258 bulletin: museum of comparative zoology. 



paper for cranial ventral nerves in Selachii, leads to the same conclusion. 

 The visceral cleft which defined anteriorly the splanchnic part of the 

 fourth segment is not ontogenetically evident in Squalus. KupfFer ('93) 

 has possibly seen evidence of a rudimentary cleft between the mouth and 

 the hyomandibular cleft of Acipenser. And possibly this cleft may be 

 represented in the "Pseudobranchialrinne " of Amphioxus. 



e. Relations of Excephalomere III. 



As in the case of the four posterior hindbrain segments, the study of 

 the development of the nerves connected with encephalomere III (Hinter- 

 him) gives the clue to the primitive relations of this primary vesicle. 

 The neural-crest cells proliferated from it pass ventrally into the man- 

 dibular arch. From a part of these a large ganglion is formed (the 

 Gasserian), through which pass the motor fibres, whose nucleus is, at 

 least in part, in encephalomere III, to innervate the musculature of the 

 first visceral (mandibular) arch. We have thus the splanchnic elements 

 of a cranial segment. In the Table of Nerve Relations (p. 253) the troch- 

 learis has been given as the ventral (somatic) nerve of this segment. 

 The evidence in favor of this view has already been stated, and consists 

 in the facts that it innervates musculature derived from dorsal (somatic) 

 mesoderm, that its fibres develop as processes of neuroblasts in the neural 

 tube, and that its histological relations and structure in the adult show 

 it to be a purely motor nerve with motor nucleus in the ventral horn of 

 encephalomere III. I regard the mouth as representing the fused visce- 

 ral clefts which bounded anteriorly the splanchnic portion of this seg- 

 ment. We have thus all the essential elements of a head metamere. 



/. Relations of Encephalomere II. 



From the simple dorsal expansion of encephalomere II are proliferated 

 cells which pass ventrally and fuse with the skin to form the meso- 

 cephalic ganglion ^ lateral to the 1st somite (Figs. 17 to 20). Although 

 this ganglion never becomes connected with the midbrain (encephalo- 

 mere II), since its fibres enter the brain through the r. ophthalmicus 

 profundus V, it must in my opinion be regarded as a segmental gan- 

 glion comparable with those of the following cranial nerves ; the oph- 

 thalmicus profundus must likewise be considered as a dorsal nerve 

 horaodynaraic with the succeeding cranial nerves. Its want of motor 

 fibres may be explained as resulting from phylogenetic loss, since in 



1 This ganglion is homologous with the first trigeminus ganglion of Cyclostomes. 



