lary sheath and passing obliquely from the inner surface of the external sheath to the axis-cylinder (fig. 1, e). These incisures, or indentations, are the same; to the existence and regularity of which, a number of years ago, I directed the attention of histologists, though Stilling, the great pioneer in the histology of the nervous system, had observed and mentioned them before me. The segments of the medullary sheath, formed by the incisures, have been termed by Ranvier the "cylindro-conical segments;" they have of late years been the subject of much speculation and discussion, though their true anatomical nature, which I shall treat directly, seems to have been, as yet, not properly understood. The incisures, which in truth only represent the vacant spaces between each two of these segments, are not directed in a regular order toward one or the other extremity of Ranvier's segments, but, on the contrary, are arranged irregularly. Thus, in many places, where, in this manner, the narrow parts of two cylindro-conical segments meet each other, they are surrounded by a segment in the form of a ring, or true hollow cylinder (see fig. 1, f).

In subjecting now these elements composing the double-bordered nerve-fibre to a special consideration, we shall begin with the axis-cylinder which represents the true nerve-fibre. Formerly, when the nerve-fibre under discussion was still supposed to consist of a simple membranous tube filled up with a homogeneous nervemedulla, the axis-cylinder was regarded as the axial part of this substance, remaining unaltered after the coagulation of its outer portion. In the course of time, however, in spite of the high authorities, who, without substantial reasons, most tenaciously defended this erroneous view, the axis-cylinder gained its proper position as a pre-formed, integral element of the nerve-fibre, representing, as already mentioned, its most essential part. But like the whole nerve-fibre, the axis-cylinder, also, was for a long time looked upon as homogeneous in its anatomical composition, until Max Schultze discovered, about 1867, that the processes of the ganglion-cells, as well as the axis-cylinders derived from them, were composed of fine, smooth fibrillæ, in the interspaces of which a number of minute dark granules were distributed. Finally, without a previous knowledge of this discovery, and not quite two years 1 Stilling. "Ueber den feineren Ban der Nerven-primitivfaser und der Nervenzelle.'

1850.

FIG. 1. Diagramatic representation of a double-bordered nerve-fibre, treated with a solution of osmic acid; a, axis-cylinder; b, medullary sheath; c, external sheath; d, annular constriction of Ranvier; e, incisures; f, cylindro-annular segment.

FIG. 2. Diagramatic representation of a double-bordered nerve-fibre (osmic acid preparation) highly magnified and showing the funnel-shaped subsegments, as well as the fibrillar arches passing from one side of the fibre to the other; a, incisures; b, fibrillar loops; c, external sheath.

FIG. 3. Schematic diagram illustrating the probable arrangement of the fibrilla; a, axis-cylinder; b, fibrillar loops; c, external sheath.

afterward, I discovered myself the true anatomical composition of the ganglion-cell processes and their respective axis-cylinders, consisting of fibrillæ representing rows of minute granules united to one another by an intermediate substance; in other words, granular fibrillæ (fig. 1, a). The correctness of this observation I verified by my investigations upon the development of the nervous tissues of the human embryo, in which I found the whole nerve consisting of distinct rows of still ununited elementary granules, which, in the course of the development, are formed into fibrils by adhering to one another. The axis-cylinder in the fully formed nerve-fibre, therefore, consists, in proportion to its diameter, of a greater, or smaller, number of granular fibrillæ, held together by an interfibrillar substance, and surrounded by a sheath, which, though distinguished by a double contour, is very likely identical with the inter-fibrillar substance. When the axis-cylinder approaches its termination at the periphery, it divides successively into the individual fibrillæ, of which it was originally composed.

As regards the medullary sheath of the double-bordered nervefibre, I will simply mention that in my paper "On the Construction of the Double-bordered Nerve-fibre," published May, 1874, in the "London Monthly Microscopical Journal," I also minutely described the complicated anatomical composition of the nervemedulla, as consisting of two parts, or layers, the inner one of which, surrounding directly the axis-cylinder, represented a finely granular, amorphous semi-liquid substance, while the outer showed a structure composed of very delicate and smooth fibrils, arranged, in the fresh unaltered nerve-fibre, parallel and very close to each other. I furthermore stated that these fibrils, on account of their interspaces being filled by the semi-liquid part, could not be well detected in the fresh fibre, but that with the addition of water, this liquid, penetrating through the external sheath, would dissolve the connecting element, and render the fibrillæ of the nerve-medulla in the interior of the nerve-fibre very distinct in the form of wavy or arched bundles. In the same form they are also observed to emerge from the open ends of the fibres, or, in the form of a so-called hernia, through the orifice produced by a small rupture in the external sheath.

