IMAGE EVALUATSON TEST TARGET (A^ ' 3) 1.0 I.I m ^ lis lio 1.8 125 1.4 1.6 ^ 6" ► V Hiotographic Sciences Corporation 23 WEST MAIN STREET WEBSTER, N.Y. MS80 (716) S73-4S03 •^ m\ ^< i\ "^ ''^" . (meaning "CON- TINUED"), or the symbol y (meaning "END"), whichever applies. Maps, plates, charts, etc., may be filmed at different reduction ratios. Those too large to be entirely included in one exposure are filmed beginning in the upper left hand corner, left to right and top to bottom, as many frames as required. The following diagrams illustrate the method: Les images suivantes ont M reproduites avec le plus grand soin, compte tenu de la condition et de la nettet6 de i'exemplaire film6, et en conformity avec les conditions du contrat de filmage. Les exemplaires originaux dont la couverture en papier est imprimde sont filmis en commen^ant par le premier plat et en terminant soit par la dernidre page qui comporte une empreinta d'impression ou d'illustration, soit par le second plat, selon le cas. Tous les autres exemplaires originaux sont film6s en commengant par la premiere page qui comporte une empreinte d'impression ou d'fMustration et en terminant par la dernidre page qui comporte une telle empreinte. Un des symboies suivants apparaitra sur la dernidre image de cheque microfiche, selon le cas: le symbols — ► signifie "A SUIVRE", le symbols V signifie "FIN". Les cartes, planches, tableaux, etc., peuvent dtre film6s A des taux de reduction diffirents. Lorsque le document est trop grand pour 6tre reproduit en un seul clich6, il est film6 d partir de Tangle supdrieur gauche, de gauche A droite, et de haut en bas. en prenant le nombre d'images ndcessaire. Les diagrammes suivants iilustrent la mdthode. 1 2 3 ; "• 2 3 ^■■::* -^ 5 6 /-.. .^ r .... ^— -..o : ON THE STRUCTURE, MICRO-CHEMISTRY AND DEVELOP- MENT OF NERVE CELLS, WITH SPECIAL REFER- ENCE TO THEIR NUCLEIN COMPOUNDS By F. H. Scott, B.A. [■ ii I II 31 1^: A ^ I I L'ni\krsity of Toronto, Juih- 3, xSqq. To llu A'r:,'is/rar IJiiiTi-isi/y of Toronto Sir, I hi'ji to report that Mr. F. II. Scott, B.A., lias passed satisfactorily the examination in Physioiog;-, his major subject for the dejjree of Doctor of Philosopliy. I beg' (o report also that Mr. Scott's thesis "On the Structure, Micro-Chemistry and Development of Nerve Cells with Special Reference to llieir Xuclein Compounds ' Is of distinction as a contribution to Physiology, and I recommend that it bo acceptcc! for the degree of Doctor of Philosophy. Vours truly, A. B. MACALLUM, Pn}/rssor of Physiologist , I IIKRBBV certify that the thesis above mentioned has been accepted by the Sena'? of the University of Toronto for the degree ot Doctor of Philosophy in accordance with the terms of the Statute in that behalf. Univkrsitv of Toronto, June 5, iSgg. JAMES BRKBNER. Ri-Kisliii 1; y ON THE STRUCTURE, MICRO-CHEMISTRY AND DEVELOP- MENT OF NERVE CELLS, WITH SPECIAL REFER- ENCE TO THEIR NUCLEIN COMPOUNDS* I' ,?, iScK). ( Repi'inted by permission from the Transactions of the Canadian Institute, i8g8-gg. } isfai-toril^- tlie Philosophy. :ro-Chemistry Compounds " t Ije accepted PhysiolO),') • the Sena''^ •rdance with Rixisfra)', The finer structure of the nerve cell has attracted a great deal of at- tention in the last few years, chiefly because the cell body contains masses that have a peculiar affinity for certain nuclear stains. These masses were first observed in 1 882 by Flemming,^ who was not certain whether they wore nodular thickenings of the ordinary protoplasmic fibrillae or independent structures. His preparations, however, were, for the most part, from material that had been fixed in chromic or osmic acid, and stained in haematoxylin or carmine, and for this reason the bodies in question did not exhibit any distinctive staining properties. It was reserved for Nissl" who examined the < ells of the cerebral cortex of mammals after fixation in alcohol and staining in basic aniline dyes, to show that these bodies stain differently from the remainder of the cell protoplasm, and in fact resemble in this respect the large nucleolus. For this reason these structures are commonly called Nissl granules or " Schollen." Some observers have employed other names, such as tigroid bodies, chromophile corpuscles, basophile or basic substance, cytoplasmic chromatin, etc. The variable form exhibited by nerve cells from different sources with respect to these granules makes the selection of a suitable name based on morphological data difficult, but for the purposes of this memoir as •A shc-t account of some of the facts recorded here was given for me hy Prof. Macallum before the Fourth International Physiological Congress, Cambridge, 1898, and the British Medical Association, Edinburgh, i8g8. See Journal of Physiology, XXIII, supp. p. 33, and British Medical Journal, September 17th, 1898. I Flemming, W., "Vom Bau der Spinalganglienzellen," Festgabc fiir J. Henle, p. la, 1882, Also: " Zellsubstanz, Kern and Zelltheilung," p. 41, i88a. 3 Nissl, Fr., " Ueher die Untersuchungsmethoden der Grosshirnde," Tagebl. der Versammlung dcutschr Naturforscher, Strasburg. p. 506, i88j. ^ 4 SCOTT : STRUCTURE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS far as it pertains to the nerve cells of adults, the name of Nissl granules will suflfice. It will be shown later that this name, in some cases at least, implies an incorrect inference as to the mode of occurrence of the chromophilous substance in the cell. The chemical properties of the Nissl granules have been studied by Held,' Eve,^ Mackenzie,' Biihler'* and others. Held found the granules were soluble in dilute alkalies, did not digest in pepsin and hydrochloric acid, were not acted on by acids and gave no reaction with Millon's reagent, Adamkiewicz's or the xanthoproteic tests. Held, however, obtained a positive reaction for phosphorus by the employment of Lilienfeld and Monti's test for that element. He concluded from these reactions that the Nissl granules were of a nucleo-albuminous nature. Eve, however, was doubtful whether the granules were really dissolved in the alkali or were merely altered in their staining powers ; and found that after treatment with acids or salt solutions, the granules stain more diffusely. Blihler found the granules were soluble in salt solutions as well as in alkalies. It has recently been observed that nuclear chromatin gives with the Millon reagent a definite reaction, and Macallum' has shown that the reaction of Lilienfeld and Monti does not differentiate the phospho-molybdate, formed by the combination of the molybdate employed and the phosphorus of the cell, from the ammonium molybdate which has simply been absorbed and retained. The only undisputed evidence, therefore, adduced by Held in favour of the nucleoproteid nature of these granules is their resistance to digestion. His conclusion is, however, further supported by the obser- vation of Mackenzie who obtained, after treatment with acid alcohol, a reaction for iron in the granules. In the present research the micro-chemistry of the nerve cell has been reinvestigated by the more recent methods and the results indicate that Held's conclusion is correct, although, as we have seen, based on insuf- ficient grounds. 3 Held, Hans. " Beitrage zur Structur dcr Nervenzellen und ihrer Fortsatze." Erste Abhandliing, Arcliiv. f Anat, u. Phys., Anat, Abth., p. 396, 1895, Zweite Abhandlung, ibid, p. 304, 1897. 4 Eve, F, C. " Symp,ithetic Nerve Cells and their Basophil Constituent in Prolonged Activity and Repose," Journal of Physiology, XX, p. ^34, 1896. 5 Mackenzie, J. J., " Investigations in the Micro-chemistry ot Nerve Cells," Report British Assn., Toronto Meeting, p. 832, 1897. 6 Buhler, Anton, " Untersuchungen iiber den Bau der Nervenzellen," Verhandlungen der Phys. Med. Gesell. zu Wiirzburg, XXXI, p. 285, Veriag von SUihel, 1898. 7 Macallum, A. B., "On the Detection a.id Localization of Phosphorus in Animal and Vegetable Tissues," Proceedings of Royal Society of London, Vol. LXIII, p. 467, 1898. VE CELLS SCOTT : STRt'CTtRE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS issl granules :ases at least, encc of the I studied by the granules hydrochloric ith Millon's Id, however, ployment of 1 from these nous nature. Ily dissolved ; and found s stain more solutions as ir chromatin icallum' has differentiate : molybdate n molybdate Id in favour sistance to the obser- d alcohol, a ill has been idicate that on insuf- :ed Activity and t British Assn., ngen der Phys. and Vegetable The mode of occurrence of these granules in embryonic and foetal cells has evoked considerable interest. Vas" and Eve found the chromo- philous substance uniformly distributed in the nerve cells of foetal rabbits, and Szczawinska'' observed the same for embryonic cells of selachians. Blihler noticed that the granules were entirel)' absent from the nerve cells at an early stage. Timofcow*" observed that in the inter- val between the fourth and sixth day of incubation in the chick, the chromophilous substance increa.sed markedly in amoinit and was uni- formly distributed. None of the above observers seem to have suspected any other than a cytoplasmic origin for this substance and none of them have followed out in detail the appearance of this substance in the cell. The nucleo- proteid nature of these bodies suggested the nucleus as