CRANIAL NERVES AND SENSE ORGANS OF FISHES. 241 



nerves. The occurrence of the blood vessels is readily 

 explicable, but why one arch should be supplied by two 

 nerves, when a compound trunk would serve the purpose 

 equally well, is a matter which does not seem to have 

 been much thought about. Then again, a branchial nerve is 

 not confined to the supply of one arch, but sends branches 

 to parts of two. The distribution of these branches at 

 once suggests that the gill arches are intersegmental, 

 i.e., if the branchial nerves themselves are segmental. But 

 however this may be, each branchial nerve consists of at 

 least two parts, which are morphologically as distinct 

 from one another as the fifth nerve is from the seventh. 

 A third division, which is often but not invariably present, 

 is concerned largely with the skin of the b>ack. .It would 

 be an interesting enquiry to determine, in the light of 

 recent investigation into the nature of branchial nerves, 

 why the latter fork over a cleft and not over an arch. This, 

 as above stated, is at once explained if the arches were 

 developed intersegmentally, but it assumes that the nerves 

 themselves are segmental (which is, however, possible 

 enough), whilst it may also be explained, in virtue of the 

 compound nature of the branchial nerves, by the fusing 

 of contiguous bundles of nerve fibres — one sensory and 

 the other motor. This, on account of the completely 

 distinct character of the main components of a branchial 

 nerve, seems to me to be the more probable view, but 

 such a view necessarily makes the solution of the problems 

 concerned more difficult than might at first have been 

 supposed. 



A typical branchial nerve (by which I mean one that 

 includes a somatic sensory component) arises from three 

 nuclei in the brain — from the "motor nucleus," "the 

 spinal fifth tract," and the nucleus of the fasciculus 

 communis system. Leaving out the more specialised tri- 



