are a few secondary fissures. 1 Untersuchungen an Walthieren, 1889. 72 JouRNAL OF COMPARATIVE NEUROLOGY. On the meso-ventral surface a fissure corresponding to the collateral is also present, but, unlike Ca//orjinus, it has connec- tion with the postrhinal. Between the collateral and the post- splenial there is another we!l marked but unnamed fissure which is parallel to the former. It corresponds perhaps to the fissure in Callorhinus which I have spoken of tentatively in connection with the occipital. The genualis and rostralis are represented but the latter differs from that in Callorhinus in being much less developed and occupying a more ventral position at a more or less acute angle to the genualis. URSUS AMERICANUS. This brain, while fairly well preserved, had been consider- ably mutilated in removal, so that for purposes of illustration and reference, a specimen from Ursus thibetianus, kindly loaned by Prof. B. G. Wilder,’ was utilized ; so that while the figures of the lateral and mesal aspects are from the latter specimen, the description is based almost entirely upon the former. The general arrangement of the fissures is similar and the minor de- tails need not cause misapprehension. The fissural plan of the brain is much like that of the canine, minus the first circumsyl- vian arch. The olfactory bulbs and crura are far superior in size to those of either of the seals. The olfactory fissure is likewise well marked. The rhinal fissure passes into the Sylvian and continues, after forming an angle delimiting a well developed pyriform lobe, as the postrhinal and ending freely. The subfissure (postica ?) in the caudal wall of the Sylvian extends to and, on one side, actually appeared to communicate with the postrhinal. Lateral aspect. The Sylvian is directed in the usual dorso- caudal direction at the bottom of which is a small and simple area representing the insula. There is no appearance of a trans- 1 Papers, chiefly anatomical, presented at the Portland Meeting of the A. A. A. S., August, 1873, are devoted largely toa description of the brains of Carnivora, Fisu, Brain of the Fur Seal. 73 insular fissure although the presence of a subgyre and subfis- sure (postica ?) in the caudal wall of the Sylvian might super- ficially indicate it. . The supersylvian fissure forms a complete arch around the Sylvian. There is no indication of a separation of a postsuper- sylvian except near the free end of the Sylvian where a branch from the supersylvian extends into the adjacent gyre. The lateral fissure forms a curve approximately parallel with the supersylvian. As compared with Phoca and Callorhinus it is much shorter. If the conception of the ectolateral is cor- rect, the latter is continuous caudally with the lateral, a slight spur indicating the place of probable separation. The ectolat- eral extends parallel with the postsupersylvian but its ventral end does not reach so far in Uysus americanus, while in the Thibet bear the reverse is the case. The ansate fissure is a cephalo-ventral continuation of the lateral, a small spur of the latter indicating a point of separa- tion. The ansate describes a curve, the convexity pointing toward the Sylvian. The coronal fissure continues from the ansate and ends freely near the superorbital. The convexity of its curve like that of the ansate points toward the Sylvian. The point of its separation from the ansate is indicated by a spur more marked than that between the ansate and the lateral. The superorbital, unlike Poca and Callorlinus, has a very distinct connection with the rhinal fissure at about half of the distance from the Sylvian fissure to the olfactory bulb. It curves cephalo-dorsad with its convexity pointing cephalad. The cruciate fissure is more highly developed than in either of the seals. It appears slightly upon the mesal aspect and ex- tends obliquely cephalo-laterad on the dorsal surface. Around its free end the coronal fissure demarcates a well-formed sigmoid gyre. The appearances found in Poca approximate the condi- tions regarding the gyre more than in Cal/lorhinus. Between the cruciate and ansate lies the postcruciate fis- sure. On the left hemicerebrum it is well marked, on the right it is smaller and superficially connected with a minor fissure, 74 JoURNAL OF COMPARATIVE NEUROLOGY. On the right hemicerebrum a branch is given off from the cruciate extending cephalo-mesad. It is the precruciate fissure. On the left hemicerebrum it is an independent fissure. In neither case does it reach the mesal surface. The precruciate with the cruciate forms a well-defined triangular area—the ‘‘ ur- sine lozenge’’ of Mivart. On the dorsal surface between the lateral fissure and the intercerebral cleft there is a well marked fissure but it is not as deep as the other fissures. It is the con- finis. On the right hemicerebrum a short fissure connects it with the lateral. The medilateral fissure arises at the caudal end of the ce- rebrum near the mesal margin, in much the same position as in Callorhinus and continues down the ventral aspect close to the caudal margin. Mesal Aspect. The splenial fissure does not reach the dor- sal margin as in the case of Poca and as on one side in Cadllo- rhinus. Its cephalic end is, also, nearer the caudal end of the cerebrum than in either of the other two forms. In this respect the fissure occupies an intermediate position in the Phoca. It arches around the splenium of the callosum and courses along the tentorial surface of the cerebrum as far as the caudo-lateral margin, ending eight millimeters from the free end of the post- ‘supersylvian. Two or three short branches are given off along its course. Beyond the presence of a slight spur there is no evidence of a postsplenial fissure, nor of a supersplenial or mar- ginal as in the case of the seals. A well developed presplenial or fissura sublimica of Kikenthal is present, resembling that of Phoca more than Callordinus. No distinct fissura sublimica was present except in the case of the Thibet bear where a small minor fissure held the proper position. The genual and rostral fissures were present and occupied the same general relation to the cephalic end of the callosum as in Callorhinus. The callosal and hippocampal fissures have the same general relations as in other forms. Fiso, Bram of the Fur Seal. 75 ZALOPHUS CALIFORNIANUS. Through the kind permission of Professor Wilder I was permitted to remove the brain from this young sea lion. Its mother came originally from the Pacific coast and the present specimen was found dead in the cage with her while in transit to the East and was presumably not far from ‘‘ term.” It meas- ured 43 centimeters long and has been in the Cornell museum of Vertebrate Zoology for some years. The brain was ina fairly good state of preservation and was photographed soon after its removal. It was too delicate to permit of much manipulation and some of the fissures were not sounded as thoroughly as in the other specimens. The cere- brum of this specimen does not show the same degree of com- plexity relative to the fissuration as indicated by Murie’ in Ofaria jubata. A direct comparison of the fissures, however, is not easy as the latter author attempts to homologize them with those of the human cerebrum. | The olfactory apparatus is well developed. Not as largely as in the bear, however, but greater than either of the seals. The rhinal fissure, as in the other forms, is well marked and passes caudad into the mouth of the Sylvian fissure. The postrhinal is formed from the subfissure (postica ?) and has no connection whatever with either the rhinal or Sylvian. Lateral Aspect. The Sylvian is prominent and occupies its usual position. In its caudal wall is a subfissure (postica ?) and subgyre which as in Cadllorjinus is continuous on the ventral aspect with the pyriform or temporal lobe. The supersylvian with its cephalic and caudal portions, the pre- and postsupersylvian, is more nearly in accord with the condition found in the bear than in either of the seals. It rep- resents an intermediate condition between the two. The pre- supersylvian lies very close to the Sylvian but does not actually enter it as in the seals. Its average distance from it is about 4 millimeters ; while the distance from the Sylvian to the post- 11874. Transactions of the Zoological Society of London. 76 JouRNAL OF CoMPARATIVE NEUROLOGY. supersylvian is four times as great or 16 millimeters. There is no sign of disconnection between either the supersylvian and the postsupersylvian, or the supersylvian and the presupersyl- vian. The supersylvian forks or sends out a branch cephalad connecting with the ansate fissure exactly as in Phoca. The lateral fissure is relatively to the length of the cere- brum shorter than in any other forms. Its cephalic end and its relation to the ansate is again exactly the same as in Phoca. On the left hemicerebrum the lateral is disconnected at a little more than half of its length, by a narrow isthmus. The coronal fissure corresponds with that of Phoca, con- necting, superficially at least, with the ansate and thus indirectly with the cephalic branch of the supersylvian and the lateral. The ansate fissure, as has already been intimated, like that of Phoca is irregular in its form and connects with the fissures above mentioned. The ectolateral fissure is quite well down toward the ven- tral portion of the cerebrum and as in Cal/orfinus appears upon the ventral aspect. The medilateral fissure is scarcely perceptible on the lateral aspect; it lies exactly along the caudal margin of the hemicere- brum as in Callorfinus and is better seen in a mesal view. The cruciate accords, in position and relation, more closely with the conditions found in the bear and Ca//lorjinus; but while it reaches to the mesal surface of the hemicerebrum it does not cut it as far as in the bear and Phoca. The precruciate and the postcruciate fissures are likewise present and have exactly the same relations as in the bear and Callorhinus. Mesal Aspect. The callosal fissure is well developed. On the right hemicerebrum it does not continue around the genu as in the left. The splenial fissure does not extend as far cephalad as in Callorhinus, nor as far dorsad as in Phoca. It is situated more closely to the splenial half of the callosum than in either of the preceding or in the bear. A branch is given off in the region of the splenium proper which seems comparable to the Fisu, Brain of the Fur Seal. 77 postsplenial in the seals. A slight spur in this region in the bear may indicate an analogy. The presplenial is not represented asa distinct fissure on the left hemicerebrum, the only possible suggestion of it being a forking at the cephalic end of the splenial. On the right hemi- cerebrum a small but distinct fissure lying cephalad of the sple- nial may represent the presplenial. The marginal fissure is well represented and on both hemi- cerebrums cuts the dorsal surface as in Cadlorhinus. In Phoca although relatively long it does not reach the dorsal margin at all. In the bear the marginal fissure is not represented. The genual and rostral fissures are but slightly developed in this specimen and bear the same relations as in other forms. The cruciate fissure shows slightly on the mesal aspect and in its relations to the other parts resesembles that of the bear more than any of the others. FISSURAL INTERPRETATIONS OF OTHER WRITERS. The Sylvian fissure, in Phoca at least, has been located as a vertical fissure (presupersylvian) which has, for a portion, only, of its length, been submerged in the cephalic wall of the true Sylvian. Numerous writers have also described this condition as the anterior and posterior branches of the Sylvian. The two fissures morphologically are entirely distinct. In Hyrax Krueg does not represent any indication of a Sylvian fissure whatever. The supersylvian is very commonly called the suprasylvian. Leuret et Gratiolet have confused this fissure with the lateral in Phoca. Following Krueg the fissure which is designated as the post- supersylvian, is commonly known as the postsylvian of Owen. What I have designated as the presupersylvian and which is only exceptionally independent, is usually described as the an- terior or frontal portion of the supersylvian. A fissure corresponding to the coronal is represented by Krueg as the presylvian in Phoca. Kikenthal makes a similar representation. Turner in Macrordinus represents a correspon- ding fissure as the presylvian and a branch connecting with it 78 JourNAL OF CoMPARATIVE NEUROLOGY. as the coronal. In Z7zchecus (walrus) he figures as the presyl- vian an apparent continuation of the lateral, and represents as the coronal an apparent continuation of a third arched fissure designated by him as the medilateral. The superorbital fissure in carnivora generally is designated as the presylvian by many writers. The cruciate fissure is shown by Krueg, in Phoca, as exist- ing only on the mesal aspect, occuppying the position of the presplenial, or anterior sublimica of Kiukenthal. Leuret et Gratiolet represent the fissure as seen on the ventral aspect at the cephalic end. Other writers place it as usually seen in car- nivora at the cephalic end of the dorsal aspect where it may or may not reach the mesal surface. THE LATERAL VENTRICLE (PARACOELE. ) On removing the dorsal portion of the hemicerebrum just dorsal to the callosum the lateral ventricle is revealed. In the bear the cavity bends cephalo-ventrad to form the precornu and caudo-latero-ventrad to form the medicornu. The striatum is a well defined body forming a portion of the floor of the ventricle in the cephalic region. Parallel with the oblique margin of the striatum is the fimbrial margin of the hippocamp. Between these two margins—the rima (great transverse fissure) the chor- oid (para) plexus—a continuation of the velum enters the floor of the cavity. The hippocamp pursues its usual curved direc- tion in the medicornu. In Phoca the lateral ventricle is relatively very much larger than in the bear and the parts present quite different relations to each other. The striatum is the same as in the bear; along its margin is a well developed plexus, but between this and the fimbrial edge of the hippocamp there is an area equally as large as the striatum; this is the optic thalamus, but that portion of of it represented in the floor of the cavity presents the same general appearance as to its surface (endymal) as do the other parts. The supposed delicate endymal membrane extending from the plexus to the fimbria has been designated as the para- tela by Wilder. The hippocamp, then, is removed some little Fisn, Brain of the Fur Seal. 79 distance from the striatum and arches around the surface of the thalamus in a ventral direction. Caudal to the hippocamp, the cavity is about as largely represented, and in size forms a dispro- portionately large postcornu. Along the mesal wall just caudal to the hippocamp is an ental ridge correlated with an ectal de- pression—the splenial fissure. This is comparable to the calcar or hippocampus minor of the anthropoid and human brains. It is larger in proportion than either of the above. The splenial in this case for a part of its course at least is, therefore, a total (Wilder) or complete (Cunningham) fissure since the whole thickness of the parietes is involved; the ental elevation being correlated with the fissural depression. In this specimen of Phoca, then, we have two total fissures, the hippocampal (al- ways) and a portion of the splenial. The conditions just described might naturally suggest a homology with the ape and human calcar and that the splenial fissure, in this seal possessing a postcornu, might be homolo- gized with the occipital or calcarine fissure in man. A question might properly arise here as to which fissure it might be homol- ogized with. In the human foetus the occipital is a total fissure, but loses its totality (ental elevation) in the adult. Its position might favor its homology with the splenial, for if the latter were rotated farther caudad it would come to occupy approxi- mately the same position as the occipital. To homologize with the calcarine we would have to imagine a still farther rotation of the splenial. The calcarine is a total fissure throughout life and is the correlative of the calcar. Some doubt may therefore be expressed, assuming the homology to be reasonable, whether this hippocampus minor represents the occipital eminence—a foetal condition in the human brain, or the calcar—a structure persistent in the adult. The relative disproportion in the growth of the caudal or occipital portion of the cerebrum may have some bearing in accounting for the presence of the postcornu. Tiedemann in his figure of the lateral ventricle of Phoca gives no indication whatever of a postcornu. In Callorhinus the conditions resemble more closely those 80 JouRNAL OF CoMPARATIVE NEUROLOGY. in the bear; the rima is narrow and the thalamus does not ap- pear at all in the floor of the ventricle. A slight caudal spur of the cavity at the medicornu represents the postcornu. The splenial fissure, so to speak, just escapes the cavity, lying im- mediately caudal to it. In the walrus Turner' represents a dissection of this cavity but shows no indication of a postcornu, but in the text he states: ‘where the cavity of the ventricle curved downward and out- ward into the horn, an indication of a recess was seen in its pos- terior horn, but it did not amount to a cornu and there was no elevation which could be called a hippocampus minor. ” Murie,” on the form and structure of the Manatee, figures a well developed postcornu. He states that, ‘‘ there is an un- doubted posterior cornu, a fully developed hippocampus minor and an eminence I am inclined to recognize as eminentia collat- eralis.’’ The same author, On the Anatomy of the Sea Lion, Otaria jubata, figures a more extensive postcornu than is repre- sented in the manatee and describes it as ‘‘ stretching backwards and outwards with a very regular sweeping arch, and goes well back into the occipital lobe, terminating in a shallow tapering extremity. The eminentia collateralis is not distinctly defined ; but what appears to represent the outwardly bulging hippo- campus minor has a length of 0.7 of an inch, and at widest is 0.3 to 0.4 broad.” Wilder in the Anatomical Technology, in indicating the lines of inquiry likely to be most productive of results in the homology of the human and feline fissures, states that ‘‘be- tween the ordinary carnivora and the monkeys are two groups whose brains should be studied with especial care; the seals have a rudimentary postcornu and occipital lobe, and these parts are said to be developed in the Lemurs which have affinities with both the carnivora and the primates.” In none of the accounts have I seen any direct mention of 1°84. Turner, Report on the Seals collected during the Voyage of H. M. S. Challenger in the years 1873-1876. 21874. Transactions of the Zoological Society of London. Fisn, Lrain of the Fur Seal. 8I the correlation of the splenial fissure with the calcar in these aquatic forms. This fact, even if it be of no direct use for ho- mology, is, at least, interesting. MONACHUS TROPICALIS. In August, 1897, I was fortunate to obtain through the courtesy of Dr. A. H. Hassall, Washington, D. C., two brains, from male and female specimens of the West Indian Seal Mon- achus tropicals. They arrived at an exceedingly opportune time for comparison with the other brains dealt with in this ar- ticle. A study of their form and fissural relations throw much light on some of the points which seemed quite aberrant in fhoca when compared with Callorhinus alone. The general form of the brain would suggest a position in- termediate between the fur seal and Poca particularly in the frontal region which is somewhat foreshortened and broader than in Callorhinus. The caudal portion of the cerebrum is much elongated, noticed particularly upon the mesal aspect when measured from the splenium of the callosum; as if, perhaps, to compensate for the foreshortened frontal region. The cerebrum also shows a slightly greater overlapping of the cerebellum. The olfactory bulb and crus resemble the corresponding parts in Phoca, but show a slightly greater development. Fissures. Postica. In all four hemicerebrums, this fissure sends a branch to the surface, thus appearing superficially as a branch of the Sylvian. The postica is less easily distinguished in Monachus than in any of the other forms, as it is submerged practically to the bottom of the Sylvian fissure. In Callorhinus there is a branch corresponding to that of MWonachus but it does not extend deeply enough to connect with the postica. The postrhinal appears as the merest trace of a fissure and has a very superficial connection with the postica. The Sylvian fissure. It is in the Sylvian region that we get numerous clues to the intermediate position of MWonachus. In the brain of the female the Sylvian has practically the same direction as in Callorhinus. In the male, the true Sylvian really branches cephalad, although there is a superficial extension in 82 JourNAL oF CoMPARATIVE NEUROLOGY. the usual dorso-caudal direction. Apparently some unusual conditions exist here, which may perhaps be accounted for by the nearly complete disappearance of the postica. The presupersylvian resembles the corresponding fissure in Phoca regarding its extreme vertical position and apparent union with the Sylvian for only the ventral third of its course. It differs from Phoca in not being disconnected from the su- persylvian. The supersylvian fissure resembles that of Poca in extend- ing a branch of good size to connect with the ansate. Postsupersylvian. In the two hemicerebrums of the male there was a connection between the supersylvian and the post- supersylvian much as in Poca. In the hemicerebrums of the female there was an entire disconnection of these fissures. The cruciate fissure more than in any of the others resem- bled that of Phoca. It forms a good intermediate stage be- tween Callorhinus and Phoca. As with Phoca the fissure is rep- resented on the mesal surface as much, if not more than upon the dorsal. In the left hemicerebrums of both brains the cru- ciate is apparently continuous with the splenial. Upon the right hemicerebrums there is no such connection. Precruciate. In all four hemicerebrums the precruciate extends over upon the mesal surface for some little distance. It is more largely represented upon the dorsal surface and its lateral end makes a very decided curve toward the coronal fis- sure. There is almost a superficial connection between the cruciate and the precruciate. The conditions in Poca indicate that such a connection has occurred even to the extent of their almost complete mergence into eachother. ‘‘Ursine Lozenge.”’ This area is, with the exception of Phoca where it is undistinguishable, smaller than in any other forms. It is nothing more than a narrow gyre, situated at a slightly lower level than the adjacent gyres, suggesting a prob- able preparation for the loss of its identity in Phoca. Postcruciate. In MJonachus this fissure was the least satis- factorily represented than in any of the other forms. In the two hemicerebrums, it does not seem to be represented at all, Fisu, Bram of the Fur Seal. 83 unless we interpret a slight branch from the cruciate as repre- senting it. In the right hemicerebrums the fissure is distinctly present but is very small. The splenial accords more closely with Phoca in its posi- tion, reaching the mid-dorsal region instead of extending farther cephalad as in Callorfinus. It sends off a branch corresponding to the postsplenial as in other brains. The presplenial is well represented in the two right hemi- cerebrums, but in the two left it appears to connect the true splenial with the cruciate. The interlocking of submerged but- tresses at the proper points indicates a superficial connection merely. The marginal fissure is more poorly developed than in any of the other forms except the bear. A series of short inter- rupted fissures takes its place. A well marked collateral fissure is present and resembles the corresponding fissure in Ca//orhinus very closely. Postcornu. Perhaps the most important point in connect- ing Monachus with Pioca, is a very well developed postcornu. Callorhinus shows the merest trace of one and in the bear it is absent. In Monachus it does not go so far as in Phoca, a great portion of the caudal wall being solid. The floor of the postcornu in Monachus is quite distinctly convex This convex- ity of the internal surface is found to be correlated with an ex- ternal depression, the lower or ventral portion of the splenial fissure. At the more vertical portion of the fissure, namely, opposite the caudal end of the callosum, the splenial fissure loses its totality and and becomes an ordinary fissure for the remainder of its upward course. The postcornu stops at the level of the depth of the splenial fissure in the callosal region. We have not, therefore, as in Phoca, a’ well developed calcar (hippocamus minor). The internal convex surface already spoken of in connection with the ventral portion of the splenial fissure, offers a suggestion as to the inception of the calcar which finds its fulfillment in Phoca. 84 JournaL oF ComparAtIvE NEuROLOGY. GENERAL CONSIDERATIONS. The average canine brain, as a matter of convenience, may be accepted as a simple type of a carnivore brain. The fissures are clearly demarcated and there is an absence of much branch- ing or secondary fissuration. Around the Sylvian there are three arched fissures separ- ating the cortical substance into four distinct folds or gyres. In the brains of cats and occasionally in dogs we find that the arched fissure nearest the Sylvian is not a complete one; that only the pillars are represented, the keystone being absent. In Hyena and Proteles the frontal portion of this arch is wanting (Krueg) but the caudal portion, fissura postica, is well represented. Correlative with this state of affairs the postsuper- sylvian, as compared with the presupersylvian, is situated at least twice as far from the Sylvian fissure. In certain others of the carnivora no trace of the first arch or Sylvian gyre with its its limiting fissure (anterior-postica) is at all present. The first arch with its fissure has disappeared, apparently swallowed up by the Sylvian. There are represented then on the lateral aspect only two arched fissures, the supersyl- vian and on the lateral aspect only the three gyres which they separate. In those forms in which only the two arched fissures are present, if the distance from the frontal portion of the super- sylvian to the Sylvian be compared with the distance from the latter to the postsupersylvian, it will generally be found to be less in the former, and this becomes much more emphasized in the case of some of the bears, where the frontal portion of an undoubted supersylvian almost enters the Sylvian fissure. In his desciption of the brain of the Polar bear, Uvsus maritimus, Turner says: ‘‘On opening up the Sylvian fissure I found to my surprise that a definite arched convolution was completely concealed within it. It was separated from the convo- lution which bounded the Sylvian fissure by a deep fissure which was also concealed. Its anterior limb, not quite so bulky as the posterior, was continued into the supraorbital area immediately external to the rhinal fissure and to the outer root of the olfac- Fisu, Brain of the Fur Seal. 85 tory peduncle. Its posterior limb reached the postrhinal fissure and the Jobus hippocampz. I could not but think that we had here, more completely than either in the walrus or seals, a sink- ing into the Sylvian fissure of the convolution which ought to have bounded it, so that both the Sylvian convolution properly so called, and the suprasylvian fissure were concealed within it. If this be a proper explanation of the arrangement, then the three convolutions on the cranial aspect would be saggital, me- diolateral, and suprasylvian ; whilst the two complete curved fis- sures between them would be the mediolateral and lateral.” The question quite naturally arises if the fissure concealed in the Sylvian may not be the equivalent of the anterior-postica of Krueg and the two remaining visible on the cranial surface, the supersylvian and lateral. The medilateral of other authors does not attain the size nor continued length in the frontal direction as ascribed to the mediolateral by Turner. And furthermore there is in some forms, as in the seals, a well defined medilateral in addition to the two principal fissures. In a specimen of Ursus americanus, I had the good fortune to discover a stage one step beyond that described by Professor Turner. On opening the Sylvian fissure I found in its caudal wall a completely submerged fissure, with a remnant of the Syl- vian gyre which might possibly be mistaken forthe insula. A true insula, although small, is present. This submerged fissure I take to be the disappearing vestige of the ectosylvian (Owen) or anterior-postica (Krueg). A study of foetal bear brains with reference to the distinct appearance of the first circumsylvian arch (anterior-postica) would be most important in this condition. It would seem then that the condition thus described in the polar bear and American bear would represent the method of disappearance, rather than the appearance, of the first circum- sylvian arch and prepare us for the conditions that we find in the sea lion (Zalophus) and the seals (Poca and Callorhinus). In the sea lion the conditions regarding the frontal portion of the Sylvian gyre are intermediate between the bears and seals. The presupersyluian fissure approaches very closely to 86 JouRNAL OF COMPARATIVE NEUROLOGY. the Sylvian fissure and the intervening portion of the Sylvian gyre, besides being narrower than in the bear, has also sunk slighly lower than the adjacent surface as if prophesying the conditions found in the seals. In the seals there appears to be some evidence, if the inter- pretation as to the frontal portion of the supersylvian fissure be correct, that after breaking up into branches with perhaps some disconnection of its parts, it shows a tendency to follow the ex- ample of the anterior-postica fissure, because in Phoca, at least, the supersylvian bifurcates a little beyond the free end of the Sylvian, one branch forming a well ‘defined arch around it, the other branch passing on in the frontal region. The branch, however, which forms the arch is not a long one but it extends to and superficially connects with a vertical fissure which for half its distance is submerged in the frontal wall of the Sylvian, and crops out again on the ventral aspect of the brain. This con- dition holds for both hemicerebrums of Phoca. Callorhinus throws a little light on this matter. In the right hemicerebrum the supersylvian is clearly continuous with the vertical fissure submerged in the frontal wall of the Sylvian but gives off a very short frontal branch. Superficially it is continuous with the post- supersylvian but a shallow at this point indicates a partial sep- aration. The direct continuity in the depth of the supersylvian with the vertical fissure would seem to point to the fact that the latter, after all, was nothing more than the frontal portion of the supersylvian, namely the presupersylvian. _ In the left hemicerebrum the parts are a little more com- plicated. The postsupersylvian is entirely separated, the super- sylvian is entirely distinct from the frontal portion and is quite irregular and branching in its course, but mainly vertical in its direction. Thus, taking the canine brain as exemplifying a simple fis- sural pattern and passing through the Felidae and Ursidae and sea lion to the seals where the fissures are more numerous and complicated by the presence of branches of considerable size, and more or less disconnection of some of the principal fissures, we may arrive at some understanding of the relationship and Fisu, Brain of the Fur Seal. 87 changes effected in passing from simple to complex conditions. In the general form of the brains that of the sea lion seemed to bear closer resemblance to that of the bear than either Callorhinus or Phoca—the latter the least of all. The elongated and narrow frontal portion of the brain as seen in the bear is represented in Phoca by a foreshortened and broadened region, less marked in Callorhinus and still less in Zalophus. The development of the olfactory lobes is also interesting. They attain their highest growth in the bear, next in Zalophus, then Callorhinus and least in Phoca. The triangular area on each hemicerebrum located between the cruciate and precruciate fissures and the intercerebral cleft, designated by Mivart as the ursine lozenge and believed by him to be of considerable importance in indicating a phylogentic relationship between the Pinnipedia and the ursine group of carnivora, was developed equally well in Zalophus and Cadllor- hinus. In Phoca it was not observable, althougn Turner states that in this form it is present but rudimentary and concealed in . the mesal fissure of the cerebrum. The length of the lateral fissure in Cal/orhinus is somewhat unexpected and in relation resembles a continuous lateral and ectolateral of the bear. In the sea lion and Pfoca the lateral is a relatively short fissure. In all but the bear there is an inde- pendent ectolateral fissure but it is not so satisfactorily devel- oped in Phoca. The postrhinal fissure shows an interesting variation in the different forms. In Callorhinus and Zalophus it has no connec- tion with the rhinal or Sylvian, but it is a direct continuation of the subfissure—postica. In Ursus the subfissure may occasion- ally reach it but as a rule it is distinct and the postrhinal con- tinues as an elongation of the rhinal. In Poca the separation of the subfissure and the postrhinal is still more marked, so that the rhinal and the postrhinal are practically different parts of one ‘and the same fissure, differentiated from each other by the presence of the Sylvian. The presupersylvian fissure is directly continuous with the supersylvian in Ursus, it is likewise continuous in Zalophus and 88 JouRNAL OF CoMPARATIVE NEUROLOGY. in Callorhinus except upon the left hemicerebrum of the pup. In Phoca the two fissures are distinctly separated The postsupersylvian is continuous with the supersylvian in Ursus and Zalophus but separated in Callorhinus. They are apparently continuous in Phoca, but a dorso-caudal branch and the presence of submerged buttresses at this point of junction would indicate that there was some attempt at separation. In the bear there is no elongation of the paracoele to form a postcornu; in the sea lion Murie finds a distinct postcornu present ; in Callorhinus it is quite rudimentary ; in Phoca Tiede- mann represents the paracoele with no appearance whatever of a postcornu. My own specimen, which so far as I know is nor- mal, shows a postcornu relatively as large or larger than in the primate brain with a distinct calcar or hippocampus minor in which a portion of the splenial appears as a total fissure. With the exception of the bear, concerning which I have no ‘data, and the additional brain from an adult Cadllorhinus and Monachus all of my material was from specimens not more than one year of age. It is believed, judging from a comparison of the brain of the young with that of the adult Ca//lorhinus as to bulk and complexity of fissuration, that comparatively little or no change occurs, especially in the latter respect. Mr. Lucas, who had casts of the cranial cavities prepared from the male and female fur seal, finds but slight difference in the size of the cavities, notwithstanding the fact that the bulk of the body of the male is about four times as great as that of the female. Of the representatives of the five groups examined, the brain of Cal/lorhinus shows a greater number of minor fissures and a more intricate arrangement and branching of larger fis- sures. With regard to the ground plan of the fundamental fis- sures, and allowing for the difference in the shape of the brains, that of the eared seals, Callorhinus and Zalophus, approximates in general more closely to that of the ursine carnivora than does Phoca. The latter, or earless seal, in some respects, appears aberrant. The arrangement of the cruciate and postrhinal fis- sures would seem to link it with the canine and feline carnivora ; while the peculiar development of the occipital region and the large development of the postcornu with its calcar point toward primate conditions. The group of lemurs is also said to pos- sess a postcornu and to have affinities with both the carnivora and the primates. Asa matter of convenience a table of the more interesting regions in the representatives of the different groups examined is herewith appended. Bde qwysys eurysjsod yqq1M pe}deuu09 yng yuesasd juosoid sainssy pojou “UODSIP [| BUS JO SalIas B younrsip A19A adaey AIA juesqe snojawnu 931nb yuasoig [esom pus [esi0q uMmoys Ay1va]9 JON MO|[BYS B JO UOTeOIpUt auos jnq snonutju0D peoeuu0dsiqg [euryqa Jo uonenunuos juasaid “DIOUT ySrs juesaid soinssy paydna -19}UI JOYS JO Satias B yuasaid Uy sIpuUy adit] snojetminu 331nb AreyusMAIpN yy [Bseul pure [esio(y [esiop pu [esoyy worjo9uU09 -SIP B OM} 19430 94} pur uoldeTU0D B MOYS OM} suUInIgs1991IMIeY INO; UT uvia|{s 131s quasoid quesoid jusasoid juasqe ]jeus jussoid sno1swnu yuasoig esi0q ]esi10q pejoouuoosiq snonuijuoo Ayjensn ynq pe}99uuO0DsIp dnd jo WSrs Axeyuourpni juasoid juasaid juasqe [ems juasoid yuasoig SIs es yuasqe yuasqe eynsuy|S1 aINSSY [V19}¥]][OD|F1 aanssy [eulsieyy| £1 yuasoid| oanssy [e19j}e[Ipey|z1 juasqe Juasqe yuasoid yuasaig UISIVUI [BSoul esiog| Juiyno ysnf jes1oq ayeionasa4g|$ [esioq snonuiju0s uvrayAs -ladns ywIM pejyoeuu0D| wWnAqsI90IMIEY Je] UC -19dNs YIIM SNonUIZUOD vorjsod YI uor9auUad [eIDy10d ns Ar9a ‘90813 a9 W|evonsod jo uonenuuoD' voysod jo uonenurju0D Jourjstp Aaa Jou “SHY IDUC/T juasaid *SNULYAOYIDI *snyojvz juosoid esiop AT}sour uvraAsiodns YIM snonuayuod uerAyAsiodns quIM snonuiuos vo1jsod Ayjeuondaoxe ‘jeurys jo uorenuiju0) — IVIBOI|II Nu1093s0q |O1 (,28U9Z0T IUISIE,,|6 soinssy — 1OUTT|g ajerons9jsog |Z ayeroni9|9 uvtayAsiadns}sog \¥ uvraAsiadnsaig|€ jeary.ysog)z yuasaid](¢) eo1sod aanssyqns)1 "SUSAQ | *U01say — gO JouRNAL OF COMPARATIVE NEUROLOGY. DESCRIPTION OF PLATES. REFERENCE LETTERS. ans.—ansate fissure. 6.—buttress. cal,—callosum. calc.—calcar. cf.—confinis fissure. c/.—callosal fissure. col.—eollateral fissure. cor.—coronal fissure. er.—cruciate fissure. e/,—ectolateral fissure. J.—fimbria. £-—genual fissure. hip.—hippocampus. ?.—lateral fissure. marg.—marginal fissure. me.—medicornu. ml.—medilateral fissure. pc.—postcornu, per.—postcruciate fissure. pl.—plexus. pre.—precornu. prer.—precruciate fissure. prh.—postrhinal fissure. prsp.—presplenial fissure. prss.—presupersylvian fissure. psp.—postsplenial fissure. pss.—postsupersylvian fissure. r.—rostral fissure. vh.—rhinal fissure. so.—superorbital fissure. sf.—splenial fissure. sty.—striatum. Syl.—Sylvian fissure. J§s.—supersylvian fissure. th,—thalamus. ur.—ursine lozenge. PLATE X. Fig. 7. The ventral aspect of the brain of the fur seal Ca/lorhinus ursinus. On each side of the cerebellum is a depression into which fits the petrosal por- tion of the temporal bone. Fig. 2. The dorsal aspect of the brain showing the cerebellum largely concealed by the cerebrum. Fig. 3. The left lateral aspect of the cerebrum of a young specimen. Fig. 4. The right lateral aspect of the cerebrum of an adult male Callo- rhinus, Pig. 5. The mesal aspect of the right hemicerebrum. Fig. 6. The mesal aspect of the left hemicerebrum. PLATE XI. Fig. 1. The ventral aspect of the brain of the haired seal, Pheca vitulina, slightly modified from Tiedemann’s figure. Fig. 2. The dorsal aspect of the cerebrum of Phoca vitulina, after Tiede- mann. Fig. 3. The left lateral aspect of the cerebrum. Fig. g. The right lateral aspect of the cerebrum. Fig. 5. The mesal aspect of the right hemicerebrum. Fig. 6. The mesal aspect of the left hemicerebrum. Fisu, Brain of the Fur Seal. gI {PLATE XII. fig. 7. The left lateral aspect of the cerebrum of the sea lion, Zalophus californianus, fig. 2. The right lateral aspect of the cerebrum of Zalophus. fig. 3. The mesal aspect of the right hemicerebrum. Fig. 4. The mesal aspect of the left hemicerebrum. Fig. 5. The left lateral aspect of the cerebrum of Ursus thibetianus. Fig. 6. The mesal aspect of the right hemicerebrum of Ursus, fig. 7. Dissection of the left hemicerebrum of Cal/lorhinus, showing the lateral ventricle with a very rudimentary postcornu. Fig. 8. Dissection of the left hemicerebrum of Phoca vttulina, showing the presence of the calcar and large postcornu in the lateral ventricle. Fig. g. Dissection of the right hemicerebrum of Sonachus tropicalis show- ing a postcornu of intermediate size. PLATE XIII. Fig. z. The ventral aspect of the brain of a female Monachus tropicalis. fig. 2. The dorsal aspect of the brain of a female Monachus. Fig. 3. The left lateral aspect of the brain of a male Monachus. fig. 4. The right lateral aspect of the brain of a female Monachus. fig. 5. The mesal aspect of the left half of the brain of a female Monachus. THE CORTICAL MOTOR CENTRES IN LOWER MAMMALS. By C. L. Herrick. WITH PLATE XIV. Two recent papers relating to the excitable zone of the cortex of the opossum recall an extended series of experiments made by the writer prior to 1892 but which have been pub- lished only in part and seem to have failed to reach the public for whom they were intended. R. H. Cunningham!’ remarks: ‘‘To be sure, the micro- scopical as well as the macroscopical anatomy of the opos- sum brain has been minutely described by Herrick, who regards the precrucial lobe as typically motor in its micro- scopical structure and the parietal and occipital portions as composed of motor and other nerve cells, but this writer does not state whether this view has been corroborated by a physiological investigation of the cortex with the electrical current.’”’ The writer must admit that he has often failed to take the usual steps to bring his papers to the attention of fellow workers (though reference is made to experiments in the paper quoted) and it may be that the fact that Professor Cunningham as also Professor Ziehen? overlook the observations referred to, is due to in part to this negligence. That part of the series not heretofore published has been withheld in the hope that an opportunity might yet arise for the completion of the contemplated series. The preliminary account appeared in the paper prepared jointly by Professor W. G. Tight and the writer and printed in the Bulletin of the Laboratories of Denison Univer- 1 The Cortical Motor Centres of the Opossum. Journ, Physiology, XXII, 4. * Ueber die motorishe Rindenregion von Didelphys. Centralblatt fiir Phys- tologie, XI, 15. Herrick, Cortical Motor Centres. 93 sity, V. 1890.1 The paragraphs immediately germain are the following : ‘Previous to the sectioning, as already said, several localiza- tion experiments were made both by electrical stimulation and extirpation. The first specimen was a male of Avctomys monax, the same specimen which furnished the sections most frequently described and figured beyond. Ether was employed as an an- esthetic, and the skin was parted down the median line of the head and the skull removed over the anterior and middle parts of the left hemisphere. The current used was from one Grove cell and was just enough to operate the induction coil, producing an irritation easily endured by the tongue. When the electrodes were introduced ata, Fig. 4, Plate V (at about the anterior one-third, near the median line and corresponding approximately to Munk’s region C of the dog) a forward and outward motion of the right fore leg was produced. A stronger current pro- duced an an electro-tonic contraction of the muscles of the whole right side. At the point 4, about 5 mm. behind anda little outward from the above (corresponding to about the pos- terior margin of Munk’s region D), the stimulus produced a straightening of the right hind leg. At the point e, about 8 mm. behind J, and near the median fissure (corresponding to about Munk’s region F, near the median line), the stimulation resulted in a sharp contraction of the orbicularis palpebrarum and orbicularis oris of the right side and some feeble contrac- tion of the facial muscles of the left side, probably due to super- ficial irradiation. ‘‘At the point d, about 8 mm. behind c, and farther from the median fissure (corresponding to the anterior margin of Munk’s region A), the insertion of the electrodes produces no motor disturbances nor did any point back of d. By a series of trials it was found that the electrodes produced some motor disturbance of the fore leg at all points within the area marked A, but not beyond it. ‘The area B likewise marks about the limits of the hind leg 1 The Central Nervous System of Rodents, etc. 94 JouRNAL OF COMPARATIVE NEUROLOGY. region. An area of about 5 sq. mm. was then removed from the cortex of the left side of the fore leg region at about a. The wound was then dressed and the animal allowed to recover. The power of abduction of the right fore leg was lost. After some. time another portion of the cortex was removed, a little back of d, on the left side. After recovery it was found that the animal was blind in the right eye. These experiments serve to locate some of the motor and sensory regions of the cerebral cortex for the subsequent histological works. ‘It may be of interest to note also one of the series of ex- periments of electrical stimulation upon the Raccoon, Procyon lotor, which has been employed for comparative study. The animal was a male about three-fourths grown. Just enough cur- rent was used to drive the coil. Ether and choloform mixed were the anesthetics employed. ‘‘Nearly the whole upper surface of the cerebral hemis- phere of the left side was exposed. The loss of blood was very moderate. ‘‘The electrodes were introduced at point 1, Fig. 2, Plate XI, about 9 mm. from the median fissure and a little in advance of the line passing through the anterior angle of the eye. The result was a forward and inward motion of the right fore leg. “At point 2, the introduction of the electrode gave an un- defined movement of the right fore leg. “At point 3, about 7 mm. from the median line and sep- arated from I by a faint sulcus, the stimulation produced a flex- ion of the pes on the crus and elevation of right hind leg. ‘“‘At 4, a movement of the right hind leg, as at 3, anda slight rotation of the fore leg inwards. “At 5, about 3 mm. from the median line, the stimulation produced an extension and divarication of the digits of the right foot. “‘At 6, the fore leg was elevated and flexed, and with a little stronger current the hind leg was also elevated and flexed. ‘“‘At 7, there was an extension of the toes corresponding to the movements produced at 5, of the opposite side. Herrick, Cortical Motor Centres. 