A simultaneous coupling azo dye technique has been used to reveal the distribution of cholinesterase activity in the musculature of the developing tadpole of Xenopus laevis. The use of inhibitors and a less convenient but more specific histochemical technique confirmed that only true cholinesterase distribution was being demonstrated; and a study of silver-impregnated material proved that this azo dye technique provides a very convenient method of following the development of the patterns of myo-neural junctions in the striated muscles of this tadpole. A wide variety of patterns is seen in the various muscles; in the axial musculature the muscle-fibres become innervated at their ends from myocommatal plexuses and never acquire endings along their length; broad muscular sheets, as in the walls of the branchial and abdominal cavities, are also first innervated terminally from the septa but later acquire secondary patterns of innervation over their surfaces; while in long narrow muscles the main innervation is along the lengths of the fibres. These different patterns of innervation are correlated with the functions of the various types of muscle. It is suggested that terminal innervation may be a special adaptation to permit rapid establishment of neurogenic activity, the pattern of endings of the more usual type forming when the need for precisely co-ordinated reflexogenic activity arises. In some muscles, the azo dye technique reveals a profuse multiple innervation of the fibres which are assumed to be of the so-called ‘slow type’ known to exist in some amphibian muscles.

The developing patterns of myo-neural junctions in the voluntary musculature of the tadpole of Xenopus laevis are described in this paper. The

survey is based largely on the study of whole larvae or of dissections in which the azo dye coupling technique described by Lewis (1958) has been performed. Such preparations reveal the distribution of esterases throughout the tissues, which within the developing muscles may be assumed to be cholinesterases. The results have now been correlated with the examination of section series of tadpoles impregnated with silver by Bodian’s method. Comparison of the results obtained by these two means has shown that the pattern of esterasepositive sites in the developing muscles is identical with that of the myo-neural junctions.

Preliminary accounts of these observations have already been published, both in the technical paper (Lewis, 1958) and elsewhere (Lewis and Hughes, 1957). In these papers, attention was drawn to the fact that the motor innervation of the larval myotomes is confined to the ends of the fibres. These observations are now confirmed, and this type of innervation is compared with that of other larval muscles.

The methods used to obtain and culture the Xenopus larvae have been briefly described by Tschumi (1957).

In a preliminary series of experiments most of the histochemical methods said to reveal cholinesterase were tried. The simultaneous coupling azo dye technique with α-naphthyl acetate as substrate appeared to be the most suitable for our purpose and unless otherwise stated all the observations recorded in this paper were obtained with some modification or other of this general technique. It soon became clear that the accepted frozen-section technique was quite unsuitable for a general embryological survey; so a special modification was devised which made it possible to make esterase preparations of thick specimens, even of whole tadpoles. Full experimental details of this modification as applied to Xenopus tadpoles have been published elsewhere (Lewis, 1958). A closely spaced series of esterase preparations of whole tadpoles was made covering the period from before hatching right up to metamorphosis. Preparations of individual muscles and of frozen sections were also made to elucidate particular points which required finer definition than could be obtained with this main series. Small pieces of muscle were teased out from some of these esterase preparations and were treated by the Bodian technique so that both nerve-fibres and cholinesterase distribution were shown in the same specimen.

Tadpoles were also fixed and embedded in paraffin, and serial sections silvered by Bodian’s protargol method.

Something should perhaps be said about the specificity of the esterase method used. Several of the visceral organs gave intense staining which was clearly due to an ali-esterase rather than a cholinesterase. These tissues were usually removed by dissection before the esterase technique was applied to whole tadpoles. A positive reaction was also given by occasional blood-cells, which were, however, easy to identify under the microscope. With these few exceptions all the sites of activity seen appear to have been due to true cholinesterase. Both eserine and 1:5-bis (4-trimethylammoniumphenyl) pentane-3one diiodide (a highly specific inhibitor for true cholinesterase (Bayliss and Todrick, 1956)) were used on representative specimens at a concentration which should produce approximately 90% inhibition of true cholinesterase, and the marked reduction of staining produced was quite consistent with such a degree of inhibition. Added to this is the fact that it was usually possible to demonstrate nerve-fibres in association with the deposits of azo dye. Both the indoxyl acetate method of Barrnett and Seligman (1951) and the thioacetic acid method of Crevier and Bélanger (1955) gave results quite consistent with those of the α-naphthyl acetate method but were used only a few times because they appeared to be technically less satisfactory for our particular purpose.