In the above mentioned paper, at the time I directed the attention to the incisures of the medullary sheath, the idea of the spontaneous coagulation of the nerve-medulla, and the inclina

tion to explain by this process all the phenomena observed in this substance, still prevailed in the minds of the leading European histologists; and, regarding, at that time, the views of these authorities with a certain amount of reverence, though I was surprised at the regular order in which these incisures appeared, I really dared not form a strictly independent opinion of their signification. Thus, I happened to consider them as folds in the external sheath and nerve-medulla. Another reason of my failing to distinguish these segments as fully as I do to-day was, because my examinations were confined to the fresh nerve-fibre, or to fibres kept for one or two days in a very weak solution of chromic acid; for osmic acid, the most valuable re-agent to render the incisures distinct, was then beyond my reach.

The examinations, which I have since made of this subject, not only confirm the statements of other observers as to the existence of the cylindro-conical segments, but will, moreover, throw some additional light upon their true anatomical structure. Convinced, from my former studies, of the fibrillar nature of the normal nerve-medulla, I also endeavored to demonstrate its fibrillæ in nerve-fibres treated with weaker or stronger solutions of osmic acid. For this purpose I prepared and examined, from time to time, the fibres of a considerable number of nerves, or nervous fasciculi, of man and different animals, of which the former offers the most advantages for the investigation.

The results of these investigations corroborated my former statements as to the fine fibrillæ forming a part of the medullary sheath. The only error to be found in these statements— which was solely due to the examinations having been chiefly made on fresh nerve-fibres, or such as were treated with water, or with very weak solutions of chromic acid-relates to the particular arrangement of these fibrilla within the nerve-fibre. For, though I observed and described their loop-like arrangement within the fresh fibre, or when escaping from its open ends,—such as the accompanying figures 2 and 3 will show,-I erroneously interpreted these forms as being produced by the action of water entering the nerve-fibre through the external sheath.

As a minute description of the true arrangement of these fibrillæ, and of the particular method of examination by which they may be demonstrated, would extend this paper beyond its proper limits, I shall, for the present, limit my remarks to the more prominent points

concerned, leaving the details to form the subject of another paper, purely histological in its nature. A few remarks concerning the action of the osmic acid solutions upon the nerves in different conditions, however, may serve to elucidate the subject more fully. When a perfectly fresh nerve has been left in a one per cent. solution of osmic acid for several hours, or is dissected in this solution, as Ranvier particularly recommends, the cylindro-conical segments will be found upon microscopical examination as they are generally described; that is, in the form of hollow cones in which no structure can be detected, for the reason, that the reagent firmly coagulates the entire segment, preventing the fibrillæ, of which it is composed, from separating as they are observed to do when acted upon by water, or by slight decomposition. If, however, the nerve is taken from the body of a man, or an animal, a few hours after death, a time when decomposition, if ever so slight, has commenced, a slight separation of the fibrillæ from one another will have taken place without a disturbance of their normal arrangement in the cylindro-conical segment, and their relative position will become fixed by the subsequent action of the osmic acid. The same object may be obtained by leaving fresh nerves, or their fasciculi, in a weaker solution of osmic acid for a longer time, in which case the water of the solution has a chance to slightly separate the fibrillæ, before they are fixed by the action of the acid.

In such preparations, it will be found that the cylindro-conical segments, as they are observed in the firmly coagulated nervemedulla of a perfectly fresh fibre, consist in reality of a number of subsegments, representing a system of loops formed by the fibrillæ in the manner represented in the diagrams, figs. 2 and 3. Each subsegment resembles in form a hollow cone, or a funnel, placed with its narrow end, or apex, into the base, or trumpet-like opening, of the following segment, and receiving on the other side its preceding neighbor. While the margins of the bases of these fibrillar funnels are in contact with the inner surface of the external sheath, the margins of their narrow extremities embrace the axis-cylinder. Judging from the length of the fibrillar bundles observed to emerge from the open ends of a fresh nerve fibre, the fibrillæ, forming the funnels, must be continuous throughout the whole length of the medullary sheath of one segment of Ranvier, and must be bent twice upon themselves in order to form a sub