95 ‘‘Perhaps more important in its bearing on the present sub- ject is a set of experiments upon the opossum, of which, unfor- tunately, no very exact data have been preserved. The con- figuration of the hemispheres as well as the details of structure resemble very closely those of rodents. Moreover, such hints as we have of the development of the Rodentia indicate a com- mon origin for the two groups and comparatively slight subse- quent differentiation. It, then, would not be surprising if a considerable similarity of distribution in the cortical elements should be proven to exist. On the other hand, the existence of an apparent homologue of the crucial sulcus near the front of the cerebrum would lead one to expect the aggregation of the motor elements near this sulcus. The experiments in this case were made with a Grenet cell and DuBois-Reymond coil, with the secondary coil at about 8 cm. the current being applied by a pair of platinum electrodes separated by about 3 mm. . Stimulation of the region about the crucial sulcus (so-called) resulted in movements of the anterior extremity, but the diffi- culty in controlling the flow of blood interfered with close anal- ysis. The area on either side of the median fissure responded with various poorly localized contractions of the trunk. About 6-8 mm. posterior to the crucial sulcus and 4-5 mm. from the medial line is an ill-defined area governing the hind leg. These motor reactions were, in the main, crossed as usual, but in sev- eral instances similar motions of the muscles of both sides re- sulted when superficial irradiation appeared to be excluded. The areas thus roughly mapped in the opossum coincide in gen- eral with those of the ground hog and weare forced to conclude that the crucial sulcus of the opossum is not strictly homolo- gous with the fissure so named in carnivora.”’ Cunningham locates the motor centre for the fore limb along the caudal boundary of the so-called crucial sulcus and doubtfully reports an area for movements of the ear in the lat- eral region, pretty well caudad. He states that there is no def- inite centre for the hind leg or, at least, that if the animal is thoroughly under the influence of the narcotic no motion can be evoked by electrical stimulation. Ziehen, on the other hand, 96 JOURNAL OF COMPARATIVE NEUROLOGY. locates the hind-leg region behind the ‘‘crucial sulcus”’ about at the point devoted by Cunningham to the flexion of the toes of the fore foot. He adds however that this region is less easi- ly excitable than the others. Ziehen is also struck by the fact that his results reverse the relation between the fore and hind leg regions as compared with the rodents and insectivora. On the other hand, our own results, so far as they go, are in har- mony with what we know of other groups, although this method of localizing the cortical areas is, we believe, open to grave ob- jections from the theoretical as well as the practical standpoint. t is no unusual thing for an animal that seems to be fully under the influence of the narcotic to respond in the most distinct and apparently localized manner to stimuli at the most unexpected points, as a result probably of general irritation. An illustra- tion of this is afforded by the raccoon brain of figure 2. The cortical area indicated on the right side was first extirpated. Before recovery from the narcosis there was a tendency to ro- tate toward the right, i. e., the side of the injury. After re- covery the animal seemed quite blind on the opposite side. He would snap at a stick or flinch from a threatening blow when within the field of the right eye, but quite ignored the same movements on the other side. A day later there seemed to be no irritation and the pupil reflexes were normal. The loss of sense of touch of the left fore leg was quite evident, though hard to differentiate because of the loss of vision. Subsequently the left hemisphere was exposed and stimulation experiments were attempted in the usual manner but which proved quite uncer- tain. Nevertheless, the areas in which are found the points numbered 1, 4, 8, and 13 presided over the motions of the hind leg and foot. This area is somewhat farther caudad than the hind leg area described in the raccoon in the experiments quoted above but in the same general region. Regions further laterad both in front and behind the area mentioned occasioned mo- tions of the fore foot and leg. Fig. 3 gives a few data from stimulation experiments with a kitten. Here points 4 and 5 are devoted to the hind leg, 3 Herrick, Cortical Motor Centres. 97 and 2 to the fore leg and 1 is on the border between the two regions. A large series of stimulation experiments on the musk rat (Fiber) are passed over as wholly ambiguous or contradictory. Fig. 4, 6 illustrates a case of extirpation in thissubject. Previ- ous electrical stimulation of the cephalic part of this area re- sulted in motions of the fore foot for the most part. After re- covery from the operation there was very evident impairment of the power of abduction in the manus and pes of the right side. When pushed towards the right the animal offers no resistance and may be turned upon its back, while the effort to push it to- ward the left is met by decided resistance. Tactile sensation is also weakened on the right side, as shown by the reaction to the needle. A second specimen was deprived of the cortex within the fore-leg region and, after recovery, some impairment of the motion of the opposite fore-leg was apparent. The ani- mal seemed unable to extend the fingers and the fingers double under while walking. It is easy to push the animal over to- ward the injured side. On the other hand the coordinated mo- tions like those of washing the face are carried out without dif- ficulty. In another specimen a larger area was removed extend- ing further caudad. Fig. 4, c. The symptoms were much the same except that there is a greater apparent involvement of the hind limb. Loss of muscu- lar sense is suggested. The power of adduction seems to have suffered most. After 10 days recovery is nearly complete, at least so far as locomotion is concerned, but there is a curious awkwardness, hard to describe but easily recognized. When sitting up it tends to fall toward the side opposite to the lesion. Vision was permanently destroyed in the eye of the opposite side. After 20 days all efforts to provoke motion result in ro- tation to the right. This is due, in all probability, to the ex- tension of the effect of the injury by inflammatory processes. In yet another case in which the same area was removed the rotation began soon after the operation and continued till it ter- minated in convulsions, The animal was blind in the opposite eye. 98 JouRNAL oF CoMPARATIVE NEUROLOGY. EXPLANATION OF PLATE V. Fig. 1. Dorsal aspect of brain of Didelphys virginiana; 1, point where fingers of left hand were extended and divaricated ; 2, left leg moved cephalo- dorsad ; 3, fingers of left manus extended and divaricated ; 4, extension of right manus; 5, same as I. Fig. 2. Dorsal aspect of brain of Procyon lotor. Extirpated area of right hemisphere shaded. See text. The numbers on the figure not referred to in the text indicate places the stimulation of which was followed by no definite reaction. fig. 3. Dorsal aspect of brain of half-grown Felts domestica; 1, flexure of pes and contraction of shoulder muscles ; 2, extension of left manus and for- ward motion of arm; 3, inward rotation of left manus; 4, forward motion of left hind leg; 5, flexure of toes of right pes. Fig. 4. Brain of Fiber zibethecus ; 6, areas extirpated. See text. Fig. 5. Localization in Opossum brain according to Ziehen. J, hind leg; F, foreleg; J, mouth-facial region. SHES NERVE, CELL ASA. UNIT.’ By Pierre A. Fisu, D.Sc., D.V.S., Ithaca, New York, WITH 7 TEXT-FIGURES, The sciences of morphology and physiology, perhaps more than any others, were of slow development. Their early years were enshrouded in mysticism and magic. Progress was retarded largely by theological opposition associated with superstition. The ancients believed that the soul was slow in leaving the body and that the latter should not, therefore, be used for dissection at once. The period allotted for this migration of the soul, left the body in anything but a fit state for investigation. This opposition did not extend to chemistry and other sciences, which, at that time, were in a flourishing condition. With the rexazssance there came a renewed interest in anat- omy, and in Italy it was decreed that oxe body should be dis- sected annually at the universities. This, curiously enough, was done by a barber’s assistant with a razor. There was a time when it was the custom to administer to the inner, as well as to the outer, ills of mankind. Barbers were particularly adept at bleeding, and combined the science of phlebotomy with that of shaving. To advertise this pro- fession they erected signs in the form of poles wrapped around with red and white bandages—the red to indicate the bleeding, and the white, the soapy lather. We must, doubtless, look upon our modern barber poles as heirlooms of this ancient and honorable profession, deprived, to some extent, of their old time significance. Because the sphere was accepted as the symbol of perfec- tion by the ancients, Plato regarded the more or less globular 1 Read at the quarterly meeting of the Cayuga County Medical Society, Auburn, N, Y., Feb. 10, 1898. 7 100 JournAL OF CoMPARATIVE NEuROLOGY. head as the seat of intelligence and perception. With slow gradations the apparent fantastic and irregular form of the wrinkled brain surface has been systematized into a general ground-plan. Segments have been differentiated and a fissural pattern for the cerebrum has been formulated. Deeper than the surface, however, there is encountered a bewildering maze of cells and fibers, the intricate arrangement and complex relations of which have at the present time, only begun to be under- stood. In the achievement of a great discovery, many are prone to overlook the factors by means of which it is made possible. The discovery of a new planet very justly brings great renown to the discoverer,—we usually stop at that and take no cognizance of the wonderful mechanism of the telescope, the laws of astron- my, and other accessories that co-operate in the grand result. And so it has been with our knowledge of the structure of the nervous system, great as it is today but at the same time inade- quate. The results of the last ten years which have so com- pletely revolutionized our conceptions of the nerve elements were possible only through improvements in microscopical ap- paratus and technique, and the improvement of histological methods. With the additional knowledge gained from the new methods, there must of necessity occur change in the termin- ology. The old notion of a nerve cell (justified by the old methods) that it consisted merely of a cell body with its en- closed nucleus and nucleolus is no longer tenable. Important as are these parts to the nutrition and’ activity of the cell, no less important to the full attainment of its function is the pres- ence of its various appendages. The Golgi-Cajal method is too well known to require any description. The formation of a silver-bichromate deposit in or upon the nerve cell and its processes has furnished us with pic- tures of these elements, which for beauty and clearness of out- line surpass anything that has preceded it. The results furnish us with at least a workable hypothesis regarding nervous phe- nomena which before was merely conjecture. This method has shown, and accumulated evidence seems Fisu, Zhe Nerve Cell as a Unit. 10! to confirm it, that there is complete morphological indepen- dence of the nerve elements, with perhaps certain exceptions, in rare cases, where a direct anastomosis of one nerve cell with another has been described, as in the battery of the torpedo and also in certain of the sense organs, as noted by Dogiel, Ayers, Masius and others. This morphological isolation of the ele- ments does not preclude the idea of physiological continuity which must of necessity exist. This isolation of elements has led to the production of the term neuron (Waldyer ’91), neurone (Van Gehuchten ’93), neurocyte (Fish ’94, after an unknown French writer), neura (Rauber ’94) and neure (Baker ’96), to signify the nerve unit, including the cell body with all its appendages. The term neurocyte has been suggested in this connection because its lit- eral meaning is a nerve cell and includes not merely the cell body, which from custom we regard as the equivalent of a nerve cell, but all of its appendages as well, just as in speaking of the leucocyte, we include the various extensions from the cell mass. The analogy may be carried still farther for under certain specia! conditions we may conceive that the pseudopodia of the leuco- cyte may be considerably extended and attenuated and from the juxtaposition of numerous other elements lose, or partially lose, their powers of retraction and movement; under such condi- tions we may consider the neurocyte comparable with the leuco- cyte so far as form is concerned. _ The appendages of the cells, with perhaps the exception of those of the spinal ganglia, appear to fall naturally into two categories ; those which collect or convey the impulse to the cell, cellipetal processes or dendrites, and those which discharge or carry impulses away from the cell, the cellifugal processes or neurites (axis-cylinders). Along with our increasing knowledge of the form of neuro- cytes there have been contributed new facts bearing upon their activity. For our purpose, we may consider the neurocyte as made up of a mass of granular protoplasm, with more or less branching appendages, containing a large nucleus of a reticu- lated character enclosing, usually, a prominent nucleolus. We 102 JouRNAL OF CoMPARATIVE NEUROLOGY. have a bit of material protoplasm similar to that of other body cells, and yet for a long time any structural change due to the activity of the nerve cell eluded the keen vision of investigators. It has been said that the secretion of a gland cell is of a ma- terial character; that of a muscle cell, mechanical energy and we might naturally expect to find in these tissues, changes demon- strable by the microscope; but the secretion of a nerve cell is consciousness which is not exactly material, and its effect upon the cell is too subtle to leave a trace. Hodge'in his fatigue experiments extending over a period of four or five years, has shown the fallacy of this view. His experiments dealing with artificial and normal fatigue were performed in a most faithful and conscientious manner on a wide range of forms with conclu- sive results, the most of them having been confirmed by later investigators. Fig. 1. Fig. 2. Fig. 3. Figs. 1,2 and 3, after Hodge. Only the cell bodies areshown. Fig. 1, represents the normal cell body with its large reticulated nucleus and the chro- matin diffused throughout the cytoplasm. Fig. 2, shows the effects of fatigue, the nucleus having become shrunken and irregular in outline, with a surround- ing area devoid of chromatin. The peripheral portion of the cytoplasm is also poor in chromatin. Fig 3, showing vacuolation of the cytoplasm as the ef- fect of fatigue. For the artificial fatigue experiments the spinal ganglion cells were chosen and the nerve connecting with the ganglion was subjected to a weak electrical stimulation for a given length of time. The spinal ganglion of the opposite side was removed without stimulation and used as acontrol in the experiment, the treatment of the two ganglia being identical after they were removed from the body. In the fatigued cells he found slight shrinkage in size, with vacuolation of the protoplasm. In the nucleus there was a marked decrease in size, nearly 50%; a change from a smooth and round toa jagged, irregular out- 1 Jour. Morphology. Vol. VII. Pages 95—164. 1892. Fisu, The Nerve Cellas a Unit. 103 line ; and a loss or condensation of the open reticulated appear- ance. The control ganglion showed none of these changes. If, after stimulation, the cells were permitted to rest and then examined, these changes were not apparent. For the study of normal fatigue, certain birds and bees were examined, some of them were killed before entering their daily routine, while their cells were presumably as yet ina state of rest; others were killed just at night-fall after the completion of their day’s work. A comparison of those killed in the morning with those killed in the evening, showed in the latter changes as marked as those produced by artificial fatigue. To make the evidence still stronger, and to show that the effects were not the result of histological reagents, it remained for Dr. Hodge to study the living cell. For this purpose he chose the cells of the sympathetic ganglia of the frog. Two frogs were pre- pared in exactly the same way, except that one received a weak electrical stimulation while the other did not. The unstimu- lated frog showed no change, while the nucleus of the cells of the stimulated frog showed very marked shrinkage and irregu- larity of outline. Certain well defined changes in the constitu- tion of the nerve cells of very old persons as compared with the newly born have also been demonstrated. Hodge has shown that fatigue effects occur in brain cells as well as those of the spinal ganglia. As early as 1884, Flesch noted differences in cells in their reaction to staining reagents due to internal modifications as an effect of their functional activity, and according to this affinity for color he designated the cells as chromophile and chromo- phobe. Vas (92) has demonstrated changes in the cells of the cer- vical ganglia, due to their functional activity, and confirms, in the main, the points that have just been stated. Asa prelim- inary result of this activity Vas has further noted that there is, at first, a swelling of the cell. This has also been confirmed by Mann (’94) who has extended the observations to the motor cells of the spinal cord and the sensory cells of the retina of the dog. From his researches, Mann concludes that during rest, 104 JourRNAL OF CoMPARATIVE NEUROLOGY. several chromatic materials are stored up in the nerve cell and that these materials are used up by it during the performance of its function; that activity is accompanied by an increase in the size of the cells, the nuclei and the nucleoli of the sympa- thetic, ordinary motor and sensory ganglion cells; that fatigue of the nerve cell is accompanied by the shriveling of the nucleus and probably also of the cell and by the formation of a diffuse chromatic material in the nucleus. Lugaro ('95 ) confirms the observations of Mann. Cellular changes of such a radical character, as has been shown above, may be the result of perfectly normal functions and disappear after a period of rest. _Howimportant is it, then, before discriminating between that which may be perfectly nor- — mal and that which is abnormal, to know thoroughly the effects incident to natural activity. In connection with the matter of electrical excitation of the nervous system, the question naturally arises, since we have such complete evidence from an experimental standpoint, what will be the result of the application of a fatal current of electric- ity? Will a very strong current applied for a few minutes af- fect the structural character of the nerve cells in a manner sim- ilar to those stimulated by a weak current for a very long time ? As opportunities have presented there have come to me por- tions of the brain and myel of four persons executed by elec- tricity, as well as from a horse struck dead by a live wire. In the first case, designated as W, a portion of the oblongata, the location of so many vital centers, was carefully studied. The number and size of the vacuoles in the cytoplasm were astonish- ing. In all, however, the nucleus appeared full and regular, al- though the cytoplasm in some of the cells seems to have become completely transformed into vacuoles. In the second case, L, examination was made of the same region and here no abnormal change of any kind could be de- tected. The cells were full and plump as were also the nuclei and the nucleoli, and the cytoplasmic chromatin showed no evi- dence of disintegration or disappearance. A portion of the cortex was also examined and in both the large and the small Fisu, Zhe Nerve Cell as a Unit. 105 pyramidal cells a very considerable amount of vacuolation ap- peared, especially in the apical dendrites, and occasionally in the body of the cell. In the third case, B, only a small portion of the cerebel- lum was studied. It required considerable search and patience to find in these sections any distinct structural change of the cells. After the examination of many sections two Purkinje cells were found, each of which showed the presence of a small vacuole. The fourth case, C, required more than one current to cause his death. The pyramidal cells of the cortex were also examined and those of the oblongata to a lesser extent. Here also there was evidence of vacuolation in the apical dendrites of the pyramidal cells, while the others, including those of the ob- longata, so far as examined, were perfectly normal. In the case of the horse the injury was inflicted at the shoulder, differing thus from the others in point of contact with the electricity. No unusual appearances were detected in the neurocytes. I have ventured to present these results, incomplete as they are. If they do nothing more, they will, I think, empha- size the importance of a working knowledge of the changes that may occur in a neurocyte as a result of its legitimate processes. The vacuolation of the cell body and of the nucleus is described by many to be due to pathological causes of various kinds, among which may be mentioned, insanity, alcoholism, epilepsy, as well as the action of various toxins and alkaloids. As has been shown by Hodge and others, many of the described path- ological changes may be duplicated by normal processes, and these, so far as possible, should be eliminated before rendering a decided opinion. Bearing upon the matter of the rapidity with which effects may be produced upon the nerve cells as a result of shock are recent experiments of Parascondolo (’98 ) ' who produced upon guinea pigs a condition of shock by striking some of them upon 1 Arch, de Physiol. norm. et path. XXX, 5th series, X. No. 1. p. 138. 106 JOURNAL OF COMPARATIVE NEUROLOGY. the thorax and some upon the abdomen. If the animal died im- mediately there were no results detected in the nervous system If, however, the animal lived thirty or forty hours, as some of them did, well marked lesions were demonstrated. By Nissl’s method he found in the motor cells of the myel a perinuclear, as well as a peripheral chromatolysis, also vacuoles in the cyto- plasm and an eccentric position of the nucleus. By the Golgi method he found deformation of the cell body but not ‘to the extent of atrophy, and a distinctly moniliform appearance of the dendrites. An inference derivable from the above experiments, is that changes of a structural character do not occur instantaneously in the neurocyte, especially if the injury be not directly applied to the nervous system. Parascondolo’s experiments are of in- terest in showing how soon the lesions may be zzduced through the inter-dependence of the tissue systems. A comparison be- tween these experiments and the results of electrical excitation shows that fatal currents of electricity may zzduce changes in the dendrites of the nerve cells in a practically instantaneous period of time, under unfavorable conditions, as the current is prevented from direct action upon the brain by the presence of the meninges, bones of the cranium, and scalp. With the weaker Fig. 4. Fig. 4. After Cajal, showing the transformation of the bipolar into the unipolar spinal ganglion cell. currents practically the nervous tissue alone was dealt with, un- der the most favorable conditions. Other things being equal, we may expect that a current of greater intensity will produce given results in less time than a current half as great. Pugnat! has demonstrated this in his experiments, finding that it required 1 Bibliog. Anat. VI, pages 27-32, 1898. Fisu, Zhe Nerve Cell as a Unit. 107 twice as long to produce certain results with a weak current as when one of twice the strength was used. We must avoid the danger of regarding the cerebro-spinal axis as a rigid and unyielding mass of substance. The action of the brain is molar as well as molecular, as evidenced by its general movements due to inspiration and, expiration. In the earlier stages of development there are migratory movements of the neuroblasts of an amoeboid nature in order that they may reach their destined positions in the adult structure. The so- A Fig, 5. Fig. 5. After Cajal, showing the changes undergone by the cerebellar granule cells, reading from left to right. called bipolar spinal ganglion cells are the permanent condition in such low forms as the ‘‘fishes;” those of higher forms pass from this stage in early development to the unipolar condition of the adult. ween’ oo” - Fig. 6. Fig. 6. After Fish (Central Nervous System of Desmognathus fusca), show- ing the changes in the form of the neurocytes as they pass from the ental to the ectal boundary of the layer of nerve cells. Cajal has shown that during their growth the granule cells of the cerebellum pass through even more elaborate changes than those of the spinal ganglia. Changes in the form of the cells and their appendages are also apparent in the central ner- i108 JouRNAL oF CoMPARATIVE NEUROLOGY. vous system of certain salamanders, as the neurocytes reach the boundary of the cellular layer. Here are evidences of the plasticity of the nervous ele- ments. Dothey lose this property entirely after they have reached maturity? It has been pretty well demonstrated by modern histological methods that these elements are morpho- logically independent, and the hypothesis of contiguity or over- lapping of the parts is now very generally accepted, instead of the older view of continuity or direct anastomosis of one cell Fig. 7. Fig. 7. After Berkley, showing a nerve cell with its processes (human); #, neurite ; ¢, collateral; @, d, d, dendrites; g, gemmulze. Illustrating Berkley’s hypothesis of the way in which the nervous impulse may pass from one nerve cell to another by contact of the gemmule. with another. Contact of one element with another is sufficient, it is believed, for the transference of a nervous impulse. The Fisu, Zhe Nerve Cell as a Unit. 109 idea has been advanced that even in the adult state the neurocyte has not completely lost its power of amoeboid movement, but that this property is still retained at the terminals of its appen- dages. This view is not accepted by Kolliker, nor entirely by Cajal, who thinks that the neuroglia cells aré more mobile than the nerve cells. The experiments upon the activity and fatigue of the nerve cell indicate that a change of volume may occur, a turgescence as a result of activity, and a shrinkage when carried to the ex- tent of fatigue. Situated in the lymph spaces and constantly bathed with the lymph for nutritive purposes, we may expect to find certain osmotic processes going on between the contents of the cell and its surrounding medium and that these processes may be influenced by the activity of the cell and that certain of them may occur coincidently with the transmission or origina- tion of the impulse in the cell. Along the dendrite, and especially well pronounced in the cortical cells, are slight lateral spurs known as gemmule. The condition of these, as well as certain irregularities in the form of the dendrites, have been noted by Berkley and others as the result of pathological causes. Berkley has shown that in cer- tain diseased conditions gemmule have been missing. He be- lieves that the cell and its dendrites has a delicate limiting mem- brane through which the gemmule protrude, as naked bits of protoplasm, coming into contact with similar uncovered masses of protoplasm from the neurite or its collaterals, or in contact with the gemmulz of other dendrites and that at these points the impulses are transferred. Any destruction or abnormality of these gemmulz would of course, interfere more or less seri- ously, with the normal conveyance of the impulse. The transference of nervous impulses from one element to another through contact, due to amceboid movement, would be of material importance in the explanation of the phenomena of sleep, intellectual processes, and pathological conditions. Before pathology has spoken its final word we may hope to know more of the remarkable chemical complexity of nervous tissue, com- posed, as it is said, of some three hundred or more different ele- R10 JournAL oF ComPpaRsTIVE NEUROLOGY. ments and compounds, If, in closing, I could have one fact shine out beyond any other it would be the idea that, while there is a morphological independence of the nervous elements, there is a physiological dependence; that, although there is unity there is community; and that a healthy psychic life is the: result of the summation of the individual activity of all the ner- vous elements. EDITORIAL. OUR COLLABORATORS. Since the publication of our last issue three additional col- laborators have been added to our editorial staff. Negotiations are pending with several others both in this country and abroad. We feel that our readers are to be congratulated that they are to have the services of men so eminent in their several depart- ments and representing so many leading institutions of this country and Europe. The staff of collaborators at the present writing includes the following: Henry H. Donaldson, Ph.D., Professor of Neurology, Uni- versity of Chicago; Growth and regeneration of nervous organs. Professor Ludwig Edinger, fvankfurt, a-M., Collaborator for Germany. Professor A. van Gehuchten, University of Louvain, Bel- gium ; Collaborator for France and Belgium. G. Carl Huber, M.D., Asszstant Professor of Hrstology and Embryology in the University of Michigan ; The sympathetic system and the peripheral nervous system. B. F. Kingsbury, Ph.D., Justructor in Microscopy, Histol- ogy and Embryology, Cornell University and the New York State Veterinary College ; Morphology of the lower vertebrates (Ich- thyopsida). Frederic S. Lee, Ph.D., Adjunct Professor and Demonstra- tor of Phystology, College of Physicians and Surgeons, New York City ; Physiology of the nervous system. Adolf Meyer, M.D., Docent in Psychiatry, Clark Univer- sity, and Assistant Physician to the Worcester Lunatic Hospital ; Human neurology. That the literary departments of the JouRNAL may be maintained at the highest efficiency, authors are requested to send books, monographs and other matter meriting editorial no- tice either to the Editor-in-chief directly or to the Collaborator 112 JouRNAL OF COMPARATIVE NEUROLOGY. in the appropriate department. It is felt that the purposes of this publication will be served best by issuing the matter as promptly as possible after its receipt and, accordingly, the pub- lisher will not feel limited to the strictly quarterly form but will issue fascicles at such times as may seem best in the interests of all concerned. REPORT ON NEURONYMY BY THE ASSOCIATION OF AMERICAN ANATOMISTS. It is much to be regretted that the results of the various attempts to secure harmonious and consistent usage in the terms employed in anatomy and especially neurology have resulted in emphasizing personal differences and producing very unscientific bitterness. It does not increase the attractiveness ofa field whose inherent difficulties are only too obvious to discover that its lan- guage is broken up into dialects the use of any one of which brands one at once with some ‘‘eponymic”’ adjective of re- proach. These remarks are suggested by the appearance of an elaborate report from the Association of American Anatomists accompanied by a caustic minority report involving personal criminations and complaints. The results of the questionaire recently reported in this Journal are such as to indicate that the system of Professor Wilder is not ‘‘ repulsive generally to educated men” and such a statement in the organ of a society of national importance betrays an amount of heat without light not calculated, to say the least, to avert the danger that Amer- ican anatomy should fall into ‘‘disgrace.”” It may be expected that the result of the discussion will be to cause many bewild- ered writers to adopt zz toto the only consistent and complete system at present at command while others will react ‘against every idea of reform and thus the breach may become impassi- ble, and for this result we shall have thank the committee of the Association of American Anatomists. For ourselves, we can only advise patience anda careful weighing of the claims of each term apart from any question of source and associations, with reference solely to the interests of our science. Too many problems of first class importance are pressing for solution to permit the student to fritter away time in nomenclature discus- sions, C, L. HERRICK. JOURNAL OF COMPARATIVE NEUROLOGY, VoL.AVIII. PLATE I. 4 SCHAPER, Dev. DENISON ENG. DEPT « wee oe : est ; JOURNAL OF COMPARATIVE NEUROLOGY, Vo. VIII. PLATE Il. JOURNAL OF COMPARATIVE NEUROLOGY, VoL. VIII. PLATE Ill - Pe Eve SCHAPER Dru. JOURNAL OF COMPARATIVE NEUROLOGY, VoL. VIII PLATE Iv Li ‘ JOURNAL OF COMPARATIVE NEUROLOGY. VoL. VIII. PLATE V. SS “a JOURNAL OF COMPARATIVE NEUROLOGY. Vot. VIII. PLATE“VI. va WAN —t PLATE VII. * VoL. VIII. JOURNAL OF COMPARATIVE NEUROLOGY. 4 é BG ate z h CUO arr f aN % 5) (\’ RGR 7 4 \ te ¥ ‘g ff ON.) ake AIA Nasa diy) 7 Sc NS f ayn rhs ie iw \ herman f NAD, Pang ) AN @LI SV ) aby ~ SS LE CED YD DNs Z On x hina, My 4 YN Ly HALE ran A Cem CO ath: UL Z a LE — LLY tole ” %. JOURNAL OF COMPARATIVE NEUROLOGY. Vox VIII PLATE Vill PLATE Ix, JOURNAL OF COMPARATIVE NEUROLOGY. Vot. Vill. JOURNAL CF COMPARATIVE NEUROLOGY. Volt. VIII. PLATE X. PLATE Xl. JOURNAL OF COMPARATIVE NEUROLOGY. Volt. VIII. al EY ON ES JOURNAL OF COMPARATIVE NEUROLOGY. Volt. VIII. PLATE XII. JOURNAL OF COMPARATIVE NEUROLOGY. VoL. VIII. PLATE XIll. JOURNAL OF COMPARATIVE NEUROLOGY. Volt. VIII. PLATE XIV. CRITICAL REVIEW OF THE DATA AND GENERAL METHODS AND DEDUCTIONS OF MODERN NEU.- ROLOGY. By Dr. AvotF MEYER, Worcester Lunatic Hospital, Worcester, Mass. Part I, With Plates XV to XIX. Neurological research seems to have struck a rich and im- portant vein for progress in the shape of the ‘neurone theory.’ No less than a dozen pamphlets have during the last few years, proclaimed this new fertile standpoint. The old division of ‘elements of the nervous system’ into nerve-cells, nerve-fibers and neuroglia is replaced by a simpler one—nerve-elements (neurones) and neuroglia elements. A nerve-element, or neur- one, consists of a cell body with nucleus and protoplasm and processes, either of the character of the protoplasm—dendrite, or of the character of a nerve-fiber—neurite. Nerve-fibers do not exist by themselves ; a fiber is always a part of a cell only, a process of a nerve-cell. Moreover, it is considered to be an established fact that a fiber splits up into thin branches on its course and that the ‘collaterals’ and ‘arborizations’ terminate blindly without forming anastomoses with other cells. Like all progress of science, this new conception, called neurone-theory, came to light in connection with a whole series of new observations, in neurogenesis and neurohistology. We shall see that it was slowly accepted in medicine, that it re- ceived various definitions and that it is in reality almost as if it meant simply the ‘modern views of the nervous system.’ Just now it seems as if a careful review of the fundamental data and problems would hardly be out of place. Dogmatizing without adequate basis has led to discrediting the ‘neurone-the- ory; certain facts seemed to militate seriously against its current 114 JouRNAL OF COMPARATIVE NEUROLOGY. presentation ; it is easy to see that much ‘ modern neurology ’ is merely old concepts in new words and that the best advantages of modern neurology are missed by those who think that the neurone-theory pure and simple will bring the much needed sal- vation. Many important facts are forgotten over the generali- zations drawn from them. His and Forel are hardly read by those initiated in the necessary generalizations from their publi- cation and their eminently sound methods of work are pushed into the shadow by the schematic silhouette work. The following sketches partly review certain facts not usually considered in the already numerous reviews on the ‘neurone-theory,’ partly they outline a standpoint suggested by data similar to those which led to the neurone-theory, a point of view from which the study of the nervous system re- ceives a certain order, without persistently ignoring the valuable lessons which His and Forel have given. The replacing of Meynert’s time-honored plan of the brain by one more in _har- mony with modern views, especially those suggested by work in pathology of the nervous system, has proved very stimulat- ing both for instruction and for a working-hypothesis; and the methods of observation and of reasoning which it suggests may prove fruitful to others. The Historical Development of the Neurone-Theory. In view of the number of accounts of the historical devel- opment of the neurone-theory, among which I mention Wald- eyer’s, Lenhossék’s, in the English literature Schaefer’s, and in our own Baker’s, Minot’s, and especially Barker's, we might limit ourselves to the enumeration of the principal facts which constitute the difference from earlier views and help to estab- lish the new conception. It seems, however, well to consider certain sides of the history of the neurone-theory frequently overlooked ; and to give careful summaries of the publications in question ; after that we point out the objections which are raised against the current formulation of the theory. As a typical statement of the views just before the neu- Meyer, Data of Modes Neurology. I15 rone-theory was advanced, we may quote the short summary of the text-book of physiology of Landois, edition of 1887. ‘The nervous elements present two distinct forms: nerve- fibers, non-medullated or medullated, and nerve-cells of various forms and functions. An aggregation of nerve-cells constitutes a nerve-ganglion. The fibers represent a conducting apparatus and serve to place the central nervous organs in connection with peripheral end-organs. The nerve-cells, however, besides trans- mitting impulses, act as physiological centers for automatic and reflex movements, and also for the sensory, perceptive, trophic, and secretory functions.” After a detailed description of the histology of the nerve-fibers, the writer says concerning the development of nerve-fibers: ‘‘ At first nerve-fibers consist only of fibrils, i. e. of axis-cylinders, which become covered with connective substance, and ultimately the white substance of Schwann is developed in some of them. The growth of the fibers takes place by elongation of the individual interannular segments, and also by the new formation of these.” No hint is given of the origin of all the fibers from nerve-cells. ‘‘ The ganglionic or nerve-cells have partly been considered as cells, partly as more complicated structures. We distinguish multi- polar and bipolar nerve-cells, nerve-cells with connective tissue capsule and ganglionic cells with spiral fibers. The large cells of the spinal cord have among their processes one non-ramified ‘axis-cylinder process’ which becomes the axis-cylinder of a medullated nerve-fiber. Whether the cerebral nerve-cells have such processes is still doubtful, etc.”’ This statement is repeated in the later editions. The American one of 1892 gives merely a few editorial remarks in parentheses, to the effect that ‘His and Forel claim that the protoplasmic processes do not anastomose but are merely in contact with one another,’ and a statement that ‘it is now cer- tain that the cerebral cells too have processes.’ It is very grat- ifying on the other hand, that W. T. Gowers, in the first edi- tion of his ‘‘Manual of Diseases of the Nervous System,”’ 1886, gives a view which has not received due attention in its day but was in many ways a perfect anticipation of the present 116 JouRNAL OF CoMPARATIVE NEUROLOGY. one, without sufficient generalization though, and without a full discussion of all available facts in its favor. The description of the pyramidal tract and its connection with the anterior horn cell is presented both in the text and in the drawing exactly as we would do today. He speaks of the pyramidal cell, its nerve- fiber and the terminal ramification of the latter in the spongy substance of the anterior horn, and of the anterior horn cell, the fiber proceeding from it, passing through the anterior root and nerve trunk to the muscle, where it divides and rami- fies on the muscular fiber. This was written before August 1886, but that a nerve-fiber was under all circumstances merely a part of a cell, was not even accepted by Edinger,’ who, as late as 1891, speaks of a double origin of nerve-fibers ; first, from nerve-cells, as Deiters and others had shown for the connection of motor nerve-fibers and the large cells of motor nuclei and ‘anterior horns,’ and second, from the ‘network consisting of all the processes of ganglion and glia-cells,’ a mode of origin illustrated by Gerlach and others especially for the posterior root-fibers. About August 1886, two Swiss scholars, Prof. His in Leip- sic and Prof. Forel in Ziirich, gave the medical public two stud- ies which established the conception that the nervous system consists of independent cells like all the rest of the human or- gans. Prof. His stood on the ground of embryology, Prof. Forel used the results of the method of Golgi for an analysis of the experimental work with v. Gudden’s method. The con» tribution of His? has undoubtedly furnished more direct data in favor of the new conception; and when Golgi, R. y Cajal, Kolliker, v. Lenhossék and others applied the silver method to embryonic material, a vast amount of detail sprang up enrich- ing our knowledge so rapidly as to make it difficult to follow. 1 EDINGER, Structure of the Central Nervous System, Philadelphia, 1891, P. 42. See also illustrations and discussions in Gerlach, Stricker’s Handbuch der Gewebelehre, Vol. 2, p. 679-685, 1872. 2His, W. Zur Geschichte des menschlichen Riickenmarkes und der Ner- ven wurzeln, 1886. Meyer, Data of Modern Neurology. 117 Prof. Forel’s’ classical study excels rather by the clearness and depth of critical analysis of a number of pathological and ex- perimentally produced conditions and as such it is the most suggestive publication for any one who wishes to advance the anatomy of the nervous system on the ground of the study of pathological lesions and degenerations in man, notwithstanding its containing a number of minor errors. Forel found the key to his problems in the discovery of Golgi. Before 1873, Camillo Golgi, a histologist and patholo- gist in Pavia, had discovered an extremely valuable method of impregnating nervous tissues with nitrate of silver. The great value of his procedure lies in the fact that his reagents do not affect all cells of a specimen, but only a few, and if an element is stained at all, it is so usually in its whole extent; the cell- body with all its processes and ramifications stands out in black. This peculiarity of the ‘black stain,’ the small number of cell- individuals brought out in their whole extent, furnished many un- expected data concerning the nerve-cells. About 1885 Golgi was ‘discovered’ by German histologists. Translations of a number of his contributions had appeared in this country long before that, in the ‘ Alienist and Neurologist,’ but had not fallen on fertile ground. In Germany, too, reports of his work had been published, but were held in a very sober, hardly ap- preciative tone of scepticism. The exploitation of the points which could promote the general concepts came through Prof. Forel. Golgi’s results are summarized by Forel as follows: 1. All the branches of the protoplasmic processes end blindly. They never anastomose, are uneven and show no fibrillary structure. 2. Every nerve-cell is wuzpolar, i. e. it has only one nerve process. 3. This nerve-process has very fine branches. 4. In the cells of the first category the nerve-process, after giving off a few fine collaterals, forms a medullated fiber ; 1ForEL, AuG. Einige Hirnanatomische Betrachtungen und Ergebnisse, Arch. f. Psych. XVIII. 118 JOURNAL OF COMPARATIVE NEUROLOGY. in the cells of the second category, the nerve-process dissolves completely into fibrils before it becomes a real fiber. The cells of the first type are motor, those of the second type sensory, and the terminations of the fiders of both join in the anxastomo- ses of a common network of fibrils. This is, on the main, the first exact demonstration of the standpoint of Gerlach (apart from the denial of anastomoses of the dendrites), with a greater number of corrections of detail ; but the delicacy of the new specimens suggested to Forel more than Gerlach’s views. Forel’s first criticism deals with the con- tinuity of the net-work of fibrils. The absolute absence of actually visible anastomoses in the Golgi specimens is quite striking. Further it is difficult to conceive how these thin fibrils coming from different cells would grow together at their ends and how the growth of the individual cells in the years of development of the nervous system could go on if there were real continuity instead of free end-buds. The next step is the assumption that all the fiber-systems and ‘net-works’ of fibrils throughout the nervous system are nothing but nerve-processes of definite sets of nerve-cells, and further that Golgi is wrong in calling all the cells motor whose nerve-process becomes a long fiber. The excess of fibers in the nervous system is only apparent, not real; the fibers are so much longer and larger than the cell ‘body’ to which they belong that the preponder- ance of ‘white matter’ over ‘gray matter ’ is not very surprising. The evidence which Forel adduces for the radical concep- tion that the nervous system consists of cells without anasto- moses of the processes and not of independent fibers plus cells is taken from the inability of finding anastomoses in Golgi preparations, and from the results with the method of Gudden and ‘secondary degenerations’ generally. Since the latter are relatively little studied by the physicians on our side of the Ocean, notwithstanding the summary of Séguin (Arch. of Med. X, 1892) and Spitzka’s work, I mention a few of the fundamen- tal results which Forel discusses. The first one refers to the external geniculate body, the principal end-station of the optic nerves, and the place of origin of a great share of the ‘ optic Meyer, Data of Modern Neurology. 