OF ARTS AND SCIENCES. 123 



embryonic tails of bony fishes with those of fishes preceding the Ju- 

 rassic period : Huxley, because the bony fishes of the present day are 

 not provided with a structurally liomocercal tail (as was supposed by 

 Agassiz and Vogt), but have — as he showed from Gasterosteus, and 

 as I have shown here from many other genera of bony fishes — a truly 

 heterocercal structure ; Van Beneden, because in the Plagiostomes the 

 tail of the young fish is at first truly homocercal, this condition preced- 

 ing the heterocercal one, while, according to Agassiz and Vogt's theory, 

 the young Plagiostomes should possess pre-eminently heterocercal tails; 

 and, taking it also for granted that the oldest fossil fish known pos- 

 sessed truly homocercal tails, the whole theory, according to him, falls 

 to the ground. 



Now, while fully admitting, with Huxley, that what Agassiz and 

 Vogt called homocercal, in the modern bony fishes, is only an external 

 delusion, due to a structure of the bones of the tail, which (as I have 

 shown here) is found in a large number of bony fishes of the present 

 day ; while also admitting, with Van Beneden, that the young Plagi- 

 ostomes, which in the adult have a truly heterocercal tail, yet have, 

 in the early stages, a strictly homocercal (structurally also) tail, — yet 

 I think that neither Huxley nor Van Beneden has upset the theory 

 of Agassiz and Vogt ; and that, mistaken as they were in the details, 

 the great generalization remains, of the complete accordance between 

 the embryonic growth and the paleontological development: only it 

 must be carried one step farther ; and we must, at the same time, give 

 a somewhat different interpretation of the meaning of the heterocer- 

 cality of the tail, so prevalent among the bony fishes of the present day, 

 from that given to it by Agassiz and Vogt. Let us preface by stating 

 that the heterocercal tail is not the earliest stage ; and that neither Von 

 Baer, nor Agassiz and Vogt, stated this, but merely noticed it as 

 one of the early stages in the fish embryo. In fact, as is well known, 

 the earliest stage of the tail in the egg, and immediately after hatching, 

 is nearly symmetrical ; the notochord extending in a straight line to- 

 wards the tail, with the dorsal and ventral embryonic fins forming 

 a rounded tail, the dorsal fin slightly narrower than the ventral. This 

 stage we might call the Leptocardial (PI. I. fig. 1), — the earliest form 

 of tail assumed by bony as well as other fishes, which jjrecedes that of 

 the heterocercal tail proper (PI. I. figs. 3, 4). 



So that, as far as embryology is concerned, the tail of the Sela- 

 chians is formed strictly in accordance with the law of development of 

 other bony fishes ; and it only remains to be seen how this accords with 

 the paleontological record. 



