122 PROCEEDINGS OF THE AMERICAN ACADEMY 



part of the chorda is turned up as in other fishes (PI. II. figs. 19, 20), 

 and the two principal cartilages of the tail-fin are placed below it, as 

 in other fishes, yet, owing to this, the separation between the embry- 

 onic and permanent caudals is never distinctly indicated (PI. II. figs. 

 18, 19, 20) — at least, not in specimens I have had the opportunity of 

 examining — by an indentation or a sharp notch, as in other species 

 figured in this paper. The tail of Phycis and of Gadus, therefore, 

 which at first glance are so beautifully homocercal, do not in reality 

 differ from the tails of other bony fishes; having like them a truly 

 heterocercal termination (PI. II. fig. 17), but completely disguised by 

 the great development of the accessory fin-rays of the dorsal and ven- 

 tral sides (PI. II. figs. 19, 20). 



In addition to the similarity of structure of the embryonic tail of 

 bony fishes with that of the Ganoids, we find another point of compari- 

 son in the fleshy, fringed pectoral fins, which recall Huxley's Crossop- 

 teiygians. This fleshy pectoral seems to be quite generally present 

 in the embryos of bony fishes. It is represented for Lophius, on PL II. 

 figs. 9, 11, 12. I have found similar fleshy, fringed pectorals in the em- 

 bryos of Cottus and of several other bony fishes. In fact, immediately 

 before the appearance of the rays in the pectorals, all bony fishes may 

 be said to have such fringed, fleshy pectorals. They are, however, not 

 sufficiently large and prominent to affect the general appearance of the 

 young fish, except in the genera Lophius and Cottus, and one or two 

 others ; but in these genera they are well developed, and are similar in 

 structure to the fleshy pectoral of young Lepidosteus. 



Carus has called attention, in the Leptocephalidte,'* to the peculiar 

 mode of termination of the chorda, — slightly bent upwards. The fila- 

 ment forming the tail-fin of Tilurus, the forked tail of some of 

 the species of Leptocephalus, and the peculiar ending of their tail, re- 

 calling heterocercal tails, are certainly embryonic characters, — judg- 

 ing, at least, from similar stages of many of our common osse6us fishes. 

 These characters, with others, — such as the unossified chorda, the trans- 

 parency of the body, the large prominent pigment spots, — all go far 

 towards confirming the view of Carus, that the Leptocephalidae are 

 only the embryos of other fishes, such as Cepola and Trichiurus. 



Both Huxley and Van Beneden f contend that the facts which they 

 bring forward completely refute the theory of the parallelism of the 



* Carus, J. v., Ueber die Leptocephaliden. Leipzig, 1861, p. 13. 

 t Van Beneden sur le developpement de la queue des Poissons Plagiostomes, 

 Bull. Acad, de Belgique, 3me serie, xl. No. 3. 



