The pseudobranchial circulation in Amia calva. J27 



median ventral arterial vessel, this vessel being, naturally, the truncus 

 arteriosus of later stages. A further, and most important difference is 

 found in the development of a large venous vessel practically parallel 

 to, and external and anterior to, the middle portion of the prespiracular 

 artery. The sections of this venous vessel closely resemble, in appearance, 

 those of the artery itself, and as it is connected with the artery at 

 certain places it is, if one is not familiar with the sections of older and 

 younger stages, most confusing. It has neither dorsal nor ventral 

 connection with the venous system, and I was at first inclined to 

 look upon the two parallel and adjoining vessels as the middle portions 

 of the pre- and post-spiracular arches, crowded together, and the pre- 

 spiracular arch in process of reduction and subsequent abortion. This, 

 however, seems a most iniprol)able assumption, and it entails the 

 further assumption that the vessel I have above described, in 9 mm 

 embryos, as the postspiracular artery is not that artery but is a 

 wholly new formation. It seems much more probable that the vessel 

 in question is venous, and that it is developed in relation to the pre- 

 spiracular artery in the same way and manner that the efferent and 

 afferent arteries develop from the originally single aortic vessels of 

 the branchial arches. In 10 mm embryos the venous vessel has lost 

 all apparent connection with the prespiracular artery, and in 12 mm 

 and older specimens it connects dorsally with the large venous vessel 

 that lies in the angle between the hyomandibular and the lateral 

 wall of the skull, as already described. It is perhaps not un- 

 important, in this connection, to note that this doubling of the pre- 

 spiracular aortic arch is repeated, but in a different way, in the 

 postspiracular. arch, by the development of the dorsal and ventral 

 opercular arteries. 



In 10 mm embryos (Fig. 4) the arterial connections are practi- 

 cally the same as in 9 mm ones. Th.e ventral and dorsal bends in 

 the prespiracular artery are more pronounced at this age, and the 

 hyo-opercularis artery arises from the carotid at a point much nearer, 

 relatively, to the point where the carotid has its origin from the af- 

 ferent artery of the first branchial arch. Whether branches of the 

 hyo-opercularis artery followed the branches of the truncus hyoideo- 

 mandibularis facialis, or not, could not be definitely determined. The 

 direct connection of this artery with the dorsal opercular artery could 

 also not be definitely traced. Ventrally the pre- and post-spiracular 

 arteries still arise together from a lateral paired vessel, but this 



