xeal: xekvous system in squalus acaxthias. 277 



(encephalomeres) may be considered as the primitive nervous centres of 

 nerves which coiTesponded numerically with visceral arches. If they 

 were such, then one of the encephalomeres (FV) affords evidence of a 

 lost visceral arch. 



Although the structure of myelomei-es and encephalomeres is seen to 

 be different, yet in the stages of embryonic development, where both 

 are present, the latter are seen to have segmental value from the fact 

 that corresponding with them there is an equal number of somites. 

 These somites, as exemplified in the 3d (van "Wijhe's), are morphologically 

 comparable and serially homologous with trunk somites. I conclude, 

 then, that there was a primitive correspondence between neuromerism, 

 mesomerism, and branchiomerism. 



The development, histological structure, and relationships of the 

 eye-muscle nerves (III, IV, and VI) show them to be the serial homo- 

 logues of ventral spinal nerves. Like the latter (His), they develop 

 as axis-cylinder processes of neuroblasts in the ventral horn of the 

 neural tube. 



Pre-otic and post-otic metameres, like their integral parts, are serially 

 homologous with one another. Therefore, if the latter are serially 

 homologous with trunk metameres the former must be also. Table III. 

 (p. 253) summarizes my opinion as to the primitive composition of 

 metameres I to VII. I regard the r. opthalmicus profundus as a seg- 

 mental dorsal nerve belonging to metamere II, while the oculomotorius 

 is its ventral root. The trochlearis is the ventral nerve of metamere 

 III, and the abducens represents the ventral nerves of metameres IV 

 to VII. 



There are five mesomeres alternating with six neuromeres in the otic 

 and pre-otic regions of the Vei'tebrate head. Probably eleven neuro- 

 meres are finally included in the head of Squalus. The evidence of the 

 numerical correspondence of neuromeres and mesomeres shows that 

 there is no more reason for believing that somites have been lost 

 anterior to Piatt's (anterior) somite, than that neuromeres have been 

 lost anterior to the primary forebrain. 



In agreement with van "W'ijhe, I homologize the mouth of Amphi- 

 oxus with the left half of the mouth of Craniota. The first pair of per- 

 manent visceral clefts in Amphioxus are exactly homologous with the 

 hyomandibular clefts of higher Vertebrates. The eight visceral clefts 

 possessed by Amphioxus at its "critical stage" (Willey) are exactly 

 homologous with the eight morphological clefts found in some Selachii 

 and Cyclostomes. 