Several modifications of the acetylthiocholine method introduced by Koelle and Friedenwald (1949) were tried, but all gave consistently disappointing results when incubation was carried out at 37° C. When the temperature of incubation was reduced to 22° C, however, excellent demonstration of motor nerve-endings was obtained; clearly, the cholinesterase of these tadpoles is much more heat-labile than the corresponding mammalian enzyme. This technique was not used as a routine, however, since it is not so suitable as the azo dye technique for thick specimens. Material from a few tadpoles was treated by the modification described by Shute and Lewis (1959) but with the incubation carried out at room temperature. The results obtained with this modified technique amply confirmed the view that all the sites of azo dye staining discussed in the present paper were sites of true cholinesterase activity.

Before the embryo of Xenopus leaves its jelly coat and becomes a larva, reactions apparently of a reflex type can be evoked. At stage 29 of Nieuwkoop and Faber (1956), some 5 h before hatching in our material, one tadpole responded to pinching of the tail by flexure of only the anterior region of the trunk. After the esterase reaction had been performed on this larva, significant activity was found in the dorsal region of a few of the anterior myocommata (fig. 1, A).

FIG. 1.

(plate), A, a stage 29 tadpole showing the first appearance of esterase activity in the upper parts of 5 of the anterior myocommata, indicated by dots; the other, unstained, myocommatal regions and the intervening rows of muscle nuclei are faintly visible. B, stage 45; a single myotomal fibre stained by the Bodian technique after the esterase procedure, showing the nerve-fibres and esterase reaction at both ends. c, stage 35; a 50 μ parasagittal frozen section through trunk myotomes, showing esterase activity confined to the myocommata. D, stage 50; Koelle preparation, 20 μ transverse section through trunk region, showing the spinal cord with primary motor-cells, a ventral root, and a myocommatal plane on the left. E, stage 49; Koelle preparation, 20 μ parasagittal section, showing cholinesterase activity confined to myocommatal regions. F, stage 45; horizontal section through whole brain, showing the motor-cells which innervate the pharyngeal muscles as one intensely stained and two moderately stained groups on each side in the hind-brain. G, stage 52; tail region showing change in pattern of esterase activity as arrangement of muscle-fibre alters; tip of tail to the right. H, stage 50; Koelle preparation, 15 μ transverse section through the end of an axial musclefibre, showing both a diffuse and a punctate distribution of cholinesterase activity. i, stage 55; interhyoideus muscle showing primary reaction in median raphe and at lateral edge of muscle, together with the secondary’ pattern along the lengths of the muscle-fibres. J, stage 48; constrictor brachialis stained by the Bodian technique after the esterase procedure, showing identity of motor-fibre and esterase distribution of secondary pattern of innervation.

FIG. 1.

(plate), A, a stage 29 tadpole showing the first appearance of esterase activity in the upper parts of 5 of the anterior myocommata, indicated by dots; the other, unstained, myocommatal regions and the intervening rows of muscle nuclei are faintly visible. B, stage 45; a single myotomal fibre stained by the Bodian technique after the esterase procedure, showing the nerve-fibres and esterase reaction at both ends. c, stage 35; a 50 μ parasagittal frozen section through trunk myotomes, showing esterase activity confined to the myocommata. D, stage 50; Koelle preparation, 20 μ transverse section through trunk region, showing the spinal cord with primary motor-cells, a ventral root, and a myocommatal plane on the left. E, stage 49; Koelle preparation, 20 μ parasagittal section, showing cholinesterase activity confined to myocommatal regions. F, stage 45; horizontal section through whole brain, showing the motor-cells which innervate the pharyngeal muscles as one intensely stained and two moderately stained groups on each side in the hind-brain. G, stage 52; tail region showing change in pattern of esterase activity as arrangement of muscle-fibre alters; tip of tail to the right. H, stage 50; Koelle preparation, 15 μ transverse section through the end of an axial musclefibre, showing both a diffuse and a punctate distribution of cholinesterase activity. i, stage 55; interhyoideus muscle showing primary reaction in median raphe and at lateral edge of muscle, together with the secondary’ pattern along the lengths of the muscle-fibres. J, stage 48; constrictor brachialis stained by the Bodian technique after the esterase procedure, showing identity of motor-fibre and esterase distribution of secondary pattern of innervation.

