treat of the histology of the nerve cells of Invertebrates. Among these may be mentioned the investigations of Rawiz, Nansen, Rohde, HALLER, BINET, DE NABIAS, PFLÜCKE and LUGARO. The paper by RAWITZ (31) deals chiefly with the histology of the nerve cells of mussels, which he studied by the isolation method as well as by sections. The ordinary indifferent fluids were used for the isolation method, while the picro-sulphuric solution of KLEINENBERG gave the best results for the study of sections. His general results as to the cell structure may best be summed up in his own words p. 403: "Wir ersehen also aus diesen Beobachtungen, dass die Ganglienzelle der Acephalen aus zwei Theilen besteht, von denen der eine eine netzförmig angeordnete, der andere eine zähe, unter Umständen ölartige Tropfen bildende Substanz ist, die in den Maschenräumen der erstern suspendirt ist."

He suggests the possibility that these drops may alone constitute the true functional nervous element in the cell, while the network serves simply as a support for the former.

Nerve cells that had been macerated in weak solutions of alcohol, or potassium bichromate, presented a structure which differed materially from that mentioned above. Here they appeared to be finely granular in structure, and a division of their constituent elements into two chemically and morphologically different substances, was not to be made out. He accounts for this in the following manner p. 404: "Es würde dann das zart granulirte Aussehen der Zellen verursacht durch die in der netzförmigen suspendirte zähe, tropfenbildende Substanz, in deren Innerm Gerinnungen durch die coagulirende Wirkung der Reagentien entstanden sind."

In only one instance (fig. 32) does he figure the presence of fibrils in these cells.

NANSEN's investigations (25 and 26), upon the finer structure of the nervous system of Invertebrates, cover a wide range of forms, including species of Mollusca, Vermes, Crustacea and Tunicata. His conception of the finer structure of the nerve cells of Invertebrates may be summed up as follows.

The contents of the axis-cylinder processes consist of primitive tubes which he says "are extremely slender tubes or cylinders, separated from each other, or rather formed by membranes or sheaths of a firm supporting substance, spongioplasm, very much resembling the neuroglia-substance". The viscous contents of these primitive tubes, the hyaloplasm, he regards as the true nervous substance.

In regard to the structure of the bodies of the nerve cells, he finds that they likewise consist of primitive tubes which contain hyaloplasm. He was unable to determine upon the exact mode of distribution of these tubes in the cell, except in the case of some which ran concentrically about the nucleus.

In addition to the spongioplasm and hyaloplasm, he describes a third substance in the spongioplasm-reticulum, between the primitive tubes, which stains deeply when osmic acid and haematoxylin have been used as fixing and staining reagents. He finds this substance absent from the nerve tubes (axis-cylinder processes), and regards it as partly fatty in character.

So far as I am able to make out from his description of this substance, it corresponds exactly to what I shall describe in the following pages as chromophilous granules. NANSEN entirely disregards the fibrillar theory held by FLEMMING and others.

The publications of ROHDE (33, 34, 35, 36, 37 and 38) known to the writer are six in number. The first five of these papers have for their object, the development of the theory, that the fibrils (Spongioplasma) in the cell body and axis-cylinder processes, do not act as paths along which nervous impulses travel, but serve rather as supporting structures (Stützgerüst) for the semi-fluid hyaloplasm, the true nervous substance, which is contained between them.

The material made use of by ROHDE in these investigations was most extensive, including examples of the Chaetopoda, Hirudinea, Nematoda, Gastropoda and Crustacea. Unfortunately no reference is generally made to the manner in which this material was fixed and stained.

For the general principles upon which this theory rests, I refer the reader to the following quotation from ROHDE's fifth paper (37), p. 387: "Das Spongioplasma tritt bei den Ganglienzellen in doppelter Form auf, theils als grobfibrilläres, theils als feinfibrilläres; letzteres bildet stets den Axencylinderfortsatz, breitet sich häufig aber auch, vom Grunde desselben ausgehend, über die ganze Peripherie der Ganglienzelle aus, so dass dann in dieser zwei verschieden aussehende Zonen zur Unterscheidung kommen; eine innere dunklere grobfibrilläre und eine äussere helle feinfibrilläre (Holzschn. A 2). Die Ganglienzellen liegen eingebettet in der Neuroglia, welche aus Fibrillen besteht, die im Aussehen wie in der Stärke mit den groben Fibrillen des Ganglienzell-Spongioplasma übereinstimmen und allenthalben von Kernen durchsetzt werden. An der Peripherie der Ganglienzelle

nehmen die Neuroglia-Fbrillen ein dichteres Gefüge an unter gleichzeitiger Vermehrung ihrer Kerne und gehen nach innen continuirlich in das grobfibrilläre Ganglienzell-Spongioplasma über, entweder am Rand der Ganglienzelle oder tiefer in ihrem Innern, im letztern Fall häufig unter Bildung von Bäumchen, welche besonders bei den Ganglienzellen mit heller Randzone scharf hervortreten (Holzschn. A 1 u. 2). Aus diesen Beobachtungen folgerte ich, dass das Spongioplasma der Ganglienzelle nur ein Stützgerüst darstellt und das eigentlich Nervöse das von diesem umschlossene Hyaloplasma ist, welches an frischen, unter schwachem Druck befindlichen Ganglienzellen in Gestalt grösserer oder kleinerer glasheller Tropfen austritt."

He further states that in addition to the coarse fibrils the neuroglia tissue is also made up of fine fibrils, which in structure resemble those found in the nerve cells. On account of this close resemblance which exists between the fibrils of the nerve cells and those of the neuroglia, he concludes that the former must possess the same functions as the latter (Stützgerüst).

