580 



efferent artery in each branchial arch has lost its dorsal con- 

 nection with the anterior artery of its own arch, as in most 

 selachians, and is connected by dorsal commissure with the anterior 

 artery in the next posterior arch. The posterior efferent hyoidean 

 artery is similarly connected by dorsal commissure with the anterior 

 efferent glossopharyngeus artery, but, as also in selachians, it retains 

 its dorsal connection with the epibranchial portion of the primary 

 artery of its own arch. 



The posterior efferent artery in each arch fuses ventrally with 

 the anterior artery in the next following arch, a loop thus being 

 formed around the ventral edge of the intervening gill cleft. From 

 each one of the first three of these loops a single artery continues 

 ventro-mesially and falls into an external lateral hypobranchial 

 artery, the artery from the most anterior loop apparently being in 

 process of abortion and possibly being in part reduced to a simple 

 hne of tissue. In Eaja clavata, Hyrtl shows but a single one of 

 these ventral prolongations of the efferent arteries, and it is said 

 to be a prolongation of the efferent artery of the second gill sac. 

 In Eaja nasuta, Parker shows two ventral prolongations, one coming 

 from the loop around the ventral edge of the second gill cleft and 

 the other from the loop around the ventral edge of the third cleft. 

 In Eaja erinacea, Parker and Davis show two of these pro- 

 longations on one side of the head, one coming from the loop around 

 the third gill cleft and the other from the loop around the fourth 

 cleft; while on the other side of the head there are four prolongations, 

 one each from the loops around the second, third and fourth clefts 

 and a fourth from the posterior artery in the fourth branchial arch. 

 There is accordingly considerable variation in the manner in which 

 the external lateral hypobranchial artery is supplied in the adult 

 Eaja, and it would seem as if it must have been primarily supplied 

 by branches from all of the gill loops, if not by ventral prolongations 

 of all the efferent arteries. 



In the one of my two specimens that was particularly examined 

 in this connection the communicating branches from the loops 

 around the first and second gill clefts are connected by a second 

 longitudinal commissure, and this commissure is continued anteri- 

 orly, internal to the common trunk of the first and second afferent 

 arteries, and turns upward in the ventral end of the hyoidean arch ; 

 this artery thus having the position of an internal lateral hypo- 



