THE DEVELOPMENT OF THE VENTRAL NERVES IN SELACHII. 305 



First, the first cells found in the forming nerve appear at the base of the nerve 

 where it joins the medullary wall. Other similar cells he half in and half out of the 

 medullary wall. As the cells of the nerve grow more numerous, more nuclei appear 

 in process of migration from the neural tube. The marked change in the contour of 

 the ventral wall of the tube (PI. XXIII, Figs. 20-23) indicates the great extent of 

 this migration. 



Second, like the neural tube, the forming nerve has a membrana limitans which is 

 continuous with the limiting membrane of the neural tube. Such a relation would not 

 be expected were the cells of the nerve added from the adjacent mesenchyme. The 

 presence of this membrane makes it possible to distinguish easily in all earlier stages 

 (PL XXIII, Figs. 12-19) between the cells of the forming nerve and those of the 

 adjacent mesenchyme. 



The cells of the young forming nerve are all of medullary origin. The ques- 

 tion arises whether all of the cells which form the neurilemma of the adult nerve- 

 fibres come from a similar source. I regard this as improbable and I am led to as- 

 sume the participation of the mesenchyme in the formation of the neurilemma on 

 the following grounds. The extensive migration of cells from the neural tube into 

 the forming nerve ceases at an early stage (PL XXIII, Fig. 19). At this and stages 

 immediately following I have never been able to count more than one hundred cells 

 (nuclei) in the nerve. I am not able to say with positiveness that at this stage all 

 medullary emigration has ceased. There is some evidence of an inconclusive kind 

 that cells may leave the neural tube in later stages. 



Among the emigrated cells I have seldom seen any mitoses. Hoffmann ('97) 

 says that they are very rare. If we admit a doubling of the number through mitosis 

 and another doubling through additions from the neural tube the number of nuclei 

 in a spinal ventral nerve would then be only four hundred. But there are in my 

 judgment over a thousand nuclei in the neurilemma of an adult spinal ventral nerve. 

 The estimate is difficult to make because of the union with the dorsal nerve, the for- 

 mation of the sympathetic and the impossibility of determining accurately the length 

 of the individual neuraxones. The number given above is, however, a very conserv- 

 ative estimate. It follows that the neurilemma must receive accessions from the 

 mesenchyme. Sections, indeed, show cells of the mesenchyme closely applied to 

 the growing nerve at its termination, but whether these form neurilemma cells or 

 a part of the connective tissue sheath I am unable to say. 



Thus, while I am certain of the ectodermal contribution to the neurilemma the 

 participation of the mesoderm is less certain though probable. Indeed, if the facts 



