132 



mobranchs, both anterior and posterior efferent arteries are developed 

 in the branchial arches, and in the first branchial arch the anterior 

 one of these two arteries is prolonged ventrally and forward, to form 

 the afferent mandibular artery. In an earlier work, just above referred 

 to, I concluded, from Virchow's descriptions, that it was the posterior 

 artery that was so prolonged in the adult Acipenser, but this is evi- 

 dently an error. And if it is the anterior artery that gives origin to 

 the afferent mandibular artery in Acipenser, it is evident that that same 

 artery must also give origin to the afferent mandibular in Amia and 

 teleosts, a supposition that is strengthened by the presence of what 

 are apparently remnants of the posterior arteries found in each arch 

 of Amia and described immediately below. I have accordingly, in the 

 accompanying diagrammatic representations, Figs. 2, 4 and 6, shown 

 the main efferent branchial arteries as anterior arteries in each of 

 these three fishes. Opposite or near the point where, in Amia, the 

 ventral longitudinal commissure is joined by the communicating branch 

 from the efferent artery of the second branchial arch, the commissures 

 of opposite sides are united by a cross-commissural vessel which passes 

 ventral to the truncus arteriosus, and from this cross-commissural 

 vessel a median hypobranchial artery has its origin, as described by 

 Paeker and Davis. The efferent arteries of the first and third 

 branchial arches are similarly united by cross-commissure, but each 

 of these commissures passes dorsal instead of ventral to the truncus 

 arteriosus. 



The ventral longitudinal commissure of Amia is a selachian 

 feature, and its branchial portion is wholly wanting in Esox, Salmo, 

 Ophiodon (Allen) and Lopholatilus (Silvester), while in Gadus that part 

 of the commissure that connects the first and second efferent branchial 

 arches is found. The prebranchial portion of the commissure persists 

 in all these several fishes and forms the basal portion of the afferent 

 mandibular artery. The branchial portion of the commissure of Amia 

 is evidently the homologue of the lateral hypobranchial of Parker 

 and Davis's (1899) descriptions of Carcharias and Raja. Of Amia these 

 authors say (1. c. p. 169): "In Amia calva no trace of lateral hypo- 

 branchials was discoverable"; a statement that is certainly not true 

 of the two specimens of Amia that I have had examined. 



From the second efferent branchial artery, near its dorsal end, 

 an important branch has its origin, and immediately separates into 

 anterior and posterior portions. The anterior portion, running forward, 