119 radiation.” In the external geniculate body, all the cel/s degen- erate when the occipital lobe (with the optic radiation) is re- moved from the new-born ; if however, the eyes are removed, merely the ‘ gelatinous substance’ disappears, so that the cells become more closely crowded. Forel explains this as follows: the large cells of the retina send their fibers into the external geniculate body. There they lose the meduliary sheath and split up in end-brushes, or end-arborizations, merely coming into contact with the cells and their processes, and helping to con- stitute the ‘gelatinous substance.’ The gelatinous substance is not homogenous, but consists (besides the neuroglia and blood-vessels) of these end-brushes and the protoplasmic pro- cesses of the cells. When the cells in the eye are removed by operation, their processes degenerate and are resorbed, and con- sequently the cells in the external geniculate body come more closely together. If, however, the cortex is removed, the cells of the external geniculate body are affected and degenerate, be- cause their end-brushes are cut off without a chance of regener- ation ; the gelatinous substance, as far as the terminal fibers of the optic nerve constitute it, is not directly affected; only the cells of the external geniculate body and their processes decay and are obliterated. If the condition is produced in an adult animal, or by pathological conditions in the adult man, the ‘re- trogade degeneration,’ i. e. the affection of the cells which are merely cut through at the termination of their fibers, is not so marked, and as we know now through experiments with Nissl’s method, a more or less transitory matter. (For a similar illus- tration published 1891, see the chapter on ‘motor neurones.’) Another interesting fact is demonstrated with regard to the peripheral motor nerve-elements. v. Gudden showed that tear- ing out a facial nerve from the Fallopian canal in the new-born leads to complete degeneration, not only of the remaining per- ipheral branches to the muscles as Waller had thought but also of the cells of the facial nucleus and the remaining ‘cen- tral stump’;:.the degeneration of the cells and central stump does not take place, however, when the tear or section occurs further in the periphery and a chance at regener- 120 JoURNAL OF COMPARATIVE NEUROLOGY. ation presents itself. In this case the axis-cylinders of the central stump grow out again towards the muscles to which they belong, if, at least, they can follow a bed along which to grow, ‘ especially after nerve-suture. The study of Forel contains a number of considerations on the slight differences between what is seen in the animal when operated young or old, and also be- tween various findings in man, and, further, illustrations of the problems in the anatomy of the fillet (retrograde atrophy and retrograde degeneration), of the auditory centers, trigeminus and pyramidal tract. The problem of experimental anatomy of the brain might be formulated as we have it now on ground of this early study of Forel. He did not summarize it in the following words but has all the material expressed in the various parts of the paper: 1. All nerve-fibers are merely processes of cells. They terminate blindly in end-brushes, like the protoplasmic pro- cesses without anastomoses. The interrelation of the nerve- elements takes place dy contact, not by the continuity of a net- work. The net-work of Gerlach is a false net-work [comparable not with a net, but with the appearance of a dense forest where each twig belongs to only one tree, although it may be difficult on a photograph to trace each correctly]. 2. Tocall the cell body the trophic center of a nerve-fiber is justified only in the sense that the nerve-fiber is a part of the entire cell and that it is subject to the general laws of cell-vitali- ty. (Any part of the cell which is cut off from the nucleus will degenerate). 3. The results of experimental anatomy and the so-called secondary and tertiary degenerations are satisfactorily explained on this basis; further, a discussion of secondary degenerations etc. is only complete when the questions are put with these facts in view. We should not merely study fiber-tracts but always search for the cell-bodies to which the fibers belong. Forel does not summarize all the conclusions, since his aim is rather the explanation of certain disputed facts relating to von Gudden’s atrophy method, In a later publication he furnishes » Meyer, Data of Modern Neurology. 121 splendid illustrations of the principal points just set forth, We shall refer to them in full. Wilhelm His arrived at similar conclusions on ground of embryological observations and considerations. I mention here a summary of propositions which the famous embryologist communicated to the Anat. Gesellschaft.’ 1. There isa period in the embryonic life in which no nerve fibers exist, neither central nor peripheral ones. 2. The motor nerves originate as processes of definite cells of the spinal cord and brain. 3. The axis-cylinder processes appear as the first pro- cesses of the motor cells; the ramified (protoplasmic) processes are formed considerably later. 4. The motor cells show early a fibrillary cell-body and the fibrillation is continued into the relatively broad axis- cylinder. 5. The motor cells, both of the spinal cord and brain are located in definite and constant zones of the medullary tube ( brain-axis and spinal cord). The latter consist of a floor-plate and roof-plate which connect the lateral walls in the median line. The lateral wall again is divided into a dorso-lateral (alar) and a ventro-lateral (basal) part. 6. All motor nerve-roots originate from cells of the ven- tro-lateral (basal) part of the tube; but not all the cells of that part send their axis-cylinder into the motor roots. A few of them grow towards the anterior commissure and others seem to enter into the formation of the longitudinal tracts of the cord and brain. 7. The motor nerve-fibers leave the tube in several types (See the copy of the drawing of His, Fig. 4): A. The type of the motor spinal roots and of the 12th, 6th and 3rd cranial nerves. B. The type of the spinal accessory and the motor divisions of the pneumogastric, glossopharyngeal and 5th.— origin from a lateral nucleus of the ‘basal plate’ and exit at the 1Herr Wilhelm His, Ueber die embryonale Entwicklung des Nervenbahnen, Anat. Anzeiger, Vol. III, p. 449-505. 122 JOURNAL OF COMPARATIVE NEUROLOGY. junction of dorso-lateral and ventro-lateral lamina. C. Type of the facial nerve—it originates from a lateral cell-nest of the basal plate; runs towards the median line, then turns through a ‘knee’ around the median cell-nest (abducens group) and out- ward between the dorso-lateral and the ventro-lateral part of the lateral wall of the medulla. D. The type of the fourth nerve— the cell-nest lies in the basal lamina; from there the fibers grow outward, into the roof of the tube and after decussating with the bundle from the opposite side, come to the surface. 8. The cells of the spinal ganglion have at first no long processes ; then follows a stage when every cell is bipolar, and later on, the spinal ganglion cell is characterized by the pres- ence of two fibrillary axis-cylinder-processes which leave the cell in opposite directions and to-.which the cell- body takes an eccentric position. The central processes of the spinal cells grow towards the medullary tube and remain to a great extent on its surface, forming a longitudinal tract. g. Inthe spinal cord the ingrowing sensory roots form the primary posterior column; within the ‘brain’ the so-called ‘ascending roots’, the roots of the pneumogastric, glosso- pharyngeal, nerve of Wrisberg and the fifth are, as it were, the posterior column-formation of the medulla. 10. Not all the sensory root fibers of the neural tube grow simultaneously. At first there are fewer fibers than later. This holds also for the central tracts, the anterior commissures anid the antero-lateral columns. 11. The ‘ascending roots’ are first quite short and grow successively longer. 12. The peripheral nerve-trunks too grow gradually. In the above studies we have the root and the soul of the neurone-theory. It is, however, nothing but fair to say that Ramon y Cajal, the greatest promotor of neurology that Spain ever has produced, should be mentioned as the chief confirmer of the same concept. This is certain that none has discovered and demonstrated more individual details in confirmation of the views of Forel and His than this indefatigable Madrid _histolo- gist, and as he says himself, he has done what neither His nor Meyer, Data of Modern Neurology. 123 Forel have done conclusively ; he demonstrates the blind ar- borizations or end-brushes of so many kinds of nerve-fibers and the connection of the latter with cell-bodies in the cerebellum, the medulla, olfactory bulb, retina, optic centers, the great sympathetic, the cerebral cortex, that the concept and the de- tailed data are established beyond doubt (sin sombra alguna de duda’). The principal generalizations which we can derive from Cajal are as follows: 1. There is no substantial continuity between the pro- cesses of different nerve-cells. The nerve-elements represent cell-units, for which he accepts Waldeyer’s term ‘neurone’. 2. While Golgi assumed that the protoplasmic processes of nerve-cells hada purely nutritive function, Ramon y Cajal renders it probable that the protoplasmic processes are those parts of a neurone with which the arborizations of other neu- rones come most likely into contact and that the contact with the cell-dody itself is exceptional, being found only in those cells in which there are no protoplasmic processes (spinal gan- elion, retina). 3. The spreading of impulses received is cellulipetal in the protoplasmic processes and centrifugal in the axis-cylinder (law of dynamic polarization). We must mention here that one of Golgi’s chief discover- ies is the demonstration of fibrils or collaterals which branch at right angles from the nerve-process of almost every cell soon after its origin. The conception of Deiters, that the nerve- process was characterized by having no branches before its final termination, was modified by Golgi, inasmuch as he found that not only the protoplasmic processes have branches but also the nerve-process ; the difference lies in the branching at acute an- gles in the case of the dendrite, and at right angles in the case of the collaterals ; further, it is easy to see that the dendrites ¢aper towards the periphery, whereas the collaterals are smooth and 1Dr. D. Santiago Ramon y Cajal, Nuevo concepto de la Histologia de los Centros Nerviosos, Barcelona, 1893. 124 JournaL oF ComparATIvE NEvROLOGY. even, diminishing in calibre only by splitting up into terminal brushes and ending in little nodules. Golgi had also observed that the protoplasmic processes chiefly terminated near the blood-vessels and had therefore largely a trophic function, a view strongly opposed by others. For the transmission of nerve-impulses, according to Golgi, the collaterals are sufficient; the cell-body with the protoplasmic processes stands above the neural circuit as the trophic focus of the nerve element. This same view is maintained by Vausex on the ground of studies on invertebrates. The reader will see at a glance what the three investigators give when looking at the two figures (Fig. 1, Plate XV)—-A. representing the view of Golg¢ and of Nansen (from Nansen’s work 1887), B. the view of van Gehuchten-Cajal (from the latter's nuevo Concepto, etc). Ina later chapter we shall have to speak of the views of Berkeley and Held. This short sketch may suffice for the history of the question before us, the development of-the neurone-theory and of the embryological and experimental method. While the neurone-concept was slowly accepted by the almost totality of the investigators who become acquainted with it, there remained a few opponents, open or disguised. It will be desirable to pass them in review. The strongest and most emphatic opposition comes from Golgi and his pupils. As the publication of Dr. Achille Monti is probably not accessible to most of the readers of these notes, an abstract is given here. | Golgi had published in 1875 a description of the olfactory bulbs of mammals which established the following facts: The fibers of the olfactory nerve enter the olfactory glomeruli, ram- ify repeatedly at right angles, and form an intimate meshwork. A further constituent of the glomeruli are the dentrites (proto- plasmic processes) of the large and small cells of the db, and finally also collaterals and nerve fibers from the olfactory tract. These results are reproduced from the original in the second 1Sulla fina anatomia del bulbo olfattorio. Fatti vecchi e nuovi che contrad- dicono alla teoria dei neuroni. Nota del Dott. Achille Monti. Pavia; 1895. Meyer, Data of Modern Neurology. 125 plate of the atlas of Golgi, published by Fischer in Jena, in 1894. Fifteen years later Ramon y Cajal published a remarka- bly simple sketch of the architecture of the same region, and since then v. Gehuchten and Martin’, Pedro Ramon’, Kolliker’, Retzius‘, Calleja®, and Conil’, have essentially corroborated it ; it has passed into almost all of the later text books. It estab- lishes only the following connections: 1. The olfactory nerve fibers terminate blindly within the olfactory glomeruli. There they come into contact with 2. the dendrites of the large ‘‘ mitral cells’’ whose fiber processes help to constitute the olfactory tract. The termination in the glomerult of fibers from the olfactory tract ts strongly denied by all these witters, and the whole arrangement is commonly used asa paradigm of the connection of peripheral and central afferent neurones, and also as absolute evidence for the ‘law of dynam- ic polarization,’ according to which the dendrites receive the nerve impulses and they pass through them into the nerve fibers. Monti reestablishes completely the o/der observations of Golgi (especially the presence of recurrent fibers to the glomeruli) and demonstrates a certain superficiality of the in- vestigators mentioned. He points out that their inaccurate de- scriptions do not furnish an unassailable basis both for the theory of the conductive function of the dendrites and even for the mere contact connection between nerve elements since the network in the glomeruli is too complex and recurrent collater- als are undeniably present. 1Van Gehuchten et Martin: Le bulbe olfactif chez quelques mammiféres. La Cellule, T. VII, fasc, II. 2Pedro Ramon; Estructure de los bulbos olfactorios de las avis; Gaceta sanit. de Barcelona; Julio, 1890; and El encephalo de los reptiles; III, Bulbo olfactorio, Septembre, 1891. 3 Kdlliker: Ueber den feineren Bau des Bulbus olfactorius; Wiirzburger physikalisch.-Med. Gesellschaft; Dec. 1891. *Retzius: Die Endigungungsweise der Riechnerven. Biologische Unter- suchungen, Neue Folge, III, 3, 1892. 5 Calleja: La region olfactoria del cerebro; Tesi di Madrid, 1894. ®Conil: Mémoires de la Société de Biologie, 1892, p. 179. 126 JOURNAL OF COMPARATIVE NEUROLOGY. It is somewhat difficult to see why the omissions of the writers succeeding Golgi, even when conceded, should furnish so much evidence agazust the general trend of their explana- tions. Monti aud Golgi undoubtedly demonstrate by their work a spirit of great accuracy and conservatism, but it appears almost as if it were sufficient to state the case as a warning against too hasty and schematic generalizations rather than as proof for a strict negation of the newer working hypotheses. Van Gehuchten and Cajal must look for more evidence, it is true, to make an undeniable law out of their hypothesis; in this we agree with Monti. Golgi’s chief objection rests on his interpretation of the character off the diffuse network of the end fibrils. He claims that his opponents have not the true scientific spirit of accuracy when they consider the question settled in favor of the forest simile with an absolute denial of anastomoses. He himself stops before the inextricable maze and leaves the question un- decided. He favors the presence of a real network but does not deny Forel’s and Cajal’s views absolutely, as his opponents do the view of the network with anastomosis. Another objection is raised by Dogiel who maintains the presence of anastomoses in the cells of the retina. The over- whelming denial by other investigators of the retina would, however, invalidate his evidence considerably. As a very serious objection we mention the ever recurring description of regenerating fibers in the cut-off end of a nerve, before the central nerve processes have reached the portion peripheral to the cut. Bow/by', and again Kennedy”, claim to have seen new nerve fibers formed within the peripheral stump, not coming from the central stump, and /ater growing together with the fibers of the central stump. This would imply the growth of nerve fibers from something else than nerve cells and the possibility would hit fatally the dogma that no nerve fiber 1 Bowlby: Injuries and diseases of nerves ; London, 1889. 2Kennedy: On the regeneration of nerves. Proceedings of the Royal So- ciety, March 11, 1897. Meyer, Data of Modern Neurology. 127, can exist except as a process of a nerve cell. All this is em- phatically denied by all writers outside of England. Prof. C. Huber, of Michigan University, who has studied the question with Prof. Howell, and to whom a similar statement had been ascribed, writes me that he has seen no evidence in favor of such assumptions. In connection with this we must mention a discussion which was carried on in the camp of morphologists, and which is never referred to in the discussions concerning the neurone theory. The most important contribution is undoubt- edly one of /. Beard, the histogenesis of nerve’. A. Dohrn, the head of the Zoological Station at Naples, published memoirs on this subject which fully supported Beard’s own way of see- ing things. Dohrn (quoted in Beard’s article) says: ‘‘Thus we have the picture of a nerve such as is found typically everywhere. The nuclei are Schwann’s nuclei, the light shining cylinders are the axis cylinders, the plasma is the soil of Schwann’s and of the medullary sheath appearing later. These four elements constituting the typical nerve, are exclu- sively products of ectoderm cells disposed in chains for the for- mation of individual fibers.” Beard says (p, 295): ‘‘ These chains (i. e. their nuclei) proceed to secrete, from before back- wards as fast as they are formed, nerve fibrils or axis cylinders outside of themselves, and each linear row secretes one axis cylinder.”” In the case of the motor nerves ‘‘the chains of cells leave the cord in a manner often described, and finally de- tailed by Dohrn in more than one publication. The blunted peripheral termination of the chain becomes applied to the muscle plate, and, with great certainty I can repeat what I have more than once stated, that the terminal end-plates of muscle and of the electric organs are formed from the wandering of such célls along with the nerve-forming cells sensu stricto from the anterior horn to the terminal region. These terminal cells must be regarded as ganglionic in character. In connection with these chains of cells the formation of nerve takes place just as described in the above.’’ He further draws attention to the ”) 1 Anatomischer Anzeiger, VII, pp. 290-302. 128 JOURNAL OF CoMPARATIVE NEUROLOGY. views of Vignal who considers the nuclei of Ranvier’s nodes as mesoblastic, but shows that they are concerned in the lengthen- ing of the nerve, and that to this end they give origin to inter- calary segments. Dohrn himself (Anatomischer Anzeiger, VII, pp. 348-351) practically defeated Beard’s view by stating that he is now con- vinced of the zervminal growth of the axis cylinder in the sense of His; and that the cells of Schwann’s sheath were mesoblast- ic; that he had seen fibrils develop independently beyond the ‘chain of cells;’’ which would corroborate the neurone the- ory. Beard, as far as I am aware, has not given up his hetero- dox view as Dohrn did; I give it therefore a place in this sum- mary asa possible, though not probable, objection to the neu- rone theory. Ffeld’s recent finding concerning the concrescence of nerve fiber terminations and cells (instead of simple contact) is a very vital objection to certain hasty conceptions of the neurone theory. The facts observed by him are the following: In the new born dog there is in the trapezoid nucleus a distinct limiting line where the end-brushes of a neurite and the protoplasmic body of the cell meet, such as we find wherever two different substances come into contact with each other. In the dog nine days old, however, this limiting line has disappeared and it is impos- sible to make out a boundary between the end-brushes of the other cell and the cell protoplasm itself. This speaks very strongly in favor of actual concrescence and not mere contact. It is evident that the growth of the cell processes becomes by no means more intricate on ground of this observation. The branches divide and those which find definite connection with other cells become fixed, the others remain free. The review in the Zeitschrift fir Hynotismus says nothing of other anasto- moses [see however our own summary in a later part of other discussions]. Fundamental principles, especially the axiom of the cellular theory, are just as easily understood if we have to admit this observation, as a fact. The fatal blow hits merely the hypothesis of the contractility of the nerve elements which has been exploited for the explanation of sleep and kindred Meyer, Data of Modern Neurology. 129 conditions by Duval, on ground of the considerations of Rabl- Riickhard and Weidersheim, who suggested the contractility of nerve elements by means of which one cell can withdraw from contact with another; or by Ramon y Cajal who attributed to the neuroglia an active role in the production of sleep and rest. In these conditions the neuroglia would separate the cells from one another while, during activity, it would withdraw and make contact possible. For the time being, these are reveries which have been pleasantly or unpleasantly interrupted by Held. They have really nothing to do with the fundamental concepts of the neurone theory and will not occupy us any longer here. We might however, look for danger in another quarter, namely in the recent development of our knowledge of neuro- glia, as presented by Weigert in his Bettrage zur Kenntnis der normalen menschlichen Neurogha, 1895. All the evidence of histogenesis goes to prove that the neuroglia elements are of ectodermal origin, and are derived from the same cells as the nerve cells'. For a long time views identical with the general prin- ciples of the neurone theory have been held with regard to the neuroglia, even before the neurone theory existed, and by men who deny its justification, such as Golgi. They (Fromann, Golgi and others) claimed that the neuroglia consists only of cells and their processes. Weigert admits this only for the embryonic condition. ‘‘In the full grown state the neuroglia consists of cells and moreover of fibers, and the latter prepon- derate in such enormous proportion with regard to the space taken up by them that they are to be considered the more im- portant part of the ‘neuroglia.’ These fibers are by no means processes of cells, but fibers which are perfectly differentiated from the protoplasm.”’ We witness in this a peculiar difference in the develop- ment of the fundamental conceptions concerning two sister structures. The neuroglia having conquered a position in the 1See Alfred Schaper: Die friihesten Differenzirungs-vorginge im Central nervensystems. Archiv fiir Entwicklungsmechanik der Organismen V, pp. S1-232. 130 JourNAL oF CoMPARATIVE NEUROLOGY. cellular hypothesis not only regarding its origin but its persist- ent existence, is described as being cellular in its origin but consisting largely of fibers independent of cells in its later stages. The neurones, or nervous substance in the narrower sense of the word, used to be looked upon as a mass of fibers, a few (the peripheral motor fibers at least) in distinct connection with the cells from which they grow, others originating from the network of the spongy substance and only indicating connec- tion with cells and masses of cells. Now we wish to establish for the nerve elements just that view which was held for the neuroglia, and which was dethroned by Weigert. We tried this notwithstanding certain difficulties concerning the monocel- lular character of the peripheral nerve-fibers. The peripheral nerve-fibers consist after all of more than one cell, unless we have it strictly understood that the cell-unit consists only of the nerve cell-body, dendrites and axis-cylinder process, and that the sheath and its nuclei are an additional coat, not really be- longing to the cell-unit, but formed by epiphytes. If we fol- low Vignal and Ranvier, we assume that the myelin sheath and the nuclei of the inter-nodes have nothing to do with the neurone itself, but are mesoblastic epiphytes. If, however, we follow other observers who consider the myelin sheath a pro- duct of the axis-cylinder, the difficulty with the ‘ epiphytes’ would be shifted; especially the effect of secondary degen- eration on the myelin would be free of serious contradic- tion with the neurone-concept as a ‘one-cell-concept.’ How- ever this be, we should make the mental reservation, when speaking of a motor cell-unit or neurone, that we do not in- clude the nuclei of the internodes, i. e. do not speak of all that is usually included in the description of a ‘fiber;’ further, that there still is some uncertainty as to whether the myelin be- longs to the neurone or the epiphytes. There are two points to be mentioned that will relieve our fears of the analogy with the fate of the neuroglia. The first one lies in the nature of Weigert’s arguments. According to him the neuroglia is a real intercellular substance, i. e. ‘non- nervous material belonging to the group of modified cell-sub- Meyer, Data of Modein Neurology. 131 stances which are emancipated from the cell bodies and which no longer can be considered to be immediately connected with the cell.’ ‘‘1. Because with Weigert’s new stain everything nervous and even the protoplasm of the neuroglia cells, remains unstained, the fibers of the neuroglia however are stained dark blue (conclusion per exclusionem). 2. Because the fibers contain a modified substance which is no longer protoplasmic, but emancipated from the cell body. 3. Because the fibers (and the cells belonging to them) react under pathological conditions just as connective tissue, i. e., they proliferate when the specific nervous tissue perishes ” (pp. 115-117). The clause in the third reason seems significant. It relates to the fact that the cells too proliferate [and must proliferate in order to produce fibrills]. Further the reason for the complete emancipation of the fibrils is decidedly not absolutely convinc- ing. The writer could never resist a certain comparison of the results of Weigert’s neuroglia stain with the results of his stain’ for medullated fibers. Only those parts of the neurone retain the hematoxylin which have enough myelin and kindred sub- stances, namely, in a correct stain, only the medullated part of the fiber. Yet we have reason to consider the myelin sheath a part of the neurone notwithstanding its peculiarity of chemical constitution. The neuroglia stain does not give a complete stain of the neuroglia either, but only of the parts which con- tain a definite substance. The difference lies in the greater number of the fibrils, the lesser degree of organization, and the difference in the distribution of this kind of specially stainable substance which is by no means of a known constitution as in the case of the myelin. Weigert advances ‘‘ with the greatest safety” the following theses (p. 105): I. The neuroglia fibers which, so far, have been taken for processes of the Deiters’ cells, are not structures identical with the protoplasm, but an absolutely different substance. 2. The chemical difference does not appear slowly at a more or less long distance from the cell-body in the ‘‘ process- 132 JournaL oF ComPparATIVE NEUROLOGY. es,’ but the differentiation exists from the beginning, in the immediate neighborhood of the nucleus itself. 3. Most of the so-called processes of the cells are no ‘* processes,” because two of them form one thread passing by the cell without in any way being interrupted by the cell body, as would be the case with ‘‘ processes” taking their origin in the cell bodies. We are not dealing with processes of cells but with fibers which are completely differentiated from the protoplasm. They may have been processes in the embryonic period only. This somewhat dogmatic view does not appear absolutely convincing, since the ‘‘loops”’ are not within the contour of the cell-body but form the outline, and many of them are outside the | outline and probably belong to other cells. Weigert’s view would seem to be the only possible one if in cells like those of his Fig. 1, Plate I, points of cross sections of fibrils could be seen inside of the outline, such as I have never been able to find either in sections kindly presented to me by Prof. Weigert, or in my own preparations made with his method. We are not absolutely convinced of the obsoleteness of the cell-idea in the neuroglia, even of the adult; and if we should become convinced of it by more strengthing evidence,’ we might console ourselves with the ‘ intercellular’ connective tissue nature of this inferior substance. Yet, that which was thus anticipated has since been real- ized by S. Apathy,’ who sees the unit of conduction in fibrils (as Gowers does in his Dynamics of Life) passing through sev- eral cells; between fibrils among themselves and also between cells he sees anastomoses—the completest revolution of the 1F, Reinke (iiber die Neuroglia in der weissen Substanz des Riickenmarkes vom erwachsenen Menschen. Arch. f. mikr. Anat, Vol. L, 1897) corroborates Weigert, saying that he has seen the true protoplasmic processes of neuroglia cells, but also the absolutely different and zzdefendent fibrils of Weigert. If this is true, we should of course have to bow to the evidence given, 2Das leitende Element des Nervensystems und seine topographischen Bezi- chungen zu den zellar. Erste Mittheilung. Mitth, ausder zoolog. Station zu Neapel, 1897. Meyer, Data of Modern Neurology. 533 views of the day, but only ‘demonstrated’ fully in invertebra- tes, evidence only being ‘promised’ for the vertebrates. The demonstration of the preparations in Wood’s Hole has con- vinced most men of the correctness of his claim as regards the existence of fibrils, but not quite that of the claim that these fi- brils pass from one ‘neurone’ into another. If we consider further that Lugaro and especially Becker and Bethe have bet- ter evidence than ever before of the existence of fibrils in human nerve-cells, we must admit that the problem of the nerve-unit is a greater puzzle now than two years ago when the dogma of the ‘neurone’ was almost looked upon as a finality. This short historical sketch must suffice for the present in- troduction and cannot help leaving the impression that the dog- matic inclinations have played a certain trick on those who believe the definition of the ‘neurone’ on the first page to be a perfect soother of all suspicion and skepticism concerning the units. We have certain embryological facts which we owe to His; we have some experience concerning the life of the ‘neurone’ under the influence of injuries, the explanation of which we owe to Forel; we have the charming schematic pictures of the Golgi preparations in the hands of Golgi, Cajal and others: much evidence goes in favor of the monocellular character of the ‘neurone’, so that we may justly call the neurone-theory the cell-theory, although even ina simple portion such as the peri- pheral fiber we stand before a puzzling symbiosis of many cells. Its formation in the period of development has been submitted to a fruitful study by Wlassak, but for an understanding of all the conditions difficulties increase in the fully developed state, and the clean lines of the individual cells become less plain. We come across uncertainties along the lines of the symbiosis noticed in the medullated fibres, and, concerning the cells, doubt is now thrown on the real value of the Golgi pictures which are not capable of producing all the fibrils discovered lately and which therefore would not show anastomoses of fi- brils, even if they existed. Before wading into the deep water of details, we return to some important data of His and Forel and others, and try to 134 JOURNAL OF COMPARATIVE NEUROLOGY. get an orientation of the heavy lines of architecture of the ap- paratus under study, making use of the solid data, but not expressing any opinion yet on the details of interrelation between the ‘neurones.’ We assume as a working hypothesis that the ‘neurones’ are units such as His and Forel and also Cajal describe them, but avoid any assumptions which would necessarily collide with any of the difficulties just reviewed. Outline of an Architecture of the Nervous System. In order not to move in abstract realms, we give in the following pages a short outline of the general architecture of the nervous system. We shall then be able to refer to con- crete conditions in such a manner as to avoid misunderstanding. We must necessarily take some position in the general method from which to approach neurology, and we choose the one of evolution in this sense that we take the phylogenetically oldest mechanisms as the starting point instead of proceeding from the cortex through the ‘projection-systems’ after Meynert’s fashion. We are inclined to start in a consideration of the nervous system from an assumed unit, the brain, and to look upon the peripheral nerves as its afferent and efferent wires. This meth- od has great disadvantages. It starts out with what is least known and most complicated and creates an ego-centric view of the human mechanism which stands in the way of an under- standing of many of the most useful facts acquired by neurolo- gy. In building up the following sketch, we begin at the foun- dation, and proceed towards the most differentiated mechanisms after having established the ground upon which to build them. We start from a sketch of the nervous system of a worm. The rain-worm is a distinctly segmented animal, bilaterally symmetrical, as the vertebrates. Its nervous system consists of a head-ganglion above the oral opening, a strand on each side of the oesophagus extending from it to the first ganglion of the ventral ganglion-chains and forming what the Germans call the ‘Schlundring ’, and the ventral ganglion-chain formed of a ven- Meyer, Data of Modern Neurology. 135 tral ganglion for each inner segment and longitudinal connecting strands between the ganglia. The structure of these ganglia is illustrated by the following elements: Specialized ‘sensory’ cells among the epithelia of the skin send fiber processes into the ganglion where they dissolve in an arborization, coming into contact with the branches of the cells which are connected by a process with the muscles located under the epidermis ; and further cells, the processes of which merely connect various parts of one ganglion or of several gang- lia together. This gives us the following three types of elements: 1. Afferent elements, specialized epithelia, which send a fiber-process into a ganglion where it ramifies into branches ending in contact with many cells; one or more of the branches may even join the longitudinal strand and terminate near the cells of neighboring ganglion. The fundamental point is that one spot in the sensory surface (skin) becomes con- nected with mazy cells." 2. Shunt cells or intermediate elements, cells which mere- ly connect various parts of one or more ganglia. Their pro- cesses do not leave the ‘central nervous system.’ 3. The motor nerve-elements, called ‘motor’ because they are in definite connection with the muscles. The cell body forms part of the ganglion, its fiber a part of the ‘peripheral nerve’, and the termination corresponds to the muscular end- plate. 1The afferent elements are usually called sensory; this term is however greatly misleading. If sensory is to mean ‘the bearer of sensation,’ it is wrong ; for the sensation lies not in these elements, but in a mechanism or combination of many cells. If the cord is severed in a vertebrate, the afferent fibers ‘below’ the lesion remain afferent, as the presence of reflexes shows; but they are in no manner sensory, bearers of sensation. It is the custom of carrying incompletely digested or obsolete psychological terms into physiology which leads us to this laxity of terminology. In the future study we shall rather avoid the word sen- sory aS an anatomical attribute and reserve it to psychophysical processes except perhaps where stilistic reasons seem to demand leniency in the choice of synonyms. The term ‘afferent’ isasarule more correct and preferable, because it says just what is meant and suggests no false psychical inferences. 136 JOURNAL OF COMPARATIVE NEUROLOGY. Passing over to the vertebrates, we start from a stage of de- velopment of the chick such as represented in the figures 10-13 in Dejerine’s anatomy taken from Duval’s atlas. The dorsal lamina of the embryo shows a longitudinal groove which tends to close itself. In Dejerine’s Fig. 12 we see the neural tube almost closed. Along the dorsal suture of the tube special clusters of cells are noticeable on either side which later are known as ‘sensory’ ganglia. 7s called the formation ‘neural ridge’ and we thus start with the ‘neural tube’ and the ‘neural ridges’ on either side of the tube. Comparative neurology shows that the elements of the neural ridge take the place, as we have seen, of specialized epithelia such as are found in the rain-worm’s skin. These special sensory cells of the epidermis send a process into one of the ganglia of the ganglionic chain. (See Fig. 2.) In other worms the ‘sensory’ nerve-cell has its cell-body deneath the epidermis, one process terminating in the skin and the other in a ganglion of the ganglion-chain. This is on the whole the type of the greatest number of afferent nerve- elements of the vertebrates, the only exceptions being the ol- factory and the optic apparatus, of which the former follows the type of the afferent elements in Lumbricus, the cell-body of the olfactory nerve fibre being among the epithelial cells of the Schneiderian membrane. Only few afferent elements seem to have their cell-body in the wall of the neural tube, as for in- stance the mid-brain root of the fifth nerve. The Amphioxus stands quite alone in having no neural ridges; all the afferent elements (spinal ganglion elements) have their cells in the wall of the neural tube. The vertebrate body is to a certain degree segmented. This is clearly shown in the general aspect of the nervous sys- tem. We can divide the body by transverse sections into laminz which show a certain harmony of architecture within the region of the vertebra, and, for our purposes, especially by the peripheral nerve-roots which come forth through the inter- vertebral foramina. The constant repetition of the type: vertebra-nerve-root-vertebra with the corresponding piece of the neural tube ‘belonging to’ the nerve-root constitutes the justifi- Meyer, Data of Modern Neurology. 137 cation for the term ‘segment’. In the cranial region the prin- ciples for a plan of segmentation are more varied; the origin of the cranial nerves is more complex than that of the spinal ones; the segmentation of the skeleton is indistinct (we only remind of the controversy on the vertebral theory of the skull since Goethe’s attemps of demonstrating a fusion of vertebre in the skull), and the complication of the neural tube is greater than in the spinal segments owing to the complex sensory-mo- tor mechanism of the head and owing to the centralization of certain general mechanisms which help to form the ‘brain’. We should however deprive ourselves of many useful analogies if we should give up the segmental method in the cranial part of the neural tube on account of these difficulties. From an architec- tural point of view we do better to give up the term ‘brain’ which means the entire intracranial nerve-mass and to dissolve it into ‘cranial segments’ and supersegmental parts’. In this way we obtain for the entire nervous system the following plan of elements : I. Segmental neurones—the sensory and the motor nerve-elements belonging to a segment (the ‘peripheral nerve’ neurones and their ‘nuclei’ in the neural stem). 2. The intersegmental neurones—nerve-elements which merely connect various segments among one another ( forming largely the ground-bundles and the formatio reticularis). 3. The supersegmental neurones, constituting the cere- bellar, midbrain and forebrain mechanism with their afferent and efferent connections with the segments." Tt is to be regretted that the term ‘segment’ has been used figuratively for parts which cannot thus be cut out. Gowers, for instance, speaks of a cerebro- spinal and a spino-muscular segment of the motor path. In order to avoid con- fusion, we shall, in the following, reserve the term ‘segment’ for the purpose of morphological divisions as described above. We do not, however, imply by this an accurate segmentation in the sense of the metamerism of embryology, but merely a functional topographical and ‘practical’ division. In principle the division is alike; but one of my segments may include several metameres. P, Argutinsky has shown (Arch. f. mikr, Anat., Vol. 48, 1896) that a real gan- glioform segmentation of the motor cells and of the cells of Clarke’s column as claimed by certain writers does not exist, but that certain intermediate cells 138 JOURNAL OF COMPARATIVE NEUROLOGY. A glance at the neural tube of the embryo, represented in plate II of His, zur Geschichte des Gehirns, and the figures 6-11 of the same work shows us that the cranial part is natural- ly divided into three enlargements, the hind brain (the part connected with the spinal cord, later called medulla and pons with the cerebellum covering the ‘fourth ventricle’), the mzd- brain (around the aquaeduct of Sylvius) and the forebrain (around the third ventricle and its projections). Plate II of His gives a good idea of the distribution of the morphological seg- ments of the entire zeuval stem (brain-stem and spinal cord). For an easily comprehensible description of the details of internal structure of this ‘neural tube’ composed of spinal cord plus brain, we should give a more fully illustrated résumé of the contributions of His. For the present purpose I limit myself, however, to the following sketch of a plan of function and ar- chitecture of the nervous system which we shall use as a work- ing basis for the study of the neurones. Moreover we refer to the summary from His on p. 121. The figures 3, 4 and 5 of our plates form our starting point, the former from an early stage of human embryonic life ( His ), giving an idea of the evolution of the various cell-types, the latter from an embryo of Pristiurus (v. Lenhossck ) illustrat- ing the sensory-motor mechanism of one segment. In the cross-section of the foetal oblongata (Fig. 4) the cells are found developing into various types; a certain number remain a simple endothelial lining of the neural canal-ependyma— other cells of a similar type are scattered throughout as cells of the frame-work or neuroglia; others become more and more highly developed, and form the various types of nerve cells proper. (Waldeyer’s Mittelzellen) and, to aless extent, the lateral horn cells in the thoracic cord, are arranged in bead-like accumulations but without a relation to the root-segments. An explanation is not offered. The readers of this essay will Jind that the terms ‘segmental’ as used here ts usually synonymons with the term ‘peripheral’ as it ts used generally; but tt means spinal peripheral, including the whole of the peripheral neurones and their connections within the neural tube but with the exclusion of the cerebral and cerebellar mechanisms, Meyer, Data of Modern Neurology. 