656 REPORT— 1902. 



probably these elements are prismatic and fit closely together, the cylindrical 

 t'orm being due to shrinkage. The outer wall of the myotome increases much in 

 thickness, becoming eventually considerably thicker than the inner. About stage 35 

 there can be distinguished in the myomere two layers of muscle — derived from the 

 two walls of the myotome, the inner and the outer, and differing decidedly in their 

 characters— the muscle-cells of the outer layer being long thin cylinders, as has been 

 described; those of the inner being, on the other hand, flattened parallelepipeds. 

 The bundles of fibrils of the outer layer are also separated by more abundant 

 protoplasm, with more numerous nuclei. A further difference, probably due to 

 the active metabolism involved in the multiplication of the nuclei and the growth 

 of the layer in thickness, is that the outer layer is practically free from yolk- 

 granules, while the inner is still richly provided with them. 



About the stage just mentioned, however, the blocks of fibrils of the inner 

 series become irregular in outline, as seen in transverse section, they are con- 

 stricted in places, and finally separate completely into more or less isodiametric 

 masses resembling those of the outer series. By about stage 36 + the whole of the 

 muscle-bundles in the myomere have assumed the same character, i.e., the 

 character at first possessed by only the outer series. 



The main features of myotome development recorded above are of especial 

 interest in regard to two points : first, in the close resemblance that the early 

 stages bear to the equivalent plieuomena in Amphioxus and Petromyzon ; secondly, 

 in the striking way in which Lepidosiren agrees with Balfour's view in regard to 

 Selachians, that the outer wall of the myotome takes part in muscle formation, 

 a view given up almost entirely by recent workers. 



The Neuromyoepithelial Cell. — -There remains to be mentioned a phase of great 

 morphological interest, which occurs transiently in the muscle-cells of the inner 

 wall of the myotome. It has been recognised for Amphia.vus and Petromyzon 

 (Rabl, Hertwig, Maurer) that in an early stage of development these muscle-cells 

 are definite myoepithelial cells. 



This is similarly perfectly clear in Lepidosiren ; but here I have found an addi- 

 tional important detail — namely, that at least some of the myoepithelial cells are con- 

 tinued at their inner ends into tail-like processes continuous with the motor-nerve 

 trunk. Each cell is, in fact, neuroepithelial as well as myoepithelial in character. 

 The Motor Nerves. — The method of development of the peripheral nerve- 

 trunks is one of the most disputed questions in vertebrate morphology. Of the 

 various views which have been held on this subject the one which is predominant 

 at the present day is that associated with the name of His — namely, that the 

 peripheral nerve-fibre is an outgrowth from the central nerve-cell, that it grows 

 outward from this origin, and eventually becomes joined up to its end-organ. 

 Hensen long ago suggested that, on the other baud, the end-organ and the nerve- 

 centre are joined up from the beginning. This view, though still held on mainly 

 theoretical grounds by several embryologjsts, has lacked the support of definite 

 observation. Lepidosiren happily during its development shows facts which may 

 almost he regarded as demonstrating the truth of Hensen's view in its main prin- 

 ciples as regards the motor nerves. In tracing back the motor trunks of the spinal 

 nerves I reached a stage in which the faintly tibrillated trunk was ensheathed in a 

 nucleated mass of protoplasm of mesenchymatous origin. It seemed as if the 

 nerve-trunk were becoming differentiated out of the strand of mesenchyme — that 

 it developed exactly after the fashion described by Sedgwick for Selachians. 

 Pushing back my investigation to earlier and earlier stages, however, I was 

 astonished to find the nerve-trunk rudiment well developed, while there was as )'et 

 no trace of its ensheathment of mesenchyme. Eventually a stage was reached at 

 which the myotome and the neural tube were still in close contact, and where, of 

 course, it was practically impossible to demonstrate the existence or non-existence 

 of connecting bridges between the two. Fortunately I had when in South 

 America been able to excise numerous embryos of this stage from the egf^ during 

 life, to spread them out in one plane under normal salt-solution, and in this 

 position to subject them to the action of the fixing agents. On examining trans- 

 verse sections of such embryos I found that the myotome had become slightly 