a possible source of the substance forming them, and this inference has been confirmed by a series of observations made on mammalian and avian embrjos. Further evidence in support of the nuclear origin of these bodies is found by the examination of the structureof the nerve ceils of animals in which no Nissl granules occur. These observations will form Parts 2 and 3 of the present memoir, while Part 4 will be devoted to the discussion of certain general considerations with respect to the structure of the nerve cell that have recently been the subject of much investigation. The question of a good fixing agent for nerve cells has been discussed by many writers but more particularly by Flemming,*' v. Lenhossek'" and Held." Flemming, v. Lenhossek and with them many others find that saturated aqueous sublimate is the most satisfactory fixing fluid for nerve cells. Held, believing in the foam-like structure of protoplasm, does not consider it as good as other fluids. Besides sublimate, Carnoy's fluid, Flemming's fluid and picrosulphuric acid are generally found to give good results. With all these fluids fair results were produced, but the sharpest granules and the clearest intergranular substance were obtained by using the modification of Foa's fluid as recommended by Bensley," viz., equal 8 Vas, Friedricli, "Studien iiber den Bau des Chromatin in der Syinpathischen Ganglienzellc," Arch, f. MIk. An.1t.. XL, p. 375, 1892. g Szcz.iwinska. VV., " Rccherches sur le systeme nerveux des Selaciens," Arch, de Biologie, XV, p. 463. 'i^T- 10 Timofeew, D., " Beobachtunffen iibjr den Bau der Nervonzellen des Spinalffanglien und des Sym- patheticus beim Vogel," Inter. Monat. f. Anat. u. Physiol, XV, p. 239, 1898. 11 Flemming, VV.. "Ueberden Bau der Spinalganglienzellen bel Saugethir.rcn, und Bemcrkungen iiber den der centralen Zellen." .\rch. f. Mik. Anat,, XLVI, p. 379, 1895. 12 V. Lenhossek. M., "Ueberden Bau der Spin.-ilganglienzellcn des Menschen," Arch. f. Psychiatrie, XXIX, p. 34S. 'Sgr- 13 Held, H., 1. c. and Arch. f. Annt. u. Phys., Supp., p. 273, 1897. 14 Bensley, R. R., "Mamin.-ilian Gastric Glands," Proceedings of Can.-idian Institute, Vol, I, Part i, p. II, 1897. I SCOTT STRUCTURK, MICRO-CHEMISTRV AND DEVELOPMENT OF NERVE CELLS parts of sublimate saturated in ninety-five per cent, alcohol and of a two per cent, solution of potassium bichromate in water. Small pieces were left in the freshly prepared mixture for two to four hours, washed in fifty per cent, alcohol, and then passed through the grades of alcohol. Material intended for chemical investigation was fixed in alcohol. The cells obtained from alcohol fixation are not materially different from those obtained with other fluids. The cone of origin and tlie process of spinal ganglion cells have nearly the same appearance in well preserved alcohol tissue that they have in sublimate material. Flemming's failure to get good results with alcohol may have been due to the circumstance that he did not leave his tissue in the alcohol for a sufficient time. Three days in alcohol as in Flemming's"^ method is not enough to insure complete coagulation of the proteids of the cell. After fixing and hardening the material was imbedded in paraffin, using oil of bergamot for infiltration. Sections were attached to the slide by the distilled water method and stained. I.— The Structure and Micro-Chemistry of the Nerve Cells of Mammals. It is generally believed that three substances enter into the formation of the body of nerve cells: (i) the Nissl granules, (2) a spongioplasm that is generally believed to be fibrillar but which may be reticular, and (3) a hyaloplasmic ground substance in which the two former are em- bedded. As this structure is found in the nerve cells of mammals and the nerve cells of this class have been most frequently studied, they will form the subject of this section. Material was used fromi the following animals : — man, ox, pig, sheep, dog, cat, rabbit, guinea pig and mouse. In most cases pieces from the cortex, cerebellum, cord, spinal and sympathetic ganglia were obtained and fixed in various fluids, but by preference in alcohol and the bichlor- ide-bichromate mixture. The shape and distribution in the cell of the Nissl granules are best demonstrated by staining sections fixed to the slide for a few minutes in an aqueous solution of toluidin blue or methy- lene blue, but preferably in toluidin blue, which v. Lenhossek regards as a specific stain. Afterstaining, the sections are differentiated in a mixture of aniline and alcohol, cleared in oil of bergamot and mounted in balsam. The results obtained with this method are similar in every respect to those obtained with the more laborious process of Nissl. «si c, p. 385. \ -^ p^E CELLS SCOTT : STRLCTl'RE, MICRO-CHEMISTRY AND DEVELOPMENT OF NEHVK CK1.L8 7 Nissl'" describes the bodies stained b>liis method, as having the form of larjjer or smaller, round, oval, spherical, often angular or irrcfjular masses which have tliread-like processes. These thread-like processes often unite the Hifferent masses into a true reticulum. Biihler" and Cox" for the spinal ganglion cells'" and Flemming-'" for the cells of the cord of Gadits have noticed this reticulum of chromophilous substance. The reticular nature of this substance is frequently seen in the spinal or sympathetic ganglion cells, or in the cells of Purkinj'J in the cerebellum, and is occasionally .^een in the cells of the cord and cortex. In sections stained with toluidin blue alone, the nucleus is seen as a clear space in the cell containing a large, round, deeply-stained nucleolus. There is usually nothing el.se stained in the nucleus, but occasionally there may be a faint bluish tint along certain lines. If instead of employing toluidin blue alone, wc use a cytoplasmic stain with it, we get the intergranular substance stained as well as the granules. The combination of eosin and toluidin blue, as employed by Mann, was the one used most frequently, although crythrosin and methy- lene blue, as employed by Held, give good results. Using these dyes we find the Nissl granules are stained blue, while the intergranular substance appears red. (Figs. I and 2). The nucleolus is also blue, but the blue is not the same as that of the Nissl granules, nor is the blue uniform throughout, for in many cases one can see a distinct red centre having a blue-stained layer on the outside. (Fig. 21). Probably the greatest change the addition of eosin to the stain has made in the appearance of the cell is in the nucleus. Here, instead of finding an un.stained sub- statKe, one sees stretching from the nucleolus to the nuclear membrane a network of eosin-stained material. This substance is generally abundant near the nucleolus arid adjacent to the nuclear membrane, while extending across the intervening space is a loose network of the same material. Sometimes, however, this material is found scattered throughout the nucleus in a finely granular form. This eosinophilous substance is generally more abundant in the nuclei of spinal and sympa- thetic ganglion cells than in the nuclei of cells of the central nervous 16 Nissl, F., " Mittheiiungen zur Anatomie der Nervenzelle," Allgemeinp ZeiUchrift fiir Psychiatrie, L, p, 37a, 1894, 17 Btihier, l.c., p. gS. 18 Cox, W, H., "Die SdbstanJigkeit der Fibrillcn im Neuron," Internat. Monat. f. Anat. und. Phys,, XV, p. 309, i8<)8. 19 Lugaro (Lo sperimentale, 1895), also observed the reticular nature of this substance. Quoted from Robertson, Br.iin, 1899, p. aia. ao Flemming, W., " Ueber die Structur centraler Nervenzellen," Anat. Hefte Heft, XIX, p. 563, 1896. (Original inaccessible. Quoted from Buhler). 8 SCOTT : STRLCTIRE, 5les, but not the nucleolus or the oxyphile nuclear substance which are related sub- stances. The slight degree of alkalinity necessary to alter the granules suggested that the blood, which is really more alkaline than some of the solutions used, might act in a similar way. This was tried and found to be the case. After loose sections of a spinal ganglion that had been fixed in alcohol had been in fresh defibrinated ox-blood for twenty hours, the granules were altered in the same way as if they had been in potash or soda solutions for the same time. We thus find that the gran- ules, as they occur in the cells after fixation, are altered by the animal's own blood. Eve observed that salt solutions had little action on the granules, but Biihler found the granules were soluble in physiological salt solution in twenty-four hours, leaving vacuoles in the cell. My results coincide with those of Eve, for when fresh spinal cord and ganglia were left in salt solutions for as long as three days at room-temperature the substance of the Nissl granules was still present. The cells contained vacuoles, forcing the granules into distorted shapes, but the substance stained normally with toluidin blue, and contained iron. In one ca?e, after material had been in the salt solution for three days, the granules were so altered that they would not stain with toluidin blue. On examina- tion the salt solution used was found to be distinctly alkaline, but in all cases where neutral salt solution was used the substance of the granules was not removed. Leaving fresh material in distilled water for five days at the tempera- ture of the room does not alter