139 The median line on the ventral side is usually called raphe ; the portion next to it, the ventral or basal lamina of the tube, develops the motor neurones, the cells which send their fibers into the muscles, aud cells of an intersegmental character, the processes of which do not leave the neural tube but grow ina longitudinal course into other segments ( ground-bundle-cells ). The dorso-lateral part of the tube-wall (also called wing-plate ) receives the central termination of the afferent neurones, the cells of which are located in the ganglia outside. This dorso- lateral part, or posterior horn, contains cells which belong to the order of intermediate cells or shunt-cells. We shall see later on, how these intermediate cells become more specialized. Looking at an entire row of segments of the neural tube, we find the following general arrangement: a small point of the skin is connected by an afferent nerve-element with the corres- ponding segment of the neural tube (spinal cord or brain-stem). One process of the cell reaches the skin; the other process grows as a fiber of the posterior root into the dorsal part of the neural tube. Directly after entering, it divides into a branch which runs towards the head and one which runs towards the caudal segments. Each branch gives off collaterals which terminate in various parts of each segment: some of them in the dorso-lateral plate, ending among shunt-cells, others in the ventral or motor plate, among the motor neurones. As in the worm, we see one afferent neurone reaching many motor cells and many shunt-cells. This is of great importance as is readily seen from this consideration: Each motor neurone is con- nected with certain definite miuscle-fibrils on which it ends as end-plate. If these muscle-fibrils belong to a flexor muscle, the neurone might be called flexor-neurone, if the muscle is an extensor or rotator or abductor or adductor, the neurones be- longing to each respectively are extensor, rotator, abductor, or adductor neurones. Now it is very probable that a sensory neurone supplying the volar side of a thumb gets into contact with sets of motor neurones connected with the various groups of muscles of the thumb. You might suppose that, if this were really the case, a stimulation of any part would call forth 140 Journat or Comparative NEuROLOGY. a contraction of all of these muscles. This is indeed more or less true in abnormal conditions as I have seen in a patient, who went into a diffuse spasm of all the muscles as soon as he was startled by atouch. In the normal however we find that certain forms of stimuli call forth certain movements. You touch the thumb with a feather, the natural result will be that the thumb moves towards the index finger to press the ob- ject between the two fingers. This means that a certain quality of stimulation throws the sensory neurone into such a state of activity as will appeal to, and arouse, the motor neurones connected with the muscles which bring the thumb and index together. If however a needle or another cutting or pricking object is held against the thumb, the same sen- sory cells are put into a qualitatively different state of activity, to which the motor neurones of the above muscles have un- learned to react, but which arouses the antagonists, those which draw the finger from the object. The fact that so many motor cells are directly or indirectly connected with each sensory neurone, makes sucha great variety of movements possible after different kinds of irritation of one and the same sensory neurone. In reality, far more complicated movements are pos- sible. For the great variety of combinations of the muscles of even one segment, the help of the intermediate cells becomes essential; for a sufficient working together of all the segments in the body this is even more evident. In order to give an idea of the complication of all the ne- cessary mechanisms of the whole organism needed for a satis- factory cooperation of all the muscles we pass in a hasty review the principal segments of the vertebrate. They are not all of the same dignity and importance. The segments of the neural tube supplying the tail are necessarily built differently from those supplying the extremities or the trunk or the head. Morphologically there is a striking harmony among the seg- ments behind the skull, as far as the vertebral column extends. The function of these segments is relatively uniform, represent- ing the locomotion, the movements of the trunk, and the extremities. But in the head, greater diversity prevails. As Meyer, Data of Modern Neurology. 14i we have seen, there is a primary morphological division into three ‘vesicles’ (the division into five vesicles had better be abandoned, since it is partly artificial and because it cannot be carried out with advantage). We recognize the rhombenceph- alon or hind brain, the midbrain and the forebrain (thalamus and hemispheres), From a physiological and architectural point of view, we recognize in this ‘brain’ elements of segmen- tal connections, and further the special supersegmental mechan- isms, the cerebellar, midbrain, and forebrain apparatus. In the human brain, we can conveniently outline the fol- lowing cranial segments : 1. Those of the mechanisms of respiration, of articula- tion and of deglutition. The hypoglossal nerve supplies the muscles of the tongue; the pneumogastric, the viscera of the neck, thorax and abdomen, and the glossopharyngeal makes the connections for the reactions to stimulation of taste. These mechanisms are located in the body segments belonging to the lower part of the medulla oblongata or hind-brain. 2. The auditory-facial-abducens segment. Here we find on the ‘sensory’ side the auditory nerve in connection with the cochlea of the inner ear, and the equilibration (?) nerve, in connection with the semicircular canals, the sense organ for cer- tain auditory qualities and ‘appreciation of position in the space.’ We know that destruction of the two produces deafness and dizziness and inability of equilibration. The motor side of this segment is represented by the facial nerve which moves the skin and muscles of the face, and, especially in animals, the exter- nal ear; and the abducens nerve, which moves the eyeball out- ward. It is easy to remember the function of this segment in in this way; you hold a watch near the ear of a dog and he will prick up his ear and turn the eye to the side; the addition- al movement of the head depends on an association mechanism with other segments. 3. A little further forward we come to the segment of mastication. There we find the motor neurones for the muscles moving the jaw, and on the sensory side the large Gasserian ganglion which supplies most of the head with sensory fibers, 142 JouRNAL OF COMPARATIVE NEUROLOGY. not only the mouth but also the other parts of the face and the mucous membranes of the head. Just this afferent nerve of the face teaches us a good lesson for the general arrangement of the sensory-motor mechanisms. We find that it spreads over the neurones for the movement of the jaws, the movement of the facial muscles, the tongue and even into the segments of the neck (by the ‘ascending’ root). Thus a prick of the cheek can be responded to directly by a movement of the jaw, of the muscles of the face, and of the entire head by the muscles of the neck. These segments constitute the hind-brain, an important ac- cessory organ of which we shall recognize presently in the cer- ebellum. 4. The mid-brain enlargement contains as a segmental mechanism the optic nerve and the nerves for the remaining muscles of the eyeball: the optic segment. 5. The forebrain contains only an afferent nerve-apparatus, the olfactory which is only in indirect communication with the various motor neurones. This is a summary of the neuro-muscular segments which constitute the human body. How are they connected? A study of the lowest vertebrate, the Amphioxus, may show how the various segments of the body can be very simply united for conjoint action. There are of course first the ground bundle-elements or intersegmental neurones. In the head end of the neural tube there are a few very large cells sending big fibers towards the caudal segments ; and in the caudal end there are also a few large cells making connection with the segments of the head end. The cells are sufficient, together with the intermediate cells between the neighboring segments, to repre- sent the coordination of movement of the Amphioxus. Such a simple mechanism would not suffice for the higher vertebrates. The number of such long connecting cells would be immense, considering the variety of complex movements of which we are capable. On the plan of the Amphioxus our nervous system would consist of the anterior horn cells and their fibers to the muscles, of the afferent ganglion cells and their fibers into the a Meyer, Data of Modern Neurology. 143 skin and into the posterior horns. Each segment is connected with its neighbors by ground bundles. The immense multi- plicity of sensory-motor combinations would require an im- mense number of long fibers running back and forth and the nervous system would have the form of a fairly uniform §thick- walled tube. The function would be slow and complicated. In reality the results of coordination are obtained much more easily by centralization of the mechanisms which represent spe- cial functions. We have seen that the auditory-facial-abducens segment contains an apparatus for equilibration. The sense-organ is formed by the semicircular canals of the internal ear, destruc- tion of which leads to disorders of equilibration. Over this segment the mechanisms relating to the appreciation of coordi- nation and equilibration of all the segments are united in a spe- cial structure, the cerebellum. It varies in size and complexity in the vertebrate series and is most developed in the fish, the birds and in man. Mechanisms which are scattered all over in the Amphioxus are thus centralized and can become more elaborate by short association elements within the special organ. Further we find the optic segment (mid-brain) with a pecu- liar system of connections with the afferent elements of the rest of the body. In many animals which depend very largely on vision, as the trout, certain reptiles and especially birds, the mid-brain is very large and ‘sensory’ paths are connecting all the rest of the neural tube with it, so that all the afferent im- pulses can be elaborated into one harmonious entity. Between the olfactory segment and the optic segment a further highly complicated mechanism developes, especially from the reptiles up, the thalamus and the fore-brain, most highly organized in man, and the organ of the highest reactions of which a living being is capable, among others the mental ac- tivity, thought, and reasoning. Without entering on the detailed structure of the cerebel- lar, mid-brain and fore-brain mechanisms, I merely illustrate the principle of the cerebellum and the fore-brain. Each segment of the neural tube has certain connections 144 JOURNAL OF COMPARATIVE NEUROLOGY. with the cerebellum. Certain of its (intermediate) cells send their fibers into the cerebellum. These cells are most numer- ous where the nerve specially concerned in equilibration comes in; they are next in number in the segments of the lower part part of the back (columns of Clarke) and in the external nu- cleus of Burdach. All the fibers end near the surface of the cerebellum, where the coordination elements of all the segments are brought near one another. The cerebellum then contains cells which influence directly or indirectly the motor elements of the various segments, and establish the necessary coordina- tion. The neurones which have their cell-body in the segments and their fiber arborization in the cerebellum, are called afferent cerebellar neurones ; the ones which have the cell-body in the cerebellum and the fiber arborization in the segments, efferent cerebellar neurones. The supersegmental part of the mid-brain, on the main the corpora quadrigemina, has a great number of afferent neu- vones; in lower animals almost every segment sends fibers to meet the optic apparatus in the mid-brain; but in man the afferent mid-brain neurones are limited largely to the auditory segment (lateral fillet) and few fibers of Gowers’ bundle (Mott). Efferent neurones of the mid-brain are not known with certainty (see however Bechterew’). The most important extrasegmental mechanism is however that which grows up between the olfactory and the optic seg- ments. We know it as cerebral cortex and basal ganglia, or cerebral mechanisms. Here too we find afferent neurones from each segment of the neural stem. These neurones are how- ever already ‘centralized.’ It is commonly known that the affer- ent cerebral fibers for the spinal segments have their cells grouped together in the Nuclei of Goll and Burdach, at the point where the head segments go over into those of the neck. The peripheral sensory cells of the more caudal segments send their fibers all the way to meet them; they form the posterior 1'W. Bechterew, Ueber centrifugale aus der Seh- und Verhiigelgegend aus- gehenden Riickenmarksbahnen. Neur. Centralbl. No. 23, 1897. Meyer, Data of Modern Neurology. 145 columns of the spinal cord, the fibers from the segments of the lower extremity next to the median line, those from the brach- ial segment to their side, and finally those of the head yet fur- ther to the side as is seen from the drawings of the cord and medulla oblongata. After an interruption and sifting in the optic thalamus, secondary afferent elements meet in the cortex the cerebral effer- ent neurones which form the pyramidal tract, the so called voluntary motor path, and their equivalents within the optic, auditory, etc., region. The forebrain is an exceedingly com- plicated mechanism ; as one would expect from the tremendous complication of all the conscious activity of which we are capa- ble, its differenciation apparatus is very elaborate. There remains to be mentioned, that the higher mechan- isms, cerebellum, midbrain and forebrain have connections among one another, not drawn in the chart (Fig. 6) in order not to complicate the drawing. Further we must say that this pre- liminary sketch of three supersegmental apparatus will require subdivisions and perhaps even additions in number. They are units only in the most general way, and given here as the types now most important. This outline will, I hope, make clear the general point of view. It can perhaps be more forcibly Illustrated in the follow- ing manner: We saw that Forel formulates the laws of v. Gudden’s de- generation experiments as follows: if a cell body is removed, the fiber belonging to it will degenerate, if a fiber termination in the central nervous system is cut, where regeneration is im- possible the whole cell will atrophy slowly, and, in the case of the new-born at least, degenerate and disappear. In accordance with the fundamental laws of the pathology of the central nervous system, we would therefore formulate our general point of view in this manner : The phylogenetically oldest mechanisms are the sensory- motor apparatus constituting the purely segmental nervous sys- tem as defined above. Over them, lifted out from them for topographical centralization, there are specialized mechanisms, 146 JoURNAL OF COMPARATIVE NEUROLOGY. the cerebellum, ‘midbrain’ and ‘forebrain’ supersegmental mechanisms. v. Monakow has been the first to speak of the forebrain and its dependendent elements (Grosshirnanteile). When he cut a part of a forebrain, he killed not only the cells which he removed, and the fibers growing from them, but also cells located in other parts which send their fibers into the piece cut out. These latter elements he calls Grosshirnanteile with the same right as for instance the pyramidal tract; they are the af- ferent elements of the part, while the pyramidal tract is the efferent. I would generalize this principle and search for the ‘Cerebellaranteile, ’ afferent and efferent elements of the cere- bellum etc. For purely anatomical purposes it is indeed the most stimulating principle. We speak of the corpus albicans and its ‘Anteile’ etc. (A thoughful reader will see between the above lines a definite concept of the ‘meaning’ (i. e. interpretation) of the interrelation of cell-elements, the discussion of which does not properly belong here. In the neurological cant, we are accus- tomed to speak of connections of neurones for the purpose of association. According to the above, we rather think of inter- relations of neurones for the purpose of dissociation and read- justment. A neurone reaches with its processes many individ- uals of many types of neurones and the interrelation with these takes place in order to make possible the reflection of different reactions in response to different states of excitation of the neurone. This assumption becomes plain when we try to explain the ‘interruption of the fiber-tracts by gray nuclei,’ as when we speak of the ‘interruption of the ‘sensory’ path by the nuclei of Goll and Burdach.’ These ‘nuclei’ contain cerebral and cerebellar afferent neurones. The cerebellar neurones pick out special elements of excitation; the cerebral afferent neurones pick out other elements of excitation, as it were, selecting out those which belong together; otherwise, there would be no ‘need’ for an interruption. The same holds for the thalamic nuclei; in fact, for all accumulations of cells, after the para- digma given on page 141. From a physiological point of view, Meyer, Data of Modern Newology. 147 the differentiation, or dissociation, becomes more prominent ; the association is naturally also implied by the anatomical ar- rangement. ) The practical and didactic value of the above plan of the nervous system is quite evident when we take as an instance a section of the nervous system. The first question after a gen- eral orientation as to the presence of a central canal and the distribution of gray and white matter is this: Are there any segmental elements—motor neurones and afferent neurones ? Which segment do they belong to? Cord, medulla, midbrain, or which part of the neural tube? Are there any cells and fibers of the ground bundle (intersegmental) formation ? Cere- bellar dependent parts? Midbrain dependent parts? Cerebral dependent parts? Any non-classified elements ? [To be Continued.] DESCRIPTION OF FIGURES. Plate XV. fig. z. A. Interrelation of afferent and efferent cells, according to Nansen (and Golgi). B&B. Interrelation of afferent and efferent neurones in the cortex, according to the idea of van Gehuchten and Cajal. From Ramon y Cajal, Nuevo con- cepto, etc., 1893. Plate XVI. Fig. 2. Diagram of the nervous system of Lumbricus. From Schaefer, Brain, XVI, p. 154. Plate XVII, Fig. 3. Human embryo of 10 mm. length (see the cross-section of rhomben- cephalon). Development of the ‘* segmental nervous system” perfectly plain. Supra-segmental mechanisms barely indicated as Anlage of the cerebrum and cerebellum. Afferent neurones blue, efferent neurones red. His, Geschichte des Gehirns, Plate II. Plate XVIII, Fig. 4. Cross-section of the rhombencephalon of a human embryo lomm, long, x 40. Motor pneumogastric and hypoglossal nuclei with ‘‘ ascending ” 148 JoURNAL OF COMPARATIVE NEUROLOGY. afferent pneumogastric root. To the left, isolated presentatlon of n. X and XII. Neither the olives and cerebellum nor the fillet and pyramids developed. Stage when only the segmental apparatus is plain. Roof of fourth ventricle (hind- brain cavity) membranous. Lateral lamina receiving the ‘‘sensory ” (afferent) fibers of the nerve, and containing intermediate cells, cerebral afferent cells and especially the Anlage for the cerebellar apparatus (olives and cerebellum). Basal lamina with nucleus and nerve and numerous intermediate elements for the formation of the substanta reticularis. Adapted from W. His, Zur Geschichte des Gehirns, Fig. 21, p. 360. Fig. 5. Central part of a spinal segment of Pristiurus. From Lenhossék. Afferent elements blue, efferent elements brick red. Intermediate elements omitted. Lateral lamina with the afferent terminations; basal lamina with the segmental motor neurones to the side of the raphe. Plate XIX. Fig. 6. General plan of the nervous system. Afferent segmental system blue, efferent (motor) segmental system brick red; cerebral mechanisms, afferent green, efferent carmine; cerebellar mechanisms, afferent yellow, efferent not shown, A REPORT OF THE NEUROLOGICAL SEMINAR OF THE MARINE BIOLOGICAL LABORATORY, WOOD’S HOLL, MASS. SEASON OF 1868. The Neurological Seminar was organized during the sum- mer of 1896, with the object of bringing together in an infor- mal manner those engaged in the investigation of the Morphol- ogy, Physiology or Pathology of the Nervous System. Meet- ings one hour in length, were held twice each week. Reports were presented embodying the results of personal research or the critical review of the literature of the subject under investiga- tion. Demonstrations and drawings were used to illustrate the points presented and prepare the way for discussion. During the first season the attendance was restricted to the active members, of whom there were twenty. The second season, 1897, there was no increase in the number of members; about half the entire number had participated in the meetings of the previous season. The subjects discussed can be grouped under a few main heads. The investigation of the lateral line of vertebrates, its innervation and the relation of its sense or- gans to the organs of special sense has held the first place. Only second to it was the question of metamerism in the ver- tebrate head and in the nervous system of annelids. The sub- ject of equilibration has received much attention from the physiologists. The present season the meetings have been made public and the neurologists at the Biological Laboratory of the United States Fish Commission were invited to take active part. The membership has advanced to thirty six, although only twenty nine presented reports, the others being prevented from doing so by various causes. 150 JOURNAL OF COMPARATIVE NEUROLOGY. Through the kindness of Dr’s. Parker and Montgomery the Seminar enjoyed tne privilege of studying Professor Apa- thy’s slides, illustrating his paper on the finer structure of the nervous elements, (Mitt. Zool. Sta. Neapel, Bd. 12, Heft 4.) The great scientific value of the slides and importance of the facts demonstrated were fully appreciated by both the neurol- ogists and investigators in other departments. An entire after- noon was given to the study and comparison of the slides and a hearty vote of thanks was tendered to Professor Apathy by the Seminar. This action was cordially endorsed by all who examined the slides. A demonstration was also given by Dr. C. F. Hodge of the structural differences between the pyramid cells from the brain of a sleeping puppy and the corresponding cells from the cerebral cortex of a puppy of the same litter which had been fatigued before killing. PROGRAM. July 14. C, JupDson HERRICK, Denison University and Pathological Institute of the New York State Hospitals. The Cranial Nerves of the Bony Fishes. T. W. GALLOWAY, Brownsville, Penna. Some Nervous Changes Accompanying Budding in Dero vaga. July 19. Miss C. M. Crapp, South Hadley, Mass. Review of Allis’ Paper on the Cranial Nerves of Amia. H. V. NEAL, Knox College, Galesburg, ll. The Problem of the Vertebrate Head. July 21. T. H. MontGomery, JR., Philadelphia, Pa. The Elements of the Central Nervous System of the Nemertians. U. DAHLGREN, Princeton, N. /. The Giant Ganglion Cell Apparatus. A. D. MorriL1, Clinton, NW. Y. Innervation of the Olfactory Epithelium. MorriLy, WVeurological Seminar. 151 July 26. Mrs. M. L. NICKERSON, University of Minnesota. Epidermal Organs of Phascolosoma gouldii. Miss ANNA Moore, Poughkeepsie, NV. Y. Review of Papers on the Nervous System of Dinophilus. W. W. NorMAN, Austin, Texas. Bethe on Forced Movements in Arthropods. August 2. Demonstration of Apathy’s Slides. T. H. MonTGoMERY, JR., Philadelphia, Pa. Review ot Apathy’s Paper on Primitive Fibrils. G. H. PARKER, Camébridgs, Mass. Influence of Apathy’s Conclusions on the Neuron Theory. August 4. Miss J. A. Haynes, 7voy, NV. Y. Review of Literature on Sympathetic Nervous System of / saree PorTER E. SARGENT, Cambridge, Mass. The Giant Ganglion Cells in the Spinal Cord of Ctenolabrus cceruleus (Storer). W.C. Jones, Evanston, 111. Report, Huber’s Paper on the Sympathetic Nervous System of Vertebrates. CRESSWELL SHEARER, Westmount, Montreal, Canada. On the Nerve Termimations in the Selachian Cornea. August 5. F.C. Watt, Cambridge, Mass. Variations in Lumbro-sacral Plexus of Necturus maculosus. G. W. Hunter, Jr., Hyde Park High-school, Chicago, [ll. Notes on The Peripheral Nervous System of Molgula manhattensis. L. E. GriFFIN, Baltimore, Ma, Tentacular Nervous System of Nautilus. S. R. WiLLiAmMs, Cambridge, Mass. Review of Hamaker’s Paper on the Central Nervous System of Nereis. August 6. C. F. Hopce, Worcester, Mass. Demonstration showing Differences between resting and fatigued pyramidal Cells in the Puppy. 152 JouRNAL OF COMPARATIVE NEUROLOGY. August 9. O. S. STRONG, Columbia University, New York. Review of Johnston’s Paper on the Cranial Nerves of the Sturgeon. C. R. BARDEEN, Johns Hopkins Hospital, Baltimore, Md. On Variations in the Distribution of the Spinal Nerves Entering the Lumbar Plexus. C. W. Hareitt, Syracuse, N. Y. Review of Conant’s Paper on Cubomedusz. E. W. BERGER, Johns Hopkins University, Baltimore, Md. The Histological Structure of the Eyes of Cubomeduse. August II. E. P. Lyon, Peorza, J71. Functions of the Otolith. IrA VAN GIESON, Pathological Institute of the New York State Hospitals. Effects of Starvation on the Nucleus of Ganglion Cells. H. R. Fiinc, State Normal School, Oshkosh, W%s. A Contribution to the Nervous System of the Earthworm. F, L. LANDACRE, Ohio State University, Columbus, O. Demonstration of Preparations of Teleost Brain. E. RyNEARSON, Central High-school, Pittsburgh, Pa. Review of Paper on the Nervous System of Arenicola marina. Among the members of previous seasons are Miss Fanny E. Langdon, Miss M. L. Nichols, Miss M. M. Sturgess, Drs. Hy Ayers; Cl) Bristol,’ G. P. Clark, .C) WW: , Greenolisoma. Gerould, A. Graf, B. F. Kingsbury, W. A. Locy, P. C. Mensch, W. A. Patten, S. Paton, and A. Schaper, Messrs. G. L. Houser, J. B. Johnston, V. E. McCaskill, J. E. Peabody and O. H. Swezey. A. D. Morritt, Chairman. NEAL, Problem of the Vertebrate Head. 153 THE PROBLEM OF THE VERTEBRATE HEAD. By. Hi Wi NEAL. Knox College. Two of the most important morphological conceptions of the nineteenth century are attributed to the poet Goethe—one, that a flower is a modified branch and its organs metamorphosed leaves—the other, that the head and trunk of vertebrated ani- mals were once composed of like segments which by slow adaptive change have become to a considerable degree unlike. After a century of probation no morphologist of today ques- tions the truth of the former conception. The truth of the lat- ter, however, is still debated and the attempt to compare a head segment with a trunk segment in vertebrates constitutes what is now known as the ‘‘head problem.” Since neither head nor trunk can be regarded as primutve in their present condition, probably a more correct statement of the problem would be as follows; Was the vertebrate head like the trunk, primitively segmented; if so, were these seg- ments serially homologous with those of the trunk; and how many have entered into the composition of the head? So far as I am aware, no one doubts that the vertebrate head is seg- mented. That it is so, is indeed clearly evinced by such seri- ally repeated organs as neuromeres or segments of the central nervous system, nerves both dorsal and ventral, somites, vis- ceral clefts, visceral arches and aortic arches. But while the great majority of the morphologists who have expressed an opinion on the question have concluded that Goethe’s conception is true and that head segments are serially homologous with trunk segments, a few have been led during recent years to regard the head, or at least its anterior or pre- otic part, as one saz generis. This conclusion has been reached partly by the recognition of the considerable differences be- tween head and trunk metameres and the organs of which they are composed—differences which seem too great to be merely differences in the degree of specialization and partly also by the 154 JOURNAL OF COMPARATIVE NEUROLOGY. ee —_ ao —_, = ton] = ee eis ee io = E SS e Ww Oo SA oe = © 2 oe. i a eee es + on PRES ge te ae ts ORE eae = ates REE ee: a “3 B52 ec ae a oe a Qe Ss i ai oo v a ee Ww I. hE geen } } | ee | ' ' ’ sypyyosiy Ay iy ay | Sey eo ee byt! Pees ane wel sTuh A Sar My DPS fis fa) fats ap c) C) @ = Awe Qe Nt Woh Wy 2 S) S ’ a < - Cae S| at Trig é , ols TPTPW” eS en eta as estas To ‘ANSA0(TO reas ; : — : = ner Ny soya pan wae pe EAT a np wm DiI Para se nese : U ‘ 1. + ae ay WIA! debe SS eS pene Giz) 1 (eS i ! i) ) ees > Pee ee - | eae es heel (er pet ' + sa Reetae tre rT ht WG sayoay Pov ae ne Caos \ aaSacc Ser BES ; | He ) 1 1 7 = ° 4 ‘4 5 ae ae tS ae : 4 age ees peer : ; oy see ' ! 1 ’ ' oe ae es ees } ee OS feet re : =a cs ort \ % = &- “oO. wo oO XS ex g S S (s Semere 3 G Uw. Neat, Problem of the Vertebrate Head. 155 conflicting evidence and conclusions, both as to composition and number of cephalic segments, of those who have advocated the prevalent morphological opinion. Of the differences stated by them I shall speak later. The confusion in, and, as must be admitted, generally unsatisfactory condition of the literature bearing on the head problem, is in my judgment attributable in great part to the fact that the observations of investigators have been confined often to a single species, often to a single organ system, while their conclusions deduced from such limited ob- servations have applied to the phylogenests of the entire vertebrate head ! That such methods are inadequate for the solution of such a difficult problem seems in view of the many divergent opinions too obvious to need insistence here, and I venture to predict that some time, if not now, it will seem strange that a morphologist should assume, or seek to demonstrate that the serial parts of any single organ system, whether neuromeres, or nerves, or somites or visceral arches, or epibranchial ‘‘ sense organs,” or what not, are the essentzal criteria of head seg- ments. In my opinion, phylogenetic conclustons concerning the metamertsm of the head based upon the study of a-single animal or a single organ system need to be ‘‘ controlled” and confirmed by the study of other organ systems in the same animal. The solution of no problem requires a broader knowledge of compar- ative embryological and anatomical facts. fig. 1. Diagrammatic representation of the cephalic metameres in Selachii, showing the component organ systems and their relations to one another. J-XT, cephalic neuromeres (segments of the central nervous system) ; a, Miss Platt’s ‘‘ anterior” somite; 7-72, van Wijhe’s first to twelfth somites ; 7'-§!, first to eighth visceral clefts ; aéd., abducens ; aol-*, aortic arches, first to eighth; ch., chorda; dors. nv., dorsal nerve; ef, vag., epibranchial portion of vagus nerve; fac., facialis nerve; glossuph. (gts.), glossopharyngeus nerve; 2y., hypophysis; #., mouth ; med. /a¢. Z., mediolateral line; eur, (#.), neuromere ; ocm., oculomotorius ; o/f., olfactorius; ophth. prof. (pf.), ophthalmicus profundus nerve; of., otic capsule (ear) ; posttrem., posttrematic branch ; praetrem, praetre- matic branch; 7. /at. vag., ramus lateralis vagi; 7. zt. vag., ramus intestinalis vagi; som., somites (van Wijhe’s) : sf.'-*, spinal ganglia first to third; ¢voch., trochlearis ; vent. ao., ventral aorta; vent. mv., ventral nerve; wésc. clefts, viscer- al clefts; vag.1-3, vagus ganglia first to third (dorso-lateral series); vsc.* third visceral arch. The arrow marks the posterior termination of the cranium in Squalus. All neuromeres anterior to this point are included in the cranium. 156 JOURNAL OF CoMPARATIVE NEUROLOGY. Holding this view, I have recently’ made an attempt to solve the head problem, and while my observations were made primarily upon the nervous system in Selachian embryos, my theoretical conclusions have been controlled by the study of the actual relations of other organ systems and also by the study of embryos of all other classes of vertebrates except Reptiles. |. Whether or not I have come nearer a solution of the head problem than have many of my predecessors, depends, I am convinced, on whether or not I have adhered with greater fidelity than they to the principle above enunciated. I regard my results as in great part a confirmation of those of van Wijhe (’82) and valuable as such. First, as regards the ature of cephalic metameres, I con- clude with the majority of investigators that they are serially homologous with trunk metameres, although the homology is today but partial. To my mind, the differences which have been considered as objections to this view by certain morpholo- gists, such for example as the fact that (@) visceral elefts and arches are confined to the head region (Gegenbaur) ; that (4) excretory organs are confined to the trunk region ; that (c) there are no somites in the head, at least in its pre-otic portion, (Kastschenko, Rabl, Froriep); that cephalic nerves and spinal nerves cannot be compared by reason of the fact that (d) ceph- alic dorsal nerves receive cellular material from the skin, while spinal dorsal nerves do not; that (e) cephalic dorsal nerves are mixed, while spinal dorsal nerves are sensor in function; that (7) cephalic dorsal nerves extend lateral, and spinal dorsal nerves median, to the somites; that (g )—at least some—ceph- alic dorsal nerves have component sensor fibers which innervate lateral line organs, while in spinal nerves these are wanting ; that (#) in one and the same occipital metamere there can be found (1) a cephalic dorsal nerve, (2) a spinal dorsal nerve, and (3) a spinal ventral nerve and that therefore spinal and cephalic 1 NEAL, H. V., ’98. The Segmentation of the Nervous System in Squalus acanthias—A contribution to the Morphology of the Vertebrate Head. Bull. Mus. Comp. Zoél. Harvard Univ., Vol. 31, No. 7, pp. 145-294, with nine plates. NEAL, Problem of the Vertebrate Head. 157 dorsal nerves cannot be of the same kind; and other less im- portant differences by no means outweigh the evidence of simi- larity of head and trunk segments. As a mater of fact some of the differences alleged above do not actually exist. Many, is is noted, apply to the nerves, and these have seemed so great that even Gegenbaur, the early champion of the present morphological conception of the ver- tebrate head states (87) that he is no longer able to consider cephalic and spinal nerves as homodynamous. With our pres- ent knowledge, however, that in Amphioxus two kinds of nerves, viz. dorsal mixed nerves whose motor fibers innervate splanchnic musculature, and ventral motor nerves which inner- vate somatic musculature, are found in each segment of the body except the first; that in Craniota both of these kinds of nerves appear in the head as well as in the trunk ; that a pair is to be found in each trunk metamere (in Petromyzon unconnect- ed as in Amphioxus), and in some head metameres, I am una- ble to regard the actual differences between cephalic and spinal nerves as fundamental in character. * The differences which appear are, in my judgment, to be expected in the case of the nervous organs in such highly differ- entiated structures as head and trunk. Furthermore, the fact that the bounds of head and trunk in the vertebrate series are not definitely fixed; that they are variable; that there is an un- broken continuity throughout head and trunk of such essential components of metameres as neuromeres, nerves, somites, vis- ceral arches, visceral clefts, and aortic arches, is evidence suffi- cient to warrant the general belief in the serial homology of the segments in these two regions. So far as I can see, no objec- tions to this view apply to the pre-otic region which are not equal- ly applicable to the post-otic region. If the segments in the one region are serially homologous with trunk metameres, those in the other region are also. I shall be obliged to refer the 1 The evidence both histological (Lenhossék, Kélliker, Ramon y Cajal) and physiological (Steinach and Wisner) given in the last decade, seems to establish conclusively the fact (rendered @ frior¢ probable by the evidence from Amphi- oxus) that spinal dorsal nerves are like cephalic dorsal nerves mixed in function. 158 JouRNAL OF COMPARATIVE NEUROLOGY. reader for further grounds for my conclusions concerning the nature of head segments to the more extended paper referred to above. x MI. ‘OR W: /N jn " 2 4 ‘Fy . 5 | Ms tt ae Mt hy aoe ’ TED t i s ¢ " id LI wea ele Yserrhyoids Fig. 2. Diagram of Selachian head, showing the cephalic metameres and their components, lateral aspect, based upon the study of Squalus acanthias, Upon the basis of the results of Kupffer, Miss Platt and others a distinction is made in the representation of dorsal nerves between dorso-lateral and medio- lateral (epibranchial) ganglia. Secondly, as regards uumber and composition of cephalic metameres my conclusions have been summarized in Figs. 1 and 2. There are in vertebrates five pre-otic, one otic and (in Squalus) five post-otic cephalic metameres. The number of post-otic segments whose vertebral components fuse into the occipital region of the cranium of vertebrates is variable. The estimate of the number of pre-otic segments is based chiefly upon the evidences that in this region of Squalus embryos neuromeres and somites zumerically correspond, and are in some cases con- nected by motor nerves. Fora more extended presentation of this evidence I again refer to the longer paper (’98). Briefly sum- marized, the composition of cephalic metameres from the first to the last is as follows: MeEtTAMERE I. euromere, neuromere I (primary forebrain vesicle) ; dorsal nerve, olfactory (motor component lacking—in correlation with the want of splanchnic musculature) ; ventral nerve, absent in correlation with the absence of somatic muscu- NEAL, Problem of the Vertebrate Head. 159 lature ; somite, ‘‘anterior’’ (Miss Platt’s); w7sceral cleft and arch, hypothetical ; aortic arch, hypothetical. MeraMereE IJ. Neuromere, neuromere II (primary mid- brain vesicle) ; dorsal nerve, ophthalmicus profundus (motor fibers absent in Squalus, but present in some vertebrates); ven- tral nerve, oculomotorius; som7te van Wijhe’s Ist; vzsceral arch and cleft, hypothetical ; aortic arch, hypothetical. MeraMeErE III. Neuvomere, neuromere III (Hinterhirn) ; dorsal nerve, trigeminus ; ventral nerve, trochlearis; somite, van Wijhe’s 2nd; wsceral arch, first (mandibular); wsceral cleft, (bounding anteriorly the ventral portion of the segment) usur- ped by mouth ; aortic arch, first (mandibular). METAMERE IV. Neuromere, neuromere IV ; dorsal nerve, hypothetical (absence correlated with the absence of a visceral arch) ; somite, van Wijhe’s 3rd; ventral nerve, abducens; visce- ral cleft and arch, hypothetical ; aortic arch, hypothetical. METAMERE V. Jeuromere, neuromere V; dorsal nerve, facialis (the acusticus a specialized sensor branch); ventral nerve, abducens; somzfe, van Wijhe’s 4th (which together with the 3rd forms in Torpedo the m. rectus posterior, Sewertzoff—rudimen- tary in Squalus); wzsceral cleft, first (hypobranchial, spiraculum); visceral arch, second (hyoid); aortic arch, second (hyoid). MeETAMERE VI. euvomere, neuromere V1; dorsal nerve, glossopharyngeus ; ventral nerve, abducens; somite, van Wijhe’s 5th (myotome absent in Squalus; forms first myotome of the lateral trunk musculature in Petromyzon); wsceral cleft, 2nd visceral (ist branchial) ; wsceral arch, third (ist branchial) ; aortic arch, third. MeraMereE VII. Neuromere, neuromere VII (the last of the neuromeres having a lateral thickening. See Fig. 1); dorsal nerve, vagus’ ; ventral nerve, abducens; somite, van Wijhe’s 6th (myotome rudimentary in Squalus) ; wvesceral cleft, third (second branchial); wsceral arch, fourth ; aortic arch, fourth. METAMERE VIII. Jeuromere, neuromere VIII; dorsal nerve, vagus’; ventral nerve, hypoglossus (anterior root, rudi- 160 JouRNAL OF COMPARATIVE NEUROLOGY. mentary) ; somite, van Wijhe’s 7th (myotome, first myotome of lateral trunk musculature in Squalus) ; vzsceral cleft, fourth; ves- ceral arch, fifth; aortic arch, fifth. METAMERE IX. Neuromere, neuromere 1X; dorsal nerve, rudimentary (unites with vagus in Squalus) ; ventral nerve, hy- poglossus, second root; somite, van Wijhe’s 8th (forms first segment of hypoglossus musculature) ; vesceral cleft, fifth; ves- ceral arch, sixth; aortic arch, sixth. METAMERE X. Leuromere, neuromere X; dorsal nerve, first spinal (represented by a rudimentary ganglion in Squalus embryos) ; ventral nerve, hypoglossus ; visceral cleft, sixth; ves- ceral arch, seventh ; aortic arch, seventh. METAMERE XI. Neuromere, neuromere XI; dorsal nerve, second spinal (rudimentary ganglion in Squalus embryos); ven- tral nerve, hypoglossus ; somzte, van Wijhe’s oth; vesceral cleft, seventh ; visceral arch, eighth; aortic arch, eighth. List OF SOME OF THE MORE IMPORTANT PAPERS BEARING ON THE HEAD PROBLEM. AHLBORN, F. 84. Ueber die Segmentation des Wirbelthierkérpers. Zeitschr. f. Wiss. Zoél., Bd. 40, pp. 309-337. BALFourR, F. M. ’78. A Monograph on the Development of Elasmobranch Fishes. London, XI + 295 pp. 20 plates. BEARD, J. ’85. The System of Branchial Sense Organs and their Associated Ganglia in Ichthyopsida. A Contribution the Ancestral History of Vertebrates. Quart. Jour. Micr. Sci., Vol. 26, pp. 95-156, Plates 8-10, Donen, A, ’75. Der Ursprung der Wirbelthiere und das Princip das Functionswechsels : Genealogische Skizzen, Jeipzig, XV + 87 pp. FRORIEP, A. P ’92. Entwickelungsgeschichte des Kopfes. Anat. Hefte, Abth. 2, Ergeb- nisse Anat. u. Entwg., Bd. 1, pp. 551-605. FUERBRINGER, M, ’97._ Ueber die Spino-occipitalen Nerven der Selachier und Holocephalen Neat, Problem of the Vertebrate Head. 161 und ihre Vergleichende Morphologie. Festschrift zum siebenzigsten Geb. von Carl Gegenbaur, Bd. 3, pp. 349-788, 8 Taf. Leipzig. GEGENBAUR, C. 87. Die Metamerie des Kopfes und die Wirbeltheorie des Kopfskeletes. Morph, Jahrb. Bd. 13, pp. 1-114 HATSCHEK, B. ’92. Tie Metamerie des Amphioxus und des Ammocoetes. Verh, Anat. Gessellsch. VI. (Wien), pp. 136-161, 11 Figs. HoFFMANN, C. K. ’96. Zur entwickelungsgeschichte des Selachierkopfes. Anat. Anz., Bd. 9, pp. 638-653, 5 Figs. Huxtey, T. H. 58. The Croonian Lecture—On the Theory of the Vertebrate Skull. Proc. Roy. Soc. Lond., Vol. 9, No. 33, pp. 381-457, 10 Figs. KUPFFER, C. ’91. Die Entwickelung der Kopfnerven der Vertebraten. Verh. Anat. Gesellsch. V. (Miinchen), pp. 22-55. MARSHALL, A. M. 82. The Segmental Value of the Cranial Nerves, Jour. Anat. Physiol., Vol. 16, Pt. 3, pp. 305-354, Pl. 10. PLATT, J. B. ’91. A Contribution to the Morphology of the Vertebrate Head, based on a Study of Acanthias vulgaris. Journ. Morph., Vol. 5, pp. 79-112, Pls. 4-6. RABL, C. ’92. Ueber die Metamerie des Wirbelthierkopfes. Verh. Anat. Gesellsch. VI. (Wien), pp. 104-135, Taf. 2, u. 4 Abbildg. SEWERTZOFF, A. N. ’95. Die Entwickelung der Occipitalregion der niederen Vertebraten im Zusammenhang mit der Frage iiber die Metamerie des Kopfes, Bull. Soc. Imp. Nat. Moscou, Année 1895, No. 2, pp. 186-284, Pl. 4 et 5. WIJHE, J. W. VAN. 82, Ueber die Mesodermsegmente und die Entwickelung des Nerven der Selachierkopfes. Nat. Verh. d. K. Akad. Wissensch. Amsterdam, Deel 22, 50 pp., 5 Taf., 1883. Also separate, Amsterdam, 1882, 50 pp., 5 Taf. 162 JOURNAL OF COMPARATIVE NEUROLOGY. THE CRANIAL NERVES OF THE Bony FISHES. By C. Jupson HERRICK. The cranial and first spinal nerves of Menidia have been plotted by reconstruction from serial sections in order to ex- hibit the relations of the nerve components both proximally and distally. In most cases the several components have been traced from their nuclei of origin or termination in the brain through the ganglia to their peripheral termination. Throughout the gnathostome vertebrates we now common- ly recognize four components in the typical spinal nerve—(1) somatic motor, from the ventral horn cells; (2) somatic sensory (general cutaneous), terminating in the dorsal horn; (3) vis- ceral motor; and visceral sensory. The central relations of the last two components are still obscure. They are probably both related to the ‘‘intermediate”’ or lateral horn zone, the sensory fibers coming in by the dorsal root and the motor fibers (in inframammalian groups, at least) going out by both dor- sal and ventral roots. Now in the bony fish the cranial nerves exhibit these four components and in addition a fifth, the acustico-lateral. The somatic motor is represented by the eye-muscle nerves; the somatic sensory by the general cutaneous component of the V and X nerves, terminating in the spinal V tract, which is the continuation of the dorsal horn of the spinal cord; the viscero- motor by the motor fibers of the other cranial nerves, going out near the sensory fibers by dorsal roots to the branchial musculature. The viscero-sensory system, like the viscero-mo- tor, has been hypertrophied and is represented by the com- munis system of the X, IX and VII nerves, terminating, either directly or through the mediation of the fasciculus communis in the vagal lobe (chief sensory vagus nucleus of higher forms). The communis system of the head, unlike the corresponding visceral sensory system of the trunk, receives fibers from taste buds and other sense-organs not belonging to the lateral line Herrick, Cranial Nerves of Bony Fishes. | 163 system. The acustico-lateral system receives fibers from the ear and lateral line organs and no others. These fibers all ter- minate together in the tuberculum acusticum. In the cranial nerves the motor fibers for the unstriated visceral musculature (with sympathetic connections?) are, as in the trunk, very small, while those for the striated visceral mus- culature of the branchial arches and for the somatic eye-muscles are large. The general cutaneous fibers are small or medium, the communis fibers are all very small, and of the acustico-lat- eral fibers those from the lateral line organs are for the most part very large, while the auditory fibers are of medium size. The accompanying diagram exhibits the relations of the sensory components in the cranial nerves of Menidia and some of the more important points are reviewed in the following summary. 1. The ramus medius (r. lateralis of authors) of the spinal nerves usually anastomoses with a twig of the n. lateralis vagt ;, but in all cases the spinal fibers go to the skin around the lateral line, and never to a lateral line organ. 2. The frst spinal is obviously a fusion of two segmental nerves, possibly of more than two. 3. The vagus nerve contains gencral cutaneous fibers (rami cutanei dorsales), which have a special ganglion (jugular g. of Shore and Strong) and which terminate in the spinal V tract. 4. The vagal lobe is mainly, at least, the terminal nucleus for visceral sensory fibers and hence is to be regarded as the continuation into the head of the ‘‘intermediate zone” of the spinal cord, rather than of the dorsal horn, as some have main- tained. These fibers, which will be termed communzs fibers, are in part general visceral sensory fibers from mucous surfaces and in part fibers from more highly specialized organs—taste buds etc. 5. The nucleus ambiguus, giving rise to the motor root of the vagus, has been specialized away from the general viscero- motor center in correlation with the development of the striated visceral musculature of the branchial arches. The central ner- 164 JOURNAL OF COMPARATIVE NEUROLOGY. vous connections of fibers for the unstriated visceral musculature of the vagus region are obscure. 6. The X/ nerve may be identified in teleosts. It arises by probably from thé caudal part of the nucleus ambiguus and — my ree FIT eta, = Ses F rstX » rcut dors X--} 4 nn a § t hm Uh é ? 7 < DESCRIPTION OF THE FIGURE. A diagrammatic view of the sensory components of the cranial nerves of Menidia, as seen from the right side. The diagram is based upon a projection of the cranial nerves upon the sagittal plane made by reconstruction from serial sections. The general cutaneous component is indicated by the single cross- hatching, the communis component by double cross-hatching and the acustico- lateral is drawn in black. REFERENCE LETTERS. 6. c. 4 to b, c. 5.—The five branchial clefts. br. g. X.—The ganglia of the four branchial rami of the vagus, the last one containing also the ganglion of the r. intestinalis. d. 1. g. ViZ,—The dorsal lateral line ganglion of the VII nerve. f. ¢.—Fasciculus communis. Gas, g.—Gasserian ganglion. gen, g. VII,—Geniculate ganglion of the VII nerve. IX.—The glossopharyngeal nerve and its ganglion. Jug. g.