Feb. 7, 1878] 



NA TURE 



287 



homocercal fishes first appear above that formation and 

 gradually predominate until, as in the present period, the 

 heterocercal bony fishes are almost limited to a single 

 ganoid genus (Lepidosteus)." " Indeed," writes Prof. 

 Owen in another place, "it [the heterocercal] was the 



Fig. I. 



fashion of tail which prevailed in fishes throughout the 

 palaeozoic and triassic periods." It never seems to have 

 been settled whether the fish with the homocercal tail was 

 or was not better off than the fish with the heterocercal 

 tail. If the more recent fishes have improved in this matter 

 of tails upon the more ancient fishes, as was to have been 



FiG. 2. 



expected, certain it is that the shark of to-day can wheel 

 quickly enough about in pursuit of his prey, and that the 

 sword-fish can come thundering against a ship's timber 

 with a vigour not easily matched by any fish with a sym- 

 metrical tail. Be this, however, as it may, the structure 



of fishes' tails has engaged the attention of most of 

 our comparative anatomists, and the student will find 

 large stores of facts collected and arranged for him 

 by Agassiz, Vogt, Owen, KoUiker, Hteckel, Huxley, 

 and Lotz. The latter four anatomists have plainly 

 shown that \yhi'e the external appearance of the 

 tail of modern bony fishes is, as we have seen, homo- 

 cercal, their real structure is only a modified heterocercal- 

 one, so that, as far as we now know, the tail of all fishes is 

 built upon modifications of the same type, and in a paper 

 just published by Alexander Agassiz, " On the Young 

 Stages of some Osseous Fishes," he proves still further that 

 this tail fin does no': differ in its mode of development from 

 the primitive embryonic fin, or from that of the back 

 (dorsal) fin. He describes the gradual change of the 



Fig. 3. 



embryonic tail in several species of bony fishes, and he 

 calls attention to the remarkable presence of an embryonic 

 caudal lobe, which has, to this, apparently escaped the 

 attention of naturalists, and which shows remarkably well 

 the identity of growth between the tails of ganoid and of 

 bony fish. 



Alexander Agassiz traces the changes gradually takiag 

 place in the tail of the common flounder, from the time 

 the little fish leaves the egg until it has nearly assumed 

 the final shape of the adult. At first (Fig. i) the caudal 

 end of the chorda is straight. The caudal fin is rounded. 

 In the next the caudal extremity of the chorda has become 

 slightly bent up vards, and there will be found the first 



Fig 4. 



trace of the division line between the embryonic and the 

 permanent caudal fins. In further stages this indentation 

 between these two becomes more marked the chord 

 becomes more arched, and the permanent caudal at 

 length projects well beyond the outline of the embryonic 



fin fold, 60 that antecedent to the ossification of any of 

 the vertebral column, the tail has assumed a hetero- 

 cercal form. 



In the stage (Fig, 2) in which the embryonic caudal 

 assumes the shape of a large independeat lobe, while the 



