482 



an error, for the corresponding vessel in both Carcharias httoriaHs 

 (Parker and Davis, 1899) and in Musteliis canis (Ferguson, 1911) 

 lies internal to the afferent arteries, and I so find it not only in Hept- 

 anchus but also in a specimen of Galeus that I have had examined 

 for this especial purpose. In Carcharias, the vessel is called by Parker 

 and Davis the lateral hypobranchial, and in a recent work (Allis, 1912) 

 I adopted this name for the corresponding vessel in Amia and teleosts. 

 But this name is not sufficiently distinctive, for both Parker (1884) 

 in Eaja nasuta, and Parker and Davis (1899) inEaja erinacea, show a 

 so-called lateral hypobranchial that lies external to all the afferent 

 arteries; and inChlamydoselachus(x\LLis, 1911) there are both internal 

 and external longitudinal vessels. It accordingly seems best to call 

 these two vessels, respectively, the internal and external lateral 

 hypobranchial. 



The external lateral hypobranchial is represented, in Heptanchus, 

 by some basal portion of the artery that I have described as the coro- 

 nar}' artery, and there is no median hypobranchial in this fish. In 

 Chlamydoselachus, the external lateral hypobranchials of opposite 

 sides have met and fused in the mid-ventral line, throughout a part 

 of their length, and so given rise to a short median vessel, as they 

 have also in Mustelus canis (Ferguson, 1911); and this fusion of two 

 external lateral vessels must have been the method of development of 

 the median hypobranchial in both Mustelus antarcticus and Carcharias 

 littoralis, in both of which fishes there are internal lateral and median 

 hypobranchials but no external lateral vessels. In Eaja erinacea 

 (PARKERand Davis, 1899), the external lateral hypobranchials extend 

 the full length of the branchial arches and there is no median vessel, 

 and also no internal lateral ones. 



The internal lateral hypobranchial, in Heptanchus, extends 

 only from the sixth to the second efferent gill loop, but the nutrient 

 branch, on one side of the head of my specimen, that extends forward 

 to the coraco-branchiales of the first branchial arches may be a remnant 

 of the hypobranchial that formerly extended to the first gill loop; 

 for in Mustelus, Carcharias and Chlamydoselachus this hyoideo- 

 glossopharyngeus section of the vessel is found. 



From the efferent loop around the ventral end of the first gill 

 cleft the afferent mandibular artery has its origin. This artery runs 

 ventrally and forward, internal to the afferent hyoidean artery, and 

 after sending three or four branches to the tissues and muscles of the 



