126 EDWARD PHELPS ALLIS jr., 



hind end of the communicating branch from the prespiracular artery. 

 At the point where it turns upward in the base of the gill cover it 

 sends a branch backward and then upward in the ventral and ventro- 

 posterior edge of the gill cover, this branch being large, and being a 

 direct continuation of the artery itself anterior to the point where 

 the branch is given off. This branch is thus the ''ventrales Rand- 

 gefäss des Kiemendeckels" of Müller's description of Lepidosteus. I 

 have called it the ventral allèrent opercular artery. Dorsal to the 

 point where the postspiracular artery joins the communicating branch 

 from the prespiracular one there is no direct dorsal continuation of 

 the postspiracular artery. There is, however, here, a large branch, 

 directed backward, which quite unquestionably represents the ventral 

 portion of the dorsal section of the continuous vessel of younger 

 stages. This branch is undoubtedly the homologue of the "dorsales 

 Randgefäss des Kiemendeckels" of Friedrich Müller's descriptions. 

 It soon breaks up into several branches the more ventral ones of 

 which undoubtedly have capillary connection with the ventral afferent 

 opercular artery. The more dorsal branch turns upward and probably 

 connects directly, though it may be only by capillary vessels, with the 

 hind end of a branch that arises from the carotid in the transverse 

 plane of the truncus hyoideo-mandibularis facialis. This latter branch 

 of the carotid is, as its position sufficiently shows, not only the hyo- 

 opercularis artery of my descriptions of older stages, but also the 

 dorsal portion of the postspiracular aortic arch of younger ones. That 

 section of this latter arch that lies between the carotid and the com- 

 municating branch from the prespiracular artery, and which I have 

 referred to as the dorsal section of the artery, has thus been pulled 

 backward, at this age, considerably out of its original position, and 

 has become, perhaps, capillary in part of its length. This entire 

 section of canal is said by Friedrich Müller to wholly disappear in 

 older embryos of Lepidosteus. 



The conditions found in 9 mm embryos of Amia thus differ con- 

 siderably from those found in 8 mm ones. There seems, however, 

 no reason to doubt that the main vessels are the same at the two 

 ages. The only important difference is in the origin, ventrally, of 

 the pre- and post-spiracular arteries from a lateral paired trunk at 

 9 mm, instead of independently from a median unpaired trunk, as at 

 8 mm. This difference is doubtless connected with the shifting, from 

 before backward, of the point where the heart connects with the 