The spinal cord itself first becomes esterase-positive in the preceding stage, N. & F. 28, by which time white matter is already present in the anterior half of the cord (Nieuwkoop and Faber, 1956), but the individual fibres do not yet show any affinity for silver. The youngest larva in which we could recognize simple nerve processes in the myocommata with certainty in a Bodian preparation was one at stage N. & F. 30, and even here the assistance of the phase microscope was necessary to reveal them. The establishment of reflex behaviour before this stage, however, shows that the rudiments of the peripheral nervous system must be present before we were able to identify them in serial sections.

At stage N. & F. 32, Bodian series show a well-silvered myocommatal plexus in the anterior two-thirds of the trunk (fig. 2), the constituent fibres of which loop around the ends of the muscle-fibres. These nervous elements are the terminals of the spinal ventral roots, the development of which has been described in a previous paper (Hughes, 1959).

FIG. 2.

Sagittal section through axial musculature of trunk of Xenopus larva at stage 32, showing plexus of motor-fibres on each side of myocomma. Bodian. me, myocomma; yg, yolk granules.

FIG. 2.

Sagittal section through axial musculature of trunk of Xenopus larva at stage 32, showing plexus of motor-fibres on each side of myocomma. Bodian. me, myocomma; yg, yolk granules.

A strong esterase reaction was always observed in the whole of the more anterior myocommata in tadpoles at stages 33— 35 (fig. 1, c). At the immediately preceding stages (approximately 30— 32) some reaction could always be seen here in all tadpoles which executed swimming movements in response to pinching the tail. The spread of esterase over the anterior myocommata must be a very rapid process since so few specimens were seen in which the process was only partially complete: either activity was confined to the spinal cord or else the most anterior 6 to 10 myocommata showed a reaction over their whole surface.

The myotomes are made up of uninuclear elements which extend between adjacent myocommata. Within each muscle-cell, usually on the mesial surface, is a bundle of myofibrillae which have already become cross-striated in the more mature myotomes. The remainder of the cytoplasm of the cell still contains numerous yolk granules (fig. 1, c).

With the subsequent development of the tail-bud, the larva rapidly grows in length, and the number of myotomes concomitantly increases. In esterase preparations of the whole larva, the number of densely stained myocommata is continuously augmented, but up to stage N. & F. 43, posteriorly there is a gradual fading out of the reaction in the last quarter of the whole series of myotomes. In longitudinal sections at these stages, it can be seen that the youngest myotomes are made up of myoblasts which as yet lack cross-striated myofibrils. Bodian preparations show that an irregular plexus of ventral rootfibres run over the mesial surface of the youngest myotomes, without as yet sending fibres into the myocommata.

Esterase preparations from stage N. & F. 44 onwards show a further development in the pattern of innervation of the axial musculature. A positive reaction now extends through the whole range of myotomes, but in the new and posterior members of the series, the esterase-positive sites take the form of a row of dots on each side of the myocommatal plane, and not of a thin continuous sheet as is seen elsewhere. This more scattered distribution of esterase about the myocommatal planes is due to a different pattern of arrangement of the muscle-fibres. These are now shorter than the inter-myocommatal distance, with long strands of connective tissue attached to each end. The fibres in each myotome are arranged in an overlapping array, each fibre having a cap of esterase and a motor nerve-termination at each end. The change in pattern, which is very abrupt, can be clearly seen in esterase preparations of whole larva which have reached the feeding stage (fig. 1, G). The corresponding Bodian preparations show in the tail region a plexus of motor fibres, extending to 30 or 40p on each side of the myocommatal plane (fig. 3). In tracing through the series of motor innervations, one sees an abrupt change from the trunk region where the motor-endings are restricted to a layer of io p in width, to the broad zone of fibres in the tail musculature. This distinction is found throughout larval life; in the trunk the motor-endings remain as simple terminal fibres applied to the bases of the muscle-fibres (fig. 4), while in the tail a branched and plexiform innervation persists. The tip of the tail in the living larva of Xenopus continuously exhibits a rapid flickering movement, while the trunk remains immobile for long periods in the undisturbed animal. The spinal ventral roots of the tail region contain a number of giant fibres (Hughes, 1959), the calibre of which is presumably correlated with the conduction of rapid impulses necessitated by the flickering motion of the tail, with which its more elaborate pattern of motor innervation must also be related.

FIG. 3.

