139 



embryos of Pristiurus, however, it is shown by Dohrn (1885) arising 

 in the angle formed where the efferent mandibular artery joins the 

 lateral dorsal aorta, if not even from this latter artery itself slightly 

 anterior to this point, and in an 11 mm. embryo of the trout it is 

 shown by Dohrn (1886) definitely arising from the lateral dorsal aorta 

 slightly anterior to the point where that artery is joined by the ef- 

 ferent mandibular artery. Of the trout Maurer also says (1888, p. 215) : 

 "bei Forellen, die gerade ausschlüpfen, das Auge noch direkt vom Circu- 

 lus cephalicus aus mit Blut versorgt wird". The origin of this artery from 

 the lateral dorsal aorta may, accordingly, be the primary condition, the 

 ophthalmica magna quite probably being, in that case, the persisting 

 efferent premandibular artery. This possibility was considered by me 

 in an earlier work (Allis 1908a) but rejected in favor of the sup- 

 position that the artery is a dorsal commissure between the mandibular 

 and premandibular arches. This latter supposition now, however, 

 seems to me less probable than the former, for, as will be shown 

 below, the external carotid would seem to be developed, in part, from 

 the prehyoidean dorsal commissures; and if that be the case the 

 ophthalmica magna evidently could not then be developed from those 

 same vessels. The ophthalmica magna artery is accordingly here 

 considered, in the diagrammatic figures given, as the persisting ef- 

 ferent premandibular artery. 



The hyo-opercularis and external carotid arteries now remain to 

 be considered, and it seems quite certain that they are both developed 

 from the efferent hyoideau artery, from some part of some prehyoidean 

 artery or arteries, and from portions of certain of the dorsal and 

 intermediate commissures that must primarily have connected these 

 arteries with each other and with the first efferent branchial artery. 

 In Mustelus and in the Scyliidae the posterior efferent hyoidean 

 artery, the dorsal commissure that connects that artery with the first 

 efferent branchial artery and an intermediate commissure that connects 

 it with the mandibular artery are all found both in larvae and in 

 the adult, while the anterior efferent hyoidean artery and the dorsal 

 commissure that connects that artery with the mandibular aortic arch is 

 either never developed or not retained even in larvae. Comparing these 

 conditions in these selachians with the conditions found in Amia, it 

 is evident that the posterior efferent hyoidean artery of selachians is 

 represented, in Amia, in the continuous vessel formed by the hyo- 

 opercularis and the secondary afferent hyoidean arteries, this continuous 

 vessel traversing the facialis canal through the hyomandibular. The 



