CEANIAL NERVES AND SENSE ORGANS OF FISHES. 243 



the visceral and pre-branchial (pre-trematic) branches of 

 the nerve, whilst the motor fibres form entirely the post- 

 branchial (post-trematic) branch. This can, to an extent, 

 be verified by dissection, when it is found that only the 

 fibres of the post branchial nerve can be traced into the 

 muscles of the gill arches (cp. Allis's last Amia paper*). 

 Both pre- and post-branchial nerves are, as a rule, con- 

 tinued beyond the gill bearing portions of the arches on 

 to the floor of the pharynx, and it is these branches of 

 the piscine branchial nerve which undoubtedly have 

 formed the lingual branches of the posterior cranial 

 nerves of the higher vertebrates. 



The pre-branchial nerve, therefore, is a visceral sensory 

 nerve, and the post-branchial a visceral motor nerve, both 

 belonging essentially to the sympathetic system. It is 

 now easy to see why each arch should be supplied by two 

 nerves, since both bundles differ essentially in nature — 

 the branch edging the anterior surface of the arch being 

 motor and that coursing along the posterior edge sensory. 

 Both bundles, although they accompany one another for 

 a time, are absolutely distinct in every other respect, so 

 much so that it appears to me probable they originally 

 pursued different courses, and that each visceral arch was 

 supplied by two independent nerves. 



We have seen that it is unphilosophical, at present at 

 any rate, to arrange the cranial nerves into a fixed number 

 of serially occurring pairs, and that even in the case of the 

 branchial nerves, where the serial arrangement is most 

 marked, it would be hazardous to definitely assert that 

 each branchial nerve represented a metameric unit. That 

 the cranial nerves form an association of structures whose 

 present form is due practically to physiological necessities, 

 and not to a survival of an originally formal serial arrange- 

 * Jour. Morpli., 1897. 