the staining powers of this substance, although the cell may contain vacuoles. Hardening material, however, by putting it into boiling water, has an action on nerve cells somewhat similar to the action of dilute alkalies. If the boiling has been continued long enough the granules will not stain with basic dyes and the iron cannot be detected in them with the acid alcohol method. The distri- bution of phosphorus is, however, normal throughout the cell. Held failed to obtain a Millon reaction in the granules. A Millon reaction may, however, be obtained throughout the cell body, the nucleolus and oxyphile nuclear substance, if sections of material fixed in alcohol are left in freshly prepared Millon reagent for several hours at room temperature. Besides the granules, the nerve cells frequently contain r. yellowish 14 SCOTT : STRUCTURE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS i i i! pigment in their body. The pigment has been found to be especially common in man and monkeys. (Warrington V The pigment present in the cells of a thoracic sympathetic ganglion of an ox, after it had been hardened in alcohol, gave the following reactions. It was still present after a one per cent, solution of potash had acted on loose sections for three days at room-temperature. It was not removed from the free sections by the action for a week of one per cent, hydrochloric acid .solution, nor did it give, after the use of acid alcohol, any reaction for iron, which confirms what Warrington found for the pigment present in the nerve cells of man. It did, however, give a positive reaction for phosphorus, using Macallum's test. Before leaving this section I would like to discuss the structure of the nucleolus. There is always one, and there may be several, nucleoli present in the nucleus of the nerve cells of mammals and in most other clas-ses of animals; but there is rarely a nucleolus in the nerve cells of the Urodela and if present it cannot be distinguished with certainty from the remainder of the nuclear chromatin. The nucleolus is considered by most observers to consist of a single substance which may be vacuolated. Several observers, however, have described the nucleolus as consisting of fine grains embedded in aground mass. This view is supported by v. Lenhossek,*- Held," Ruzicka," Obersteiner,*'"' but more particularly by Timofeew" who says the nucleolMs consists of basophile grains embedded in an oxyphile ground substance. The nucleolus consists of two substances, but the relation of these two is different from that usually described. I find the nucleolus is a vesicle with an oxy-centre and a basophile covering.^' This relation is often seen in sections stained with eosin and toluidin blue, or in material fixed in Flemming's fluid and stained with his orange method. A somewhat similar structure has lately been described by Heimann*", who noticed the periphery of the nucleolus had a great affinity for stains. This structure is best seen in the nerve cells of rodents but occurs in 4> W.-irrington, W. B. "On the Structural Alterations observed in Nerve Cells," Journal of Physiology, XXIII, 1898. 41. V, Lenhossek, I. c. 43. Held, Archiv f. Anat. u. Physiol,, p. 207, 1897. 44. Ruzicka, Zeit. i. Wiss, Mikroskopie, p. 452, 1897. 45. Obersteiner, Zcit f. Wiss. Mikroskopie, p. 60, 1898. 46. Timoteew, l,c. 47 Mackenzie also observed this relation in the nucleolus. Ora Communication, British Association Toronto Meeting, 1897. 48 Heimann, I. c. »VE CELLS be especially ment present after it had It was still ed on loose amoved from hydrochloric my reaction lent present reaction for cture of the ■ai, nucleoli most other rve cells of h certainty of a single 'ever, have n aground Ruzicka," says the ile ground these two s a vesicle aften seen il fixed in somewhat o noticed occurs in f Physiology, SCOTT : STRUCTURE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS •5 all animals and in well-stained sections is easily observed. Vacuoles are also quite frequent in the nucleolus, a fact which has attracted the notice of several observers. That this is quite correct is shown by the action of alkalies or of digestive fluids on the nucleolus. The action of digestive fluids in sometimes leaving a shell of undigested material has been referred to, but the effect of alkalies is more convincing. Held found that after prolonged treatment in the alkali the nucleolus no longer stained with methylene blue, and he thought that this showed that the nucleolus was formed of fine grains embedded in a ground mass. Alkalies have an altering action on the nucleolus similar to that on the Nissl granules but the action must be prolonged. If tissue which has been fixed in sublimate is used the action is very slow and one can often find the outer covering of the nucleolus broken, between the portions of which the oxyphile centre may be seen. This structure can be seen in sections stained with eosin and toluidin blue, or in iron-alum haematoxylin, but the clearest way of demonstrating it is the gold method of Apathy.^" Figs. 9 and lO are the nuclei of cells that have been treated with potash and then stained with this methou. The oxyphile centre can be seen between the pieces of the basophile covering which has undergone frag- mentation. The above considerations render it clear that there are at least three distinct nuclein compounds in nerve cells, the Nissl granules, the baso- phile covering of the nucleolus and the oxyphile nuclear substance. Each of these bodies contains iron and phosphorus, the usual constituents of many nucleo-proteids. Van Gehuchten'" and CajaP believe the nuclein is coiidensed into the nucleolus, while v. Lenhossek maintains that the nerve cell does not contain true nuclein or chromatin. There seem to be many different nuclein compounds in different cells, but we shall see that for the nerve cells these different nuclein compounds are genetically related, and that intermediate substances are found in the nerve cells of different animals. 49 Apathy, Stefan, "Das leitende Element des Nervensystems, etc," Mitth. aus der Zooi. Station zu Neapel, XII, p. 49,1;, 1897. $0 Gehuchten, A. van, " L'anatomie fine de la cellule nerveuse," La Cellule, XIII, p. ,113, 1897. SI C.ijal. S. R., Revisto Trimensal Micrografica, 1896. (Original inaccessible, quoted from van Gehuchten). Association |6 SCOTT : STRUCTURE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS [ ' n.— Tut: Development of Nerve Cells with Special Refer- ence To THE Development of the Chromatic Substance of the Cell Body. Several attempts have been made to determine the origin of the Nissl granules, but all have failed to detect it. Vas made some inter- esting observations on the chromatin of foetal sympathetic ganglion cells but did not attempt to ascertain the origin of the chromatic sub- stance of the cell body. Eve found the chromatic substance completely filled the cell body at an early date. The cells of the vagus ganglion were the first to show an appearance like the adult cell with regard to the distribution of this substance. Szczawinska, working with selachian embryos, did not trace it further than the stage in which the cells were uniformly stained. Buhler" states that foetal cells are devoid of granules, but does not ascertain the origin of the granular substance. He did, however, notice that the nuclei of young nerve cells are baso- phile and gradually become oxyphile as development proceeds. Timo- feew observed that in chick embryos the basophile substance increased markedly in amount in the cells of the spinal ganglia, between the fourth and sixth day of incubation. He says nothing of its origin and evidently considers it cytoplasmic. The chromatic substance which forms the Nissl granules is undoubt- edly derived from the nuclear chromatin. A series of pig embryos from 7mm. onward to birth was the chief material used for these obser- vations, but calf, sheep, rabbit and chicken embryos were used to confirm the results. The embryos were fixed in the bichloride-bichromate mixture or 'a picro-corrosive fluid. Material intended for chemical methods was fixed in alcohol. The development of the chromatic substance whicli forms the Nissl granules is closely connected with the morphological development of the cell. His^ showed that the neuroblasts are derived from cells, lying, in mammals, next the medullary canal, which he calls germinating cells. These cells have a protoplasm which may be divided into an outer clear 52 L. c, p. 46. 