Horizontal section through axial musculature of tail of Xenopus larva at stage 45, showing broad plexus of motor-fibres between adjacent myotomes. Bodian. ec, ectoderm; d, dermis; me, site of myocomma.

FIG. 3.

Horizontal section through axial musculature of tail of Xenopus larva at stage 45, showing broad plexus of motor-fibres between adjacent myotomes. Bodian. ec, ectoderm; d, dermis; me, site of myocomma.

FIG. 4.

Horizontal section through axial musculature of trunk of Xenopus larva at stage 49, showing simple character of nerve-endings at myocommatal boundary. Bodian. me, myocomma.

FIG. 4.

Horizontal section through axial musculature of trunk of Xenopus larva at stage 49, showing simple character of nerve-endings at myocommatal boundary. Bodian. me, myocomma.

Experiments with the modified Koelle technique confirmed that all the activity seen in esterase preparations of the axial musculature was due to true cholinesterase (fig. 1, D, E). The Koelle technique was in general more limited in application than the azo dye technique, but it did sometimes give much finer detail in sections through particular regions. Thus in transverse sections through the ends of axial muscle-fibres (fig. 1, H) it revealed a punctate distribution of enzyme activity which is highly consistent with the electronmicrographs of this region obtained by Muir (1959). A paper on the fine structure of this type of ending in a number of species, including Xenopus, is due to be published shortly by MacKay, Muir, and Peters (personal communication).

The musculature which can be seen in esterase preparations of the whole larva was studied in a series of specimens from stages N. & F. 40 to 47, and the individual muscles identified as far as was possible from the descriptions given by Edgeworth (1930), Weisz (1945), and Nieuwkoop and Faber (1956). Figs. 5 and 6 illustrate the pattern of esterase-positive sites in the muscles of a 14 mm larva at stage N. & F. 47 in the pharynx, and in the branchial and abdominal cavities. Such preparations were again compared at each stage with Bodian-stained section series.

FIG. 5.

Esterase preparation of whole larva of Xenopus at stage 47. Trunk and part of tail seen from below’, ee, extrinsic eye-muscles; gh, geniohyoideus; ih, interhyoideus; im, intermandibularis; Ih, levator hyoideus; Im, levator mandibularis; me, tail myocommata.

FIG. 5.

Esterase preparation of whole larva of Xenopus at stage 47. Trunk and part of tail seen from below’, ee, extrinsic eye-muscles; gh, geniohyoideus; ih, interhyoideus; im, intermandibularis; Ih, levator hyoideus; Im, levator mandibularis; me, tail myocommata.

FIG. 6.

Lateral view of esterase preparation of Xenopus larva at stage 47. The tadpole was divided into two through the median vertical plane. The innervation of the abdominal wall shows through the pigmented coelomic wall, cb, constrictor branchialis; gh, geniohyoideus; ih, interhyoideus; im, intermandibularis; Ih, levator hyoideus; Im, levator mandibularis; me, tail myocommata; tv, trans versus ventralis.

FIG. 6.

Lateral view of esterase preparation of Xenopus larva at stage 47. The tadpole was divided into two through the median vertical plane. The innervation of the abdominal wall shows through the pigmented coelomic wall, cb, constrictor branchialis; gh, geniohyoideus; ih, interhyoideus; im, intermandibularis; Ih, levator hyoideus; Im, levator mandibularis; me, tail myocommata; tv, trans versus ventralis.

From stages N. & F. 34 to 39, nerve-branches of the trigemino-facialis group can be traced to the surface of the levator mandibularis, and along the ceratohyal, to each side of which are attached the levator hyoideus and the interhyoideus respectively. All these muscles are in an early myoblastic stage at the beginning of this period. By stage N. & F. 40, myofibrils have developed within the constituent cells of the pharyngeal muscles, while in the levator hyoideus, cross-striations are already apparent. Within all of these muscles, nerve-fibres can now be traced, yet none of them as yet show a positive esterase reaction, although in such preparations a longitudinal band of coloured cells can be seen on each side within the hind-brain in the position of the primitive motor tract, both in front of and behind the otic capsule (fig. 1, F).

In esterase preparations at stage N. & F. 41, the coloration within the hindbrain is stronger than before, and now for the first time a positive reaction is seen in the external eye-muscles, and in some of the pharyngeal muscles (fig. 7), namely, the levator mandibulae, the levator hyoideus, the geniohyoid, and the interhyoideus. At this stage, the coloration in the latter is confined almost entirely to the lateral margin and to the median raphe. By now motor-fibres have entered the lateral margin of the interhyoideus, and have grown mesially to end in the mid-line on each side. The esterase preparations show that only at their lateral entry into the muscle and at the median termination is the enzyme as yet to be detected.

