neal: nekvous system in squalus acanthias. 271 



true, is easier for occipital than for pre-occipital segments. The serial 

 homology of occipital with trunk segments is not generally questioned 

 at present. A comparison of the integral parts of occipital and 

 trunk metameres shows that the helief in their serial homology is well 

 founded. It must, however, he admitted that occipital metameres 

 show no evidence of either excretory or reproductive organs. ISTever- 

 theless we may readily believe from the evidence of these organs in the 

 gill region of Amphiosus that this is a coenogenetic loss in the Ver- 

 tebrate series. The chief grounds for belief in the homology of trunk 

 aud occipital metameres are these : (1) Occipital somites with their 

 (2) ventral nerves are undoubtedly the serial homologues of trunk 

 somites with their ventral nerves. This evidence alone has convinced 

 most morphologists. But there are still other reasons. With our 

 present knowledge, we may, I think, affirm that (3) dorsal occipital (or 

 cranial) and doi-sal spinal nerves ai'e serial homologues. One by one, 

 since the discovery by Schneider ('79) of ventral nerves in Amphioxus, 

 the differences between dorsal spinal and cranial nerves, which were 

 at one time or another maintained, have been with increased compara- 

 tive embryological and anatomical knowledge shown to be unessential. 

 The evidence given by Schneider (79), Hatschek ('92), and van Wijhe 

 ('93) shows that dorsal nerves, as seen in Amphioxus, are mixed in 

 function, innervating the skin and splanchnic musculature, while ventral 

 nerves are motor in function, innervating somatic musculature. The 

 typical cranial nerves of Craniota, viz. V, VII, IX, and X, are mor- 

 phologically comparable with the dorsal nerves of Amphioxus, and are 

 therefore to be regarded, as Balfour for other reasons regarded them, 

 more primitive than the spinal nerves, which lack the lateral and dor- 

 sal (except in Cyclostomes) cutaneous branches.^ The recent researches 

 of von Lenhossek ('90), Ramon y Cajal, and Kolliker, by demonstrating 

 the existence of non-ganglionic fibres in the dorsal spinal nerves of 

 Craniota, which by their relations must be regarded as motor in func- 

 tion, have shown that in this respect spinal nerves do not differ from 

 cranial. Moreover, in view of the evidence given by Goronowitsch ('92), 

 Sewertzoff ('9.5), Neal ('97), and Miss Piatt ('97), it can no longer be 



1 The place of these branches has been usurped by the lateral branches of tlie 

 vagus, as I believe lias been suggested by Eisig. The advantage in greater cen- 

 tralization is obvious. If it be true, and it is generally admitted, that cranial nerves 

 receive cells from the skin while the spinal nerves do not, an explanation of 

 this also is seen in the extension of the vagus and the concomitant loss to 

 spinal nerves. 



