THE SEGMENTAL VALUE OP THE CRANIAL NERVES. 163 



Firstly, we might suppose that in some way, and for some reason, 

 the sensory and motor portions of the originally single root became 

 completely separated from one another, and that while the sensory- 

 portion of the nerve retained the primitive mode of development and 

 position of attachment of the root, the motor portion acquired a new 

 mode of development and a new position of attachment, and then 

 united farther on with the posterior root to form a spinal nerve. On 

 this view the motor and sensory roots of a spinal nerve correspond to 

 the motor and sensory portions of the single root of Amphioxus. 



Or, secondly, we might imagine the anterior root to be, not the 

 motor portion of the original root, but an altogether new development, 

 an independent outgrowth from the spinal cord to supply the more 

 complicated system of muscles that would necessarily accompany the 

 gradual perfection and complication of the internal skeleton • that this 

 new root was at first completely independent of the original or dorsal 

 root, and for a time coexisted with a dorsal root of mixed function ; 

 that in the case of the spinal nerves the whole motor function gradually 

 got transferred to, or usurped by, the new root ; while the two 

 roots, originally separate along their whole length, became united to 

 form the mixed trunk of the spinal nerve. 



Now, although there are very considerable and obvious difficulties 

 in the way of accepting either of these alternatives, yet it appears to 

 me that the second is far more in accordance with the actual facts 

 than the first, and that it offers a ready explanation of many points 

 unintelligible on the first hypothesis. Thus, the second view explains 

 why in actual development the anterior spinal roots appear later than 

 the posterior, and why they are for some time quite distinct from 

 these latter ; it also explains such cases as Petromyzon, in which the 

 anterior and posterior roots of the spinal nerves are said to remain 

 distinct from one another throughout life. 



By far the most important argument, however, in favour of the 

 second hypothesis is afforded by the explanation it yields of the con- 

 dition of the cranial nerves as compared with the spinal j and in 

 connection with this point I would direct special attention to the 

 statements already made concerning the sixth and seventh nerves. 



It has been shown above that the seventh nerve in Elasmobranchs 

 develops in a manner precisely similar to the posterior roots of the 

 spinal nerves; that it arises as an outgrowth from the neural crest 

 (fig. 2, VII), the nerves of the two sides being at first directly and 