FIG. 7.

Esterase preparation of larva of Xenopus at stage 41, partly dissected, and seen from ventro-lateral aspect, gh, geniohyoideus; ih, interhyoideus; h, heart; I, lungs; Ih, levator hyoideus; Im, levator mandibulae; me, tail myocommata.

FIG. 7.

Esterase preparation of larva of Xenopus at stage 41, partly dissected, and seen from ventro-lateral aspect, gh, geniohyoideus; ih, interhyoideus; h, heart; I, lungs; Ih, levator hyoideus; Im, levator mandibulae; me, tail myocommata.

In both the genio-hyoid and the external muscles of the eye, a positive esterase reaction appears as soon as their respective motor-fibres have entered. At stage N. & F. 40, the external eye-muscles are cross-striated while the genio-hyoids have not yet reached this stage of differentiation. Again, at this stage nerve-fibres have already reached the transverse ventralis muscle, which meets on each side below the bulbus cordis, but this muscle does not become esterase-positive until stage N. & F. 44, some 30 h later, by which time cross-striations are apparent in it. There is thus no uniformity in the order in which the various muscles receive their innervation, become cross-striated, and give a positive esterase (table 1). Concerning the spatial distribution of esterases within the various types one generalization, however, can be made. In long strap-like muscles, such as the external eye-muscles and the geniohyoideus, only a slight terminal concentration of the enzyme is evident, while in broader muscles a positive reaction is at first confined to the ends of the fibres, a distribution which recalls its myocommatal location within the axial musculature. The transverse ventralis muscle consists of four bundles of fibres on each side attached end-to-end. The enzyme reaction first appears at these nodal points and subsequently continues more intense there than elsewhere within the muscle.

TABLE 1.

Nieuwkoop and Faber stages

Nieuwkoop and Faber stages
Nieuwkoop and Faber stages

In thin muscle sheets, such as that of the abdominal wall and the constrictor branchialis, in both of which a positive esterase reaction is first seen at stage N. & F. 45, a single motor-fibre, along which an intense coloration is evident, runs in the septum between adjacent muscle segments. The nerves course in parallel lines within the abdominal muscles, and also in the lower and lateral part of the constrictor branchialis. Over the dorsal surface of the branchial chamber the arrangement of muscle and nerve is less regular, though even here the esterase reaction is at first confined to the ends of the muscle-fibres.

In many muscles the primary innervation is later augmented by another. The constrictor branchialis form a thin sheet, consisting only of a single layer of fibres. The development of the secondary pattern of innervation can here be seen particularly clearly. Motor-nerves spread gradually across the musclefibres, and repeated patches of esterase appear along their course (fig. 1, j). From these elements an elaborate tracery is constructed. Again, in the interhyoideus, the primary terminal reaction of the fibres is supplemented within about two days by an intermediate zone of distribution of the enzyme. The motor-fibres which have grown mesially through the muscle acquire new junctions with the muscle-cells, which in esterase preparations are evident as a delicate network between the original sites of reaction (fig. 1, 1). Although that within the median raphe remains undiminished in intensity, nerve-fibres in this zone become less conspicuous than elsewhere within the muscle. In general the esterase reaction is most intense at points where muscle-fibres are inserted in the mid-line. In long, narrow muscles, the motor-endings themselves become more complex as development progresses. By stage N. & F. 46, the neuromuscular junctions within the external eye-muscles take the form of a tangled plexus of fibres, some elements of which return upon themselves to form closed loops (fig. 8). Similar configurations are also to be seen within the long genio-hyoids. Thus with respect to their type and distribution of nerveendings, these muscles stand at one end of a series, the opposite extreme of which is represented by the myotomes.

FIG. 8.

From a horizontal section through a larva of Xenopus at stage 46, to show relatively complex nerveendings on an extrinsic eyemuscle.

FIG. 8.

From a horizontal section through a larva of Xenopus at stage 46, to show relatively complex nerveendings on an extrinsic eyemuscle.

