MUSTELUS L.EVIS. 143 



depending upon the peripheral reUitions of its end organ 

 with reference to that structure." This seems to me too 

 dangerous a principle to be applied without definite facts to 

 support it in each particular case, and such facts do not here 

 exist. Pushed to a legitimate extreme, a nerve might even 

 be assumed to change its relations to such an obstacle as a 

 visceral cleft. In any event a pure assumption is here 

 invoked by Herrick to explain certain facts that I had sought 

 to explain by what was, at the time, an equally pui'e assump- 

 tion ; that is, that the internal recti muscles of vertebrates 

 are not all homologous structures. Recent work relating to 

 this especial subject supports, however, m}^ assumption. For 

 both Hoffmann (34) and Sewertzoff (58j say that the rectus 

 superior and rectus internus of Acanthius and Torpedo re- 

 spectively arise from the dorsal wall of the head cavity from 

 which all the muscles innervated by the oculomotorius arise : 

 while Rex says (63, p. 235) that in the duck the rectus 

 internus and the rectus inferior arise from a single embryonic 

 muscle-mass that has its origin from the ventral wall of the 

 same cavity, the rectus superior alone arising from the dorsal 

 wall of the cavity (p. 238). The bundle of fibres that I found 

 in one specimen of Amia hanging festoon-like between the 

 rectus internus and rectus inferior muscles thus receives an 

 evident explanation, and, inversely, tends to confirm for Amia 

 the origin ascribed by Res to the corresponding muscles in 

 the duck. This thus seeming to be established, the conditions 

 presented in Mustelus seem to be such as might have neces- 

 sitated or led to the development of another and different 

 internus. The internus in this fish is so obstructed in its 

 effort to obtain an attachment on the skull, that a con- 

 siderable part of it remains for a long time wholly func- 

 tionless. Had it been still more obstructed the muscle 

 might not have been able to maintain its separate and 

 independent existence, and a new and wholly different in- 

 ternus would have been developed. 



Herrick calls attention (32, pp. 234 — 236) to two apparent 

 errors in my diagrammatic representation of the eye-muscles 



