In the frog's muscle the nerve ending has no ground plate in which the branches ramify. The ramifications are not localized but are spread over a relatively large and apparently variable area of the muscle fiber. Usually but one medullated nerve ends in a muscle fiber, though two medullated nerves may be seen at times and three have been described; occasionally two medullated nerve fibers and one non-medullated and much finer fiber may go to the muscle fiber. When more than one nerve goes to a muscle fiber it is often possible to trace the origin of these to the same nerve stem. In no case was there even the suggestion of one of these fibers coming from a nerve whose course lay distinctly apart from the others. If more than one nerve go to the muscle fiber, the places where the nerves enter into contact with the fiber are, if not always, at least most frequently, in close proximity to one another. In short, on the muscle fiber the area to which the entering nerves apply themselves relative to the entire length of the muscle fiber is limited. Several varieties of motor endings have been described. Thus, there are the four or five types of CUCCATI which RETZIUS would reduce to two; the one exemplified by the branching plate, the other by the broad band. At present, any classification must be but temporary; with equal justification several varieties may be classed as typical by one and rejected by another. For example, one might well wish to add to the types of RETZIUS that more strictly localized variety which DOGIEL has described and which occurs not unfrequently in certain muscles—a variety which closely resembles the branching of mammalian nerve endings only without an end plate.1 There is, however, one type which is generally recognized as predominating-das Stangen-geweih of KÜHhne. In it the nerve after losing its medullary sheath divides more or less dichotomously and spreads itself along the length of the muscle fiber. The band variety is less common; one notes that the better stained the preparation the less frequently it appears. 1 Arch. f. mikrosk. Anat., 1890, Bd. xxxv. In the frog the end-arborizations do not terminate necessarily on one muscle fiber. It is not unusual to find that, while the majority of the terminal fibers confine themselves to one muscle fiber and to a certain definite area, one or more fine non-medullated fibrillae pass far beyond the area or to a neighboring muscle fiber. So common is this that in well stained preparations one is ever expectant of finding at least traces of such fibrillae. They are often difficult to observe; but the cause of this is not so much that they do not readily stain, as that it is hard to fix the dye and easy to have it extracted during the stages subsequent to fixation. Such non-medullated fibrillae may be termed ultraterminal. A convenient way to describe them is to classify them according to the manner in which they end, so far, at least, as such can be traced at present. With this in view, we might describe them as follows: (1) Relatively thick, non-medullated fibers which pass to adjacent (3) Very fine non-medulated fibrillae which detach themselves at a) by getting so faint and so fine that it becomes impossible to follow them farther (Fig. 1; Fig. 2; Fig. 3). b) by terminating at what appears as a much thickened knob (similar to Fig. 1, B and C) c) by breaking up into a plexus from which some at least of the fibrillae disappear in the muscle fiber while others continue on (Fig. 2) d) by forming a plexus which enters into close relationship with a typical nerve ending (Fig. 3) e) by breaking up after a relatively long course to form a small localized ending, each termination of which is furnished with a knob (Fig. 6). At no time have I seen any appearance of an intermuscular nerve plexus formed by these fibrillae. I am inclined to regard the motor fibrillae which pass into the intermuscular connective tissue and there disappear, as either broken fibrillae or fibrillae only partially stained. The relation of the nerve ending to the muscle fiber.-The relation of the nerve endings to the sarcolemma has been much disputed. Of recent writers who have discussed this question with reference to the muscle of the frog, the following only need to be referred to: HUBER-DEWITT,' after a careful investigation, come to the conclusion that the terminals lie under the sarcolemma and are devoid of any sheath. SIHLER,2 on the other hand, using a method which he finds particularly applicable to this research, considers that the endings lie over the sarcolemma and that the end of fibrils are covered "down to their tips with the sheath of SCHWANN;" further that the sheath of HENLE is open (verwächst mit nichts) and does not cover the end fibrils, but that the nerves emerge from it as an arm from a sleeve (wie der Arm aus dem Aermel); at the same time, he does not deny that there are points where the nerve substance and the muscle fiber may come into contact. In fresh muscle fibers in which the nerves have been stained by the intra-vitam methylene blue method, I have found that, while the majority of the terminal branches lie in close relation to the muscle fiber, at times a terminal fibril is seen to rise some distance above the muscle fiber, and occasionally such a nerve fibril sends down to the muscle fiber little rootlets comparable to those described by SIHLER. But sooner or later even in such nerve fibrils the ultimate terminals come to lie on the muscle fiber. While the occurrence of such nerve fibrils may be held to prove that the larger fibrils may be epilemmal, there is nothing to show what the relation of the terminal fibrils 1 Huber-DEWITT: Nerve Endings in Muscles. J. Comp. Neurol., 1897, VII, p. 185. 