The muscles of the abdominal wall also acquire a secondary pattern of innervation. This first becomes noticeable in esterase preparations at about stage 48 as occasional isolated groups of endings, usually near the midline (fig. 9, A, B). The secondary pattern spreads quite slowly; a week after its first appearance it is complete only over the most anterior segment (stage 50) and only after a further two weeks is it fully developed over the more posterior segments (about stages 53 to 55). At this stage of development this secondary pattern takes the form of a series of endings along the course of each musclefibre (fig. 9, c). Immediately after metamorphosis, however, most of the abdominal muscles have a different secondary pattern in which the endings run in groups obliquely across the muscle-fibres (fig. 9, D) with each musclefibre receiving one, or at the most two, endings along its length. The septal esterase activity at the ends of the muscle-fibres is still present (fig. 9, D). This apparent change in the secondary pattern is probably due to the fact (Nieuwkoop and Faber, 1956) that there is a progressive development of new muscles in this region during late tadpole stages with disappearance of some of the original abdominal musculature at metamorphosis.

FIG. 9.

(plate), A, stage 48; part of abdominal muscle stained by the Bodian technique after the esterase procedure, showing the primary reaction at the septum (on the right) and the earliest development of the secondary pattern of innervation from a bundle of nerve-fibres branching off from the septum. B, a higher power view of part of area reproduced in A showing some of the myo-neural junctions of the developing secondary pattern. c, stage 55; part of abdominal muscle, showing secondary pattern of motor-endings along course of muscle-fibres. D, soon after metamorphosis; part of abdominal musculature, showing groups of motorendings running obliquely across the muscle-segment and persistent esterase activity at the ends of the muscle-fibres. E, stage 55; part of the lateral segment of the transverse ventralis muscle, showing the profuse pattern of motor-endings. F, a higher power view of part of the same muscle-segment reproduced in E. G, stage 48; part of the lateral segment of the transverse ventralis muscle stained by the Bodian technique after the esterase procedure, showing a single nerve-fibre forming 5 esterasepositive junctions (arrowed) with three muscle-fibres over a total length of about 150 μ.

FIG. 9.

(plate), A, stage 48; part of abdominal muscle stained by the Bodian technique after the esterase procedure, showing the primary reaction at the septum (on the right) and the earliest development of the secondary pattern of innervation from a bundle of nerve-fibres branching off from the septum. B, a higher power view of part of area reproduced in A showing some of the myo-neural junctions of the developing secondary pattern. c, stage 55; part of abdominal muscle, showing secondary pattern of motor-endings along course of muscle-fibres. D, soon after metamorphosis; part of abdominal musculature, showing groups of motorendings running obliquely across the muscle-segment and persistent esterase activity at the ends of the muscle-fibres. E, stage 55; part of the lateral segment of the transverse ventralis muscle, showing the profuse pattern of motor-endings. F, a higher power view of part of the same muscle-segment reproduced in E. G, stage 48; part of the lateral segment of the transverse ventralis muscle stained by the Bodian technique after the esterase procedure, showing a single nerve-fibre forming 5 esterasepositive junctions (arrowed) with three muscle-fibres over a total length of about 150 μ.

A number of workers have assayed the cholinesterase of developing Amphibia by chemical methods (Youngstrom, 1938; Sawyer, 1943 a, b\Boell and Shen, 1950). All have used Amblystoma; Youngstrom in addition worked on two anuran species. Each of these investigators has demonstrated a rapid rise in the content of the enzyme from the time when the larva begins to swim, though in some points there are differences in detail between their findings. Sawyer found that some cholinesterase was present in the embryo as early as the neurula stage in ‘presumptive nerve and muscle’. Boell and Shen, however, found that the ‘first appearance of cholinesterase in detectable amounts in the spinal cord coincides with the ability of the embryo to respond neurogenically to tactile stimuli’.

Sawyer (1943) measured cholinesterase in the somatic musculature of larvae from which the spinal cord had been removed. In such myotomes with which no motor nerve-fibres had ever made contact, the enzyme was still found to be present, though in much less amounts than in the normally innervated muscle of the corresponding age. Sawyer’s result thus indicates that the cholinesterase of the axial musculature of Amblystoma has a dual origin, partly intrinsic, but with the greater part originating in some way through contact with motor nerve-roots.