53 His, W., "Die Ncuroklasten und deren Entstehung im embryonalen Mark," Arch. f. Anat. u Phys,, Anat. Abth., 1889, p, 249. Also, " Histogenese und Zusainmenhang der Nervenelemente," Arch. t. Anat. u. Phys., Supp. 1890 p. 9S- IVE CELLS SCOTT : STRUCTIRE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS I 7 and an inner granular layer. The next distinct stage which His dis- tinguishes in the development of the nerve cell is the neuroblast phase. Here an oval nucleus bears a conical cell body, and this in turn is con- tinued into a long process. The nucleus is moderately rich in chromcitin of which there are several mas:^es united by a filament. There are no protoplasmic processes and the protoplasm around the nucleus is very scanty. The neuroblasts arise in the iimer layer from the germinating cells and oass out secondarily into the mantle layer of the wall. In the transformation of the germinating cells into neuroblasts His distinguishes five stages : (i). Germinating cells of round form with a broad protoplasmic body. (2). Germinating cells of round form with initial point and broad pro- toplasmic mantle. (3). Intermediate cells of pear shape with little protoplasm around the unclosed nucleus. The cytoplasm is continued into a long process and the cells may still lie close to the internal membrane. (4). Intermediate cells of pear shape with closed nucleus, deeply staining, outer cone, but little protoplasm around the remainder of the nucleus. (5). Finished neuroblast. My observations confirm those of His on the origin of the neuroblasts and in addition show the fate of their chromatin, a point not touched upon by His. Germinating cells occur in the pig from the earliest stage procured by me (7mm.) to that of i8mm. length. If a section of a cord of, e. g., a lomm. pig, is stained with eosin and toluidin blue, one finds that all the blue-staining substance in the germinating cells is confined to the chromatin of their nuclei. The reactions for iron demonstrate that the cytoplasm is devoid of substance containing this element. At this period all iron-holding material is confined to the chromatin of the nucleus. These cells are of round or oval shape, (Fig. 12), and are in some stage of mitosis. The cytoplasm is free from iron-holding material, or from material staining with toluidin blue. The cells are sometimes in the loose-skein phase, sometimes in the dyaster stage but most frequently in the equatorial-plate phase. The cells are generally in the equatorial plate phase when the process begins to be formed. As the cone in- creases the chromatin becomes excentric (Figs. 13 and 14). There is I l8 SCOTT : STRUCTURE, MICRO-CHEMISTRY AND TEVELOPMENT OF NERVE CELLS Still no nuclear membrane and still no iron-holding nuclein compounds in the cell body or process. The cell bodies appear, quite frequently, reticulated. The equatorial-plate stage is soon passed, and the chromatin begins to distribute itself in the nucleus. Various steps in the distribution of the chromatin may be followed until a stage, such as is represented in Fig. IS, is reached. By this time a nuclear membrane has been formed. Usually several masses of chromatin are found touching the nuclear membrune, while others are found towards the centre of the nucleus, but all seemingly connected by filaments. There is, as yet, no oxyphile substance in the nucleus. These cells are usually found under the membrane of the medullary canal but in very young embryos they occur in the mantle layer. The nucleus is entirely excentric and the cell body runs out into a long process. The cell body and process are still entirely free from iron-holding nuclein compounds. As the cells pass outward into the mantle layer and become older, the substance having affinity for toluidin blue in their nuclei disappears and a substance with more affinity for eosin takes its plac^. Syn- chronous with this change, a substance with great affinity for toluidin blue appears in the cell around the nucleus (Figs. i6 and 17). With the appearance of this .substance in* the cell body iron may be de- tected there for the first time. In this stage, which would correspond to stage 4 of the series described by His, there are several granular masses in the nucleus with marked affinity for toluidin blue, but the most of the nuclear chromatin stains intermediate between the red and the blue. As development proceeds (Figs. 18 and 19) the basophile substance in the nucleus continues to decrease, while the basophile substance in the cell body increases, and as it does so the affinity of the nuclear chro- matin for eosin also increases correspondingly. One part of the chro- matin does not alter but remains basophile and constitutes ultimately the peripheral portion of the nucleolus. Figs. 18 and 20 represent cells from the same embryo. The one indicates the distribution of the oxyphile and the basophile parts, while the other shows that both contain iron. In Figure 21 is represented a cell from the medulla of a 32mm. pig embryo. The basophile substance forms a homogeneous mass filling the cell body. The cell may be said to have undergone at this time complete development of its chromatic substance, for now the nuclear oxyphile substance is completely digestible, and it stains like the substance found in the nuclei of adult mammalian nerve cells. I RVE CELLS SCOTT : STRUCTURE, MlCRO-CHEMISl RY AND DEVELOPMENT OF NERVE CELLS •9 ;ss are still These facts indicate that the three nuclein compounds of the adult nerve cell, the Nissl granules, the nucleolus an' the vyphile nuclear substance, are derived from the chromatin of the nucieus of the ger- minating cell. This chromatin divides into two parts, each containing iron and phosphorus, but the one is oxyphileand remains in the nucleus, while the other is basophile and diffuses into the cell body and becomes the Nissl granules. The nucleolus seems to correspond in character to the chromatin of such a stage as is represented in Fig. 15, where little change has occurred irom the equatorial-plate phase. The cell body is filled with duTu.scd chromatin before the protoplasmic processes are formed, but as the cell grows and the protoplasmic pro- cesses ari.se, the diffused chromatin is formed into masses and these in turn into smaller pieces until the size observed in the adult is reached. The breaking up of the diffused chromatin into masses is probably due to growth, and not to functional activity as one might think from the results obtained by Vas and Eve, for the ganglion cells of the retina of a foetal calf of 60 cm. were distinctly granular. If the process of fragmen- tation proceeds far enough, the masses will be isolated in the cell, but if not they will constitute a reticulum. No evidence was observed of a connection persisting between this diffused chromatin and the nucleus. One criticism of the observations of His is necessary. His stage 4 in the development of the neuroblast should succeed his stage 5, for the description of the latter stage is of a cell in which the basophile substance has not yet diffused from the nucleus, while the description of the former stage is of a cell in which this diffusion has taken place. That such a mistake might arise is seen from the fact that the basophile substance is not distinct in the cell body for a long time after neuroblasts are formed and have migrated into the mantle layer. Thus germinating cells are found in all the stages from the earliest procured (7 mm.) to that in which the embryos are about i8mm. long, while neuroblasts of the type that His describes as developed are abundant in the mantle layer at 7mm. and continue to be so until the formation of neuroblasts ceases. The chromatic substance, however, is not abundant in the cell body till the embryo is about 15mm. long. Consequently the process of transformation and diffusion of the chromatin is going on while the embryo grows from at least 7mm. to 15mm. In an embryo of 15mm. there are still many neuroblasts in the mantle layer that have not a distinct colourable cone but only a thickened mass of basophile sub- stance around the nuclear membrane. A single section of an embryo pig from 15 to 1 8mm., since it contains Ir :' \ 20 SCOTT : STRICTURE, MICRO-CHEMISTRV AND DEVELOPMENT OF NERVE CELLS germinating cells and neuroblasts with diffused chromatin, ^hows all stages in the process of transformation and diffusion of the chromatin. This circumstance enables one to be sure that the gradual loss of affinil)' of the nucleus for basic dyes is not due to overstaining in eosin, for after the chromatin begins to change, an overstaining with eosin will make it red, but a shorter time in the eosin will colour it pur- plish. If one examines a section in which all stages of the diffusion of the chromatin are seen, one can easily seethe great affinity the nuclei of the nerve cells next the medullary canal have for toluidin blue. The nuclei of ceils lying farther from the canal have less and less affinity for basic dyes, but one can observe that with this lo.ss of affinity on the part of the nucleus for these dyes, a substance with marked affinity for them appears in the cytoplasm. This substance is first seen as a thickened nuclear membrane, but as more of it diffuses from the nucleus it takes the form of a distinct cone in the cytoplasm forming a cap on the nucleus. These observations have been confirmed on the cells of the cortex, cerebellum and spinal ganglia of mammals and also on the cells of the chicken. The chick embryos are not as suitable as mammalian embr}os for following this process, on account of the general distribution of yolk nucleo-proteids, but the stages can be distinctly observed. The chromatic material appears in the cytoplasm of the cells of the medulla before it appears in those lower down in the cord, but my observations on the rate of the development of the material in the different centres are incomplete. The action of dilute alkalies on the cell varies with the degree of its development. At a stage such as is represented in Figure 21 the action of alkalies is similar to their action on adult cells. After treatment with alkalies the cell body would not stain with toluidin