—The general cutaneous ganglion of the vagus nerve—jugular g. of Shore and Strong. Herrick, Cranial Nerves of Bony Fishes. 165 supplies the trapezius muscle. It is apparently a viscero-motor nerve. 7. The x. lateralis vagi terminates in the tuberculum acus- ticum, crossing all of the other vagus roots without, however, being connected with them in any way. It, however, receives a small bundle of communis fibers from the IX root. The latter, apparently go out with the first three or four branches of the n. lateralis (the first of these being the r. supra-temporalis vagi), accompanying the proper lateralis fibers, and ultimately anasto- mose with the r. recurrens VII. 8. The sensory IX is composed exclusively of communis fibers. They enter the lobus vagi by way of the fasciculus communis. Neither lateralis nor general cutaneous fibers are received during any part of its course, nor is there any connec- tion with any other nerve save the sympathetic chain and the intra-cranial anastomosis with the root of the n. lateralis vagi already mentioned. The IX nerve lacks the r. pre-trematicus and the r. supra-temporalis. 9. The motor IX arises from the cephalic part of the nu- cleus ambiguus, runs for a considerable distance along the lateral lob. X.—The lobus vagi. . 7.—The olfactory nerve. . [f,—The optic nerve, . cut. dors. X.—Ramus cutaneus dorsalis of the vagus. . tntest, X.—Ramus intestinalis of the vagus. Jat. X.—Ramus lateralis of the vagus. oph. sup. V—Ramus ophthalmicus superficialis trigemini. . oph, sup. V77.—Ramus ophthalmicus superficialis facialis. ot,—Ramus oticus. . pal —Ramus palatinus facialis. vec. V77.—Ramus recurrens facialis. os a Re ee ea eee | st. X.—Ramus supratemporalis vagi. r. VII p-t —Ramus pre-trematicus facialis. sp. V. t.—Spinal V tract (** ascending root of the trigeminus ”’). t, a.—The tuberculum acusticum. t. hm.—Truncus hyomandibularis of the tacial nerve. t. inf.—Infra-orbital trunk, containing the r. mandibularis V, the r. maxil- aris V, and ther. buccalis VII, together with communis fibers. Vi/7,—The eighth nerve. v. l. g. ViZ,—The ventral lateral line ganglion of the VII nerve. 166 JouRNAL oF CoMPARATIVE NEUROLOGY. surface of the fasciculus longitudinalis dorsalis and before leaving the latter contributes a considerable bundle of fibers to it. 10. The auditory nerve terminates in the tuberculum acus- ticum, and its fibers are internally so mingled with the lateralis fibers from the X and VII nerves that analysis is impossible in Weigert preparations. 11. The sensory VII roots contain two components. The communis portion, enters the fasciculus communis, com- prising the whole of that tract except the fibers received from IX. It terminates in the lobus vagi, a lobus trigem- ini not being developed. This communis root enters the geniculate ganglion and distributes to (1) the r._ palati- nus (comprising the whole of that nerve) for the mucosa and taste buds of the roof of the mouth; (2) the truncus hyo-man- dibularis VII for the mucosa and taste buds of the inside of the lower jaw and lip; (3) the r. maxillaris V to taste buds within the upper lip; (4) the r. recurrens VII. The latter fibers pass dorsally into the cranial cavity, forming in the men- inges an elaborate plexus, finally to combine into the r. recur- rens which runs the length of the body superficially near the dorso-median line. These communis fibers supply some terminal buds on the top of the head and some others probably run for- ward with the ophthalmicus superficialis. In addition to the above there is (5) a small twig which leaves the geniculate gan- glion between the truncus hyo-mandibularis and the r. palatinus running directly ventrally to the roof of the mouth, supplying its mucosa in the region between the areas supplied by the IX and palatine nerves. In its course it passes along the cephalic face of and innervates the very large pseudobranch. This is the only nerve supply which that organ possesses, and this nerve is accordingly, I think, to be regarded as the pre-trematic VII nerve, the pseudobranch representing a spiracular gill and the truncus hyo-mandibularis the post-trematic VII. 12. Thesecond sensory component of the VII nerve is represented by two dateral line roots. (a) The ventral lateralis root has a separate ganglion and supplies organs of the opercular and mandibular canals, via the truncus hyo-mandibularis. (b) Herrick, Cranial Nerves of Bony Fishes. 167 The dorsal lateralis root also has a separate ganglion and sup- plies organs of the infra-orbital and supra-orbital lines, via the r. buccalis and r. ophthalmicus superficialis VII respectively. 13. The nucleus of the motor V/J corresponds in position and structure to the n. ambiguus and the root is related to the dorso-median fasiculus exactly like the motor IX root. It is, at its origin, distinct from the sensory roots of VII and supplies the mm. levator operculi, adductor operculi, adductor hyo-man- dibularis, adductor arcus palatini and the branchio-stegal mus- cles, as usual among the teleosts. It does not, however, supply the m. genio-hyoideus, as usually stated. 14. The sensory V is composed exclusively of general cutaneous fibers. It receives the whole of the pre-vagal spinal V tract. The Gasserian is its proper ganglion and this term should not be applied to any other cells of the V + VII gan- glionic complex. From the Gasserian g. are given off general cutaneous fibers into (a) r. maxillaris V, (b) r. mandibularis V, (c) r. ophthalmicus superficialis V, (d) fibers running back into the truncus hyo-mandibularis VII for the operculum, also (e) a very small r. profundus V. The latter accompanies the sym- pathetic fibers of the radix ciliaris longa of the ciliary ganglion to that ganglion after which they can no longer be separately followed. The relations of this nerve, which has not before been described for teleosts, indicate that the embryonic profun- dus ganglion has fused with the Gasserian. 15. The motor V nucleus resembles that of the motor VII, but lies farther laterad and dorsad. The fibers enter the r. mandibularis V and supply the mm. dilator operculi, levator arcus palatini, adductor mandibularis, inter-mandibularis and ‘genio-hyoideus. The innervation of the latter muscle has hith- erto been usually assumed to come in teleosts from the VII. This muscle is almost certainly not homologous with the muscle in the corresponding position of other vertebrates which is sup- plied by the 1 spinal or XII nerve. 16. The sympathetic chain has ganglia on nearly all of the cranial ganglia and probably sends fibers into all of the rami from 168 JOURNAL OF CoMPARATIVE NEUROLOGY. the latter. In passing from IX to VII ganglia the sympathetic runs external to the ear capsule. The various components can be followed with great pre- cision proximally in the root portions and through the ganglia of the cranial nerves. Throughout the peripheral courses of the nerves the analysis is somewhat more difficult, but has been satisfactorily accomplished in all but a very few cases. The naked organs of the lateral line series (pit lines) and the terminal buds of the skin (communis system) are sometimes hard to dif- ferentiate because their nerve fibers are intermediate in size be- tween the exceedingly large fibers typical for the lateralis system and the very small communis fibers. The general cutaneous system of nerves is, however, as clearly separable from the others peripherally as it is centrally. And this is important in many ways. For example, it will materially assist in the at- tempt to homologize cranial and spinal nerves to know that not all sensory cranial roots are comparable with spinal dorsal roots. It is, e. g., no longer legitimate to homologize lateral line roots with dorsal spinal roots. The latter are represented in the brain mainly by the spinal V or general cutaneous system, and the special cutaneous systems (terminal bud and lateralis) are probably neomorphs in the head, as Strong has maintained. If cranial and spinal nerves were derived from a common type, the common ancestral nerve probably contained two kinds of sensory fibers, viz. general cutaneous and general visceral. Both of these kinds of fibers appear to be present in the dorsal roots of Amphioxus and of the spinal nerves of Craniota. Two of the cranial nerves retain the general cutaneous fibers (viz. X and V); the others seem to have lost them. The vis- cero-sensory fibers have either been lost or rendered unrecog- © nizable on account of their extreme specialization in all but the X, IX and VII nerves. In these nerves they have been cen- tralized to form the communis system and hypertrophied to serve a double purpose: (1) The viscero-sensory nerves of the trunk seem to have been in large measure supplanted by the r. intestinalis vagi. (2) In the cephalic end of the digestive tract more highly specialized sense-organs (taste buds) have been de- Herrick, Cranial Nerves of Bony Fishes. 169 veloped in response to an obvious functional need. The advan- tage to be derived from such a centralization of the sensory ap- paratus of the entire digestive tract is obvious. The acustico-lateral system is apparently phylogenetically the youngest of the cranial systems. Its relations to the other sensory systems are still problematical. ADDENDUM. Since this paper was read there has appeared the very suggestive paper on the cranial nerves of the sturgeon (Anat. Anzeiger, XIV, 22-23) by J. B. Johnston. His conclu- sions, which differ somewhat from my own, I shall examine critically at another time, merely mentioning a few of the sali- ent points here. Johnston identifies the general cutaneous and acustico-lateral systems, regarding them both as representing the dorsal horns of the spinal cord. The acustico-lateral is the more highly specialized part and it possesses a spinal portion running parallel with the spinal V, which he calls the spinal VIII. The close internal connections between these two sys- tems and their close parallelism in many other respects certain- ly favor the belief that the acustico-lateral has been differen- tiated from the general cutaneous, in spite of the complete dis- creteness of the two systems peripherally. And it should be noted that this does not imply that the lateralis rami from the head can ever be directly homologized with any rami of spinal nerves ; for the former are none the less neomorphs in the head, even though their precursors were in the spinal nerves, as Cole has so ably argued. It is interesting to note that the latter au- thor also regards the acustico-lateralis system as the derivative of the general cutaneous, the evidence in this case being embry- ological. Now, Johnston regards the communis system as_ peculiar to the head, having no spinal representatives. He even goes so far as to state that ‘‘no sensory fibers of the spinal nerves supply visceral structures.’’ This, I think, is erroneous, even in the higher forms, though the great reduction and profound modification of the viscero-sensory system of the trunk under the influence of the r. visceralis vagi are freely granted. Further- ? 170 JOURNAL OF COMPARATIVE NEUROLOGY. more, Johnson regards this communis system as exclusively vis- ceral, i.e. entodermal, and opposes to it the other sensory system, fi viz, the general cutaneous and acustico-lateral,as related to strictly ectodermal sense-organs. This, however, seems to lead us into serious difficulties, for, in the first-place, the terminal buds of the outer skin, which are very numerous in some fishes and which can hardly be other than ectodermal, are apparently all innervated from communis system. Again, the taste buds of the mouth of fishes all or nearly all lie in the region of the . stomodzum and are therefore probably of ectodermal origin. These among other facts seem to forbid the employment, in the present state of our knowledge, of any such morphological cri- teria of the components as Johnston adduces. Indeed, the basis for the seggregation of the components may be funda- mentally physiological, as Cole and Kingsbury seem inclined to believe. August 1, 1898. REVIEW OF JOHNSTON ON THE CRANIAL NERVES OF THE - STURGEON. | By O. S. Srrona, Columbia University. This communication contains a résumé of the results of the author’s investigation on the hind brain of Actpenser rudbt- cundus, Le Seur. The investigations were made by means of the method of Golgi on the brains of fishes 25 to 40 cm. in length. Only a few of the most striking results will be noted here, leaving a more detailed review till the appearance of the final paper. The work is of a character much needed in this field at present and though surprising in some respects, the re- sults will doubtless be very valuable. 1 Hind Brain and Cranial Nerves of Acipenser, by J. B. Johnston (Univer- sity of Michigan). Anatomischer Anzeiger, XLV Band, Nr. 22 and 23, 1808. STRONG, Johnston on Nerves of the Sturgeon. 171 The ventro-lateral tracts of the medulla appear to be made up of the neurites of commissural and tract cells, as has been described in the spinal cord of Selachians and Teleosts.. The smaller number of the fibers of the fasciculus longitudinalis pos- terior come from the central diffuse nucleus of the thalamus, the greater number from the motor cells of the ventral horn along the course of the fasciculus in the medulla. Especially interesting is the statement that ‘‘it gives immediate origin to the VIth., to the larger part of the ventral root of the VIIth., to the whole of the ventral roots of the [Xth. and Xth., and to the XIIth. nerve. A part of the ventral root of the VIIth., all of the ventral Vth., as well as the IVth. and IIIrd., arise from cells lying in the latero-dorsal portion of the ventral horn whose neurites pass more or less directly out into these roots without entering the fascicuius longitudinalis posterior.”’ A considerable part of the sensory portion of the trigem- inus after it enters the medulla descends as the spinal V and principally terminates in an enlargement of the dorsal horn of the cord at the point of transition from cord to medulla (nucleus funiculi). The spinal V also receives contingents from the IX- X group as described by Strong and Kingsbury. The larger part of the fibers of the sensory Vth. (deep portion), however, en- ters the tuberculum acusticum where they form a distinct bundle running both forward and backward. The descending portion of the fibers of the Vth. entering the acusticum become arcu- ate fibers and most of them, possibly all, reach the opposite side of the medulla. This is a remarkable course for dvect root fibers of the V, it may be remarked here, and should rest on a firm basis of observation. The ascending portion of the deep Vth., enters a nucleus at the anterior end of the medulla which is immediately continuous with the body of the cerebellum. A part of the fibers of the VIIIth. end in relation to the Zwischenzellen of Goronowitsch, the remainder, together with the lateral line roots, also form ascending and descending bun- dles. One portion of the descending fibres (called the spinal VIlIth.) partly terminates in the nucleus funiculi and partly in a smaller nucleus mesad of it; the remainder of the descend- 172 JOURNAL OF COMPARATIVE NEUROLOGY. ing fibers forms arcuate fibers like those of the deep Vth. The ascending fibers of part of the VIIIth. form a slender bundle close to the central cavity and terminate in relation to cells closely investing it. ‘‘ The remainder of the VIIIth. fibers and all the ascending lateral line fibers run up to the cerebellum, most of them spreading out in the lateral lobes and the re- mainder entering the body.”’ Some fibers of the ventral lateral line root of the VIIth. enter the so-called lobus trigemini which also receives the dorsal lateral line root of the VIIth. Owing to its entirely misleading character, Johnston proposes to sub- stitute the name Jodus linee lateralis for the term lobus trigem- ini. Its structure is similar to that of the tuberculum acusti- cum. In both are found three types of cells, (a) cells with short neurites like those in the cerebellum, (b) minute cells comparable with the granule cells of the cerebellum, and (c) a large number of cells comparable with the Purkinje cells. Kingsbury’s view as to the non-identity of the lobus trigemini of Acipenser with the structure of the same name in Teleosts is confirmed. The sensory fibers of the IX and X break up in the dorsal part of the lobus vagi. ‘‘ These fibers end in relation with cells of the II type whose neurites break up mostly in the ventral and lateral parts of the lobe. This part of the lobe is made up of cells whose neurites take a ventro-lateral course to the lateral part of the medulla. Here the fibers either pass anteriorly or posteriorly without dividing, or they divide, one branch going anteriorly, the other posteriorly. The smaller number of fibers turn posteriorly. They form a distinct bundle of non-medul- lated fibers ventral to the spinal Vth. and continue into the cord. The anteriorly directed bundle runs ventral to the acus- ticum to the anterior end of the medulla, where it ends ina large nucleus, forming the antero-lateral limit of the medulla, lateral to the median Vth. nucleus. This is the Rindenknoten of Mayser and Goronowitsch.”’ missure. They are connected by a com- The anterior part of the lobus vagi ‘‘ which corresponds to the L. trigemini of Mayser is very much smaller in Acipen- | | | 4 | . Stronc, Johnston on Nerves of the Sturgeon. 173 ser than in Teleosts and is composed chiefly of the root fibers of the dorsal VIIth.” which enter it and turn caudad. John- ston very properly recommends that the name lobus trigemini be entirely dropped. It is amusing that all recent writers are agreed that the one nerve with which either species of ‘‘lobus trigemini’’ has nothing to do is the trigeminus. The destination of the neurites of the Purkinje cells of the cerebellum was not determined but it is considered probable that they run through the acusticum to the base of the medulla. The fibers entering the cerebellum come chiefly from the med- ulla, the tectum and optic thalamus, also from the lobi inferiores. Johnston emphasizes the structural continuity of the cerebellum with the acusticum and is of the opinion that this fact points to the conclusion that the cerebellum is the enlarged anterior end of the center for the sensory nerves of the integument. The facts above mentioned in connection with the cranial nerves lead Johnston to agree with Kingsbury in his analysis of the sensory centers in the main, but, in view of the partial mingling of the V-VIlII-lateral line group in the medulla, to think that Kingsbury separates the spinal Vth. and the cere- bellum from the acusticum more than is warrantable. On the other hand, Johnston emphasizes the distinction between the centers for the above group and the center—the lobus vagi— for the VII-IX-X group. After a brief consideration of the peripheral structures innervated by these nerves, he comes to the conclusion that ‘‘all sensory structures of ectodermal origin are supplied by components of the Vth. (including spinal Vth. components running in other nerves), VIIIth., and lateral line nerves, and that all fibers supplying such structures have their central endings in the nucleus funiculi, the |tuberculum acusti- cub, or the cerebellum, except such as pass through the acusti- cum as arcuate fibers. On the other hand, all sensory struc- tures of entodermal origin are supplied by VIIth., [Xth. and Xth. components, and all fibers supplying such structures find their central endings in the lobus vagi.”’ In coming to this general conclusion the writer appears to have overlooked the important fact, which seems to be quite 174 JouRNAL OF COMPARATIVE NEUROLOGY. certain, that fibers innervating end-buds scattered over the sur- face of the head and even the body of the Teleosts have their central termination in the lobus vagi. Such end-buds are of course ectodermal. The association of such fibers with visceral fibers is puzzling in any case and it has occurred to the writer of this review—as well as to C. Judson Herrick—that it might be accounted for on the supposition that the end-bud organs originated on or near endotermal surfaces. Such a sup- position is, of course, merely speculative and unless it could be established, the fact that fibers innervating end-buds on ecto- dermal surfaces have their central termination in the lobus vagi would appear to constitute a fatal objection to Johnston’s gen- eralization that there are two distinct sensory systems—ectoder- mal and entodermal. The partial confusion of centers for the V-VIII-lateral line group which Johnston describes and which led him to group these nerves together, merits a careful consid- eration, however, by those who are attempting the analysis of the cranial nerves in fishes. Johnston also comes to the conclusion that there appears to be ‘‘no structure in the cord with which the lobus vagi can be considered homologous.”’ Consequently such nerves are not available in determining segments of the head homodynamous with trunk segments. He finally concludes that in using the cranial nerves to determine segmentation, it is probably best to disregard the sensory nerves altogether and use only the motor series. The fibers of Meynert’s bundle are traced beyond the gan- glion interpedunculare. Most of the coarse fibers after decus- sating ventral to the ansiform commissure turn dorsad, pierce the above commissure and terminate in cells near the fasciculus longitudinalis posterior whose neurites cross the middle line in the ansiform commissure and join the ventro-lateral tracts. The fine fibers could not be traced with certainty, but there is reason to believe they terminate in a nucleus of small cells on the lat- eral surface of the lobus linez lateralis. Crapp, Alls on Nerves of Amia. 175 REVIEW OF ALLIS’ PAPER ON THE CRANIAL NERVES OF Awmia.! By Miss Cornetia M. Crapp. Mt. Holyoke College. Mr. Allis seems now to have accomplished the task set for himself in 1886. This is the third paper which has appeared ; the first being the admirable account of the ‘‘ Anatomy and Development of the Lateral Line System of Amia calva;”’ the second, a short paper in ’95, preliminary to the third which ap- peared about a year ago. It is, as it was intended to be, ‘‘an ‘accumulation of facts and references grouped so as to be con- veniently used as a basis for further work.”’ In the eighteen beautiful plates illustrating the paper, Mr. Nomura, the artist, has faithfully reproduced the dissections, many of which were made by himself. The thoroughness of the work is indicated from the state- ment in the introduction, that ‘‘nothing is shown in the adult that was not controlled in larve, and everything found in larve has been sought for until found or accounted for in the adult.”’ The chief merit of this voluminous paper lies in its accur- ate detail, making it invaluable to the anatomist as a reliable source of information. From the study of the eye-muscles and their innervation, Allis thinks that the muscles of the eye in vertebrates are not homologous structures, the want of homology being found en- tirely in those muscles that are innervated by the oculomotorius and abducens, i. e. those muscles that are said to arise from van Wijhe’s first and third somites. As the eye-muscles have been developed from the muscle masses in these two somites, differ- ent arrangements have arisen in the various groups of Ichthy- opsida. On this basis Mr. Allis has constructed a number of prototypes, and has added another to the forest of genealogical trees. 1The Cranial Muscles and Cranial and First Spinal Nerves in Amia calva, by Edward Phelps Allis. /ourn. of Morph., Vol. XII, No. 3, 1897. 176 JouRNAL oF COMPARATIVE NEUROLOGY. . ' Plate XXII shows the relation of the eye-muscles and the nerves innervating them, to the ophthalmicus profundus or naso- ciliaris trigemini. There are two main lines of descent: on the one hand, by the splitting of the large inferior oblique muscle, a new rectus internus is formed, the earlier muscle of that name disappearing or fusing with the rectus superior; this gives the condition seén in Petromyzon. On the other hand, the new muscle formed by the split becomes the rectus inferior and the earlier inferior rectus fuses with the rectus internus; thus arises the proto-urodele type, from which there are three lines of des- cent. One of these lines leads to the selachians, another to the fishes (ganoids and teleosts) and a third to the Amphibia and higher vertebrates. The amphibian branch shows two diverging lines of development, the one leading to Urodeles and the other to Anura, where a new rectus internus is formed and the old one disappears or is fused. Allis remarks in general that ‘‘the lines leading to the higher types of each class resemble each other in that the su- perior branch of the oculomotorius in such lines innervates but one of the muscles of the eye, while in the lines leading to the lower types it always innervates zwo.”’ Healsosays: ‘‘ There has been but one zmpulse, if it may be so called, leading to the formation of the arrangements found in higher types and not repeated ones.”” We may suppose that the splitting of the large inferior oblique muscle spoken of above, is that impulse. Allis states that this ancestral tree of the muscles and nerves of the eye-ball are based on ‘‘insufficient and perhaps inaccurate data,” but it is interesting to note that the same grouping of orders of Ichthyopsida is shown by Hasse and Maurer ; the former basing his conclusions on the study of the development and structure of the vertebral column and the lat- ter on the development of the muscle cells and muscle fibers. The innervation of sense organs is discussed at length in this paper, and much of the ‘‘review and comparison of nerves” has great interest for those working along these lines. Since ’88 Allis has discovered that the glossopharyngeal nerve takes no part in the innervation of the canal organs, the ——. Ow | DAHLGREN, Grant Ganglion Cell Apparatus. 1a so-called dorsal root of that nerve receiving its fibers from the root of nervus linez lateralis and from the ramulus ampulle posterioris. If the terminal buds, found in such numbers and so irreg- ularly distributed over the head of Amia, represent a stage in the development of the canal organs, they and the nerves inner- vating them should arise in connection with sensory ecto- dermal thickenings, as do the canal organs and their nerves. Allis has shown that those trigeminal and facial nerves, in Amia, that are known to innervate terminal buds or regions where those buds abound, all arise fromthe median, fasciculus communis portion of the main trigemino-facial ganglion. He thinks it probable, as Strong has suggested, that ‘‘ the fasiculus communis tract of the brain is largely or entirely concerned in the innervation of terminal buds.” The chorda tympani belongs probably to the fasciculus communis nerves, and is represented in Amia, by the mandibu- laris internus trigemini. There is no true ramus ophthalmicus profundus in Amia, but the ganglion is distinct, anda rudiment of the nerve is sometimes seen. The ‘‘hitherto undescribed cranial nerve’”’ of Pinkus in Protopterus is found in Amia. THE GIANT GANGLION CELL APPARATUS. By Uric DAHLGREN, Princeton University. The following report, read before the Neurological Semi- nar of the Marine Biological Laboratory in July, 1898, is an outline of some recent work on the problem presented by the ‘‘Giant Ganglion Cells”? found in the median dorsal fissure of the myelon of a number of fishes. This report includes several recent advances made by the writer on the problem in question, 178 JOURNAL OF COMPARATIVE NEUROLOGY. The name ‘‘ Giant Ganglion Cells’ is objectionable, as it not only describes nothing but their size, but is used for other large nerve cells in many scattered forms of the invertebrate animals as well as the vertebrates. ‘‘ Median Cells” or ‘‘ Dor- sal Cells’? would seem better, but the writer will refrain from committing himself to a new term until more is discovered con- cerning the apparatus. These cells were first described in Lophius piscatortus by Fritsch. In this fish they number about 200, are very large and are massed in the anterior portion of the dorsal median fis- sure. Each one gives rise to a single large neurite, according to Fritsch, which dips down into the cord and becomes a con- stituent fiber of one or the other of two lateral, symmetrical fiber-bundles. The fibers thus forming part of these bundles pass cephalad into the brain and out of that through certain roots of the 10th and 5th nerves. Fritsch considers them to be sensory nerves. Tagliani found the same large cells in a similar position in the myelon of Balstis and of Orthogoriscus, in which latter animal he found the fibers running both cephalad and caudad. In my article of 1896 (Anat. Anz., Bd. XIII, p. 281) I described these giant cells in a large number of Heverosomata, or flounders, and found that the neurites entered the usual fiber- bundles but ran caudad in them instead of cephalad as Fritsch described for Lopiius. At the same time I gave certain reasons for believing that this apparent difference in neurite-distribution did not prohibit the homology of the cells in the different forms (p. 291). Other and more satisfactory proof that the giant cells of these fishes are homologous has been found since and consists of the following facts. The giant cell apparatus of Pseudopleuronectes Am. was further studied by means of Pala- dino’s Palladium Iodide method (Lee, 1Vth Ed., p. 413) by means of which it was possible to trace the neurites with more ease and certainty than before. It was then found that in many of the cells the large neurite bifurcated just before entering the fiber-bundle and one branch ran cephalad, while the other ran caudad, both in the bundle. A Pediculate fish, nearly allied to . 2e am — DAHLGREN, Grant Ganglon Cell Apparatus. 179 Lophius was then examined and the missing caudal fiber was immediately discovered. This fish, Pterophryne histrio possesses only 24 giant cells which is the smallest number yet found in an adult teleost fish. These cells are very large and closely re- semble those of Lofizus in every particular. Having demon- strated this bifurcation of the giant cell neurite in these two forms, it is probable that the same will be found to hold true for the other fishes possessing this apparatus. The number of species in which the giant cell apparatus appears, has been found to be far greater than supposed. A vast number of teleost fishes possess it and the writer does not hesitate to estimate that this number is over one half of the living species. It is found less frequently in the lower orders, the NEMATOGNATHI, for instance, not showing one instance as yet in which it has been found; while the PepicuLatr and HETERO- SoMATA fail to yield one species which does not possess it in a highly differentiated and developed form. Where found ina more primitive form it is characterized by the large number, small size, and simple structure of its ganglion ceils, while in the specialized forms the ganglion cells are fewer and of mon- strous proportions and uncouth structure. Its relation to the ‘‘Transient Ganglion Cells” or ‘‘Trans- ient Nerve Apparatus,” described in Salo, Raja, etc. by Ro- hon, Beard, Van Gehuchten and others, is clearly, in the writer’s mind, one of identity; the giant cell apparatus being composed of some or all of the transient cells which have re- mained. Van Gehuchten’s discovery of the bifurcation of the neu- rite of the transient ganglion cells of Sa/zo must therefore be considered the first exposition of this fact for the giant cell ap- paratus as well. Many embryonic fishes have been found to possess the transient apparatus and the writer doubts if any form of teleost is without it. Work will be continued on Pseudopleuronectes Am. and other forms. 180 JOURNAL OF COMPARATIVE NEUROLOGY. INNERVATION OF THE OLFACTORY EPITHELIUM. By A. D. Morri t. Hamilton College. The innervation of the olfactory organ is of interest as it is the only sense organ, in vertebrates, where continuity of the nerve-fibers with the sensory cells exists, according to the large majority of investigators. The statement by Dr. Ayers that the relation of the nerve fibers to the sensory cells was the same in the cochlea of the pig as in the olfactory organ of vertebrates led me, after study- ing the relation of the fibers and sensory cells in the ampulle of the ear of the smooth dog-fish, to try the same methods with the olfactory organ of this fish. In the ampulle, with Ehrlich’s method, I found contact and free endings but no case which seemed to be true continuity. In the olfactory organ I found continuity and some some free endings but nothing that resembled the relations found in the ampullze. I found the three types of cells described by Dogiel in his paper on the olfactory epithelium of the sturgeon: Ist, spindle- shaped; 2nd, cylindrical with a slight constriction near the middle; 3rd, conical, cells. All are ciliated and continuous with nerve fibers, which extend toward the brain. Between these olfactory cells are long irregular shaped supporting cells. Whether the difference in shape of the sensory cells is due to difference of function or to mechanical causes was not deter- mined. The development of the olfactory nerve and epithelium have been much studied in connection with the problem of metamerism in the vertebate head. Some investigators, with Marshall, consider the olfactory nerve as the first cranial and consequently of segmental value, while others regard it as sim- ilar to the eye in being a modified portion of the brain. The recent paper of Disse on the development of the ol- factory nerve and epithelium in the chick is of considerable in Morritx, /unervation of Olfactory Epithelium. 181 terest on this account. He found, with the Golgi method, in the olfactory pit of the third day chick neuroblasts with taper- ing processes pointing toward the central portion of the epithe- lium but found no trace of anerve. Supporting cells were found between the neuroblasts (Fig. 1). This had been ob- served by His and others. ~- supporting cells. neuroblast: fig. 7. After Disse. At the fifth day the nerve processes from the neuroblasts were found to have extended quite a distance from the epithe- lium toward the brain but had not reached it. During the sixth day the fibers were found penetrating the outer portion of the brain which was still free from cells and the processes easily fol- lowed as had been previously observed (Fig. 3). brain. fibers of olfactory nerve, gan elion cell- Fig. 2. After Disse. The ganglion cells found quite early in the developing ol- factory nerve between the fibers and considered by His as bi- polar Disse claims to be unipolar and since he finds them first in that part near the olfactory epithelium and of the same shape as the neuroblasts and later finds them nearer the brain which 182 JOURNAL OF COMPARATIVE NEUROLOGY. they finally enter and since only one process was ever observed and that on the side toward the brain, he concludes that some of the neuroblasts have migrated into the nerve and finally reach the brain (Fig. 2). oo ona 7 7g cell. LP ~~ =--o}fagtory nerve brain capsule. k- olfactory : pit. fig. 3. After Disse. His found cells which he described as bipolar in the devel- oping nerve and on this account thought that the term olfactory ganglion was more appropriate than olfactory nerve. The main points claimed by Disse are: Ist, confirming previous investigators that the olfactory nerve develops from neuroblasts in the olfactory ‘epithelium; 2nd, That a part of the neuroblasts wander into the olfactory nerve and pass event- ually into the brain but remain unipolar ; 3rd, He considers the olfactory nerve as embryonic in character with cells of origin remaining in the olfactory epithelium. The fibers are non-med- ullated and in these respects quite different from the cranial nerves. ss 1. TS Sl SS ee SARGENT, Giant Ganglion Cells of Etenolabrus, 183 THE GIANT GANGLION CELLS IN THE SPINAL CoRD OF CTEN- OLABRUS ADSPERSUS (WaALB.-GOODE). By Porter E. SarGEnt, Harvard University. During the winter of ’97-’98 while engaged in the study of Golgi preparations of the central nervous system of the common cunner, Ctenolabrus adspersus, my attention was at- tracted to large bodies lying in the median dorsal fissure of the spinal cord. A little study showed them to be nerve cells of gigantic proportions, giving off a ventral process. The following preliminary paper is a summary of studies made during the spring of 1898. No attempt at a critical dis- cussion of the subject will be attempted in this paper, that be- ing reserved for the final article. I wish here to express my ob- ligations to Prof. E. L. Mark for kindly advice and assistance, and to Mr. Alexander Agassiz for opportunities enjoyed at his Newport Laboratory, where the material for this study was collected and prepared. Colossal ganglion cells in the spinal cord of certain Ichthy- opsida have attracted the attention of a large number of obser- vers during the past forty years. Upward of sixty articles in the literature deal with the subject to a greater or less extent. The greater number of these papers have to do with a transient nervous apparatus existing only in the embryos and larval stages. The most recent papers on this subject are those of Studnicka ’95 and Beard ’96. In adult fishes giant ganglion cells occurring in the dorsal portion of the cord have been noted by many investigators from Miller ’44 to Kolster ’98. Most of these observations are frag- mentary and all are very incomplete, so that as yet little is known of the occurrence, distribution and stucture of these cells, and almost nothing of the course of their fibers, while their function isa mere matter of conjecture. In only one in- stance have the neurites been traced. Fritsch ’84 and ’86 found that the giant cells imbedded in the anterior part of the cord of 184 JouURNAL oF CoMPARATIVE NEUROLOGY. Lophius, sent their’axis cylinders cephalad to the roots of the Trigeminus and Vagus nerves. Three recent papers bearing more directly on this subject deserve notice here. Dahlgren ’97 finds in the embryos and adults of the order Heterosomata certain giant ganglion cells lying in the median dorsal fissure or in a double row on either side of the dorsal fissure of the cord, and varying in number from 69 to 500 in different species. These cells give off neu- rites all of which run caudad in two fiber bundles lying bilaterally in the dorsal part of the cord. The neurites were followed but a short distance through the bundle, and their termination was not made out. The suggestion is made that they are connected with sense organs in the fins. Kolster ’98 describes giant ganglion cells lying in the dorsal fissure of the cord of Perca fluviatilis. The cells are stated to have no den- drites and the neurites were followed but a short distance, the direction which they take not being stated. The hypothesis is advanced that they have the function of raising the spines of the dorsal fin. Tagliani’97 has described the occurrence of similar cells in Orthagoriscus and Tetrodon. There has been a tendency among writers on this subject to consider as homologous all the colossal ganglion cells occur- ring in the dorsal part of the cord in the various groups of Ich- thyopsida, or to make wide and sweeping generalizations as to their homology, although at the same time the greatest diver- sity of function has been hypothetically ascribed to them. A comparison of the very diverse conditions described in various fishes and the utter lack of harmony in the homologies made by different writers, taken in connection with my own observa- tions on many different species, justifies the conclusion that the conditions are much more diverse and complex than has yet been recognized, and that these varied elements are not homol- ogous throughout the Ichthyopsida, or even throughout the group of fishes. Though they may have had a common origin in the ancestral giant cells of worms and crustaceans, they have assumed such very different form, position and function that they cannot be said to be homologous; and it is perhaps more Sarcent, Giant Ganglion Cells of Ctenolabrus. 185 probable that they have been independently derived from less conspicuous elements as the occasion for great size has arisen. _Methods.—The brain and spinal cord was carefully removed and immediately fixed in one of the following fluids, — (1) 10% solution of Formol. (2) Saturated aqueous solution of Corrosive Sublimate. (3) Flemming’s stronger chromic-osmic-acetic fluid. (4) Potassic Bichromate, gradually raised from 2% to 5 % solution. Many stains fail to bring out clearly the giant cells and their neurites though staining other parts of the nervous system well. This is particularly true of the carmine stains. The fol- lowing in the order named proved the most valuable: 1. Kenyon’s Copper Sulphate Phosphomolybdic Acid Hematoxylin, following formol preservation. 2. Heidenhein’s Iron Hematoxylin, used on formol or sublimate material. 3. Sahli’s Methylene Blue Acid Fuchsin Axiscylinder Stain, used on Bichromate material. 4. Ehrlich’s Acetic Acid Alum Hematoxylin double stained with Congo red, or Acid-fuchsin. The first stain proved of the greatest value, and as this is the first time, I believe, that it has been used on vertebrate material, deserves a word of comment. Material fixed in 10% formol and preserved in 5 % was washed and put in a 5% solu- tion of copper sulphate for 24 hours, by which time it had as- sumed a green color. After cutting in paraffin and mounting in the usual way, they were stained on the slide from 15 to 30 minutes, in the following :— 10% Phosphomolybdic acid, , 1st ek Hematoxylin crystals, . I gm. Chloral hydrate, , i Ose ms, Water, 400 ¢.Cc. They were then rinsed in water, dehydrated, cleared and mounted in the usual way. This is an excellent differential stain for neuroglia, the dendrites of ganglionic cells and espe- cially the axis cylinders, the myelin being left wholly unstained, 186 JouRNAL OF COMPARATIVE NEUROLOGY, The giant ganglion cells of Ctenolabrus form a single me- dian longitudinal row in the dorsal portion of the cord, lying within the dorsal fissure with their upper surfaces flush with the dorsal limit of the cord, and covered by the membrana prima, (fig. 3). Each cell lies within a capsule formed of three ele- ments,—(1) the membrana prima which is arched above each cell, (fig. 5, mb. p. 190), (2) the neuroglia fibers (7.f), which hy?, Fig. 1—Diagrammatic parasagittal section of the medulla and anterior part of the cord of Ctenolabrus, showing the arrangement of the giant ganglion cells and the course of their neurites. The lateral bundle (27.4.) and the lateral cells (2) are projected on the median plane. The canalis centralis is shown in dotted lines, #s.7d. fissura rhomboidalis ; cé/. cerebellum ; 2yf. hypoaria. Fig. 2.—Diagrammatic frontal section of same, the cells and lateral fiber bundles projected on a plane. O#¢. /. optic lobe. come off from the membrana prima and extend downward to the canalis centralis, (3) the fine neuroglia network, (w.7.). The capsules have an internal diameter of one anda half to two times that of the cell, so that each cell is surrounded by a space, SARGENT, Gzant Ganglion Cells of Ctenolabrus. 187 in which it is partly supported by the numerous dendrites which run off from the cell to the surrounding neuroglia. The giant cells extend from the posterior end of the fissura rhomboidalis caudad through the anterior end of the cord, (fig. 1). The largest cells are toward the anterior end of the series, and there is some diminution in size posteriorly. They are as a rule more closely set anteriorly, being separated from each other by intervals of from one-fourth to one-half their diameter. The intervals between the adjacent cells increases posteriorly to three and four times the diameter of the cells, the last few being irregularly placed at perhaps greater intervals (figs. 1 and 2). The cells tend to become aggregated in groups of three or four, separated from other groups by wider intervals. In the anterior portion of the series mutual crowding may influence the form of the cells, or may result in pushing some of the cells to one side of the median line, or downward below the level of the others. Occasionally two cells may be found ina transverse section lying side by side. On either side of that portion of the canalis cen. tralis where it widens out and opens into the fissura rhomboidialis there are sim- ilar cells bilaterally placed, usually two on each side but the number may vary from one to three (fig. 2, Z),) in He, “P, \ these, ate : shown projected upon the fig. 3. Diagrammattic transverse sec- tion of the anterior part of the cord. The median plane (2). Rarely outlines of the grey matter are shown in similar giant cells are found dotted lines. 7.d. radix dorsalis; ¢.c. cam- which do not lie in the alis centralis; #z.f. Mauthner’s fibers. 2 ace dorsal fissure but lie lat- erally, deeply buried in the cord, (fig. 1, 6). This condition occurs inabout one case out of 300. Size.—In the youngest fishes 3 cm. in length the cells have an average diameter of 7 or 8 ,. In the various stages 188 JouRNAL OF COMPARATIVE NEUROLOGY. from the half grown individual 10 cm. long, to the full grown fish of 20 cm. there is a difference in the size of the cells keep- ing pace with the growth of the body. In the adult there is considerable variation in the size of the cells. The smaller hav- ing a minimum diameter of 40 ,, the larger of 70 », with an extreme length to the beginning of the axis cylinder of 150 ». The number of cells in a single fish is between 35 and 4o, and seems to be fairly constant regardless to the age or size of the fish. fig. 4. Five Giant Ganglion Cells showing variation in form and branch- ing of the neurites. The form of these giant cells, though always characteristic, is very variable. In young specimens of Ctenolabrus 3 cm. in length the cells show much greater regularity in form and dis- tribution than in the adult. In the anterior part of the cord they lie closely together, the intervals increasing regularly pos- teriorly. The cells are usually rounded, but anteriorly mutual pressure may give them a somewhat angular outline. Oc- casionally in the smallest specimens examined, the cells are uniformly dorso-ventrally flattened, approaching a discoid or lenticular form. In the adult the simplest form is approximately spherical (Fig. 5), but this form grades off to the pyriform which is the most typical (Fig. Ic, Fig. 4 4). The tapering end is ventral and from it comes off the axis-cylinder. This may pass off PORE a — SARGENT, Giant Ganghon Cells of Ctenolabrus. 