In the embryo of Xenopus the first signs of a positive esterase reaction is in the anterior region of the spinal cord at stage N. & F. 28. A myocommatal reaction beyond the cord is first seen in the succeeding stage. The correlation of these physiological events with the development of the spinal motor-roots is hampered by the fact that nerve-fibres at these stages have as yet no affinity for silver, either in Xenopus or in Amblystoma, where Coghill (1913) found that impregnation methods were not of assistance until the early swimming stage had been reached.

However, our observations on all other muscles which develop subsequently to the axial musculature have shown that motor-nerves arrive either before or at the same time as an esterase reaction first appears. Thus a group of pharyngeal muscles which are innervated at stage N. & F. 40 show a positive esterase reaction in the following stage. The cell-bodies of their motor-fibres within the motor tract of the hind-brain are first esterase-positive at stage N. & F. 40, from which time the reaction increases continuously in intensity. Here, then, the evidence points clearly to the interpretation that the enzyme, or some precursor, is synthesized within the perikaryon, and reaches the nerve-muscle junction by efferent flow along the axon (Lewis and Hughes, 1957).

Both in the axial musculature and elsewhere there is a close correlation between the first appearance of cholinesterase and the onset of reflex behaviour in each muscle. This correlation was particularly striking in a tadpole at stage N. & F. 29, which responded to stimulation by a simple flexure of the forepart of the trunk, and was found to show enzyme activity in only a few anterior myocommata. In the pharyngeal musculature, enzyme activity and reflex behaviour develop simultaneously, for when the enzyme is first detectable the response of the muscle is rudimentary, and co-ordinated movement only begins at about the stage when the pattern of the distribution of the enzyme approaches completion.

Very different patterns of innervation are seen in the various muscles and it is possible to suggest relationships between each particular arrangement and the type of activity demanded of the muscle. In the transverse ventralis muscle, for instance, each lateral segment consists only of a few muscle-fibres of relatively small diameter, in which an enormous number of separate patches of esterase activity are seen (fig. 9, E, F). A study of segments dissected from esterase preparations and then stained by the Bodian technique showed that each of these patches of esterase activity was a genuine myo-neural junction, though simple in form. They are distributed along almost the whole length of each muscle-fibre at intervals of 30 to 150 p (fig. 9, G), a branch of the same nerve sometimes innervating several consecutive junctions on the same muscle-fibre. Although it was not possible to make an accurate count of both muscle-fibres and myo-neural junctions in a single segment, rough counts suggest that each muscle-fibre possesses some io to 20 junctions along its whole length. It seems likely that this pattern is in fact that of the ‘slow muscle innervation’ known to be present in amphibians (Kuffler and Vaughan Williams, 1953). These slow muscle-fibres have a multiple innervation which does not cause a twitch but only a slow contracture which is induced by electrotonic spread of depolarization from the individual myo-neural junctions. A similar pattern of innervation is seen in some of the abdominal muscles before metamorphosis (fig. 9, c). The muscle-fibres of the constrictor branchialis also have a multiple innervation though here the nerve-fibres run mainly across the muscle-fibres (fig. 1, j).

In the extra-ocular muscles the profuse distribution of separate motor end-plates may be correlated with the need for precisely-graded reflex responses in these muscles. At an opposite extreme is the pattern seen in the main axial musculature, where each muscle-fibre is innervated only at its ends from plexuses in the myocommata (fig. 1, B). This innervation of the myotomes calls for some comment since it is so unlike the usual type of myoneural junction in vertebrate muscle. Myocommatal endings of nerve-fibres in lower vertebrates have previously been described by Giacomimi (1898 a, b), who, however, regarded them as wholly sensory in nature. In the early development of the motor system in Amblystoma, Coghill (1926, pp. 101,102) recognized that some branches of spinal motor-nerves end at the myocommata. Again, in Harrison’s classical paper (Harrison, 1924) on the respective contributions of neural crest and cord to the development of the peripheral nervous system in this urodele, he showed that the first motor-fibres to the abdominal musculature end at the septal planes, while the sensory nerves spread generally over the surface of the segmental units. Terminal motorendings to muscle-fibres in selachians have recently been described by Couteaux (1955). Although the existence of myo-neural junctions at both ends of the muscle-fibre would not seem to be a suitable general method of obtaining well co-ordinated reflex activity, it might have a specific advantage in the larva of Xenopus, which passes through the early stages of development at great speed; a sharp avoiding-reaction to tactile stimulation appears within considerably less than 48 h of fertilization. Nerve-fibres are believed to grow most readily along surfaces (contact guidance; Weiss, 1941) and hence functional contact between the motor nerve-fibres and the myotomal muscles might be most rapidly achieved along the myocommatal planes—i.e. at the ends of the muscle-fibres. Some observations on the regenerating tail-bud of the lizard, Sphaerodactylus, are of particular relevance here (Hughes and New, 1959). The muscles in the normal tail of this lizard have large classical end-plates along the course of each muscle-fibre. When muscle-fibres first develop in the regenerating tail, however, they are first innervated only at their ends and give a pattern reminiscent of that of the myotomes in Xenopus tadpoles. At a later stage the nerve-fibres grow out over the surface of the muscle to form the normal vertebrate type of end-plate. Innervation at the ends of the muscle-fibres may thus be a special adaptation to permit of the earliest possible neurogenic activity, where speed of development is important and accurate control of response is not. This idea might also explain the first appearance of enzyme activity at tendinous septa in several muscles, e.g. interhyoideus, the abdominal muscles, c., where the more normal type of myo-neural junction along the length of each muscle-fibre is formed later, when presumably the need for co-ordinated reflex activity arises.