blue or did not con- tain iron, but the body of the cell still remained. If the view of Held and Biihler regarding the nature of the Nissl granules is correct, then the whole cytoplasm should have disappeared. Alkalies are very slow in altering the staining powers of the chromatin in mitosis. The nuclei of germinating cells and of the nerve cells in which the kinetic chromatin was only slightly altered, contained a large quantity of material that stained with toluidin blue and proved to be iron-holding, after small pieces of tissue, that had been fixed in alcohol, had been treated with potash (0.2%) for six days. In this same solution the Nissl granules had been altered so that they would not give their ERVE CELLS SCOTT : STRUCTURE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS ordinary reactions in six hours, but the nucleolus of the nerve cell and the nuclei of the ncjroglia cells manifested their ordinary reactions, although more diffusely, at the end of the six days. On digestion of material which had been fixed in alcohol there was no appreciable effect on the nuclein compounds until a stage of which Figs. 19 and 21 are representations. In the stage illustrated in Figure 21, which is from the cord of an embryo of 32mm., all the oxyphile nuclear substance digests as in the adult, but in stages before this one, there is always some substance which does not disappear on digestion. After digestion the periphery of the nucleolus always remains. In this respect as well as in all its other reactions it resembles the chromatin found in primitive nerve cells. In order to facilitate reference to the different stages through which the chromatic substance passes in getting into the cytoplasm of the nerve cell I shall adopt the example of His and divide the process (arbitrarily) into different stages. Stage I. Germinating cells (Figs. 12, 13, and 14), stages I, 2, and 3 of His. The chromatin is confined to the nucleus and is in mitosis- Weak alkalies alter this substance very slov>?^ly. Digestion dissolves the cell body but does not alter the staining power of the chromatin. Stage 2. Neuroblast stage (Fig. 15), stage 5 of His. The chromatin is confined to the nucleus but is broken into masses. A nuclear mem- brane has been formed and the greater portion of the chromatin is dis- tributed around the membrane. Alkalies and digestive fluids have little or no power to alter the reactions of this substance. Stage 3. (Figs. 16, 17, 18, 19 and 20), stage 4 of His. Some of the kinetic chromatin is transformed into two other kinds, an oxyphile and a basophile. As the chromatin alters, the basophile part diffuses into the cytoplasm but the oxyphile substance remains in the nucleus. Most of the chromatin alters at the same rate but there may still be masses in the nucleus with marked aflinity for basic stains. Alkalies have an altering action on the diffused-out chromatin in extracting the iron from its substance, thus changing its staining reactions. Digestive fluids affect the nucleus but still leave the basophile parts behind. Stage 4. (Fig. 21). The transformation of the kinetic chromatin into the oxyphile and basophile kinds is now completed and the diffused basophile part fills the whole cell body. Alkalies alter the chromatin, especially the diffused part. Digestion dissolves the oxyphile substance I ill »J SCOTT : STRUCTURE, MICRO-CHKMISTRY AND DRVBI-OPMENT OF NKRVE CELLS completely but leaves the diffused substance and the periphery of the nucleolus unaffected. Stage 5. Adult cell. Differs from stage 4 only in the distribution of the chromatin which has diffused into the cytoplasm. III.— On the Structure of the Nerve Cells in Other Classes of Animals. r^! When such a remarkable cliange as that described in the development of the Nissl r mammals 5 this state- isek's paper large nerve lion cells of obable the ganglion. udied with From Reptilia : in the ganglia of Testudo, Emys, Uromastix, and Agama!* From Anura: in ^aw^j,**'* "* and ^«>." From Pisces : in the electric lobe of Torpedo^* in Gadus,^" in LeuciscHs^^ in CyprinusJ^ in Alopias?^ From Invertebrates: in the crayfish," earthworm,"" molluscs" " and insects." With the exception of Szczawinska, who states that granules are not present in a few cells of some rays, all these observers find a substance analogous to that of the Nissl granules present in the nerve cells of the different animals examined. I think the reticulum, in the cells without granules which Szczawinska figures, is made up of this chromatic material, for one frequently finds the chromatic substance in such a reticulum in some of the nerve cells of the earthworm, Limax and Limncea, while neighbouring cells have the chromatic substance in a granular form. In any case the observations on the occurrence of this substance in the nerve cells of the pigeon, frog, earthworm, crayfish and various molluscs were confirmed. The nerve cells of other forms which have not been studied, so tar as I know, by others, were examined, and this chromatic material was found in the following forms : From Reptilia : cells of cord and cortex of Chrysemys picta. From Ganoidei : cells of cord of Awi'a calva. 64 Pujfnat, Charles Amedee, " RecherchcH siir l.i structure des cellules dcs g.inglions spinaux de quelqucs reptiles," Anat. Anz,, XIV, p. 80, 1898. 6s V. Leiihossek. M., "Centrom unj Spliilre in den Spmalg.inglienzellcn des Frosches," Arch. f. Milt. Anat., XLVI, p. 345, 1895. 65 Dehler, AJolf., " Beitrag. zur Kenntniss von feineren Bhu der sympathotischen Ganglientellen dcs Frosches," Arch. f. Mik. Anat., XLVI, p. 724, 1895. 67 Biihler, I.e. 68 Rohde, I.e. 69 V. I^unhossek, M., "Der feiaere Bau, etc.," p. 159, 1895. 70 Fleinming, I.e. 71 Biihler, I.e. 72 Szczawinska, I.e. 7J PnlaJino, G., "Sur la constitution morphologique du protopl.-isma des cellules nerveuses ' Arch It. de Biol., XXIX, p. 60, 1898. 74 Pfliike, Max, "Zur Kenntniss des feineren Baues der Nervenzellen bei Wirbellosen," Zeit. f. Wiss. Zool., LX, p. t/Xi. 1895. 75 Eve, I.e. 76 McClure, Charles F. W., " The Finer Structure of the Nerve Cells of Invertebrates," Zcologi.che JahrbUcher, Abth. fi'ir Anatomie, XI, p. 13, 1897. 'iv.. A- l\ ■:■' .J, •si h W |ini! llMi I/,'! :l! 24 SCOTT : STRUCTURE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS From Teleostei : cells of cord of Amturus catus. In all cases this substance in the cell body, although distributed dif- ferentl}', stained with toluidin blue and gave the reactions for iron and phosphorus. In all cases tried th- substance was found to be insoluble in pepsin and hydrochloric acid but to be easily altered by dilute alkalies. The widespread occurrence of this substance in such diverse forms has been taken by some (Rohde," Marinesco'") to indicate that this material is an essential constituent of all nerve cells. This, however, is not the case, for in 1895 Biihler'-' described the cells of the forebrain o{ Lacerta agilis as frequently devoid of Nissl granules.and I find that the vast bulk of the nerve cells of the Urodela are absolutely devoid of them. It will therefore be necessary to enter into a detailed account of the nerve cells of these forms. Several specimens oi Nerturiis,A)nblystoma, Plethodon and Diemyctylus were obtained and the cord, brain and ganglia fixed in different fluids. A series of a Snlamnndra larva was also examined*" and series of larval Amblystomata of various ages were made. The nerve cells of all these different forms were found to correspond in their structure and characters. In the nerve cells of these animals the cytoplasm, instead of holding granules which contain iron and phosphorus and which stain with basic dyes, is often free from iron, phosphorus or substance staining with toluidin blue, and on the other hand, their nuclei, instead of containing very little basophile substance, abound in granules of such basophile material. This is true of ganglion, retinal and central nerve cells. If one fixes in Flemming's fluid and stains with his orange method there is no gentian-stained substance in the cell body while the nucleus is filled with granules and threads which stain deeply with the gentian. If instead of the orange method one uses safranin and light green, according to Benda's process, one finds all the substance staining with safranin confined to the nucleus. In material that has been fixed in alcohol or in sublimate, and stained 77 Rohde, I.e. 78 M.irinesco, G. " Recherehes sur la blotogie de la cellule nerveuse," Arch. t. Anat. und Phys., Phys. Abth., 1899, p. 89. 79 Bahler, Anton, " Protoplasma-Structur in Vorderhirnzellen dcr Eidectisc,' Verli. d, phys. mcd Ge»., WOraburg, Stahel, 1895. 