2 SIHLER: (a) The Nerves of the Capillaries with remark on Nerve Endings in Muscles. J. Exp. Med., 1901, V, p. 511. (b) Neue Untersuchungen über die Nerven, etc. Zeit. f. wiss. Zool., Leip., 1900, LXVIII, pp. 351 and 375 to the muscle fiber may be. To investigate this a much higher magnification and a more differential stain are necessary than at present are available in the examination of fresh tissues. By means of the orange G acid fuchsin counterstain above referred to, I have found it possible to distinguish clearly between the nerve fiber (blue), the sheath of HENLE (rose-pink), the faintly stained neurilemma (pink), and the muscle fiber (orange). In thin sections (5 to 7μ) examined by the 1-12 ZEISS oil immersion lens, where the medullated nerve was observed to lose its medullary sheath and divide into the primary branches, I often clearly saw the primary branches, especially when they were lying in the upper edge of the muscle fiber, enclosed by the neurilemma and by the sheath of HEnle. Moreover these sheaths could at times be traced for some distance on the primary divisions of the ending. The differentiation of the two nerve sheaths was at times aided by the fact that one could distinguish the attachment of the neurilemma to the node, as described by BOVERI and BETHE,' while the sheath of HENLE had no such attachment but was continuous over the node. This condition was seen in the section from which Fig. 7 was obtained. Here the neurilemma attached itself to the node where the medullary sheath ceased, and thence was continued over the fibrils. For this research it was not necessary to determine whether at the node the neurilemma was or was not interrupted. Outside of this lay the sheath of HENLE. Each sheath could be followed separately to the muscle fiber, where together they applied themselves to the sarcolemma. When so applied to the muscle fiber, it was not possible to distinguish with accuracy between the pale stained neurilemma and the strongly stained sheath of HENLE. But it was possible to see clearly that the sheath round the nerve was distinct from and was placed outside of the sarcolemma-a distinction which was aided by the fact that here the sheath of HENLE stained at times more strongly than it had previously done and appeared as if it were there thickened 1 BETHE: Anatomie und Physiologie des Nervensystems, 1903, p. 50. or compressed. As the fibrils subdivided each branch was surrounded by a sheath which became fainter as the nerve fibril became finer. Occasionally there coiled round a primary terminal branch of a nerve ending another nerve in no way related either to this nerve ending or to its muscle fiber. Thus in Fig. 5 at (2) the nerve (3) curved round a primary terminal division of the nerve termination on muscle fiber A. This is only to be explained by one of two suppositions; either both nerves at that point were over the sarcolemma or under it. The latter supposition is open to many objections; the former will generally be acknowledged to be the more probable, and to me is a confirmation of what I have already stated. Attention was now directed to two points: (1) The relation of terminal fibrillae and of the end knobs to the sarcolemma; (2) The relation of the ultraterminal fibrillae to a sheath. In sections where very fine fibrils were seen, one could trace the blending of the nerve sheath with the sarcolemma. In Fig. 9 the terminal nerve fibrilla rested on an apparently homogeneous substance which was not separated from the contractile muscle substance by any sarcolemma and which was covered by a cap formed by the blending of the nerve sheath with the sarcolemma. In short, it lay under the sarcolemma. When the section from which Fig. 9 was drawn, was carefully examined under the 1.5 mm. ZEISS apochromatic and Oc. 6, the sarcolemma appeared at (a) to split into very faint lines. Two of these were seen to project as at (a) and disappear in the covering cap. Similarly at (b) a line was seen to go off as delineated. The impression formed by the study of this and of other similar sections, was that though the cap over the terminal nerve fibrilla or end-knob was formed by the blending of the sheath which surrounded the nerve fibril with the sarcolemma, the tissue which had principally to do with its formation was the sheath of the nerve fibril. The relation of the root-like knob to the muscle fiber was sometimes particularly interesting. It seems to lie more in the |