189 from the cell abruptly as the stem from a pear, or the cell may gradually taper out into the axis-cylinder. In their variation in form the cells may approach the oval, the conical, the dis- coid, club-shape, or they may be irregular. One interesting variation in form assumed is shown in Fig. 4 ¢, where the cell is apparently drawn out into two parts. Every gradation may be observed from the gradually tapering cell (0) through forms like d and ¢, to the apparently double cell e. Numerous dendrites are given off from the cells, varying from the finest filaments to processes of considerable size. They are given off most freely from the dorsal part of the cell, and as a rule do not greatly influence the outline, but particularly in the anterior bilaterally placed cells they are occasionally so large as to give the cell a multipolar appearance. These den- drites branching freely pass through the open space of the capsule surrounding the cell (Fig. 5), and interlace and anasto- mose with the surrounding neuroglia cells, forming thus a direct protoplasmic connection between the giant ganglion cell and the neuroglia. In some few cases observed there is apparently a direct anastomosis of the dendrites of the adjacent ganglion cells. Nucleus.—The internal structure of the cell is peculiar and characteristic. The nucleus is abnormally large nearly filling the cell and having in general much the same outline as the cell itself (Fig. 5). The nucleus is eccentrically placed usually crowded close up to the dorsal wail of the cell so that occasion- ally the cytoplasm can with difficulty be distinguished between the nucleus and the cell wall at that point. The chromatin net- work can be seen distinctly in iron hematoxylin, and Ehrlich’s hematoxylin preparations, extending uniformly through the ‘cell, (Fig. 5, ch. 2). The nucleolus is large, oval or spheroidal, usually lying eccentrically in the upper part of the nucleus. (Fig. 5, 2//). It takes most stains deeply, but nuclear stains like Ehrlich’s hematoxylin leave it transparent. Rarely a second nucleolus of smaller size may be seen. The nucleolus contains from eight to twelve spherical granules which stain deeply with iron 190 JOURNAL OF COMPARATIVE NEUROLOGY. hematoxylin, but remain transparent and highly refractive in preparations double stained with Ehrlich’s hematoxylin and Congo red. Fig. 5. Diagram of Giant Ganglion Cell and its surrounding capsule. w/., nucleus; #//., nucleolus; ch.z., chromatin network partly drawn in; é.f., membrana prima, .f., neuroglia fibre; .7., neuroglia net; a@x., axis cylinder. Immediately around the nucleolus the karyoplasm often stains less deeply than in its more peripheral parts. This may be attributed to the karyoplasm in that region being greatly vacuolated. This lightly staining region varies greatly in size, and the definiteness of its outline. The chromatin net show- ing faintly may be followed from the lighter area to the darker area showing the continuity of the karyoplasm. Usually this area is relatively small and its limits indistinct, the denser karyoplasm gradually becoming lighter toward the nu- cleolus. Sometimes it is entirely absent, karyoplasm being ho- mogeneous throughout. This lighter area is often of peculiar and varied shape, sometimes crescentic occasionally sending out forked tongues toward the periphery of the nucleus. In some preparations what corresponds to this area is an empty space, and the karyoplasm can be seen to have shrunken away from the nucleolus at one side leaving a crescent-shaped’ space. SARGENT, Grant Ganglion Cells of Ctenolabrus. Ig! From the examination of a single cell of this kind a quite dif- ferent interpretation would be possible, namely that the lighter area is the nucleus and the darker substance a differentiated cytoplasm aggregated about the nucleus. The examination of “several hundred cells preserved and stained by a variety of methods shows that this is not the case. The cytoplasm has the characteristic shining appearance of a highly refractive substance. Under a 1-12 in. oil im- mersion it shows a finely granular structure. The chromoph- ilic granules are elongated and lie with their long axes parallel and concentric with the cell wall (Fig. 5). They are most con- spicuous in the dorsal and larger end of the cell, gradually fading out toward the point from which the axis cylinder comes off. The cytoplasm lies principally in the lower part of the cell, but usually may be seen to extend around the periphery of the cell. In cells having the form of those in Fig. 4 the nucleus is approximately spherical and lies in the upper part of the cell, the cytoplasm having the appearance of having been crowded downward. Neurites.—As has already been stated, the cells are in general unipolar giving off a large neurite which passes ventrally into the cord (Fig. 3). The course of the neurite may be directly ventrad, or obliquely inclined cephalad or caudad, or again it may run horizontally near the surface for a distance of five or six diameters of the cell before passing downward into the cord. Rarely a neurite is seen to pass out laterally from the cell and become lost in the grey substance. The neurites of anterior bilaterally placed cells run caudad near the surface of the cord for some distance, then curving ventrad, laterad, and cephalad pass forward through the fiber bundle. The neurite having passed down one-half or two-thirds the distance to the canalis centralis curves gradually either to the right or left, sometimes dividing and finally enters or sends one of the two branches into the lateral bundle made up of similar fibers, The neurites pass alternately to the right or left, but this does not 192 JouRNAL oF ComPARATIVE NEUROLOGY, hold strictly, sometimes several successive cells sending their neurites to the same side. Entering the lateral bundle the neurite may pass either cephalad or caudad (Figs. 1 and 2). Dahlgren ’97 finds that in the order Heterosomata the neurites a// run caudad. In exam- ining upward of three hundred cells in which the neurites were followed into the bundle, approximately one-third were found to send the neurite through the bundle caudad, the other two- thirds cephalad. This harmonizes with results obtained by counting the number of fibers in the bundle at different parts of its course, which shows that the majority of the neurites run cephalad. In approximately two-thirds of the number of cells exam- ined the neurite was found to divide into axis cylinders of equal diameter. In the other one-third no such branching could be seen. This may sometimes have been due to the imperfection of the preparations, but in a few instances at least it would seem that the neurites do not divide. There is a remarkable and interesting variation in the man- ner of this division. In the most common type (Fig. 40) the neurite passes ventrad and laterad nearly to the level of the bundle and then splits into two equal axis cylinders at least one of which enters the bundle and passes through its entire course. The division may take place higher up near to the cell, the two branches diverging as they ‘pass downward (Fig. 4,¢ and e). Or the division may occur so high up that the two processes come directly from the cell (Fig. 4,@). Ina few cases the di- vision was observed to take place after the neurite had entered the bundle, the two resulting processes continuing parallel for some distance. The axis cylinder stains deeply with iron hematoxylin, Kenyon’s or Sahli’s method, and is uniformly stained through- out its length. Frequently however the initial part of the fiber immediately adjacent to the cell takes the stain but lightly, the protoplasm of that part of the neurite staining precisely like the cytoplasm of the cell, with which it seems to be continuous and identical. The axis cylinders are throughout their course SARGENT, Grant Ganglion Cells of Ctenolabrus. 193 unmedullated, but Schwann’s sheath is present showing the char- acteristic nuclei. The neurites form two distinct and characteristic fiber bundles lying symmetrically on either side of the cord lateral and dorsal to the canalis centralis (Fig. 3,2 f. 6.) They may be distinguished throughout their course from other adjacent fibers by three characteristics, (1) the absence of a medullary sheath ; (2) their large size; (3) their aggregation into a char- acteristic bundle. Each bundle in the region of the medulla consists of from nine to twelve fibers. At the posterior limit of the series of giant cells in the cord the bundles consist of four or five fibers. The number increases cephalad of this point as the neurites enter the bundles. In their course through the cord the fibers lie within the dorsal horn of the grey substance close to its lower limit. The fibers here are loosely aggregated having a somewhat un- dulating course. In the medulla the bundles rise to the level of the floor of the fourth ventricle, and at the same time curve laterad. Forward of this they again become depressed. In the medulla the fibers are closely pressed together, so that in cross section each fiber has a more or less sharply polygonal outline. In the medulla the bundles are often abruptly deviated from their direct path in passing around the deep roots of the cranial nerves. In the region of the fifth cranial nerve the fiber bun- dles curve laterad and ventrad and pass out through the ventral root of this nerve. The fibers have been traced out zzto the nerve, and have been traced through their course in a consid- erable number of series cut in the frontal, sagittal and trans- verse planes. The course and ending of the fiber bundles posteriorly yet remains to be worked out. That branch of the neurite which does not enter the lateral bundle is difficult to follow, owing in part to the peculiar filiform neuroglia structures in that part of the cord, which strongly resemble the non-medullated fiber of the giant ganglion cells. The evidence derived from the study of many preparations indicates that this branch turns laterad 194 JOURNAL OF COMPARATIVE NEUROLOGY. and dorsad and branching finely becomes lost in the network of the dorsal horn of the grey substance (Fig. 3, @). SUMMARY. In the anterior third of the spinal cord of Ctenolabrus there is a series of from 35 to 40 giant cells lying in the dorsal fissure, each cell within a capsule. At the anterior end of this series and near the posterior edge of the fissura rhomboidalis there are two pairs of giant cells lying bilaterally near the surface of the cord. The form of the cells is variable. Numerous dendrites are given off which anastomose with the surrounding neuroglia cells. The cytoplasm contains elongated chromophylic granules arranged concentrically with the wall of the cell. Each cell gives off an axis cylinder which runs ventrad and laterad usually dividing into two neurites of equal size, one of which enters the lateral fiber bundle. This neurite follows the fiber bundle through the cord either cephalad or caudad. The fiber bundles passing forward through the cord and medulla pass out through the ventral root of the Trigeminus nerve. The other branch apparently divides and becomes lost in the network of the dorsal horn. Harvard University, Cambridge, U. S. A., July 25, 1898. BARDEEN, Variations in the Lumbar Plexus. 195 On VARIATIONS IN THE DISTRIBUTION OF THE SPINAL NERVES ENTERING THE LUMBAR PLEXUS. By C. R. BARDEEN, The Johns Hopkins University, Baltimore. Special attention has been given during the last two or three years in the Anatomical Department of the Johns Hop- kins University to the origin and distribution of the main nerve trunks in the human body. The students in practical anatomy have been encouraged to note carefully any variations from the usual descriptions of the courses of the various nerves and to record the results of their observations on printed ‘‘ tabulation charts” furnished them. Besides the filling in of the printed charts, diagrammatic drawings have been freely made as a fur- ther means of illustration. The work has been verified and the charts have been carefully controlled by Dr. A. W. Elting dur- ing the early part of the undertaking, during the sessions of 1896-1897, and by myself during the past year. In these records sex, color, apparent age, and marked peculiarities of bodily structure have been carefully noted as well as the distribution of the larger nerves. It is hoped that from these charts interesting and valuable statistics relating to the distribution of the peripheral nerves may be obtained. There is a very great variety in the modes of origin and distri- bution of the peripheral nerve trunks in certain regions of the body. This is well illustrated by the conditions found in the nerves entering into and leaving the lumbar plexus. To a con- sideration of these nerves I invite your attention to-day. In man the lumbar plexus is formed as follows: The eleventh thoracic nerve has in the main the character- istics of a typical segmental nerve. It divides into a dorsal primary division for the supply of the skin and muscles of the back and a primary ventral division for the supply of the ven- tro-lateral musculature and skin, and gives off a visceral branch, 196 JOURNAL OF COMPARATIVE NEUROLOGY. or branches, which enter the sympathetic system. The ventral division has two main branches, a lateral, piercing the muscles at the side and giving off dorsal and ventral branches, and a ventral, extending well towards the median line in front, run- ning for the most part between the muscles, but finally piercing them and supplying the skin in front. This ventral branch of the eleventh nerve usually becomes sub-cutaneous in a region somewhat below the umbilicus. It usually anastomoses freely with similar branches from the tenth and twelfth nerves. In addition, the primary ventral division of the eleventh nerve may give off near the spinal column a branch of commu- nication to the corresponding branch of the twelfth thoracic nerve. , The twelfth thoracic nerve likewise has the characteristics of a segmental nerve, supplying the region immediately caudal to that supplied by the eleventh nerve. The lateral branch usu- ally extends well down over the middle third of the iliac crest. In addition to the ventro-lateral nerve, or in place thereof, the ventral division may give rise, usually near the spinal column, to the zeo-hypogastric nerve. This nerve runs to the ventral third of the iliac crest, there divides and sends an zac branch over the lateral hip region and a hypogastric branch to the ab- domen just above the pubis. This nerve bears varied and often very intimate relations to the ventro-lateral nerves. More rarely origin is given to the z/co-znguznal nerve. This nerve has a course very similar to that of the ileo-hypogastric but supplies a more ventral region of the hip, and the skin of the pubic re- gion and root of the penis and of the scrotum and the adjacent part of the thigh. There are often anastomoses between the ileo-inguinal and the ileo-hypogastric nerves. The iliac branch of the former is often wanting. The first lumbar nerve has no true ventro-lateral branch. Its place is usually supplied by the ileo-hypogastric and ileo- inguinal nerves which have many of the characteristics of ven- tro-lateral nerves. In addition, the first lumbar usually gives rise to the genzito-crural nerve or to one of its branches. This nerve, sometimes with separate genital and crural branches, runs — ae BARDEEN, Variations in the Lumbar Plexus. 197 down through the psoas muscle and across to the pubis. The crural branch supplies the skin of the upper leg just below Poupart’s ligament, the geuzfal branch passing down the in- guinal canal and anastomosing with the inguinal nerve supplies a region similar to that innervated by the latter. The extent of distribution of the genital is inversely proportional to that of the inguinal. There is usually a branch of communication rom the ven- tral primary division of the twelfth thoracic nerve to that of the first lumbar. A similar branch from the first lumbar to the second is even more constant. There may be a similar connec- tion between the second and third and between the third and fourth lumbar nerves. There is thus a sort of ‘‘collector”’ nerve formed which may run from the eleventh thoracic to the fourth lumbar nerve. So long a course however is rare. In- deed it is doubtful if nerve fibers often pass thus over more than two segments. Certainly the majority usually enter the nerves arising from the segment immediately below. From this ‘‘collector’’ nerve and from branches springing directly from the second, third, fourth and rarely the fifth lum- bar nerves arise the obturator and the anterior crural nerves. The obturator arises from the ventral surfaces of these nerves. Passing through the obturator foramen it supplies the adductor muscles of the leg and some of the skin of the inner thigh. The anterior crural nerve arises mainly from the dorsal sur- faces of the spinal nerves and passing over the iliac crest sup- plies the extensor muscles of the thigh and the skin of the outer, middle and inner ventral regions of the upper leg and of the inner side of the lower leg. The nerve supply- ing the outer side of the thigh (the external cutaneous) usually arises separately from the main trunk; the other cutaneous nerves of the thigh may do so. Sometimes an ‘‘accessory obturator”? runs over the pubic crest to supply the pectineus muscle and anastomose with the obturator, b , 198 JoURNAL OF COMPARATIVE NEUROLOGY. The variation in the relation of these different nerves to the spinal nerves is expressed in the following tables :! TABLE OF DISTRIBUTION OF SPINAL NERVES. This table gives a summary of the distribution of the ven- tral primary divisions of the various spinal nerves entering the lumbar plexus. ‘‘Comm. br.”’ indicates a proximal communi- cating branch between the given spinal nerve and the next be- low. The percentage sign refers to the ratio of the number of the plexuses in which the given condition is found to the total number of plexuses (122). The names of the various peripheral nerves indicate adzvect origin from the given spinal nerve. ‘‘Types of distribution” refers to the various distinct modes by which the constituent elements of the spinal nerve have been found distributed. rith Thoracic. 2d Lumbar. Comm. br. each side_-__.__- 1 body Commi:) branehes-222 25-2336" 538.2% ES “* one side only--- 4 bodies Genito-crupalies-22-- 222 seen 72.9% ef Sito bal ss oc esos 4,9% External cutaneous-_----.--. 54.9% Types of distribution.__-. 4 ANTETION renural Mee oe ne SD 46.7% Obluratons a.) ese ee aeee 39.3% rath Thoractc. Types of distribution.___._- 50 3@ Lumbar. Both wee aes eee es A% Anterior cnurallt-2. sa 22" 100 % Tleo-inguinal - : 5.7% Obturatorn = srs eae 100 % Types of distribution.___.. 9 Accessory obturator_....._.. 5.7% Types of distribution._.. -- 12 13th Thoracic. Found in one body. 4th Lumbar. Comm. br. on one side. Comm. br. sacral plexus.... 93.4% Ileo-hypogastric on each side. ANteMlOn (Cruiraliueseeeoneeee 98.4% Tleo-inguinal on one side. Obturator 2.26 ee ak 96.8% Accessory obturator .-_- 5.7% rst Lumbar. Types of distribution._..___. 7 Comm. branch___._....__.- 85 % Tleo-inguinal_.-. 2.2.2.2. 86 f 5th Lumbar. Tleo-hypogastric _...._.._- 60 6% Anterior crural each side__. 2 bodies Lleo-hypog., no comm, 12d 27.8% ae “e 1 side only. 5 Genito-crural -.-._-_.__.:- 47.5% Anterior crural, total_______ 7.4% Types of distribution. _._- 26 Obturator each side__..___.- 1 body s 1 side only ___-__- 6 bedies Obtuyator, totale ==. 20 s2- 6.5% 5th lumbar in lumbar plexus 8 bodies 1 These statistics are based upon a study of but a part of the tabulation charts. They are based upon the conditions found in sixty-one bodies, or 122 lumbar plexuses. It is probable that when all of the charts have been examined the figures given will need some alteration, as expressive of the lumbar plexus. The tables, extended so asto include tabulations from a greater number of charts, will appear in amore extended article on the lumbo-sacral plexus, in the preparation of which Dr. A. W. Elting and myself are at present engaged. a eo a. BARDEEN, Variations in the Lumbar Plexus. 199 The most notable thing brought out in this table is the great variety in the distribution of those spinal nerves which supply the cutaneous nerves distributed to the regions where the thigh and abdomen meet. There is nothing especially remarkable about the distribu- tion of the eleventh thoracic nerve. It is interesting to note how- ever that in one of the cases in which there was a communica- tion between the eleventh and twelfth thoracic nerves, a thir- teenth thoracic nerve was found. It is to be noted that the ileo-hypogastric nerve arises in over 40% of the plexuses direct- ly from the twelfth thoracic. In the text-books it is said to arise from the first lumbar nerve. The first lumbar nerve is essentially the source of supply of the ileo-inguinal nerve. In a little over 25% of the specimens it alone gives rise to the ileo-hypogastric. It is intimately connected with the genito- crural and seems to be the main source of cutaneous supply to the border-land between the thigh, the genitalia and the ab- domen. While the main strength of the second lumbar is given to the obturator and the anterior crural nerves, either directly or through the communicating branch, it also contributes largely to the external cutaneous, and to the genito-crural. The great variety of the paths by which this nerve is distributed is re- markable. Of the third lumbar nothing special need be said. The fourth lumbar as a rule is divided between the two great nerves of the lumbar plexus, and the sacral plexus. In thir- teen per cent. of these plexuses it was distributed wholly to the former nerves. The fifth nerve as a rule belongs wholly to the sacral plexus. For the sake of simplicity no account has been taken in these tables of the small muscular nerves derived directly from the spinal nerves, of the relative size of the nerves, or of the peripheral distribution of the main nerve trunks. 200 JOURNAL OF CoMPARATIVE NEUROLOGY. TABLE II. ORIGIN OF THE NERVES FROM LUMBAR PLEXUS. xI X11 X11 z >. ie be. 9 6 I II Ill Iv 6A Tleo-hypogast’c | 1 1 Ant. Crural}_._-| ¢ 1 1 1 i ieee e 1 c ¢ e 1 1 1 are Sesh ao See e 21. 21 21 ae STAN WS Mt Eye ra MLE ley A TY A ee Te eee S| fae cged | a eb ema be aS Re c 21 21 21 ane ral let? SAENG e 1 1 1 e OMe) | Rt BR 2 Pa SEN OS oe | eg Se | eee Sm Rey | fererre| te Svs, fe Lead 10 10 10 ae e Cre ale Serer eA AN eS ee ES See te te TE Se eee see 2 2 2 a Nee ees iy c cd 27 27 eae. 4, Seen G c e 1 ey (eee re ee Fe e e 6 6 6 vee cei c e 31 31 ty Jes —— | —— , |! —— _ ——— | — 36.9% | 100% | 100% |98.4% Tleo-inguinal __| ¢ 1 Fo pe we eS LS) ER a (ES Ste 9 3g Bt an TE II HTX Ev: Vv Rene I} fia a 2 Sahel [PSY Dee Be) fee 2 | SS) | | | ty Ae el RE PE ae at | rhe vial U's Ses Obturator_.]--.-| ¢ 1 1 1 1 mig c Com st 1 hae gs ea a Boab veh | (Re at C e e 2 2 2 EE T= Se Eli Ome ais E Se) See ees pee ee es Pat e 20 20 20 fs 35 Si3e.5) eae fen BO? Cee a eee ae ES eS NOK 1 1 eee |e Deh ea ee One e 6 By Poke | Lee BTS ES |) 22 22 22 Ey) Inguinalionly-<|o58|jor) [oe Tate sense ste Shee Sap yy es ee ee co ee 2 2 2 eas BSA) Heol (eee Pe oe ee ae ae) | (rk eet dg Ey || fete Pee QO e ty jee ee rere SNe MPAA eee ec Pd |S ie meee ares soe ec jc e e 3 3 = Missing 4 times]____] 50% |_____. ONO Meena | Berle s bs eke c c 21 21° | eae —— | —— } -——— | ——- | ——_|—__|—___||——__itee.-}---- c c 20 20 ae. 5.4 Oy [tee @ Ga be bw I Il 7h OG (cca 1 TM Wh Re nui py Pea ES e ¢ 4 4 + —| ——_ | — —-| ——_ | ——_ Ne | ee aT ery cen | PARE Ue etal [eae a e 11 11 as BE FS RNR fees c 2 2 2 Genito-crural._| ¢ Gir eee Dively vue eet Ss | Ton SPY fees ae of 2 ee 3 3. ae Se Me Our Pree oe PO) Were eel ste PEE | ie erred tea em! eA) NS Be Oe Ze Say ie Sabha 2 2 2 Genital only. ple ees e fs RE ED de Ny HAAS ost, Beal hire e ¢ c 4 ee Crural only_-|____|.¢ f e e 2 Speen Ree ees We ia 5 e ce 1 Cenisdlionly jos lee We 1 Crural‘onlye 2h) o. juss ahi. 1 90.1% | 100% |98.4% Pee ele eh ee 12 2 — —|——__] — Genitalvonlys 224) Veo | itech 2 2 ll 1II Iv lv RD) jes Sepa ps Se Sas 11 1 —— | | —— |} ——_ | —_— | —_] —_. e CE eee e 2 c tt: | eee H EA Cy Neeees e ZUR | ee nee erst a a2 Siie c ec 1 1: ss Genitalonly <0 9s) elias e Dip fy Sa eS Py oe em cee e e 1, oe ES | be ae | RE e 27 eee Lema PE ped Pee al Pyre 1 ES) FS Memory only) re as O72 aye ole ye e FO FE TOA mE | La Sein | Ae re BS Ee |e 1 J) eee Seer [abe ae] Ne eS eee 2 RSE ay Exeie pee (eee ge | a tes gel IVES dp, 2 aus dia Eee fae fray eee ee c 1 1 SS | mre ated iL aml (2 eclnd) [Berea| Bet ahee) [arse SS, | ke i 4) eee Missing ONCE Sate Sees alee ee 98.4% CE | Egmee ps See: EKG UL | CRIT I II m1 | Iv Ext. cutaneous.|____| e@ |______ TT aap) etter Oe | [OE a HE A BAD: Pere 2 | Sak 1 Sa ee [ome Pres 7 paw eee 1 EN (eee ee c CR amour y e We ie se Sa oe Se CG Meta ere e CN} 4 (PS S| (Se Cowal Soe acl Cree « 28 eee eA ee de ers | fe | Reece! ob ALS 3 3 RE PAS: iets 3 e 5 5 oe Pes Noosecelanees e 8 8 eee ees ee Oe | Ce ea 3 3 Ves he re (ia Ce | |e c 4 13 Bare pee) (eee re ee an c e 5 Seer ES Re 5 lee) MSE EO eer) Bees 3 mae 2, Ree |e a) (eee e c 1 1 From Ant. Cru- Wal '8i times oi4s- oslo ssc ue kee 54% 170.4% 131.1% I---- In Table II the relation of the nerves named in the column at the left to the spinal nerves in the various plexuses is pointed out. Each horizontal line represents the relations found in the ie BARDEEN, Varzations in the Lumbar Plexus. 201 number of plexuses indicated by the numerals. The numbers are placed under Roman numerals representing the spinal nerves which give rise to direct branches to the peripheral nerves in question. The ‘‘c” represents a branch of communication between two spinal nerves. . Thus *! X! indicates that the ileo-hypogastric arose in one case from the twelfth thoracic nerve when there was a proximal branch connecting the eleventh with the twelfth thoracic nerve. The percentage at the bottom of some of the columns indicates the ratio between the number of plexuses in which the given spinal nerve is distributed to the peripheral nerve in question and the total number of plexuses examined. In making the percentages it is not assumed that a given spinal nerve sends fibers more than one segment below through the communicating branches. There is little need of an extended review of the conditions tabulated. The nature and variety of the relations of the nerves leaving the plexus to the spinal nerves forming it is shown in the tables. Eisler, Paterson, and others who have examined the lum- bo-sacral plexus in man and other animals have pointed out that the plexus as a whole may vary considerably in relation to the spinal nerves composing it. It may present a very high form in which a given spinal nerve plays the role usually played by the one just below or a very low form in which the reverse is true. The plexus may vary anywhere between these two ex- tremes. They have also mentioned that the various parts of the plexus may vary as well as the plexus asa whole. Not enough stress has been put upon this latter point however. Thus for instance one might expect, from the association of the eleventh thoracic with the plexus, the existence of a general ‘“‘high” form, yet this was the case to a marked degree in none of the six plexuses into which the eleventh nerve en- tered. The entrance of the fifth nerve into the lumbar plexus likewise was associated with a general marked low form in but two of the eight bodies in which the condition was found, 202 JOURNAL OF CoMPARATIVE NEUROLOGY. NoTES ON THE PERIPHERAL NERvouS SYSTEM OF MOLGULA MANHATTENSIS. By G. W. Hunter, Jr. These observations were made during the summer of 1897, at the Marine Biological Laboratory. The forms used were Molgula manhattensis and Cynthia partita (Verrill). The former was the more productive of results owing the ease with which it could be freed from its tunic. Young specimens, on account of the clearness of their tissues, could be used as whole mounts. Two methods were used for the staining of the peripheral system. Molgule were immersed in a weak solution of methy- lene blue (1-1000) for 1% to 2% hours and left in running water a few moments before dissection. A slight exposure to the air seemed favorable to the stain. Injection of a strong so- lution of (B. X. Meyer) methylene blue (1 to 4%) also gave very favorable results. In the latter case the blue was injected into the ovarian vein, from whence it reached the heart and was dis- tributed over the body. After one hour specimens thus in- jected frequently showed the whole peripheral system sharply defined. (1). Sensory cells in the endostyle. The endosyle in Molgula resembles that organ in the other Tunicates. At the bottom of the endostylar groove is found a row of flagellated cells, next come two laterally placed pads of gland cells, each pad divided by a row of deeply stained, closely packed spindle cells, the nuclei of which lie at different levels. These cells possess very short spike-like cilia and stain deeply with hematoxylin. Outside the glandular epithelium are found ciliated cells; the whole apparatus being bordered with a lip lined with cubical ciliated epithelium, which seems to be contin- uous with that of the peri-pharyngeal bands. The sensory cells are found in the lateral position occupied © by the above mentioned deeply staining spindle shaped cells. These cells are quite numerous, sometimes several hundred ap- Hunter, Nervous System of Molgula. 203 pearing at once under a fairly high magnification. They do not appear to be grouped ina regular manner, although they seem to be quite evenly distributed the length of the endostyle. They stop abruptly at the anterior end of the endostyle, not being found in the peri-pharyngeal bands; nor are they found in the anterior portion of the digestive tract proper. The bipolar sensory cell as stained with methylene blue is characterized by a distal knob or spike-like enlargement, one or more enlargements situated more proximally, one of which con- tains the nucleus, and a more or less conspicuous enlargement at the point where the nerve fibril leaves the basement membrane. This cell, however, assumes many other forms—presumably modifications—as may be seen by glance at figure 1. The nu- Fig. 2. Sensory and gland cells in the endostyle. Showing modification of the bipolar type of sensory cell, Camera drawing from several different speci- mens. I-12 1mm. Oc. 6 (Zeiss). clear enlargement may be situated very near the distal end of the cell or it may, on the other hand, have a basal position and lie close to the basement membrane. In one or two cases the nuclear enlargement appeared to show fine protoplasmic branchings. The proximal continuations of the cells after leaving the basement membrane turn sharply at right angles and run as many single loose fibrils up the endostyle. No very definite 204 JoURNAL OF COMPARATIVE NEUROLOGY. bundle of fibers is found but the fibrils seem to form a felt or basket work under the epithelium. Here and there anastomoses are found. The ultimate distribution of the fibers in the cen- tral system has not been proven. Some evidence however, points to the fibers running under the peri-pharyngeal bands and entering the ganglion by that course. No supporting cells (Sttzellen of authors) are found. Many gland cells are stained. They usually take the stain less intensely than the sensory cells and are large and irregular in shape. The sensory cell in the endostyle of the Tunicates resem- bles the general type of sensory bipolar cell as found in Oligo- chaetes, Polychaetes, Crustacea, Mollusca, etc., and described by Allen, Bethe, Gilchrist, Retzius, van Gehuchten, von Len- hossék, vom Rath, etc. It, however, more closely resembles the sensory cell in the olfactory epithelium of Myxine (Retzius).. (2). Sensory endings in the branchial basket and peri-pharyn- geal bands. The outer lip of the endostyle, as well as the outer lip of the peri-pharyngeal bands, is covered with cubical ciliated epi- Fig. 2. Ciliated cells in the branchial basket. Cells selected by the blue and show endings on the basal side. Ammonium picrate and glycerine. I-12 imm. Oc. 6 (Zeiss). thelium. Certain of these cells are selected by the methylene blue and stand outa vivid blue with the cilia also stained. Close investigation often shows contact endings on such cells. Con- tact endings of the same character appear on the basal side of Hunter, Nervous System of Molgula. 205 many other cells not selected by the blue and frequently cells are stained blue to which no nerve supply can be demonstrated sO we cannot say anything definite with regards to such selec- tive staining as mentioned above. The same type of ciliated cells with a like type of ending is found in the branchial basket on the borders of the stigmata. (See Fig. 2.) The nerve fibers supplying the ciliated cells bordering the endostyle and peri-pharyngeal bands are applied rather closely beneath the epithelium, in many cases forming a true plexus of anastomosing fibers. It may be said however, that this plexus is limited to the fiber after it breaks up under the basement membrane to form the endings. The endings appear to be disk, cup or knob like. Some- times only one knob is found in contact with the base of a cell, sometimes several ; the trefoil ending was infrequently found. Frequently the endings appear to be free. (See Fig. 3.) Fig. 3. Nerve endings on the ciliated epithelium of the peri-pharyngeal band. Rounded surface seen in optical section, whole mount. 1-12 imm. Oc. 6 (Zeiss). In the branchial basket the fibers do not always remain closely associated with the basement membrane, but may follow the supporting tissue in the sinus-like interior of the branchial bars. (See Fig. 2.) Here the endings are not limited to cili- 206 JOURNAL OF CoMPARATIVE NEUROLOGY. ated cells, but are found on other cells, of a probably glandular and sensory nature. Fibrils applied closely to the basement membrane may be seen to end on the base of certain ciliated or mucous cells in disk shaped endings, then continue their course touching other cells in like manner before finally ending onacell. (See Fig. 3.) Such endings are described and fig- ured by Peabody in the ampullz of Selachians and Bethe in the gustatory epithelium of the frog. Since this work was begun I have in continuing my obser- vations during the summer of 1898, succeeded in demonstrating: (1). The sensory nature of the buccal tentacles and dis- tribution of nerves to the same. (2). The sensory nature (in part at least) of the ciliated funnel (dorsal tubercle). (3). Sensory papillae and sensory cells in the body epi- thelium. (4). Muscle endings in the body musculature. A later paper will contain further observations on the peri- pheral nervous system and its relation to the central system in Molgula and Cynthia. Aug. 25th, 1898. THE ELEMENTS OF THE CENTRAL NERVOUS SYSTEM OF THE NEMERTEANS. By Tuos. H. Monrcomery, Jr., Pu.D. (Lecturer in Zoology, University of Pennsylvania.) The studies forming the basis of this communication were published in the ‘‘Journal of Morphology,” Vol. 13, 1897. The genera Cevebratulus and Lineus were investigated, the elements of the nervous system in the American Metanemer- teans having been found, on account of their minute size, less favorable for study. Four types of ganglion cells were found, which, in accord with the nomenclature of Birger, may be referred to as cells Montcomery, Central Nervous System of Nemerteans. 207 I, 2, 3 and 4. Of these the 4th type, or neurochord cells, are present only in Cerebratulus. All these cells are membraneless, and in all the cytoplasm has a remarkable vacuolar structure, — strands of spongioplasm bounding large hyaloplasmic vacuoles. Such a remarkably vacuolar structure does not appear to be normal in any other group of animals. In cells 3 of Lzneus occur in the cytoplasm peculiar homogeneous, rounded bodies, which have no regular arrangement, and which are not compar- parable to the chromophilic granules (Nissl’s granules) of other forms: for them the term chromoplilic corpuscles was employed ; genetically they seem to be local condensations of the cyto- plasm. A large number of fixing reagents were employed, but after them all the axis cylinder process (I was unable to find the dendritic processes described by Birger) presented the appear- ance of a single nerve tubule, and did not contain any primitive fibrils ; that is to say, the outer (alveolar) spongioplasmic sheath of the cell body is continued distad to form the outer sheath of the axis cylinder, and the hyaloplasmic, structureless substance to form its core. Spongioplasmic strands may penetrate a short distance into the proximal end of the cell process, but this is an irregulur phenomenon, and such strands are not prolonged to form fibrils. Thus the minute clear spaces in the fibrous core of the central nervous system represent axis cylinders, and bundles of such; and the larger and more irregular clear spaces, lymphatic tracts. In the brain lobes and the lateral nerve chords the follow- ing connective tissue layers may be distinguished: the outer neurilemma, which is a capsule enveloping the ganglion cell layer; the inner neurilemma, a capsule separating these cells from the fibrous core (‘‘ dotted substance’’). This tissue is the same as that forming the connective tissue basement membranes of all the epithelia of the body, and is formed of branching cells with a dense intercellular substance (cf. Montgomery, Spengel’s ‘‘ Zool. Jahrb.” 10, 1897). A different tissue forms the neuroglia proper (‘‘ Hillgewebe,”’ Birger). This neuroglia consists of branching cells, without any intercellular substance. Within the ganglion cell layer its elements envelope with their 208 JOURNAL OF COMPARATIVE NEUROLOGY. branching fibers the ganglion cells, forming a loose fibrillar cap- sule around each of the latter; continuations of these fibers along the axis cylinder produce a sheath of Schwann. Another layer of neuroglia cells lie just on the periphery of the fibrous core, and send their branches into the latter. The only fibrils within the nervous system are those of the neuroglia and hence Birger, who described the axis cylinder as a dense fibril, prob- ably mistook the neuroglia fibrils for axis cylinders. Ganglion cell I occurs only in the dorsal lobe of the brain ; it is the smallest and probably sensory. Ganglion cell 2 occurs on the ventral surface of the ventral brain lobe, as well as along the lateral nerve chords. These cells are arranged in radial clusters; and in the lateral chords these clusters have a regular alternating arrangement. Ganglion cell 3 occurs on the median surface of the brain lobes, and, more sparingly, in the lateral chords. These are large, pyriform cells, and it is on them that the structure of the axis cylinder may be best determined. Ganglion cells 4, or the neurochord cells, have been found by me only in Cerebratulus, but by Birger also in Langza, Pros- adenoporus and Drepanophorus. (Birger found in the two latter genera only a single pair of these cells, situated in the brain; and in Langza and Cerebratulus one pair in the brain, and a large number along the lateral nerve chords). In Cerebratulus lacteus I found the following distribution of these giant cells. There are 3 pairs in the ventral lobe of the brain, situated one behind the other. On the nerve chords there are none in the oeso- phageal region ; behind the latter region they are found again, and become more numerous towards the distal end of the chords; near the proximal end of the chord they are more nu- merous on the dorsal surface of the chord, but distally this po- sition is reversed. They are arranged on the chords without regularity, and there is no symmetry in the arrangement on the two surfaces’ of the same chord, nor on the two sides of the body. Zones where they are relatively numerous alternate with those where they are relatively scarce. About four-fifths of a worm six inches in legth was serially sectioned, giving the "> SHEARER, Verve Terminations in the Selachian Cornea. 209 following figures in regard to the arrangement of these cells in the nerve chords: - Right Chord Left Chord Dorsal Ventral Dorsal Ventral 68 16 55 20 Total : 84 76 The axis cylinders of the neurochord cells all pass distad in the nerve chords, divide dichotomously in their course, and at certain distances show constrictions. ” In regard to the ‘‘gross anatomy” the following discov- eries may be noted: in Cerebratulus lacteus and Lineus gesseren- sts there is a second commissure uniting the ventral brain lobes, and in an European species of the latter genus, in addition to the second, a third. There are three commissures of the oeso- phageal nerves in Cerebratulus, and four in Leneus. On THE NERVE TERMINATIONS IN THE SELACHIAN CORNEA. By CRESSWELL SHEARER, University of McGill, Montreal. Throughout the vertebrate body there is hardly an organ whose innervation has received so much attention as that of the cornea. Time and again as new neurological methods have been discovered they have been applied to the study and re-study of the corneal nerve endings. Ever since 1866 when Cohenheim (1) published his celebrated paper on the termination of sensory nerves in the cornea, as demonstrated by him with his gold chloride method, down to the more recent researches of Dogiel (2 and 3) with methylen blue, an innumerable number of pa- pers have appeared. Despite the fact, however, that so much has been done and said on this subject, little is really known about these nerve ter- minations in the cornea of vertebrates lower than amphibians, 210 JOURNAL OF COMPARATIVE NEUROLOGY. and it is surprising to find on what few types our knowledge rests, the frog, rabbit and human subject being the regular stand-bys. As nothing to my knowledge had been done on se- lachians I thought it might be worth while studying the condi- tions there presented. My results soon gave me reason to believe that this hope had not been misplaced and that the ter- mination of the sensory nerves in the selachian cornea was evidently different from that of amphibians and mammals ac- cording to the classical researches of Hoyer, Arnold, Izquierdo, Klein, Kolliker, Dogiel, and so well worthy of further study. The following remarks apply to some short and very imperfect work which I have done on the subject within the last few months. The material I have used mostly was from the ordi- nary form of ‘‘ Smooth dog fish”’ (Mustelus canis) so abundant here, although I have secured the corneas from the following sharks occasionally : Galeocerdo tigrinus, Carcharhinus obscurus, Sphyrna zygaena, Carcharias littoralis, for all of which the fol- lowing results also hold. The methylen blue method of staining was adopted and so far I have used it only. The particular modification of the blue method used was that recommended by Dogiel (3). Apathy’s (4) fixation also gave good results, but for thorough action Dogiel’s fixation is more to be depended upon. Bethe fixation (5) has been used for sections, but on account of the trouble experienced in cutting I have had few results with it, for the cornea tissue proper so hardens in the usual processes of embedding that it is nearly impossible to cut it. At first I had trouble with my fixing fluids in that they caused too much maceration. This was stopped by adding a few drops of 1-10 per cent. solution osmic acid, which was not enough to blacken the tissues, and so render them obscure. A few words as to the general histological structure of the cornea in Selachians. The anterior corneal epithelium is somewhat thicker al- together than that of the cornea stroma proper, composed of large cells, having centrally placed rounded nuclei. The epi- thelium is on an average 12 to 18 layers of cells deep, the su- SHEARER, Nerve Terminations in the Selachian Cornea. 211 perficial cells having the usual flattened scale-like appearance, the layer next the corneal substance proper tall cubical column- ar, and the cells of the middle layers present the well known ‘‘prickle”’ appearance. There appears to be no membrane of Bowman or of Descemet; the posterior epithelium consists of a single layer of cells. The cornea substance proper presents the usual clear laminated appearance composed of about 12-14 sheets with corneal cells and lymph canals. a an . . Spe tes Deg fig. z. Methylen blue preparation from the dog-fish cornea show- ing straight unbranching inter-epithelial fibers with dark bodies 6.8 mm, obj. comp. oc. 4, Zeiss, Camera. Examining one of these corneas properly stained and fixed one is struck with the great number of nerves present, their relatively straight course from the border of the cornea inwards towards the center, and their unbranching course. It is surprising to find how long some of the fibers are, going apparently in some cases right across the whole cornea. Again one notices the very regular distances they keep from one another, always more or less parallel, looking under a low power of the microscope like a series of ruled lines. This condition is very different from what Klein describes in the frog where the nerves run very irregularly, crossing one another sometimes nearly at right angles. Some of these fibers give off lateral branches which cannot be followed for any distance 212 JOURNAL OF CoMPARATIVE NEUROLOGY. and soon become lost. These lateral branches do not anastomose with one another, as can be seen from the magnified camera drawing of Fig. 3, 6. They suggest small fibrils com- ing off to form a network, but the closest examination does not show this to be the case, and I am pretty certain that a true inter-epithelial plexus is wanting in Selachians. The unbranched condition of these nerve fibrils is perhaps made more striking by comparing Fig. 2 with Fig. 1, which represents a similar preparation from the cornea of one of the osseus fishes (Prio- notus carolinus) where the irregular joining and course of the fibers is apparent, besides in the cornea of the dog-fish the fibers are much larger and thicker. These nerves are covered with a Fig. 2. Methylen blue preparation from cornea Sea Robin (Prionotus car- olinus) Zeiss 8 mm. obj. Comp. oc. 4. , Camera, sheath which in some places pulls away from the fiber axis proper, leaving a clear space, at the other points swelling up. This sheath however does not in any way resemble the half medullated sheath which Dogiel (/. c.) has described as occur- ring on the corresponding nerves of the human cornea. But is most probably a result of the changes caused in the nerve fibre by the staining process, and which is always characteristically obtained when the blue method is used. SHEARER, Weve Terminations in the Selachian Cornea. 