80 For ihis privilege ' am indebted to Dr. J, Stafford. If'.;. ,,^,-^^;^ ERVE CELLS istributed dif. IS for iron and o be insoluble •ed by dilute rse forms has this material er, is not the in o{ Lacerta the vast bulk lem. It will e nerve cells Dieinyctylus erent fluids, ies of larval cells of all ■ucture and of holding I with basic ining with containing basophile ells. fc method le nucleus e jjentian. fht green, ling with d stained . und Phys., phys. med SCOTT : STRUCTURE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS 25 with eosin and toluidin blue, there is in the bodies of most nerve cells no blue-stained substance, while the nucleus is full of blue-stained granules and threads (Fig. 5). On staining sections in the Ehrlich- Biondi mixture, cne finds the cell body is red, but all the nuclear chro- matin is greenish, and there is no difference in the staining reactions of the nuclei of nerve and neuroglia cells such as is found between these cells in mammals. The reactions for iron (Fig. 7) and for phosphorus (Fig. 6) show there is no iron and little phosphorus in the bodies of most nerve cells. In a few cases a little basophile substance was observed in the cell body. In these the cytoplasm also contained a slight amount of iron and phosphorus-holding substance, but the much greater part of this substance, or of the substance staining with basic dyes, is in the nucleus A sufficient number of specimens to determine the cause of the presence or absence of this slight amount of basophile substance in the cytoplasm have not been examined, but when it is present, it is most frequently diiTuse and not in granular form, although the latter, in rare cases, has been seen. On digestion little material is dissolved from the nucleus, but the oxyphile substance, which was present in traces previously, has now disappeared (Fig 9). Those cells which contain a little basophile sub- stance in the cytoplasm retain it after digestion. The action of alkalies on the nerve cells of these animals is similar to their action on the neuroglia cells of the adult, or on the nerve cells of embryo mammals. Thus, after six days in a solution of potassium hydrate (0.2%) the nuclei still held a large quantity of material which contained iron and phosphorus, and which stained with toluidin blue. This same solution had removed all the basophile material from the cytoplasm of the nerve cells of adult mammalia in a few hours, but the nucleolus of the nerve cell and the t auroglia cells stained with basic dyes after six days, and the same was true of the embryonic nerve cells of mammals. The nuclei of the neuroglia cells of these Urodela, as in mammals, resist the action of alkalies. There is, therefore, in the former, no difference with respect to the action of alkalies between the nuclei of nerve and neuroglia cells. The slight amount of the basophile material present in some cells is easily and quickly altered by the alkali. For some reason, the transformation and diffusion of the chromatin has not proceeded, in the cells of the Urodela, past a certain stage, cor- It I '■ Ml 26 SCOTT : STRf CTLRE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS responding to stage 2, given above for mammalian development. Com- pare Fig. IS, which is the nerve cell of an i imm. pig, with Fig. 5, which is the motor nerve cell of an adult Necturus, fixed in the same fluid and stained with the same dyes. Besides the staining reactions the effect of alkalies or of digestive fluids is practically the same in both cases. Levi"' has examined the nerve cells of different types of Vertebrata ( Vespertilio, Cavia, Cams, Bos, Testudo, Zamenis, Rana, Triton, Proteus, Spelerpes, Tinea, Raja, Scyllium, and Petromyson), and has noticed the peculiar nature of the nucleus in the cells of the Urodela. He offered no explanation of the peculiarities of these cells, nor did he draw any conclusion about the nature of the substance in the cytoplasm of other forms. A comparison of nerve cells of larval Amblystomata with those of the adult form shows them to be exactly similar. There is no transforma- tion, except to a slight degree, of the chromatin into an oxyphile and a basophile part. I have noticed that those cells of the adult that had a little basophile substance in the cell body also had .some oxyphile substance in the nucleus. There are other forms that have not yet reached the adult or mam- malian degree of differentiation. Thus in Limax and Livincea (and from the descriptions of PflUke and McClure, in all Gasteropods) the cells have stop;ied developing at a stage between 3 and 4 of the mam- malian development. There is a quantity of iron and phosphorus- holding substance in the body of the nerve cell, but the nuclear chro- matin is peculiar. It is not affected by digestive fluids, it stains green with the Ehrlich-Biondi stain, it stains with safranin, and is generally purplish with eosin and toluidin blue, although, by long action of eosin and short action of toluidin blue, it may be quite red. Other forms (earthworm and crayfish) were also noticed to vary slightly from the mammalian type. I believe that if the nerve cells of all adult animals were examined, one would find a complete series in the diffusion of the chromatic substance to the cytoplasm. 81 Levi, G., "Su alcune particolarit.n di struttiir.i Jcl luicleo delle cellule nervose," Rivist.T d! patho- logi.i nervosa e mentale, Vol. I, p. 141, 1806. Also: "Ricerche citologiche comparate siilla ccllula nervosa dei Vertebrati." Ibid., Vol. II, pp. 103 and 344, i8g7. (Both papers inaccessible, quoted from the Zoologischcr Jahresbericht for i8q6 and 1897.) !RVE CELLS "lent. Com- F'g. 5. which ime fluid and the efifect of 1 cases. f Vertebrata ■I'^ou, Proteus, noticed the He offered 'e draw any ism of other those of the transforma- fjhile and a t that had a e oxyphiJe t or mam- ^nncea (and opods) the the mam- liosphorus- :'ear chro- I'ns green generally " of eosin ' to vary ■e cells of series in st.T d; pathn. • PP- '03 niid w SCOTT : STRUCTURE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS 27 IV.— Some General Considerations on the Structure of THE Nerve Cells. It may seem strange to revert to this subject, but owing to the fact that the Nissl granules were thought to be cytoplasmic structures, sever- al views concerning the structure of nerve cells have been advanced that would not have been if the true nature of the granules had been known. The first question is, whether the Nissl granules are formed elements of the cell body, or are precipitated while the cell is dying or when it is affected by the fixing agent. Held*'- accepts the latter ex- planation, as he claims to have seen fresh cells in which there were no granules but a homogeneous cell body. On standing for a few min- utes the cells become granular, thus showing the granules were precipi- tated while the cells were dying. On adding water to the cells they become vacuolated, but the vacuoles would collapse on adding a fixing agent, thus leaving the granules around a vacuole. Held also believes that the granules are soluble in alkalies, and that the normal reaction of the nervous system is alkaline, but it becomes acid shortly after death, and that this is the reason the cells contain vacuoles in tissue hardened in alkaline alcohol. v. Lenhossek and Flemming say the granules are visible in the fresh condition shortly after death. Each animal has a typical form of gran- ule in the spinal ganglion cell, whatever fixing fluid has been used, which could not be if the granules were precipitated either in dying or with the acid reagent. Buhler maintains that the granules are not seen in a fresh state, or even in a fixed condition, but this is no argument for their non-existence in the living cell, for the nucleus is often invisible in a fresh state. Ruzicka believes the granules are only due to differentiation in stain- ing, because if you overstain you do not .see them, and if you extract too much they are invisible. I agree with Held and Buhler that these granules are invisible in a fresh condition, and with Biihler also that the granules are hardly visible as^uch in the fixed cell. If one examines an unstained section of a spinal 8i I,, c, iSqj. ii w I n 28 SCOTT : STRICTURE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS fratiglion, one cannot be sure that there are distinct granules in the cell, but the chromatic substance is responsible for its optical appearance, for the periphery of the cell is homogeneous, and does not resemble the cen- tral parts where we usually find the granules. The cone of origin, and the layer around the nucleus, described by v. Lenhossek as free of gran- ules, are also homogeneous. For observing fresh nerve cells I used the retina, because one can examine nerve cells in this organ more easily and more quickly than in any other place, and because the retina is transparent, and does not need to be crushed or removed from its normal medium for examination. The eye was generally excised immediately after death, but it was often half an hour before it was opened and the retina placed in some vitreous humour. It was laid on the slide with its nerve-fibre layer uppermost, and a cover slip placed upon it. Observing such a preparation, one can frequently detect absolutely no structure in the retina, other than the blood corpuscles in the vessels, even with the best lenses. After a few minutes, the rods and cones come into view, and then, after a consider- able time (sometimes an hour) the outlines of the ganglion cells appear, but for a longer time the cells themselves are homogeneous. Eventually the nuclei of these cells become visible, and still later the cytoplasm becomes turbid. One might quite as properly contend that the retina did not exist in life except as a homogeneous substance, and that the cells were precipi- tated in dying or by the fixing reagent, as that because the Nissl granules are not seen in a fresh condition, they are not formed elements of the cell. We have seen that the granules are not soluble in alkalies, so that argument of Held's on the present point is valueless. Since it might be argued that it was one of the properties of the retina to be transparent the cells of the cord and cortex of young animals also were examined. These were killed by decapitation, the skull or verte- brae opened and a small piece of tissue taken and put in a drop of methylene blue. A cover was placed on the preparation and gently pressed till the latter was transparent enough for observation. The cells were found to have a granular appearance, resembling what would be found if the tissue had been fixed, embedded and stained with toluidin blue, and this within two minutes of death. Held lays stress on the fact that different fixing agents produce a different form of granule. It is well known that diff"erent fixing fluids 83 Turner using methylene blue on fresh brain has observed the normal appearance of the cells shortly after death. ; Brain, part I, page 100, 1899, also Journal ot Mental Science. 