We are indebted to Drs. P. A. Tschumi and R. T. Sims for their co-operation in the supply and rearing of tadpoles, to Miss V. Day for technical assistance, and to Mr. J. F. Crane for the photography. Some of the expenses of this research were defrayed by grants from the Nuffield Foundation and the Medical Research Council.

Barrnett
,
R. J.
, and
Seligman
,
A. M.
,
1951
.
Science
,
114
,
576
.
Bayliss
,
B. J.
, and
Todrick
,
A.
,
1956
.
Biochem. J
.,
62
,
62
.
Boell
,
E. J.
, and
Shen
,
S. C.
,
1950
.
J. exp. Zool., 1×3, 583
.
Coghill
,
G. E.
,
1913
.
J. comp. Neurol
.,
23
,
121
.
Coghill
,
G. E.
1926
.
Ibid
.,
41
,
95
.
Couteaux
,
R.
,
1955
.
Int. Rev. Cytol
.,
4
,
335
.
Crevier
,
M.
, and
BÉlanger
,
L. F.
,
1955
.
Science
,
122
,
556
.
Edgeworth
,
F. G.
,
1930
.
J. Anat. Lond
.,
64
,
184
.
Giacomimi
,
E.
,
1898a
.
Atti Accad. Siena
,
10
, 105, 116, 371, 375, 560.
Giacomimi
,
E.
1898b
.
Monit. zool. ital
.,
9
,
92
.
Harrison
,
R.
,
1924
.
J. comp. Neurol
.,
37
,
123
.
Hughes
,
A. F.
,
1959
.
J. Embryol. exp. Morph
.
7
,
128
145
.
Hughes
,
A. F.
and
New
,
D.
,
1959
.
Ibid, (in press)
.
Koelle
,
G. B”
and
Friedenwald
,
J. S.
,
1949
.
Proc. Soc. exp. Biol. N.Y
.,
70
,
617
.
Kuffler
,
S. W.
, and V
aughan Williams
,
E.
M
.,
1953
.
J. Physiol
.,
121
,
289
, 318.
Lewis
,
P. R.
,
1958
.
Quart. J. micr. Sci
.,
99
,
67
.
Kuffler
,
S. W.
and
Hughes
,
A. F.
,
1957
.
In Metabolism of the nervous system,
p.
511
(Ed. Richter). London (Pergamon Press).
Muir
,
A. R.
,
1959
.
J. Anat. Lond. (in press)
.
Nieuwkoop
,
P. D.
, and
Faber
,
J.
,
1956
.
Normal tables of Xenopus laevis (Daudin). Amsterdam (North Holland Publishing Co
.).
Sawyer
,
C. H.
,
1943a
.
J. exp. Zool
.,
92
,
1
.
Sawyer
,
C. H.
1943b
.
Ibid
.,
94
,
1
.
Shute
,
C. C. D.
, and
Lewis
,
P. R.
,
1959
.
J. Anat. Lond. (in press)
.
Tschumi
,
P. A.
,
1957
.
Ibid
.,
91
,
149
.
Weiss
,
P.
,
1941
.
Growth (suppl
.),
5
,
163
.
Weisz
,
P. B.
,
1945
.
J. Morph
.,
77
,
163
.
Youngstrom
,
K. A.
,
1938
.
J. Neurophysiol
.,
1
,
357
.