In the following pages I hope to be able to disprove the existence of coarse neuroglia fibrils in the nerve cells of Gastropods. ROHDE's last paper (38) which deals with the development and growth of nerve cells will be considered further on in connection with another topic.

HALLER (13 and 14), in a manner similar to KUPFFER, divides the cell substance into "Proto- and Paraplasma". The former is arranged in the form of short fibrils which distribute themselves in the "Paraplasma" according to the functional status of the cell. Thus any functional change in the cell may produce a correspondingly different arrangement of its elements. For example under certain conditions, the fibrils may have an irregular distribution, or they may be concentrically arranged in the cell body; they may be so regularly distributed, that in sections they will appear as coarse or fine granules; or finally, they may be so distributed in the cell, as to produce the appearance of a network.

BINET (1) speaks of the difficulty experienced by him in procuring good sections of nervous tissue from certain Invertebrates, and ascribes his inability to do so, as due to the circumstance that these tissues are enclosed by dense layers of connective tissue which prevent the entrance of the fixing fluid.

The writer also experienced, at first, a similar difficulty in procuring good sections of the nerve cells of Crustaceans and worms

when sublimate was used as a reagent. This difficulty was, however, overcome by substituting bergamot oil for xylol as a clearing agent. By this procedure, scarcely, if any, shrinkage was apparent in the structure of the cell, and I agree with HEIDENHAIN, that when evidences of shrinkage are present, they are more frequently due to an improper after-treatment of the tissues, than to a lack of penetration of the fixing agent.

For fixing agents, BINET obtained the best results with sublimate, FLEMMING'S and HERMANN's fluids. For staining, he employed chiefly VIALLANES' 1) haematoxylin method with the addition of safranin. By means of this combination he obtained a double coloration of the cell, the cell body staining red, the cell process blue, the nucleus blue and the nucleoli red 2).

He says in regard to the presence of fibrils in the cell (p. 478): "la fibre nerveuse est bien constituée, comme l'a pensé et figuré REMAK, par un faisceau de fibres parallèles, et ces fibrilles, au moment où elles pénètrent dans le corps protoplasmique, pouvent être suivies pendant une certaine partie de leur trajet." He classifies the nerve cells of the abdominal ganglia of Astacus under three categories (p. 479) as follows:

1) "les cellules dans lesquelles le cylinder-axe pénètre sans qu'on puisse saisir aucun détail sur le mode de pénétration; la substance du cylindre-axe paraît être en continuité directe avec celle du protoplasma" (tab. 12, fig. 6).

2) "Dans une seconde catégorie de cellules, on observe une légère striation de lignes circulaires et concentriques; et de plus, on peut constater que ces fibrilles, disposées régulièrement autour du noyau, convergent vers le cylindre-axe et s'y réunissent, ce qui montre que ce sont bien les fibrilles du cylindre-axe qui entourent le noyau de la cellule et en sillonnent le protoplasma. Cette structure est celle qui a été le plus souvent décrite; elle a été figurée, notamment, par VIGNAL (43) chez le Homard" (fig. 8, tab. 15).

3) "Nous rangeons dans une troisième catégorie des cellules, en général de grande taille, dans lesquelles le cylindre-axe pénètre et même décrit un demi-cercle, en conservant les dimensions qu'il présente dans le nerf; ce sont des cellules à cylindre-axe intra

1) See BINET's paper for details of this method, p. 470.

2) I find this method no improvement on the methylen blue-eosin combination.

cellulaire 1). Dans ce trajet, le cylindre-axe se distingue nettement du protoplasma par la puissance avec laquelle il concentre la matière colorante."

Again on p. 480, he says: "Le faisceau est formé des mêmes fibrilles accolées ensemble que dans le cylindre-axe extracellulaire". In regard to the structure of the cell body, he describes it as consisting of two elements - namely, of fibrils, and of an interfibrillar substance. Each of these elements stain differently when treated by the method of VIALLANES; the fibrils stain blue and the interfibrillar substance red. If I understand him correctly, he considers the fibrils to be distributed throughout the cell body, but most abundant in the cortical layer; while the interfibrillar substance is most abundant in the centre of the cell around the nucleus. His figures represent th interfibrillar substance as granular in character; but he makes no reference to the row-like arrangement of these granules in the cell body, nor to their correspondence to the chromophilous substance found in the nerve cells of Vertebrates.

Unfortunately, I have been unable to see the publication by DE NABIAS (24), so that any criticism of his work is impossible. I insert, however, the reference, so that the bibliography on this subject may be as complete as possible.

A recent publication by PFLÜCKE (30) deals chiefly with the structure of the nerve cells of Astacus. He studied these cells in the fresh state, as well as by sections. For fixing agents, sublimate and FLEMMING'S solution gave the best result; while for staining, NISSL's methylen blue method, safranin, fuchsin and iron-haematoxylin after HEIDENHAIN were used.

แ

Corresponding to the discoveries of NISSL and others on the cells of Vertebrates, PFLÜCKE similarly finds in the nerve cells of Astacus, two substances which possess unlike affinities for methylen blue. The manner, in which these cells are stained, is best expressed in his own words (p. 514): "- nämlich in eine stark färbbare, structurgebende Substanz und in eine anscheinend homogene, kaum oder überhaupt nicht gefärbte Zwischenmasse." The first of these substances is represented by a large number of deeply stained granular appearing structures which lie close to each other, and which vary in size and form; some of them being spindle-shaped, others

1) I have also observed these structures in the nerve cells of Cambarus, Homarus and Astacus (see 22), but reserve a description of the same for a future paper.