1898. i.i "S in the cell, pearance, for ible the ceti- f origin, and Tree of gran- use one can ckly than in oes not need xamination. t was often me vitreous uppermost, on, one can er than the After a few a consider- -11s appear, Eventually cytoplasm 3t exist in re precipi- 5l granules Its of the 'S, so that he retina Tials also or verte- flrop of I gently n. The ig what led with )duce a g fluids ills shortly SCOTT : STRUCTURE, MICRO-CIIEMISTRV AND DEVELOPMENT OF NERVE CELLS 29 produce slight differences in distribution of all chromatin, and different fixing fluids also form a slightly different intergranular substance which would cause the granules to have a different appearance. Putting all things together we may conclude that during life the granules have the same refractive index as the remainder of the cell, but that they are formed elements in the cytoplasm as much as ordinary chromatin is a formed part of the nucleus. It is probable that all chromatin is more or less plastic, for different fixing fluids produce a slightly different disposition of chromatin in the nuclei of all cells. It seems to me to be impossible to answer F"lemming's objection that the cone of origin of the process of spinal-ganglion cells is always free of granules, if the latter are precipitated elements in the cell. Many authors, including De Quervain,"* Held, Flemming, v. Lenhossek and others, consider the Nissl granules are made of fine particles em- bedded in another substance. It is true that the Nissl granules, in the different cells, but more particularly in the spinal-ganglion cells, do not appear homogeneous. Is this due to one kind of substance embedded in another different substance, or is it due to irregularities in contour of the same substance ? 1 think the latter is the correct explanation. In sections i// thick and stained with eosin and toluidin blue, iron-alum haematoxylin or other dyes, or treated to liberate the " masked " iron, the same result was always obtained ; the granules appeared homogeneous but of different densities. The edges of the granules are never straight, a circumstance that man)' have noticed, and thus a section of the cell must contain different thicknesses of the material. The granules often contain vacuoles, which would also tend to give them a heterogeneous appearance. The vacuolated appearance is also due to inequalities of the surface of the granules, for one can see in almost every preparation how a section at right angles to the plane being examined would appear to leave cavities in the chromatic material. I do not intend to discuss in this paper the arrangement of the Nissl granules in the cell, and shall refer only to the presence or absence of these granules in the axis cylinder and cone of origin of this process. The history of development would tend to show that the Nissl granules would not be found in the axis cylinder process, and this is what all observers who have worked with material that had been fixed and then 84 De Quervain, Fritz. " Ueber die VerHnderung des Centralnervensysteins bei experimenteller Kachexia thyreopriva der Thiere," Virchow's Archiv. CXXXIII, p. 527, 1893. 11 :'«^' 30 SCOTT : STRLCTf RE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS Stained, have observed. Dogiel,'^ however, finds the cone of origin and axis cylinder itself are finely granular, and considers the Nissl granules are formed by the running together of these fine grains. Dogiel's method consists in staining the fresh material in methylene blue, fixing in ammonium picrate, and transferring to a mixture of ammonium picrate in glycerine, where the tissue remains for some time, in order to get sufficiently transparent for examination. Using this method on the retina, and spinal and sympathetic ganglia, one obtains figures of cell-structure exactly resembling the figures of Dogiel. In this way spinal ganglion cells were obtained with the cone of origin, and the process filled with bluish-black grains resembling some of Dogiel's figures. This structure must be considered entirely artificial, for these grains occur more or less uniformly throughout the whole preparation, and the examination of tissue before and after fixation shows that they are formed by the precipitation of uncombined colouring matter. Examining the cells, stained in methylene blue, but not fixed in the picrate, one sees they are either granular, i.e., the Nissl granules only are stained or they are uniformly stained, i.e., the intergranular substance is stained as well as the granules, but if one puts the same cells through the fixing process, one finds fine dots of precipitated colouring matter all over the cells. This can be most easily followed in the retina, as little or no teasing of the preparation is necessary, and errors from that source are avoided. If one stains a retina with methy- lene blue, and examines it after washing as much of the colour as possible out of the preparation, one will find the nerve fibres are uniformly stained ; but, if one puts the same retina through the fixing process, and then examines again, one .sees the nerve fibres are filled with spindles and round mas.ses resembling what Dogiel figures. The same change may be followed in sympathetic and spinal ganglia, in which uniformly stained cells become covered with precipitated colouring matter in the process of fixation. The Nissl granules have, in the fixed prepara- tions, a different tint from this precipitated colouring matter, and could not be formed by the running together of these masses, even if the latter were elements of the cell. From my observations on prepara- tions stained by Dogiel's process, I have concluded that his method is one of the best to show the morphological connections between the 85 Doffifl, A. S., 1. c. and "Hie Striicturder Nervenzelleii der Retln.i," .\rchiv. f. MiU. An.it., XLVI, p. 394. '895. Also : " Ziir Frage ubcr den feineren Bail des Sympathischen Nervensystenis bei den i^augethieren," Arch. f. Mik. Anat., XLVI, p. jus, 1895. Also: " ZurFrage ubcr den feineren Bander Spinalganglien und.deren Zellen bei Siliigethitren," Inter. Monat. f. Anat. u. Phvs., XIV, p. 73, 1897. tic gan the same cci'ijitated )llowed in isar>-, and th methy- s possible Jnifonnly cess, and spindles 2 chaiig-e nifonnly latter in prepara- nd could 1 if the prepara- method een the ^t., XLVI, Sethieren," "•" Inter. SCOTT : STRICTURE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVK CELLS .31 cells, but that it gives entirely artificial appearances in the cytoplasm of the cells. The true structure of the cytoplasm of nerve cells has been the object of much investigation by Flemming, v. Lenhossek, Dogiel, Held, Lugaro, Cajal, Marine.sco, van Gehuchten, Cox, and many others, in fact, nearly all the works mentioned contain references to it, and there are good reviews of the literature in van Gehuchten, and in Goldscheider'" and Flatau. The question is whether there are independent fibrillje, or fibrill.-t forming a reticulum in the cell, or whether the cytoplasm has a foam-like structure. In this paper I do not intend to discuss the structure of the cytoplasm but shall point out, that since the substance of the Nissl granules does not diffuse into the cell body before the structure of the cytoplasm is determined, (in other wcrds, these are superadded to the cytoplasm), they cannot be a part of the fibrillar or reticulum. Thus the Nissl gran- ules are not thickenings of the protoplasmic fibrillae, or are not the nodal points of the cytoplasmic reticulum, but are independent of the cytoplas- mic structure ; and although the fibrillae, if they exist, might even run through the granules, they would never lose their independence. Several of the above-mentioned authors have reached the same conclusion, but could give no definite proof of its truth. No definite conclusion has been reached as to whether the nucleus keeps sending new material from the nucleolus to the cytoplasm, during the life of the cell. If it does give out new material to the cyto- plasm it certainly does not do so in the manner described by Rohde. The latter has described the migration of the accessory nucleoli into the cell body to become the Nissl granules, and the migration of the ordi- nary nucleoli to become the nuclei of neuroglia cells. He used iodine green and fuchsin as stains, and found the accessory nucleoli (which are only masses of ON yphile substance) resembled in their staining power the Nissl granules. Iodine green and fuchsin form a difficult combination to differentiate exactly, and the two appearances described by Rohde" can be obtained by a little longer or shorter differentiation ; in any case, the resemblance of the staining properties of the oxyphile nucleai sub- stance to the Nissl granules is much better seen by using Flemming's orange method (vide ante). Rohde says that by staining with iron-alum hasmatoxylin and long differentiation, the accessory nucleoli retain the stain longer than any other part, and thus the process of migration of 86 GoMscheider und Flatau, "Anatomie der Nervenzellen," Berlin, 1898. 