OF ARTS. AND SCIENCES. 125 



ment of the permanent caudal from the lobes of the heterocercal tail 

 of the young fish embryo. 



In addition to the parallelism of the embryonic and paleontological 

 development of the tail, we find other embryonic characters in the Old 

 Red fishes. I have already alluded to the old-fashioned structure of 

 the pectorals in the embryo of Lophius and other bony fishes ; and 

 would call attention to the innumerable embryonic fin-rays (PI. I. figs. 

 5—9) of the embryonic dorsals and ventrals, which recall strikingly the 

 similar numerous rays so characteristic of the fins of the older Ganoids. 



So that, while Agassiz and Vogt were undoubtedly mistaken in the 

 details of their explanation and comparison of the homocercal and 

 heterocercal tails, yet the parallelism they attempted to prove, not only 

 exists, but can even be carried out far beyond any thing they conjec- 

 tured. Their very mistakes regarding the heterocercal structure of 

 what they called a homocercal tail in the bony fishes of the present 

 day being (as I have attempted to show) the best proof of the 

 existence of such a parallelism, and the clearest indication possible of 

 the uniformity of structure of the tails of the fishes of the present day 

 with those of the fishes of the most ancient geological period in which 

 fishes have as yet been found. This parallelism could, however, not 

 be conclusively made out, until it was proved that the extension of the 

 chorda dorsalis into the upper lobe of the heterocercal tail gave us the 

 explanation of the peculiar heterocercal structure still to be traced in 

 the so-called homocercal tails of the bony fishes of the present day, 

 long after the disappearance of the upper caudal lobe, which (as I have 

 shown) exists in "bony fishes only during a short embryonic period. 



