480 



commissure does not persist, that portion of the commissure thai 

 primarily connected the posterior artery in each branchial arch with 

 the main, epibranchial portion of its related artery aborting, and 

 giving rise to the condition found in the adult Mustelus (Parker, 1886) 

 and in Heptanchus. In the hyoidean arch, in all selachians so far 

 described, the posterior efferent artery retains permanently its pri- 

 mary connection with the main epibranchial portion of its artery. 

 Ventrally, in Heptanchus, the posterior efferent artery of each 

 arch is connected with the anterior efferent artery in the next posterior 

 arch in exactly the same way that the dorsal ends of the same arteries 

 are connected with each other, a series of loops thus being formed 

 around the ventral and of each gill cleft, excepting only the seventh. 

 These loops correspond to the short longitudinal commissures shown 

 in Parker's (1886) figure of Mustelus antarcticus and in my figure of 

 Chlamydoselachus (Allis, 1911), but in Heptanchus the several loops 

 are not directly connected with each other by commissure, as they 

 all are in Mustelus and as certain of them are in Chlamydoselachus. 

 From each loop in Heptanchus, excepting only the most anterior one, 

 a short branch is sent ventro-postero-mesially and falls into a longi- 

 tudional vessel that lies internal (dorso-mesial) to the afferent arteries 

 and parallel and slightly dorso-lateral to the truncus arteriosus. 

 This longitudinal vessel may be considered to begin, in Heptanchus, 

 where it receives the communicating branch from the second gill 

 loop and to end where it receives the branch from the sixth gill loop, 

 but the vessel extends both anteriorly and posteriorly beyond these 

 points. The posterior continuation goes to muscles and tissues of 

 the region and, in my one specimen, is larger on one side of the head 

 than on the other. The anterior continuation is a so-called nutrient 

 vessel and, on one side of the head of my specimen, goes to the tissues 

 of the region, but mainly to the coraco-branchiahs muscle of the second 

 branchial arch. On the other side of my specimen, this nutrient 

 branch, after supplying the coraco-branchiahs muscle of the second 

 branchial arch, is continued forward beyond that muscle and supplies 

 the two corresponding muscles, one on either side of the head, of the 

 first branchial arch. Other nutrient branches are sent from the 

 longitudinal vessel, opposite each gill loop, to the corresponding 

 muscles and general tissues of the other arches. 



Opposite or not far from the points where the longitudinal vessel, 

 above described, receives the communicating branches from the second, 