213 Where these nerves enter the epithelium around the cor- neal border small fibrils are often given off which may be traced for some distance winding in and out among pigment cells which are always collected there in considerable num- bers. Many of these fibrils enter into close relation with these cells, in every case they can be shown not to end on them although sometimes forming loops around them. Some of these pigment cells presented the appearances of contraction and expansion figured by Ballowitz, but no nerve endings as he describes in relation with them. lone the: Course of ‘the: nerves) Fig.)\ 1) 0,6; Fig: 3, -¢, and scattered throughout the field are seen dark staining bod- ies bearing processes looking like delicate nerves. These bod- ies will be seen to be of varying size and shape, in some places gathered together in clusters, in other places scattered and _ ir- regularly disposed. In some cases a nerve will send off a deli- cate fibril to one of these cells lying near to its course, in oth- ers to terminate directly in it, but generally passing on to an- other body further on. The processes coming off from these bodies which are shown in Fig. 3, d, d, wind in and out among the cells for a short distance and then become lost from view and indistinct; apparently not joining with one another. I cannot help thinking these bodies are similar to the bodies which Dogiel (2) describes as ending bodies in man. The fact that they have these processes however seems to be against this and one brings to mind the assertions of Inzani about special terminal ends situated amongst epithelial cells and which Klein (7) lays to imperfect specimens and bad technique. No bod- ies similar to Dogiel’s complicated ‘‘knaulchen’’ were met with. That these nerves and bodies are within and limited to the epithelium is easily demonstrated by transverse sections, and by macerating the epithelium off from the cornea substance proper. This method of maceration by over fixing is perhaps the best way to obtain good preparations of these nerves and bodies ; in some maceration preparations the epithelium which comes off in one piece becomes broken into several pieces by the 214 JOURNAL OF COMPARATIVE NEUROLOGY. pressure of the cover glass; these pieces will separate a little bit leaving a clear space between them. Across these spaces the nerves will be seen running from one piece to the other un- broken, showing their strength and elasticity. Some observers have described the nerves of the epithe- lium as giving off short hook-like branches which bend back Fig. 3. Methylen blue preparation of the inter-epithelial nerves of the Dog-fish under high magnification, showing dark bodies ¢, lateral fibers 4; ad, processes from the dark bodies. Zeiss homog. immers. Comp. oe. 4. Camera. and enter into relation with stromal plexus of the cornea tis- sue proper and it occurred to me that these short lateral branch- es (Fig. 3, 0) were of this nature; but after repeated examination I could not determine whether they did or not, but from the fact that I could not find branches running down between the deepest layer of cells towards the cornea tissue proper, I do not believe this to be the case. I have already stated I have been unable to find any true plexus or nervous network within SHEARER, Nerve Terminations in the Selachian Cornea. 215 the epithelium which could in any way answer to the various networks described by Klein (7). Klein states that the sub-epithelial network is situated beneath Bowman’s membrane and that he was able to remove the entire epithelium without disturbing this net-work. On examining a similar preparation from the dog-fish with the epithelium so removed no trace of this plexus is found, but in- stead we get a view of regular plexus of the corneal tissue proper (Fig. 4) so very different in appearance from the much larger nerves of the epithelium. This network, of which Fig. 4 is but a very poor represen- tation, is of the very finest texture, the fibers forming it being of the very finest in size, perfectly uniform throughout their course and at once to be distinguished from the nerves of the epithe- lium by the way in which they branch at right angles and their irregular course and all pretty much within the same plane. Every sheet of the corneal substance proper seems to have a special network of these fibers over it. When one network is within focus the the networks of the layers deeper can be faintly made out, and by properly adjusting up and down one can bring nearly all the networks into view one after the other, One pe- culiarity of these nerve fibers is their sharply granular appear- ance as if made up of a series of closely arranged dots one af- ter the other in a delicate strand. These fibers in branching and winding about amongst the corneal cells do not keep any definite relation with them, and it is needless to say no anasto- moses with them or their processes was to be distinguished. As Dogiel has found in man, the nerve fiber never comes into real relation with the cell but simply passes over it or along one border. On comparing Fig. 4, with Fig. 7 of Dogiel’s (2) pa- per the general resemblances of this network in selachians and man is very apparent. There is a tendency in selachians to greater regularity of branching of the fibers at right angles, they run in one direction for a certain distance then abruptly turn and run at right angles to their former course, while in man the change of direction is less sharp and sudden. 216 JOURNAL OF CoMmPARATIVE NEUROLOGY. As to the distribution of this network it seems to be uni- formly allover the surface of each lamella. No large trunks are ever seen to join it from the border and going to form it. The nerves composing it appear to be very continuous, where an ap- parent ending takes place the fiber seems to fade out in a manner which renders it impossible to tell whether it is a free ending or not. Fig. 4. Fine plexus of the cornea substance proper, methylene blue, show- ing the right angled courses of the fibers. Zeiss homog. immers. oc. 6. Camera. To sum up. The chief peculiarities presented by the nerves of the selachian cornea are: I. The relatively straight, thick, nerve trunks which run in the anterior epithelium and their parallel courses with rela- tion to one another. 2. The dark bodies into which these nerves run and some- times terminate. 3. The unbranching condition of these nerve fibres. 4. The lack of apparent relation between the nerves of the epithelium with those of the cornea substance proper and the lack of all nerve fibers in the cornea stroma proper simi- lar to these nerves of the epitehlium. GaLtoway, Budding in Dero vaga. 217 LITERATURE. 1, Cohnheim.—Virchow’s Archiv, Bd. xxxvili, 1867. 2. A.S. Dogiel.—Die Nerven der Cornea des Menschen. Anatomischer Anzeiger, No. 16 and 17, Jahrg. 18go. 3. A. S. Dogiel—Die Nervenendkérperschen (Endkolben W. Krause) in der Cornea und Conjunctina bulbi des Menschen. Archiv f. Mik. Anat., Bd. 37s) 60251801. 4. S. Apathy.—Zeit. f. Wiss. Mik. Bd. ix, P. 30, 1892. 5. Bethe.—Archiv f. Mik. Anat., xliv, 1895. 6. Emil Ballowitz.—Die Nervenendungen der Pigmentzellen. Zeit. f. Wissen. Zoologie, Bd. lvi, Hft. 4, P. 673, 1893. 7. E. Klein.—Termination of Nerves of Mammalian Cornea. Quart. Jour. Micr. Sci., Vol. XX, 1880, P. 464. SomE Nervous CHANGES ACCOMPANYING BUDDING IN DERO VAGA. By T. W. GALLoway. (Abstract from Paper on Non-Sexual Reproduction in Dero Vaga.') In Dero, as in several other Naidiform Oligochaeta, budding is a common method of reproduction. The process, occurring wholly within a single body segment, involves the formation, from the segment, of appropriate tail parts for the anterior zooid, and of the head structures of the posterior one. In the latter case the new formations consist of the prostomium and four cephalic segments, together with their contained structures: pharynx; ventral bristle-bundles ; muscles; blood-vessels; and nervous structures, namely, the brain, ventral cord, circum- cesophageal ring, and certain small visceral ganglia, sometimes denominated ‘‘sympathetic.’’ In the anterior individual, an anal or pavilion segment, bearing the branchial apparatus and digitiform appendages, and a preanal zone, which is concerned in the formation of new segments, are produced from the ante- rior portion of the bud-zone. 1Similar investigations have been made with essentially similar results, by Max y. Bock for Chaetogaster diaphanus, 218 JouRNAL oF CoMPARATIVE NEUROLOGY. In such a budding segment the central nervous system consists, at the beginning of the process, of the ventral nerve cord with the segmental ganglia. The cord is made up ofa cen- tral fibrous mass, surrounded in the ganglionic region by cells. These cells are more numerous on the ventral surface and on the lateral horns of the cord. (Figs. 1, 2, gv. a.) In the budding zone the ectoderm thickens, forming a girdle about the middle of the segment. Cell proliferations of fy. z. (Diagrammatic). Transverse section of Dero vaga before budding process begins. cd. /. /m., cells of the lateral line ; con't. crc’e@., circumcesopha- geal commissure; fdr, cols., giant fibers; gx. d., brain: en. v., sub-cesophageal ganglion; mz. d., dorsal longitudinal muscle band; mz. @’., detached portion of dorsal band ; mz. 7, lateral muscle band; mz. v., ventral muscle; mz. v’., detached portion of ventral band; zefk., nephridium; sy., pharynx; set. v., ventral bristle bundle; vas. sug., blood vessel; 7, 2, 7, regions where the ecto- derm gives rise to nervous elements. ectoderm break through the peritoneal lining, penetrating the muscular layers, into the body cavity. If we call the band of longitudinal muscle fibers dorsal to the lateral line cells (Figs. 1 GaLLoway, budding in Dero vaga. 219 and 2, cl. /. x.) the dorsal band, those fibers from the lateral line to the ventral bristle sacs, the /ateral bands (mu. Z.), and the remainder, the ventval band (imu. v.), we shall be able to lo- cate the ingrowths in a more satisfactory manner. A series of ectodermic ingrowths occurs on either side between the lateral and ventral bands. We may neglect these as they do not con- tribute to the formation of nervous structures, unless possibly Fig. 2. Transverse section of Dero, showing arrangement of nervous structures in head of posterior zodid. Lettering asin Fig. I. they give rise to the small visceral ganglia lying upon the phar- ynx, which may be deposited as the lateral ectodermal sulci of the buccal cavity are infolded. With regard to the origin of these ganglia my evidence is as yet unsatisfactory. The spaces between the dorsal and lateral bands are normally occupied by the cells of the lateral line, which have 220 JOURNAL OF COMPARATIVE NEUROLOGY. usually been considered nervous.’ Ectodermic ingrowths occur here, the fate of which we shall consider later. A pair of ingrowths, one on either side, penetrates the dorsal muscle band, detaching a portion (mu. d'.) equal perhaps to one- half the lateral band, in width. These ingrowths grow dorsally, overarch the digestive tract and bloodvessel, and unite in a me- dian position, producing the supra-cesophageal ganglia or brain. A similar pair of ingrowths penetrates the ventral muscle band near its margin and contributes directly to the growth of the ventral cord and to the formation of the sub-cesophageal gan- glia. The connective (con’t. crc’e@.), as its fibers grow from the brain toward the ventral cord, passes superficially to the detached portion of the dorsal muscle band, re-enters the body cavity at the break between the dorsal and lateral bands,—the place of occurrence of the lateral line cells,—passes superficially to the detached portions of the ventral muscle (ww. v'.), and joins the ventral cord by way of the ectodermic ingrowths penetrating that band. (Fig. 2.) The circum-cesophageal connective thus comes to embrace, within its circuit, four strands of musle fibers. These persist in this position and become functional in the pos- terior zooid. It seems further probable, from my studies, that there is a multiplication of nervous cells in the horns of the cord itself, which are responsible for at least a portion of its increased prominence. In the preanal, segment-forming zone of the anterior zodid, where the ventral cord is being extended caudad, there is a me- dian constituent contributed by the ectoderm, in addition to the two latero-ventral contributions mentioned for the posterior in- dividual. The cord readily shows the extent of the single me- dian and paired lateral components, in the newer segments. 1 Since the above investigations were made Brode’s excellent paper on the finer anatomy of Dero vaga has appeared. This author is convinced that the lateral line cells are not nervous in character. While my own investigations have caused me to hesitate in accepting the common view concerning them, I am not in a position to corroborate or controvert Brode’s conclusions. Nickerson, Epcdermal Organs of Phascolosoma. 221 SUMMARY. 1. The complex nervous system produced in the budding process arises either (1) from the cells of the nerve cord, or (2, and chiefly) from ectodermic ingrowths. 2. There are five regions in the ectoderm which may give rise to nervous elements: (1) a single median ventral region (1, Fig. 1) especially active in producing the median portion of the cord in the new segments of the anterior zooid ; (2) a re- gion, on either side, superficial to the latero-dorsal muscle band (3, Fig. 1), which produces the brain in the posterior zooid ; and (3) a region on each side, superficial to the latero-ventral muscle band (2, Fig. 1), concerned in the development of the sub-pharyngeal ganglia. 3. The brain arises in the region immediately contiguous to the lateral line cells; and the ectodermic ingrowths marking the point where the connective re-enters the body cavity from its position superficial to the muscle band appear in connection with these cells. If the cells of the lateral line are nervous, they are thus brought into relation, in an interesting way, with the central nervous system. The brain, in this event, is developed in connection with the lateral line cells, while the ventral cord is derived from elements much more ventral. EPIDERMAL ORGANS OF PHASCOLOSOMA GOULDII. By Marcaret L. NICKERSON. Scattered abundantly over the introvert and body of this worm are found the epidermal organs which on the introvert have the form of papillz, while on the trunk they are partially included in the large excavations on the inner surface of the cuticula. These bodies are ovoidal in shape with the smaller end directed outward, while the large base rests upon the circu- lar muscle. Each is surrounded by a delicate membrane which is probably an invagination of the membrana propria. 222 JoURNAL OF COMPARATIVE NEUROLOGY. The following is a summary of the results obtained from a study of these bodies. 1. The sensory nervous system of Phascolosoma gouldii is to be found entirely in the epidermal organs distributed abundantly throughout the body of the worm and the nerve fibers connecting them with the central nervous system. 2. These epidermal bodies may be grouped into four classes, two of which contain gland cells, the other two being non-glandular. The two types of glandular organs may be readily distinguished by the presence or absence of intracellular canals in the gland cells, while the two types of non-glandular organs are to be distinguished by the possession in one case of a bulb-like structure projecting above the general level of the cuticula. 3. All four classes of the epidermal organs possess sen- sory cells. 4. Nerve fibers are never found in continuity with the gland cells of either type of glandular organ, as has been sev- eral times asserted by different investigators. 5. The sensory cells of all these organs are bipolar, the cell body in the non-glandular organs being larger than that in the glandular organs. 6. Each of the peripheral processes of the sensory cells ends in a delicate sensory hair which in some cases at least is prolonged beyond the surface of the cuticula. In one case only, the glandular organs of the first type, the exact form of the per- ipheral ending was not made out. 7. The central processes of all these sensory cells enter the large nerves passing to the ventral nerve cord. 8. One type of glandular organ possesses a remarkable structure consisting of a communicating set of intracellular canals, each canal leading from an otherwise closed pouch. This pouch is surrounded by a zone of radiating threads. All these communicating canals finally open to the surface through a common duct. g. The intracellular sacks belonging to this type of glan- dular organ are reservoirs for the secretion from the gland cells BERGER, Eyes of Cubomeduse. 223 and show much variation in size and appearance in correspond- ence with the phases of activity of these cells. The ducts from these sacks are the channels by which the secretion is conveyed to the surface of the animal. The radiating threads surround- ing the sacks are probably continuations of the reticulum of the cytoplasm. Tue HIsTroLoGicaAL STRUCTURE OF THE EYES OF CUBOMEDUSZ. By Epwarp W. BERGER. While in Jamaica with the Johns Hopkins Marine Labora- tory, during the summer of 1897, Dr. Conant preserved mater- ial and tried experiments for the purpose of continuing his re- search on the Cubomedusz, begun the year previous and now published as his thesis by the University. Upon the unfortu- nate death of Dr. Conant this material and notes were placed in the present writer’s hands by Dr. Brooks. It is intended in the following paper to give only the principal results ob- tained by a careful study on the histology of the eyes of these medusz, leaving their fuller discussion, together with Conant’s physiological notes, fora more complete paper. The present work was done wholly on Charybdea xaymacana, while Co- nant’s own work was in part done on Tripedalia. For a complete description of the anatomy of the Cubo- meduse Dr. Conant’s thesis, ‘‘The Cubomeduse,” or the ‘‘Johns Hopkins University Circulars,” No. 132, November, 1897, should be consulted. Roughly speaking, the Cubomedusz, as the the name im- plies, are cubes with their tentacles (four in Charybdea but twelve in Tripedalia) arranged at the four corners of the lower face of the cube. These tentacles are said to lie in the interradii. Half way between any two points-of attachment of the pedalia (the basal portions of the tentacles) and a little above the lower margin of the beil, hang the sensory clubs, one on each side, 224 JOURNAL OF COMPARATIVE NEUROLOGY. four in all. Each sensory club hangs ina niche of the exum- brella and is attached by a small peduncle, whose central canal is connected with one of the four stomach pockets and in the club proper forms an ampulla-like enlargement. Each club is said to lie in a perradius, and belongs to the subumbrella, as is shown by the course of the vascular lamelle, Explanation to ##g, 7. This is an outline taken from Schewiakoff’s Fig. 7 and is placed here to show the general relations of the different parts of a club. Since this drawing represents a section the simple eyes are not indicated. C—concretion cavity; CO—cornea; CP—capsule; CSC—cavity of sensory club, £C—ectoderm; “N—endoderm; “MVC—endoderm of sensory club; Z—lens; NC—network cells; M/—nerve fibers; 7 --retina; S2A—supporting la- mella; V8—vitreous body. bands of cells, which passing through the jelly from endoderm to ectoderm all around the margin, form the line of division be- tween sub- and exumbrella. Each club has six eyes. Two of these on the mid-line of BERGER, yes of Cubomeduse. 225 the club facing inwards are called the larger and smaller (lower and upper) complex eyes because of their more complex struc- ture, while the other four simple eyes are disposed laterally, two on each side from the line of the two complex eyes. All of these eyes look inwards through a thin transparent membrane of the sub- umbrella into the bell cavity. Besides the eyes and ampulla already referred to, a concretion fills the lowermost por- tion of the club, and a group of large cells having a network-like structure and called network cells by Dr. Conant fill the uppermost part of the club be- tween the smaller complex eye and the attachment of the club to its pedun- cle. What is evidently nerve tissue, fibers and ganglion cells, fills the rest of the club. Meyer, Data of Modern Newology. 271 ation of the external geniculate body according to whether the eye or the occipital cortex is removed, and that this finding is interpreted by Forel in favor of his neurone-concept. If the eyes are removed, the end-arborizations of the optic nerve fibers in the ‘ground-substance’ of the external geniculate body will decay ; consequently the cells will come more closely together, part of the ground-substance being resorbed ; if however the cortex is destroyed the optic radiation is affected and its cells will undergo atrophy and even resorption and neuroglia forms the scar. (See Monakow’s Gehirnpathologie, Fig. 82 and 83). We must remember that the resorption is most complete where the lesion occurs in very early life. The general law is easily demonstrated in the spinal cord of cases of infantile hemiple- gia where the degenerated direct pyramidal tract is often resorbed without leaving a neuroglia scar, and the area of the degener- ated crossed pyramid is very much smaller than in the case of hemiplegia in the adult or senile. (2). Among the cerebral efferent neurones we really are familiar with the pyramidal systems only. The thalamic radia- tion and even the efferent paths to the cranial nerves are not accurately enough known to allow us to speak of complete neu- rones, although the latter have been much cleared up by Hoche. All bodies of the efferent ‘ofor’ neurones belong to the pyramidal type, a cell-form which undoubtedly owes its out- line to the peculiar composition of the cortex in layers with more or less perpendicular radiation of fibers. Among the many forms of cortical cells the large motor pyramidal cells are quite characteristic, not only by location but by structure. These are the ones which von Gudden and von Monakow have shown to be absent after destruction of the internal cap- sule in the young. These large cells of the motor region are located in the deeper parts of the fourth layer of Cajal; the most striking ones are the giant pyramids of the paracentral region, between the layer of small cells and the polymorphous. They are among the largest cell-bodies of the human nervous system, perhaps because a correspondingly long neurite comes from them. The 272 JOURNAL OF COMPARATIVE NEUROLOGY. form is usually pyramidal although it is difficult to always strike the right plane for making sections; a deviation from it pro- duces pictures like the real motor neurones in the ventro-lateral lamina of the neural-tube. To judge from numerous specimens it would, however, be wrong to say with Nissl that they were not pyramidal. In structure, they represent the ‘motor’ type. The large lumps of stainable substance are regularly arranged, leaving plain paths for the non-stainable substance which contains the fibrils. The apex-process and the lateral processes have a very neat arrangement of longitudinal spindles, The nucleus and nucleolus are very large. A number of drawings which lately appeared in the Journal of Insanity give a fair idea of both the normal and certain abnormal conditions of this type of cells, at least figure 1, 4, 6 and 7. The writer describes several alter- ations of the cell-body ; for Figs. 3 and 4 he urges the proba- bility of a lesion of the neurite, for Fig. 6 and 7 toxic pro- cesses; but we necessarily miss statements concerning the fate of the fiber-part of the neurones, since it would be impossible at present to study their fibers thoroughly in their whole extent; the real terminations of the pyramidal fibers especially being only little known, and no methods for their isolated study be- ing available. The efferent cerebral neurones which are not a part of the pyramidal system are comparatively little known. Apart from the studies of v. Monakow and Moeli there are very few data, and even these are hardly sufficient to establish any satisfactory description of these weusones. There are many well founded suppositions; but suppositions are not neurones.’ There are quite a number of intra- and inter-cortical nerve- elements, cortex-cells of which we might give descriptions here, 1 Demonstration of various types of changes in the giant cells of the para- central lobule, by Adolf Meyer, M.D. American Journal of Insanity, Vol. LIV, No. 2. 2 Through an oversight the cerebral efferent supply of the third, fourth and sixth nucleus is represented in the ‘plan of the brain’ as coming through the pyramids. It is more probable that the origin of these elements lies in the large cells of the visual area and that their course is independent. Meyer, Data of Modern Neurology. 273 _ further the cell-types of the cornu ammonis etc. The cell- bodies are very eagerly studied now both by the Nissl school and the friends of the Golgi method. It is by no means an easy task to correlate the two series of results, because the gemmules, very striking parts of the Golgi pictures, are but rarely reproduced by other stains. 3. The study of the cerebellar mechanisms has re- ceived a new stimulus through the application of the methods of Golgi and of Marchi. The field is, however, still full of contradictions. For the plan of the brain I should accept the sketch of Thomas,’ since my own investigations on this point are not yet sufficiently advanced. Only few of the elements which unquestionably belong to the cerebellar apparatus are known so as to be pictured as neurones on ground of actual demonstration. We take up only the best known elements. Since Flechsig’s investigations, the direct cerebellar tract is known as the principal afferent system of the cerebellum from the spinal segments. The'cells of Clark’s column are considered to give origin to it. They have a fairly characteris- tic structure. The stainable substance is arranged in fairly uniform medium-sized granules, filling the cytoplasm, perhaps more uniform and larger than those of the afferent cerebral cells. The nucleus is large, frequently covered up somewhat, and more or less near the center of the cell; in other cells it stands in an area where the granules have disappeared and just form a peripheral ring around the cell; and again I have found it excentrically located, even bulging beyond the natural pe- riphery of the cell, just as in the cells in which the neurite was cut; a condition which was found by Marinesco in locomotor ataxia, by Councilman and Barker in meningitis and by myself moreover in elderly people without any observed cerebellar or other affection. The dendrites are slender; the stainable sub- stance in them is scanty and in granules rather than in short streaks. The description given here is that of transverse sec- tions. In the oblique and longitudinal ones, the cells are strik- 1Le Cervelet, Etude anatomique, clinique et physiologique. Paris, 1897. 274 JoURNAL OF COMPARATIVE NEUROLOGY. ingly similar to the ‘motor cells’ of the ventral horns, except in shape and general arrangement. The cell-body is twice to three times as long as broad and the heavy terminal dendrites with the typical spindles take a longitudinal course. The lateral dendrites are not very numerous, but turn at once into a longi- tudinal course, frequently sothat one branch grows caudad and the other cephalad. The neurite originates from the side of the cell-body. This observation explains many peculiarities of the cells shown on transverse sections. The difficulty of differen- tiating them from the ‘motor type’ may become of import- ance in the final criticism of the question: to what extent are the form and structure of a cell characteristic for its functional connections? We shall show in a future article to what extent the size of the cell and the architecture of the tissue generally are of importance for the arrangement of the Nissl-bodies. The connection between these cells and the fibers of the direct cere- bellar tract are little known, perhaps on account of their oblique course ; the upper course of the fibers is in the center of the restiform body and the termination in the cortex of the upper worm of the same and of the opposite side. The cerebellar afferent neurones of the brachial region partly constitute the nuclei of Stilling (the homologue of Clarke’s column in the cervical segments), partly unite into the external nucleus of Burdach ; those of the cranial segments are much less localized (lateral nucleus, olives and red nuclei ?). The efferent cerebellar neurones are undoubtedly the Pur- kinje cells and the cells of the central nuclei of the cere- bellum. While the course of their axones is a matter of varying opinion, the structure of the Purkinje cells is so well known that we outline it here for the purpose of establishing another type of neurones. The remarkably graceful silhouettes obtained with the Gol- gi method are familiar. Nissl gives the following description of this type of structure (Z. f. Psychiatrie, Vol. 54, p. 69): ‘A careful study of a Purkinje cell not only reveals a net-shaped type of a structure but moreover stripes. The stainable bodies, embedded in the net-work so as to form, as it were, Mever, Data of Modern Neurology. 275 the nodes of the net, are more or less plainly arranged in rows. They run parallel with the surface of the nucleus and form cir- cles around it, at least along its base and sides but not towards its apex-process. These rows of bodies turn towards the apex where the meshy structure of the perinuclear part of the cell disappears and gives way toa purely striped arrangement of the stainable particles. On the pole of the nucleus situated towards the apex there are frequently rather large stainable masses like nuclear caps. These shields or crescents or irreg- ularly shaped masses of stainable substance are peculiarities of this cell-type. _ Little is to be said of the structure of the pro- cesses since they consist largely of non-stainable substance. The basal process can be followed a short distance only; the apex process however for a great length. Although devoid of plain marks it appears striped evidently on account of the pe- culiar structural make-up. It is exceedingly difficult to fix the nuclei of the Purkinje cells; they are strongly inclined to chromophily. It is not difficult to see how different they are from the nuclei of motor cells.’ So little is known of the cells of the olives, pons, red nu- cleus etc., that we dispense with their description. They hardly figure as neurones yet except in diagrams. The midbrain me- chanisms, thalami and corpora striata are not much less prob- lematic, although much light has been thrown on them by v. Monakow. The data of knowledge of these parts cannot be incorporated yet in simple diagrams of neurones known in their totality with the exception perhaps of the ganglion habenule and its ‘ Anteile’, and the fornix-apparatus. For a complete summary of the safe data of the anatomy of the nervous system, we should add an analysis of all the less known types of cell-bodies in the various accumulations of gray matter, and further a summary of the unfinished analysis of ‘white matter.’ The progress from the old carmine stain to the Weigert myelin-stain and to the Marchi stain has led to an in- crease of facilities for the study of fibers and their states of myelinization and degeneration. Indeed the study of this point has become so easy that it is almost child’s play and it is sur- 276 JouRNAL oF ComMPARATIVE NEUROLOGY. prising how many writers are satisfied with these superficial findings of tracts, without attempting to look for corresponding cell-bodies to complete the neurones. The time when ascend- ing and descending degenerations and the length of the tracts of degeneration would pass as a description fit for publication is by no means over and much splendid material is thus half-way wasted because it is not exhausted with better methods of pre- paration and study. One would think that the frequent cases of compres- sion of the spinal cord surviving from a week to several months would have been used for settling what the degeneration meth- od can settle in the anatomy of the spinal segments. In the vast majority of the few studies published, neither the cells nor the cephalic’ ‘tracts’ have been decently studied. Lately the . writer received two such specimens, without spinal ganglia and without oblongata and brain. My well-meaning friends evidently have no exact conception of the extent of neurones. Well-established ‘systems’ might well be called foundlings; we know nothing of the mother-cells from which they come. Among these are: Gowers’ tract, a system of neurones consti- tuting afferent elements from the lumbar cord-segments to the cerebellum and the midbrain; the so-called anterolateral de- scending tracts, which may come from the cerebellum, or mid- brain or from Deiter’s nucleus; the septo-marginal and comma- tracts which most probably do not come wholly from the affer- ent neurones of the spinal ganglia. It would further be impor- tant to revise the work of Flechsig from the neurone point of view ; espcially his idea of ‘systems’ which, he says, are the same as will degenerate in locomotor ataxia. I am inclined to attribute much of the common superficiality to the way stu- dents are usually taught. The phrases commonly used are: such a tract degenerates upward or downward and therefore conducts upward or downward. This may be shorter, and more easily remembered than the following attitude of mind dictated by the neurone-theory: any destruction of nerve fibers leads to a rapid decay of the part of the fiber cut off from its nucleus, while the fiber stump remaining in connection with the cell will usually Meyer, Data of Modein Neurology. My hy be found preserved in the adult and practically normal at the autopsy, apart, perhaps, from the atrophy due to disuse and from the characteristic alteration of the cell-body, the transitory, so-called traumatic, reaction. Only in the young do we find atrophy and even degeneration of the cell and remaining stump. Descending degeneration is the term used for the decay of fibers, the cells of origin of which are located ‘above’ a lesion; ascending degeneration a term used for the decay of fibers the cells of which are located in segments ‘ behind’ a lesion. Such a statement gives the facts and at the same time the problems, and can be grasped by every man who deserves to be called a medical student, if the proper illustrations and demonstrations are furnished. We cannot leave this sketch of the best-known neurones without a short statement concerning the changes observed in them zz normal and pathological conditions, and especially the t- terrelation of the neurones and their ways of interaction. For the neurone-theory the life-history of the elements is of fundamental importance, The early stages of development are well-known. It seems certain that a cell which has a spe- cific process and is thus characterized as a nerve-cell, has lost its power of reproduction, Karyokineses become rare after the fifth month of gestation and the claim has been made that at birth the entire stock of nerve-cells is present, later growth of the nervous system meaning merely a growth of existing ele- ments. The question is treated by Donaldson (Growth of the Brain, p. 163-171). Schiller’s count of the fibers of the third nerve in the young and adult cats and Kayser’s counts of cells in the cervical cord of the human foetus, child and man, show a discrepancy. Schiller found the number of fibers remaining the same. Kayser found an increase of developed cells up to adult life. It is probable that the discrepancy is only apparent; because we do not know how many wndeveloped cells Kayser started with. From the presence of centrosomes found in a few cells (Lewis, Lenhossék, Dehler, etc.) no conclusion should be drawn now; the occasional karyokinesis not only of neuroglia 278 JOURNAL OF COMPARATIVE NEUROLOGY. but of nerve-cells, in the healing wounds of the nervous sys- tem (see Valenza, p. 32 and A. Tedeschi, anat. experimenteller Beitrag zum Studium der Degeneration des Gewebes des Cen- tralnervensystems. Ziegler’s Beitr. z. path. Anat. Vol. XXI, H. 1, 1897) is hardly more than an object of curiosity. The occurrence of more than one nucleus, not infrequent in the sympathetic nervous system, very rare elsewhere, is also an uncorrolated fact. (See further the latest summary of this question: Sulla cariocinesi della cellule nervosi. Ricerche del dott. Giuseppe Levi. Riv. di Pat. nerv. e ment., Marzo, 1898. Hodge has published a little study of the appearance of human cells at various ages, only referring to the cell-body, and on ground of three individuals. A diminution of the rela- tive size of the nucleus in old age seems to be the best estab- lished result. The strongest part of the life-history of the neurone lies undoubtedly in the facts which led His to estab- lish the neurone-theory from the embryological point of view (See). pp. £27), There are a vast number of observations on experimental and pathological alterations in various forms of nerve-cells in the literature of the last two years, the effects of inanition, in- toxications, diseases, etc. After summarizing the available material, the writer finds that only a small number of leading features throw any light on the neurone-theory. The majority of studies deal merely with the changes in the cell-bodies in preparations with Golgi’s and Nissl’s stains, or their modifica- tions. The history of our knowledge of functional changes in nerve-cells due to stimulation, exercise and intoxications has been written a number of times in late years. The principal contributions begin with Hodge, Mann, Vas, Nissl, Schaffer, Pandi, Sarbo, Berkeley, etc. They have established processes of fatigue and recuperation, of toxic disorganization with ter- mination in death or recuperation, of very much the same char- acter as those known in other cells of the body. About the details there is much controversy. Meyer, Data of Modern Neurology. 279 I refer, for a summary of the literature, to the memoirs of Giambattista Valenza (cambiamenti microscopici della cellule nervose nella loro attivita funzionale e sotto l’azione di agenti stimolanti e destruttori. Napoli, 1896) and the excellent re- views of the Rivista di Patologia, Revue neurologique and Neu- rolog. Centralblatt, and especially to the studies of Nissl and Lugaro. Asa concise and easily accessible statement of most of the data, the two general reviews by van Gehuchten and by Marinesco, offered to the Congress of Moscow, deserve recom- mendation. I give here a short account of the set of experi- ments made by Goldscheider and Flatau, also reported at the Moscow Congress and accessible in abstract in the Revue neu- rologique, Vol. V, p. 525. 1. The injection of ‘malonnitrile’ (CN-CH, -CN) pro- duces violent phenomena of intoxication which disappear after the administration of hyposulphite of sodium. In connection with the intoxication one finds the Nissl bodies in the ventral horn-cells ina process of deformation and disaggregation; they be- come smaller and lose their regular and symmetrical disposition. The non-stainable substance and the nucleus take an equally in- tense stain. Under the influence of the hyposulphite of sodium all these alterations disappear within three days. It is of in- terest that the functional symptoms of intoxication disappear very rapidly, more rapidly than the morphological alterations of the cells. 2. When an animal is overheated to a temperature of 10g-112° F., the cells increase in volume, become homogene- ous, opaque, and take a light blue color. The Nissl bodies ‘are destroyed; the dendrites are pale blue, oedematus and vari- cose. The changes begin to decrease at once after the experi- ment, and disappear completely in two or three days. The same alterations, but less marked, are observed with a tempera- ture of 106-7°, if at least this elevation of temperature is kept up for at least three hours. In this experiment too the function recovers before the restitution of the anatomical changes. 3. The toxin of tetanus produces very characteristic nu- 280 JOURNAL OF COMPARATIVE NEUROLOGY. tritive changes in the motor cells of the anterior horns, especial- ly swelling and pallor of the nuclear bodies ; increase, breaking up and finally dissolution of the Nissl bodies, increase of vol- ume of the entire cell. These alterations are the more pro- nounced the greater the concentration of the toxin. The in- jection of the anti-toxin or the use of weak concentrations of the poison still allows the cells to return to their normal state, the swelling of the cell disappears and the nucleus takes on an- gular shapes and returns to its normal volume. 4. The greater the concentration of the poison, the more rapid the evolution of these alterations, and also the more rapid the restitution. Weak concentrations however produce very slow alterations which may persist two or three weeks. 5. The complete restitution of the Nissl bodies is ob- tained more rapidly than that of the nuclear body. 6. Not all the cells react exactly alike; one often ob- serves noticeable differences of intensity in neighboring cells. Similar differences in the degree of morphological changes are observed from one animal to the other (individual peculiarities). 7. There is no accurate relation between the intensity of the phenomena of intoxication and the degree of anotomical le- sions of the cells. These can manifest a tendency to regenera- tlon at atime when the symptoms of intoxication are on the increase and the reverse may be noted. The same dispropor- tion between the anatomical and the physiological phenomena has been ascertained by the writers in the experiments on ‘malonnitrile.’ This fact must be taken into consideration in the appreciation of anatomo-pathological findings. 8. The intravenous injection of the antitoxin of tetanus exercises a manifest influence on the evolution of cellular alter- ations by retarding them ; preventive injection hastens the on- set of the phase of regeneration. g. The action of the antitoxin on the cells is beyond doubt zzdirect. It consists in the neutralization of a quantity of toxin bound by the cell. 10. The morphological alterations of the cell are the ex- Meyer, Data of Modern Neurology. 281 _ pression of a chemical connection between the toxin and the cell- body. [?] 11. The injection of strychnine produces changes analo- gous with those produced by the toxin of tetanus. The first alterations set in sometimes as early as in three minutes. If the animal survives the experiment one also sees that restitution of the function precedes that of the morphological structure. 12. The analogy ot the action of tetanus and strychnine on the morphological structure of cells suggests that the ana- tomical alterations are of great importance for the exaggerations of that excitability of the cells which is found clinically in these forms of poisoning. This is a fair instance of the present position of the ques- tion. Studies of the axones and their terminations are not avail- able yet and therefore the picture of the zeurone is incomplete. To judge from what was said of the direct motor neurones (see p. 255, etc.) the great problem is now as follows: are there two fairly distinct mechanisms in a nerve-element, a vegetative element supplying the possibilities for nutrition and growth, and a functional element (the fibrils ?), the substratum of neural ac- tivity? Or we might perhaps ask more clearly: are there two elements, the functional and the nutritive, held together by the processes of metabolism and growth, and, morphologically, by the nucleus? The present methods give no satisfactory answer to this because the stainable substance is the chief element brought out; and because diffusion of the stain (deeper colora- tion of the non-stainable paths and the processes) is only a very indirect criterion of the real condition of the ‘fibrils.’ This may be the reason why the clinical and morphological phases are not quite parallel and this again might justify us in assum- ing as a great probability the existence of such a division of mechanisms into vegetative and specific activity. The promotion of these studies depends largely on the improvement of tech- nique. If Bethe’s new fiber-method can give a perfectly relia- ble demonstration of fibrils in the cells and all their processes, we may remove many objections of fundamental importance, as those made by Held. On the same evidence will depend 282 JouRNAL OF CoMPARATIVE NEUROLOGY. our knowledge of the process of regeneration. This has some practical importance. There are surgeons who maintain that they observe a healing per primam of sutured nerves and a rapid re-establishment of function absolutely inconceivable from the point of view of the neurone-theory and especially of the fiber theory. These ‘observations’ are in serious collision with the results of the most careful experiments. Before we pass to the second problem, the interrelation of the cells among one another, an interesting contribution of Achille Monti! deserves our attention on account of its funda- mental importance as a possible method in the future. Embolisms were produced by injecting powdered carbon or lycopodium in the carotid of animals. In surviving animals; killed about five days later, the Golgi method brought out small foci of alterations in neuroglia and nerve cells. The most stri- king result is that the dendrites suffer first:—become varicose and lose the gemmules; the cell-body also becomes deformed and last of all the axone is affected. We must, of course admit that the method is that of silhouettes; but it is sufficient to show that in a cell near a focus only the dendrite directed towards the focus, may degenerate and the others, with cell- body and axone, remain quite intact. It is easy to see what a vast series of important data can be obtained from such studies when the right methods are developed. In this connection we may also mention an interesting phenomenon observed by Meyer in the case of facial paralysis quoted above. The auditory nerve, being also involved on ac- count of the hemorrhagic process in the internal auditory canal, showed at its termination a remarkable increase of neuroglia cells. The terminations of the fibers had evidently suffered more rapidly than the fibers and this decay had called for the neuroglia reaction. Moreover it was evident that the cell-bodies of the central ‘auditory nucleus’ were affected slightly, al- though according to present theories, they were only in /une- 1 Sulla anatomia patologica degli elementi nervosi nei processi da embo- lismo cerebrale, Boll, della societa medico-chirurgica di Pavia, 1895. Meyer, Data of Modern Neurology. 