87 L. c, p. 70s. > ^ i Mi'^ J2 SCOTT : STRUCTURE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS these bodies can be easily followed. I find it is not the accessory nucleoli that retain the stain but the ordinary nucleoli themselves. In all places examined by Rohde (spinal and sympathetic ganglia) it is well known that several nucleoli are found, as can be easily seen by staining with toluidin blue. In the cells of the cord where one nucleolus is the rule, it is this nucleolus which retains the iron-alum ha;matoxylin stain, and not the neighbouring oxyphile substance. If these nucleoli which retain the haematoxylin stain are outside the nuclear membrane they are artificially brought there. One can some- times find, as V. Lenhossek has pointed out, the nucleolus pulled out of the spinal ganglion cell. He believes this occurs because the nucleolus is loosely attached to the linin thread, or because the nucleolus is very nard under certain conditions. I have never seen a nucleolus outside the nuclear membrane in the cells of the cord, and in ganglia that have been fixed in sublimate this appearance is far more common than it is in material that has been fixed in alcohol. The fact, (and I have carefully examined my preparations to see that it is a fact), that where more than one cell have their nucleoli displaced in the same section the direction of the displacement of the nucleoli is always the same, shows that these have been displaced in cutting. One can make the appearance of migrating nucleoli quite common, if one cuts section.s, i or 2 //. thick, of ganglia fixed in sublimate, but all the apparent migration is in the same direction. If, however, thicker .sections are cut, or if material that has been fixed in alcohol is used, the appearance may be said to be non-existent. Holmgren'^ also believes in the migration of formed masses of the nuclear chromatin to the cytoplasm. In the cells of the spinal ganglia of Lophiiis he has described the migration of the chromatin out of the nucleus to form the Nissl granules, the migration of accessory nucleoli, and the passage of the Nissl granules back into the nucleus These changes are brought about through the agency of the micro-centre with its radiating threads, and are supposed to be different stages in the activity of the cells. Some of the cells observed so differed from the usual condition that they could only be considered as dying, and yet it is from cells in the same ganglia that these changes are described. Holmgren tries to justify his position by a study of the cells of Acan- thias, Gadus, Raja, and Rana, in which similar conditions were observed. In the spinal ganglion cells of Rana I have never observed such con- ditions, except in cases which are manifestly artifacts made in cutting, as 88 Emil Holmgren, " Zur Kenntniss der Spin.-ilganglienzellcn von Lophius piscatoriiis Lin." Anat. Hcftc, XXXVIII, p. 7l,i8qq. E?: tVE CELLS SCOTT ; STRL'CTLRE, MICRO-CHEMISTRY AND DEVELOPMENT OF NERVE CELLS 33 le accessory mselves. In anglia) it is sily seen by ne nucleolus aematoxylin outside the can some- illed out of e nucleolus >lus is very lus outside that have lan it is in e carefully more than ; direction hows that 3pearance I or 2 //. ition is in f material said to be ■M described before, nor have I noticed any difference in the staining power of the nuclear membrane next the micro-centre.although I have observed many of the different conditions of the cell with respect to the distribu- tion of the jjranuies described by Holmgren. Thus, after the chromatin has once diffused from the nucleus, nothing occurs, in my opinion, to indicate the renewal of the granular substance from that organ. I do not deny that such renewal may lake place, but if it does, it is in solu- tion and not in formed masses. Further investigation, however, is necessary to decide this point. Concerning the reason for the diffusion of the chromatin from the nucleus, it may be to aid physiological action, for it is a general rule, which no physiologist would now deny, that an iron-holding nucleo-proteid is necessary for the cell to carry on its normal function. These compounds are generally confined to the nucleus, but they occur in the cell body of all gland cells. It seems to me that it would aid physiological action in having these nuclein compounds in direct contact with the cytoplasm of the nerve cells, for in this case the cytoplasmic action would not be delayed by immediate participation of the nucleus. Thus cytoplasmic impulses may pass from one process of the cell into another without going through the nucleus, which could not happen if the chromatin had remained in the latter. v.— Conclusions. I 11 lasses of le spinal natin out ccessory nucleus ro-centre ;s in the "rom the d yet it scribed, f Acan- )served. ;h con- ting, as in." Anat. 7^5 . '%< The Nissl grannies are of a nucleo-proteid nature, since they contain " masked " iron and organic phosphorus, and are derived from the nu- clear chromatin of the germinating cells. Pepsin and hydrochloric acid do not dissolve them, nor are they dissolved by alkalies or acids which, however, liberate the iron, and in consequence of this their staining re- actions are altered. Digestion with pepsin and hydrochloric acid does not affect the occurrence of iron and phosphorus in the granules. The nucleolus consists of an oxyphile centre with a basophile covering. The basophile covering seems to correspond to the original kinetic chromatin ot the germinating cell. It contains iron and phos- phorus, and alkalies extract the iron very much more slowly from it than they do from the Nissl granules. The oxyphile nuclear substance is also a nuclein compound since it contains iron and phosphorus. It is readily dissolved in pepsin and i " 'I 34 SCOTT : STRl CTL'RE, MICRO-CHEMISTRV AND nRVEI.OPMKNT OK NERVE CELLS hydrochloric acid. It is altered but not dissolved by acids and alkalies, which liberate the iron from it. The alkali acts much more slowly in removing the iron from this substance than from the Nissl granules. The three nuclein compounds of the adult nerve cell are derived from the mitotic chromatin of the primitive nerve cell. It follows from this that the Ni.ssl graiuiies are constituted of chrom^itin that has diffused from the nucleus into the cytoplasm. A substance analogous to that of the Nissl granules is found in the nerve ceils of most animals, but not in all, as it is rarely present in the nerve cells of the Urodela. Those animals, whose nerve cells are devoid of this material, have chromatin in the nuclei of such cells similar to that found in the nuclei of the cells of other tissues. The Nissl granules are morphological elements of the cell, and consist of one substance. They have the same refractive index as the cyto- plasm during life, and are not found in the axis cylinder process. All the results obtained go to support the view that all iron-holding nuclein compounds are derived from pre-existing ones, and in mitosis all the iron-holding substance of the cell is confined to the nuclear chro- matin. , ,1111-. ot CanadiHii InuUtute Voi VI ERVE CELLS s and alkalies, lore slowly in granules. ; derived from lows from this : has diffused ' found in the present in the lis are devoid cells similar I, and consist as the cyto- "ocess. iron-holding id in mitosis luclear chro- *1* ii 8 -y' *■ ■•* \ 14. r i»X HI 'JP'-^"' % t » *^ *^ ' tot '^ '•fii V % % % •t.. \ .,^. *. % >* J' 1 13. I \% 16. F HM'.h l.ith' EiJir,' |-'' y i 7^ 2. \ ■J *.T»t. !?*•»' 6. 5. r H iWolt id. I MIU. Ol Cflllililirtll i 11 L> U 'i 'li I •■• Vol '1 r-' V- • J -A 4, «^^, I -•-.*^A *' -f » X ;** ^ ./%** •>§ \ « t.® 11, 12 i 13. f '0 r 14 / K^'K^ >^ ^:; it. la . 16, m 18. K Hu.il, l.iti:' Kiim' I f «!;! I I I ;■ I it 1 j^» VI SCOTT : ST tU'CTlRK. M,.«O.CHKM.STRV AN., nKVKI.OPHKNT O. NKRVK CKU.S 35 EXPLANATION OF PLATE NOT..-A,, figures .e.e a.wn .o ...e sa.e ---;;;;- rtl^:;^^ lacida, as seen under the Lei,/ .-.homogeneous unmers.on len.. ocular 1 2 of Zeiss. F,c . -Motor nerve cell of cat. Alcohol, eosin. and toluidin hlue. the Urodela.) F.O. 3.-Motor cell of cat. Alcohol, a.nmonium molybdate in nitric acid -, hours, phenyl- liydrazin hydrochloride. F,G 4.-Motor cell of cat. .Alcohol, acid alcohol 3 hours, h.-ema.oxylin. p,a.6.-.Mo.or col, of N.,,..r,... AkoUo, „o„i,™ n,o,yMr.U- trio «cid, l.ou„, phenylhydrazin hydrochlonde. F,a. 7:-Motor cell of N..un.. Bichloride, acid alcohol 6 hours, potassiutn ferro- cyanide. chloride. F.G. 9.-Motor cell o^ Ne.X.rus. Alcohol, eosin and toluidin blue. . r . .11 r.f