THE DEVELOPMENT OF THE VENTRAL NERVES IN SELACHII. 301 



a. Is the neuraxon of a ventral nerve fibre multicellular in origin or the process 

 of a single medullary cell? Three answers have been given to this question. The 

 first and oldest one is that neuraxones are formed by the fusion and differentiation 

 of the axial protoplasm of chains of spindle-shaped cells. Harmonizing with the early 

 opinions of nerve structure as stated by Remak and Schwann, this hypothesis seemed 

 to receive embryological confirmation in the observations of the cellular structure 

 of embryonic selachian nerves by Balfour ('81) and others and in the more detailed 

 studies of selachian nerve histogenesis by Dohrn ('91, '92) and by Beard ('92, '96). 

 According to Dohrn ('91) the histogenesis of the neuraxon and neurilemma in a 

 spinal ventral nerve is as follows : One of the cells composing the growing nerve sepa- 

 rates from its envelopes, elongates into a slender structure in the middle of which 

 lies a spindle-shaped nucleus and differentiates within its cytoplasm a neuraxon, 

 one end of which forms the nerve fibre while the other penetrates between the muscle 

 fibres and effects the first connection between the nerve and its end organ, the mus- 

 cle. The remainder of the cytoplasm becomes the neurilemma. 



Dohrn confirms the opinion of Goette ('88) that "ganglion cells" have no genetic 

 relation with neuraxones which are formed by "nerve cells" in the manner just de- 

 scribed. On the basis of observations upon the structure of adult nerves, Apdthy 

 ('89-'90) makes the same distinction between ganglion cells and nerve cells. The 

 evidence of the "cellular structure of embryonic nerves" given by Balfour ('81), 

 Van Wijhe ('89), Beard ('96), Dohrn ('91, '92), Beraneck ('87), Kupffer ('94, '95), 

 Hoffmann ('97), Capobionco e Fragnito ('98), and Raffaele (:00) has been generally 

 but fallaciously thought to favor the cell-chain hypothesis of nerve histogenesis. 

 The considerations advanced by various advocates of the cell-chain hypothesis in 

 support of this view are as follows : first, embryonic nerves are cellular ; secondly, the 

 nerves of invertebrates are cell chains; thirdly, by analogy with the differentiation 

 of muscle fibres, we should expect the nerve fibres to develop in situ by a differentia- 

 tion of cell cytoplasm; fourthly, the direct observations of Dohrn ('91) of the differen- 

 tiation of the neuraxon fibre within the cytoplasm of the nerve cell; fifthly, the 

 improbability that a cell can give rise to a process twenty thousand tunes as long 

 as the diameter of the cell; and finally, the adult structure of the nerves as demon- 

 strated by Apathy's method. 



According to the second theory of neuraxon histogenesis, the neuraxon is an 

 extraordinarily long process of a ganglion cell, and every nerve fibre from beginning 

 to end is to be understood as a product of a single cell (von Lenhosse"k, '97). This 

 theory, first stated by Kupffer in 1857 and established on an empirical basis by His 

 ('79) has been confirmed most conclusively through the use of the Golgi method by 