283 tional connection with the degenerating auditory fibers, as they belong to the intersegmental or cerebral afferent type. Gold- scheider and Marinesco had formulated the opinion that for the normal vitality of a nerve-cell the normal stimuli were neces- sary. They would perhaps refer the above finding to the abo- lition of the conduction of auditory stimuli. Another explana- tion might be offered, namely, that the same condition which called for neuroglia-proliferation was also the cause of the alter- ations in the contact cells, and in this way we might avoid too generalizing theories by remaining on morphological ground. Coming to the question of the interrelation of the neu- rones we must mention the work of He/d which would if sub- stantiated draw a veil of denial over many of the statements given so far. I refer especially to Held’s last contribution’ which may well be regarded with some apprehension by certain ultraprogressive speculators. To fully understand the bearing of his view, we must re- turn to the sketches given of the transmission of nerve-impulses on plate XV. To Wansen and Golgz the cell-body and its pro- toplasmic processes appeared asa vegetative mechanism; the conduction of impulses is perfectly intelligible through the axone and its collaterals. In 1891, van Gehuchten claimed to have found inthe mitral cells of the olfactory bulb a cell in which no other conduction was possible than that through the dendrite to the cell, and through the cell into the axone. (That the arrangement of the olfactory cells does not give an absolute proof of this view has been stated in the abstract of Monts’s paper on page 124. Even here recurrent collaterals play a role). Cajal took up this statement and illustrating it in his sketch of the cortical mechanisms he established it under the name of the ‘law of dynamic polarization.’ Charles-Amidée Pugnat* gives a short account of the latest phases of this the- 1 Beitrage zur Struktur der Nervenzellen und ihrer Fortsatze. Zweite Ab- handlung, von Hans Held. Arch f. Anat. u. Entwicklungsgeschichte. Anat, Abt. 1897, p. 204-2y4, and plates IX to XII. 2 De l'importance fonctionelle du corps cellulaire du neurone, par C-A, Pugnat, Revue Neurologique, Vol. 6, No. 6, 1898. 284 JouRNAL oF CoMPARATIVE NEUROLOGY. ory. In order to do justice to the occasional origin of an axone from a dendrite instead of from the cell-body, and also to the peculiarity of the segmental afferent neurone, Cajal has, in his new work! maintained three ‘laws’ established by him: 1. The law of economy of time. Inthe spinal ganglia the cells are attached to the fibersso,as to forma T. The con- ductors are thereby placed into the very axis of the ganglion and in the direction of the shortest way between periphery and posterior root; and the current need not pass through the ex- centric cell. 2. The law of economy of matter. In the midbrain of fish, batrachians, reptiles, and birds, there are certain fusiform cells, the axone of which originates from a dendrite. In this way the axone spares the whole distance between the cedl-body and the point of the dendrite from which it originates. 3. The law of economy of space. The body, i. e. the most voluminous part, of certain neurones is (occasionally) placed in regions poor in dendrites or final arborizations of ax- ones, for instance Dogiel’s cells of the internal granular layers of the cortex. ‘On careful consideration of the physiological meaning of the cell-body, one comes to the conviction that it presents noth- ing but the convergence of the protoplasmic expansion towards the origin of the axone, enlarged by the presence of the nu- cleus.’ He adds ina note: ‘The cell-body is after all only a segment of the conductor.’ We need not comment on these exaggerations of ‘legisla- tive tendency,’ but refer to what was said of the efforts of vax Gehuchten towards making the segmental afferent neurones ap- pear lawful (see p. 266). Berkley has probably been most explicit concerning the mode of contact between cell-elements, in the Johns Hopkins Hospital Reports, Vol. VI, p. 89-93 and plate XV (the intra- cerebral nerve-fiber terminal-apparatus and modes of transmis- 15, Ramon y Cajal. El sistema nervioso del Hombre y de los Vertebrados, 1 fasc. Madrid, 1897. Meyer, Data of Modern Neurology. 285 _ sion of nervous impulses). His first claim is that the fine end- branches of fibers are endowed with a ‘protective sheath of great tenuity not easily recognized by ordinary methods of staining, which the silver method does not show atall. It is therefore more than probable that it is only at the free bulbous termination of the nerve-filaments (shown by the silver method only) that we have naked protoplasms, and from this uncovered nervous substance the dynamic forces, generated in the cor- pora of the cells, are discharged, through contiguity, on to the protoplasmic substance of other cells.’ This limits the func- tional contact to these end-bulbs. Berkley further assumes the presence of a protecting membrane around cell-bodies and den- drites. The fine stems of the gemmules of the dendrites pierce this membrane and only the tips of the gemmules show free dendritic protoplasm. ‘The number of end-bulbs (one on each terminal branch) of the association and ascending fibers from the lower regions is not numerous, seldom exceeding six or eight, and the form is that of an arborization of the nerve twig; on the other hand the terminations from the collaterals of the psychic cells are much more numerous on the final branches and show the disposition of his Fig. 1. ‘The inter- pretation of the objective existence of the terminal apparatus of the nerve-fiber can be made but in one way, namely, that the impression conveyed from external sources to central cells and from local cell to local cell, is not accomplished by a diffus- ion of the excitation through the whole cortex, or even at vari- ous points along the course of the finer branches of the axons, but at single points, perfectly definite in their distribution, and that these points are situated only at the extremities of the nerve fiber, in the form of an histologically exact formation— the bulbous ending of the nerve fiber—which in itself consti- tutes the sole and only means for the carrying over of the cel- lular force from axon to dendron, and from cell to cell, and is in entire conformity with the conception of Waldeyer of the entity of the neuron, each cell standing as an unit in the ner- vous formation, and only in continuity with others at definite points, ’ 286 JourNAL oF ComPARATIVE NEuROLoGy. After these statements we understand how broad a line of ‘facts’ Held attacks with his publications. The first attack goes against the interpretation of the structure of the cell. On sections of 1 micron, Held could not convince himself of the fibrillary character of axones or cell-plasma. He corroborated the view of Bitschli. He sees the ‘ fibrils’ connected by cross-fibrils. A thin section through a honey-comb would produce just the same ap- pearance. In the axone Held sees this axo-spongium (simulat- ing the fibrils) and the neurosomes, small granules embedded in the axo-spongium and especially numerous in the end-plates of the axones. Held has observed in the trapezoid nucleus of the medulla that the axones entering it form a sort of end-plate as they approach the cells. In the cells he distinguished the cytospongium, the neurosomes and the Nissl-granules, and he finds it possible to distinguish the axo-spongium from the cyto- spongium by a difference of stain. In an animal one or two days old, or in the new-born, the axone-ends form a basket-like surface of contact with the cell-body separated from it merely by a homogeneous line probably of ectoplasm of the nerve-cell (or axone-plasm ?). In the animal nine days old, this limiting line cannot be seen. In this and in the adult the basket branches of the axone are quite plainly distinguished from the cytoplasm by the number of neurosomes and the stain; but in many places it is impossible to deny a concrescence between some of them and the cytoplasm. Ina third part of this study (Arch. f. Anat. und Entw. 1897, Supplement Band, p. 273-312, plates XII-XIV), Held establishes the same facts with a more delicate method, insists on the Axencylinder-endflache being in contact largely with the dendrites, consisting of a real net, not of inter- lacing but anastomosing axone-terminations, even so that ax- ones from several neurones should enter into the mesh-work. He corroborates Béla Haller’s findings in principle at least, the claim of anastomoses of dendrites in the spinal cord of adult teleosts, and the findings of true meshes by Badlowitz in the electric organ of torpedo, but refuses Apathy’s net work of nerve-fibrils in worms. Why? He also examines Golgi speci- mens of the adult with the same result and attributes the gene- Meyer, Data of Modern Neurology. 287 rally adopted views to the study of embryonic material. He stands firmly on the ground of His as far as the primary origin of neurones goes, but establishes concrescences in the adult. Anyone who examines the excellent drawings of Held and the account of his technique must admit that his data cannot be refuted by the available literature. However improbable his results seem, they seem so largely because for years we have with all our efforts tried to get rid of old views, gained less on ground of minute observation on all kinds of material than on the one which did best justice to the new demands. This was the embryonic materia! concerning which Held says explicitly that anastomoses could not be observed yet. Personally, I cannot change from one view to the other as if I never had any, and I feel rather sceptical toward certain claims of Held, but certainly not to much so to not consider it unwise to disregard his findings. They appeal to me because they are at last again studies of tissues and not merely studies of ‘cell-individuals’ disregarding everything in the section that does not fit into the customary plan. It will however be neces- sary to prove the same conditions by preparatians gained with the ‘fibril-methods’ of the future, before we can call ourselves convinced.’ With this view we must close the data we have on hand for the neurone-theory. The fact that Wiedersheim once saw motion in the nerve-cells of Leptodora hyalina and that certain differences in the tissues of the retina and nervous system dur- ing rest and activity have been observed, is in altogether too problematic a state to be discussed here. The reader is referred to van Gehuchten, 2d edition, p. 218-222. 1 Since this was written (May, 1898), Nissl has repeatedly announced that the neurone-theory has died under the weight of Apathy’s and Bethe’s discov- eries. We shall take occasion to analyze his reasons for the returntoa diffuse network, which, after all, is merely a ‘fibrillary’ edition of Golgi’s réseau ner- veux diffus. 288 JouRNAL OF CoMPARATIVE NEUROLOGY. The Utilization of the ‘Neurone-Theory’ for Neuro-pathology. While these data of histology and of experimental pathol- ogy cannot help being of ultimate importance for the establish- ment of a more accurate neuropathology which would be use- ful for clinical purposes, a candid reader of the foregoing pages will recognize that not enough is really gained yet to entitle us to the proud statement that the neurone-theory has revolu- tionized neuropathology and solved its great problem, the cor- relation of physiological and histological data. In treating of the relation between the neurone-theory and disease we must recognize that the mere speaking of neurones instead af the old ‘fibers and cells’ is a relatively insignificant change; a concession which old established clinical and anatom- cal neuropathology is making to the new nomenclature and stand-point of histogenesis. We cannot even say that the cell- theory was quite a stranger in neuropathology before 1887; at least as far as the study of nerve tissue as a tissue is concerned, or when we consider Gowers’ view of the motor paths. On turning the pages of a modern text-book of pathology one might even suspect that merely a correction of details had been the result of the revolution. This is evidently not the opinion of men like Andriezen, who favors us already with a ‘complete’ outline of the patho- logical anatomy of psychoses, or of Cajal and Duval and their followers, who unveil the silhouette pictures of sleep, hypnosis, hysteria, etc. One would think that we have it all in black cells now; yet, in the main, the new diagrams are merely new editions of the old ones. The medically important recent dis- coveries are the recognition of the nature of the spinal ganglia, and the description of the fillet as a cerebral afferent apparatus. In view of the position taken by Gowers in 1886 the ‘two neurones of the voluntary motor apparatus’ are hardly a new addition and even the fillet might be claimed as a preneuronic acquisition. The further changes are largely changes in names. The connection of the direct cerebellar tract with the cells of Clarke’s columns is described by Gowers on page 121 of his Meyer, Data of Modern Neurology. 289 1886 edition, and the other tracts of degeneration in the cord received their cell-bodies and become ‘neurones’ whether the cells were known or not, simply because the neurone-theory demands it. The progress lies to a great extent in the new formulation of problems. The novelty makes itself strongly felt when we apply it to customary clinincal thought. A constant and most clinical exponent of the new standpoint, Goldscheider, says in Nothnagel’s Pathologie und Therapie, Vol. X, p. 96: ‘the at- tempt to trace the pathological anatomical phenomena of the nervous system to the neurones, seems at first sight to lead to a certain conflict with regard to the customary division in diseases of the brain, cord and peripheral nerves. For the neurones belong mostly at the same time to the spinal cord and the periphery or to the spinal cord and the brain. But we have instances of a common and universal participation of the nervous system in diseases which we classify, according to the principal localiza- tion or the clinical character, as brain, cord, or peripheral affec- tions. This customary classification will not decrease the value of the reference to the neurone.’ _ This characterizes a mixture of progressive and conserva- tive spirit justified in a work for physicians trained in the old views and gives a hint as to what might come. Leyden and Goldscheider cannot expect, even if they might feel inclined to be revolutionary, that the practioner would enter without con- fusion into the new spirit, into a completely revised system of neuropathology. But need the time be far off when the growmng generation might be shown the field in its new arrangement ? With this question we enter upon the core of the modern prob- lems in neurology and also upon the more restricted point to be discussed in this essay. The great share of the progress of neuropathology lies un- questionably in the beautiful discoveries along the lines of ‘lo- calization.’ As soon as the physicians learned how ‘motor memories’ of speech and of all voluntary movements, became flesh in the shape of ‘centers,’ all the current thought focused on the search for more ‘centers.’ Many of them had been sup- 290 JOURNAL OF COMPARATIVE NEUROLOGY. plied before and many since the cerebral localization came to honor. Around these ‘centers,’ the data of neurological teach- ings were arranged. Everything points unquestionably to a verdict that this must be in a way the ideal plan of progress, with Morgagni’s motto: ubi est morbus? if at least it takes the advice of his further statement: Nulla est alia pro certo noscendi via, nisi quam plurimas et morborum et dissectionum historias collectas -habere et inter se comparare—the only logical true method, with limitations, though, which we often disregard over the de- sire of drawing conclusions. We see both the neurological sciences work in this direc- tion, physiology and anatomy. The former reigned supreme till lately, and it has given neuropathology it method of reason- ing. With the coarse anatomy of the nervous system and the knowledge of ‘centers’ furnished by physiology, the diagnoses of nervous diseases are made. Real histology then fur- nished the ce// in which the functions of the centers were ‘ pro- duced,’ and the jiber-paths which ‘conducted’ the ‘ discharges,’ the functional energy ready made and stored up in the centers and leaving there its traces: This is the current conception of most medical men of to-day. The minute anatomy in the hands of a Meynert may have helped to strengthen this attitude; first, by making the finer anatomy of the nervous system appear as an abstruse subject, and second, by creating the idea of the ‘projection-systems ’ which has arisen from the center concept and can be manipu- lated theoretically without any real histological knowledge. This may explain why physiologists have been and for the most part are, absolutely devoid of interest in pure nervous anatomy. The blade of a knife used in the operation was the most accu- rate instrument of precision used as far as ‘anatomy’ went, until v. Monakow began to work out some brains coming from Munk’s laboratory, and Langley and Edinger a few of Goltz’s brains. Schaefer’s excellent anatomy of the brain in Quain might be held up against my statement as an anatomy written by a physiologist. It is indeed the best descriptive anatomy in Meyer, Data of Modern Neurology. 291 the English language ; but it is not written from a stand-point which would keep in view the whole field of neurology, genetic, comparative, physiological and pathological, because it is forced into the frame of an ‘anatomy.’ The work in physiology done by Schaefer has only the remotest contact with that anatomy. Horsley and Gotch, in their remarkable conjoint study come probably next to methods of precision in detailed anatomy with purely physiological methods. What v. Gudden and his pupils had long foreseen and done, has become more fashionable since the introduction of Weigert’s medullary stain and especially the popularization of Marchi’s method. There, in the production and study of lesions with secondary degeneration, is the beginning of physiological anat- omy. It is on ground of these methods that the plan of archi- tecture ot the nervous system given in the second chapter has its foundation. Further elements are furnished by com- parative anatomy combined with the degeneration method, and by the method of embryology and study of later development in man and animals (methods of His and Flechsig). Side by side with this are, of course, the physiological observations. The fact that the neurone-theory grew partly out of this combi- nation of neurological methods, and finds its natural home in them, is one of the reasons why the Golgi-method should not be praised as the backbone of modern neurology in the customary exclusive fashion. That neither the ‘neurone-theory’ nor the Golgi method can bring exclusive salvation in neurology, has been shown by the third member of the great trio of Swiss neuro-histologists, Prof. Kolliker in Wurzburg, who felt himself * justified in opposing the non-decussation of part of the optic fibers on purely histological ground. The great progress achieved through the present revival of histological research consists in the intimate union between so many methods and stand-points. The scalpel-physiology gives way to embryological, developmental, comparative, experi- mental and pathological histology and physiology and instead of centers of the old types, mechanisms are being unveiled and the deductions are more closely physiological than ‘ psycho- 292 JouRNAL OF COMPARATIVE NEUROLOGY. logical’ or ‘meétaphysical.’ The old ‘center’ is a sort of homunculus, a mysterious pigmy who acts his part as a little man would, accumulates images and energies and discharges them, sends them along the wires of fiber-paths to his superiors and inferiors who discharge again on others. Such action is decidedly more anthropomorphic-electrical than physiological. The plan offered here is one of physiological mechanisms, leaving open the ultimate question what the ‘discharge’ and the life-process of the individual neurone really is, but search- ing patiently for details of the relations of neurones and of the parts of neurones to one another. It turns against the tiresome habit of riding every detail-discovery to death by try- ing to explain through it the whole unknown and is more seri- ously bent on a large plan which shall not come into conflict with any of the possibilities of detail. It takes its base in the oldest and best-known and best-knowable, the segmental ner- vous system, and proceeds to the superstructures, the cerebel- lum, midbrain, thalamus and cerebrum, on this basis. At the same time it remembers certain general concepts of biology as not altogether out of place where general methods of reasoning are in question. The medical literature shows a great deal of the stand- point of Cartesian localization left; however modified and di- luted, it is Cartesian in principle. Nobody would, of course, search the pineal gland for the spring of all action, the ‘soul’ ; but this same soul is, though split up a little, seated in the cen- ters. A truly biological spirit trained in comparative anatomy and physiology would hardly embrace this attitude except per- haps for the carelessness with which we always use old-fash- © ioned ways of speaking. To these consistent or casual ‘ Car- tesians’ I would like to recall the famous statement of Spinoza which may lead the ‘psychologically’ inclined students to- wards a more modern biological concept: Et enim quid cor- pus possit, nemo hucusque determinavit, hoc est, neminem hucusque experientia docuit, quid corpus ex solis legibus nat- urae, quatenus corporea tantum consideratur, possit agere, et quid non possit, nisi a mente determinetur. Meyer, Data of Modern Neurology. 293 The neurologist finds as material for his studies the living being, be it an animal ora person. This living object offers three series of phenomena: (1) The morphological series, in- cluding all the facts of visible and tangible anatomy and histol- ogy; (2) the physiological series, furnishing the material for the specification of energies which become evident in the pro- cess of vital manifestation. These are the two objective series. A certain number of the physiological processes may moreover involve a third series of phenomena, subjective in character, but none the less objective in many manifestations, namely (3) the psychical series. An adequate digestion and correlation of all these three series of facts by the mental activity itself con- stitutes our neurological science. We have just seen how the older physiological school grouped its data and how it differs from the one represented by the modern anatomical-physio- logical school which the writer would favor. We take for the time being an objective point of view, leaving out the psychical because we lack sufficient experimental means now to closely outline the ‘psychical’ part of the nervous system. (For this latter problem, to use a probably justifiable simile from our plan of the brain, we should have been able to produce in a living being a state of pure automatism where all the actions, even complicated ones, could be elicited but without conscious- ness; and we should then be able to examine the brain for the cells which have been ‘ paralyzed’ for this purpose in an iso- lated manner. This seems almost inconceivable with our pres- ent methods and means, and it will therefore be a utopian task to search in this way for the pusely psychical ‘ Anteile,’ or superstructures of the nervous system. It is easy to see from this simile that there are extreme difficulties in the way of demonstrating which neurone-complexes are absolutely essential for psychic happenings, and also, that it would be premature to speak of the ‘psychic’ neurones simply because we suspect that they are involved in conscious activity. To call the neu- rones of the pyramidal tract (the motor efferent cerebral neu- rones) ‘voluntary’ motor tracts, may not be far from the truth, but their action is not necessarily conscious (hysteria, epileptic 294 JouRNAL oF CoMPARATIVE NEUROLOGY. equivalents, etc.). If we admit that practically every sensory- motor reaction, even the most complicated one, can be uncon- scious-automatic, under certain conditions mentioned, we con- clude that a further mechanism of differentiation and associa- tion must enter into activity to allow the quality ‘ psychical’ to come in as an additional biological phenomenon. These may be arranged all over the cortex, after the type of the ground-bundles for the real segments of the neural tube, or more definitely, in part at least, as Flechsig suggests by his association-centers. | This point can certainly not be called settled now. Flechsig’s data do not with necessity point in the direction of his interpretation. It is to be hoped though, that, by exclusions, we shall-learn with sufficient accuracy which cells belong to the ‘psychical’ mechanism before the above positive experiment can be looked for). In an analysis of the anatomical and physiological data we must avail ourselves of the biology of the cell, but not forget in doing so the whole organism. We do well to bear this in mind when we speak of neural mechanisms. We do not use the word mechanism in the sense of mechanics, but work with biological factors, A biological entity, animal or plant, would roughly speak- ing show us two sides of life, vegetative and reactive. The two cannot be separated absolutely ; but we can speak of these two sides or aspects, both of the organisms as a whole and of the individual cells at each moment. This is best illustrated by the following criticism of the reflex-arc. In most text-books the reflex-arc is represented by the following elements: (1) The afferent neurone, and (2) the efferent neurone in connection with a muscle. It seems more true to biology and to our actual knowledge of the nervous system, though, to think of the fol- lowing points ina reflex, illustrated in the description of the seg- ments: Suppose that it is possible to irritate one isolated sensory nerve-element, although the interlacing of the terminations in the skin would almost preclude even this. Assuming then, that we can stimulate one afferent neurone independently by a prick of the sole, we find that this cell sends many collaterals to sev- Meyer, Data of Modern Neurology. 295 eral segments of the neural tube, and towards many types of cells. Among them are cells connected with various groups of muscles (various kinds of efferent neurones, see p. 135), and intermediate cells, segmental or supra-segmental (see p. 141, etc.). Each cell is in a definite state of nutrition and by this and its connections represents the sum-total of its previous life ; it is, as it were, ‘attuned to’ a certain range of reaction, will react to certain stimuli only with a characteristic action, while other influences would leave it passive. Among the many cells which are reached by the afferent nerve only those will react which are ‘attuned to’ a definite impulse and give the ade- quate reaction. Under no circumstance is it possible to stim- ulate ONE efferent cell, but a reflex-arc consists always of (1) an afferent (at least one), and (2) many efferent neurones of one group of coordinated muscles and (3) the necessary number of intermediate cells. If a certain stimulus is complicated, the state of excitation of the afferent neurone may be of sucha char- acter as not to ‘appeal’ directly to the segmental efferent neu- rones; cerebellar and cerebral afferent neurones however may re- spond to that form, and the functional groups of cerebellar and cerebral efferent neurones respond to the excitation received in the only way in which they can respond, through specially ‘at- tuned’ segmental efferent neurones. If the reaction is complica- ted we might call it automatic, following the customary con- cepts without subscribing to the uncritically accepted dogma that whatever is ‘automatic’ must once, have been voluntary and therefore conscious. Or the stimulus may produce such a state of function in the cerebral afferent cells that ‘ psychical’ elements are the only neurones which take it up or respond to it, and from this the state of activity may, or may not, spread on to definite groups of cerebral efferent neurones, reaching either the motor apparatus of expression in the case of simple thought and speech, or other groups of segmental efferent neurones (voluntary and vaso-motor action.) In this way indefinite numbers of selections can be obtained. And we repeat, that these unconscious or conscious reactions presuppose three types of conditions. (1) A special stimulation of afferent neurones, 296 JourNAL oF ComPARATIVE NEUROLOGY. segmental, or segmental plus suprasegmental; (2) a definite vegetative condition and definite functional connections of all the elements concerned ; and (3) definite groups of efferent ele- ments. This consideration shows us that a reaction is not to be compared to a piece of meat put into one end of a sausage machine and coming out as a sausage at the other; nor is it best to compare the process with a current; it is more correct to say that a chain or complex of nerve-elements gets into a state of coordinated activity ; one element after the other takes up the state of ‘tension’ and the cooperation of the whole chain represents the neural activity in any reaction. It is a wave of agitation or of action passing over very heterogeneous material, involving some elements but not others. Or if we take the simile of a current we must realize that each link has its state of action for itself, each cell perhaps in all its branches, and it does not get the action from another cell but merely the impulse. This seems perhaps a mere speculation. I do not hesitate to call itso; the ordinary reflex-diagram, however, is also speculation, evidently less ‘hampered’ by facts, although we see it so often that we acquiesce to it as a ‘fact.’ This concept of interaction of nerve-elements and the for- mation of systematic acts, etc., is in full harmony with the concept of the neurone. It considers many kinds of neurones with many interrelations, each with its value as a cell and yet part of the whole organism, capable of entering into many chains of neural activity. Experimentally these chains can be followed as Horsley and Gotch have done for electric stimula- tion, or perhaps by poisoning certain definite mechanisms, etc. But only a combination of all neurological methods will help us to arrive at more clearness. In the mean time the above plan would seem to furnish a fair working hypothesis. In clinical neuropathology, the chains of neural activity become accessible to observation in two ways : (1) The con- traction of voluntary and involuntary muscles, (and perhaps also of other tissues, in the retina, etc.?) and (2) the subjective mental side of the process. We observe odjectively, as it were, Meyer, Data of Modern Neurology. 297 the results of the action of segmental efferent neurones only; and subjectively perhaps only the action of psychical elements in the chain with their connections with the apparatus of speech, etc.? This seems fairly probable. We do not know ‘how’ we contract muscles although we may know what we ‘do’; nor do we know how the segment feels; we only know ‘how’ the ‘personal,’ or psychical, part of the chain reacts to what is going on in the segment. From the knowledge of these two factors we build up the psychological concept of ‘sensory- motor’ mechanisms. It would lead us too far, were we to analyze here the methods of reasoning of neuropathology. What has been said must be sufficient to show that the connection of clinical with anatomical neuropathology has many difficulties apart from the difficulty of direct observation. The interpretation of findings is, indeed, very difficult as soon as one leaves the beaten tracks of the clinical parade-types. In current localization the great questions are: peripheral, spinal, central or psychic for the quantitive motor disorders; cerebellar or ‘sensory’ for the ataxias ; peripheral, spinal, central or psychic for the disor- ders of sensation. For the parade forms of nervous diseases these general types of localization may be sufficient: but there remain enough cases which will not fit, and enough diseases which give rise to a flood of perhaps unnecessary literature, simply because the localization-spirit is poorly directed by the spirit of faulty general pathology. The inefficiency lies not in the desire for localization, but in the use of an exclusively topographical \ocalization. The localization furnished with experiments of irritation or destruction of parts of the nervous system—we called it the scalpel-method—is roughly topographical. Now it lies in the very nature of the plan of the nervous system given with the views developed above that coarse morphological topography must yield to, or at least accept, the assistance of functional lo- calization. This, we have seen, drops the divisions of the ner- vous system into brain, medulla oblongata, spinal cord and peri- 298 JourRNAL oF CoMPARATIVE NEUROLOGY. -pheral nervous system in favor of: segments plus supra-seg- -mental apparatus. We should have to write a book if we tried to discuss the arrangements of the facts available in the entire neuropathology onthe ground of the concepts dictated by our discussion. We cannot give more than a rough outline here. The neurone-the- ory is only part of the whole concept. Recent investigations have been turning around problems relatively immaterial for the general point of view to be taken in neuropathology. Detail- problems have dulled the interest in larger ones. Many stu- dents are remarkably well informed on all the shades of the ‘contact-theory,’ but with the same effect which we all know to ‘come when histologists work only with oil-immersions and for- «get to get a broad frame for the details with the use of low powers. We have seen that we get our neurological knowledge from three clearly independent series of data. Our efforts go in the direction of melting them into one, i. e. into an objective con- ception of the nature and work of the nervous mechanisms. The naive realism of ‘common sense’ achieves this correlation »very rapidly, usually in the sense of the center-theory. ‘Wher- ever you can destroy it isolately, there must be its center, and in the center are the images etc.;’ this is roughly expressed its motto. I shall try to show how an uncritical use of this point of view combined with a purely anatomical one leads to habits of ‘thought not altogether safe. Without entering on the polemics concerning localization, we can state that the ‘ motor region’ alone (Bastian) or the motor area plus the parietal (Starr) form »the principal part of the ordinary highest sensory motor me- chanisms. The path usually given for a ‘sensation’ is the one represented in the plan of the brain as afferent cerebral. The ‘segmental afferent neurone sends a branch to the nucleus of Goll cor Burdach (the former in the case of a lumbar segment, the latter in the case of the cervical) ; the impulse is communicated there to elements of the fillet and from these to elements of the radiation of the nucleus ventralis thalami. If this is correct we Meyer, Data of Modern Neurology. 299° _ should think that a cut anywhere in this path and limited: to’ this path should cause anaesthesia to any irritation. This we know to be established for two portions of the path only. First, the afferent segmental neurone as far as its entrance into the spinal cord ; and second, less convincingly, in the fillet from its nucleus to the thalamus; higher up only when a large part of the thalamic radiation and the cortex belonging to it is involved. In the ‘segmental region,’ the spinal cord, etc., the matter is by no means clear. Brown-Séquard, ina criticism of the experiments of Mott (which spoke in favor of the anatomical path) gave up the well-known plan of the immediate decussation which tries to do justice to the symptom-complex of the Brown-Séquard paraly- sis. He states that not only hemi-section of the cord but also section of the posterior roots in the upper thoracic segments may produce a hyperaesthesia of the hind-leg of the same side and an anaethesia of the opposite side. Further, if by a hemi- section of the cervical cord a contralateral anaesthesia is pro- voked, it can be reversed into hyperaesthesia by a second hemi- section in the thoracic cord, whereas the hyperaesthesia exist- ing on the same side passes over into anaesthesia. The Brown-- Sequard symptom-complex undoubtedly exists, but the anatom- ical explanations are hardly adequate. If we add further pecu- liar facts, such as the paralysis of a limb when all its ‘sensory’ roots are cut while the motor are intact (Sherrington), and all the scattered data on hyperaesthesia, on muscular, tactile and temperature senses, we should begin to feel that certain activi- ties of the segments escape our attention and make no psychi- cal impression except by the final results. The result of a sum represents items, but is xot a picture of the summation, just as little as the figure 7 should be a composition of the figures 3 and 4 as an evidence for 3 plus 4 being 7, or just as little as we are conscious of which nerves and muscles are going to contract and relax when we catch a fly. We are radically wrong if we try to translate psychic phenomena uncritically into anatomical structures and believe that the plan of analysis and subdivision of psychic processes can be read piece by piece in the parts of the anatomical substratum. To return to the simile: who 300 JOURNAL OF CoMPARATIVE NEUROLOGY. would think that from an analysis of the number 121 into 1 hundred, 2 tens, and 1 unit, we could getany dzvect light on the process of getting the sum of 30 plus 46 plus 45? And is it not the same effort if we believe we must speak of temperature fibers, muscular sense fibers, tactile fibers ?—-why not of hyper- aesthesia-fibers ?—while we only know small differences aniong the cells of the afferent segmental types? Are we not forced rather to think of thermic, static etc., arrangements? Yet Bechterew claims that the broad fibers of the posterior roots serve to the muscular sense and the fine ones to the cutaneous sensibility. And since we already possess the term aesthesio- neura (Minot and Baker) we shall soon speak of myoaesthesio- neura and dermaesthesioneura! It will be the task of neuro- logical research of the near future to go to the bottom of these segmental localizations with a full recognition of the danger of ‘psychological’ neurology and the possibility in mind that many or all the qualities of ‘sensations’ and reactions are products of the function of mechanisms and that the elements of the ‘ psychical’ products are not necessarily fit to be identified with the elements of the mechanism. This cannot, however, stop our eagerness for progress and we must admit that we are in perfect sympathy with the most eager localizer. Our hope for the progress and future of neuropathology rests on localization as a first step to the research on the zature of the lesion. We must associate certain symptoms with lesions of certain groups of neurones and we can do that without in the least becoming untrue to the above principles, and, if we look closely, this has at all times been the method of the more conservative. Com- paring many kinds of clinical types with the sets of lesion found 1 Ina discussion of biological monism some one objected by saying: On monistic ground psychology becomes simply a part of neurology. To any one who shares that fear I should recommend to sacrifice at once the desire for unity and to keep at least the three series of experiences apart: 1. The morphological- anatomical, 2. The ergetic physiological, and 3. The psychical. It would bea pity if a desire for monism would obscure the necessity of critically separating these three series in biology. I hope the readers will accept the term ‘ psycho- logical neurology’ as simply meaning the uncritical mixing up of facts of neuro- logical and of the psychical series, a) Ne ee Meyer, Data of Modern Neurology. 301 in them, could not help furnishing certain rules of localizations which are clearly established. The difference of opinion refers more to the interpretations than to the facts established. The interpretation is indeed the point requiring most caution and this becomes at once obvious when we realize that the function as we know it is the outcome of activity of a whole mechanism and that the defect of function is not to be identified with the function of the diseased element by itself but as part of the functional mechanism. Jf you cut one leg of a tripod, it will fall and still you will not claim that the tripod stands on this leg only, although its loss implies inability to stand. We have already stated that clinical neuropathology has only two series of observations to build on; the subjective— the observations and feelings of the patient, and the objective which is altogether depending on the segmental efferent neu- rones for its expression. The following sketch of general diag- nosis on ground of our working-hypothesis is not essentially ‘different from the ordinary plan; it has its origin only in the desire to use methods which will promote the principle of re- search, at the same time doing justice to all the facts. For this purpose we take with us @ fundamental idea the segmental plan of the body aud the extsience of supra-segmental mechanisms. We usually start with the study of the subjective series of data by ascertaining the general mental condition and reliability the general sensation, the function of the special senses and the common sensibility, because these are necessary data and because during their systematic examination many points con- cerning the second series, the motor phenomena, will attract our attention. We proceed as is commonly outlined in the text- books, searching methodically for areas with any of the known qualities of sensory disorders, The results must be so arranged that we readily connect them with a localization of the possible lesion, using the distribution and qualities of a disorder for a guide. We begin with the sensory examination of the olfactory segment, and proceed to the optic segment, which illustrates best the general method. We need a knowledge of the gen- 302 JouRNAL oF COMPARATIVE NEUROLOGY. eral visual power, and of the field of vision of forms and colors. If there is a defect we exclude first mechanical disorders (of re= fraction etc.), then affections of the retina and optic nerve; then we examine the reflex-irritability; if ascotoma was discov- ered, especially for the area of the scotoma. This is of special importance for the discrimination of certain forms of hemian- opsia. Finally we search for subjective optic phenomena (from the phosphemata to real hallucinations). Clinical and anatomical: pathology have furnished tables for localization and we dis- tinguish the symptom complexes of: 1. Mechanical lesions (optic disorders). 2. Lesions of the retina. 3. Lesions of the afferent optic neurones. 4. Lesions of the cerebral afferent neurones. 5. Lesions of intracortical or general character.—We need not speak here of group I and 2. For group 5 are character- istic : concentric constriction of field of vision and reversion of the color-fields, and the more complex subjective phenomena. (hallucinations, fortification lines etc.). Lesions of group 3 and: 4 are distinguished largely on anatomical grounds (presence‘or: absence and kind of hemianopsia), the involvement of the ‘ re- flex-arc,’ the appearance of the disks and especially accompas panying lesions of other mechanisms, as hemianaesthesia, hemiplegia, mimic paralysis, etc. We must make here the reservations necessary on account of the peculiar course of the reflex path which is still under discussion (Redlich, Bechterew, Massaut). Next would come hearing (mechanical, -segmental af- ferent or cerebral afferent lesions, or psychic condition ?), taste (lesion limited to the segment of the 5th, 7th or oth?) and fin- ally the general sensibility. It would go altogether beyond the domain of this essay to detail all these points. I have done as much asI did in order to illustrate the method of correlation of data of clinical and anatomical pathology, and merely add, that for a diagnosis of ‘sensory’ lesions we must know the do- main of the branches of peripheral nerves, of the plexus, of the nerve-roots and spinal and cranial segments, the. hemian- Meyer, Data of Modem Neurology. 303 aesthesias with or without hysterical stigmata, or stereognostic disorders, etc. In the objective motor sphere we look out for affections of the muscles and groups of muscles in reflex and complex activ- ities (referring them in a similar manner as in the sensory sphere to special nerves, plexus, or segments, according to their dis- tribution) azd the character of the disorders present. If we find any disorder anywhere, we have to examine first for those symptons: which are associated with lesions of the segmental efferent neurones (the dect motor neurones of Goldscheider). This is necessary, because motor symptoms produced by any disorder of a mechanism will only come out clearly, tf the links between it and the muscle are in a normal condition. It is therefore of great importance that we know very valuable tests for the condition of the segmental efferent (direct motor) neu- rones. The cardinal points (flaccid paralysis with atrophy of the muscles, absence of reflexes and electrical reaction of de- generation) are to be excluded in every instance before a dis- order is looked for beyond. With a lesion of the indirect mo- tor neurones (the cerebral efferent or voluntary motor appara- tus) we associate an involvement of certain groups of muscles coordinated for special movements and joints, a tendency to rigidity, usually no atrophy, exaggerated reflexes and clonus, but normal electrical reaction, i. e. none of the symptoms asso- ciated with lesions of the direct or segmental motor neurones. Another group of symptoms is characteristic for mechan- isms of a more specialized, perhaps psychical character, com- monly called the group of psychic motor disorders, usually in- volving the entire limb (for instance both legs) or only special uses of the limbs (astasia-abasia, etc.). A further type of motor disorders which requires inves- tigation is formed by the ataxias, associated either with lesions of the motor cortex and the afferent cerebral neurones, the fillet (motor ataxia), with tumors of the frontal lobe (?), with lesions