OP ARTS AND SCIENCES. 



121 



Fig. 2. Tail of young Sjngnatlius, f in. loi 



ence of this embryonic caudal lobe, or a trace of it, are Atherina, 

 Batrachiis, Cottus, Ctenolabrus, Lophius, Gasterosteas, Fundulus, 

 Piijcis, Gadtis, Menhaden, Temnodon, Labrax, Scomber, six species 

 of PieuronectidoB, Poronotus, Lumpus, several of the genera of Vivi- 

 parous fishes (Embiotocoidte) from San Francisco, and a few oilier 

 species as yet undetermined. la the very youngest specimens of 

 Syngnathus I have been 

 able to examine (fig. 2), 

 the position of the two sup- 

 ports of the caudal rays, 

 below the upturned termi- 

 nation of the notochord 

 differed in no wise from 

 that of the other bony fishes 

 here mentioned. 



In Atherina, PL II. figs. 

 1-4, the embryonic caudal 

 does not form quite so marked a lobe as in the species figured in PL I. 

 Still, the separation between the permanent and the embryonic caudals 

 is sufficiently well-marked to leave no doubt of the existence of the 

 caudal lobe. The same is the case with Batrachus (PL II. fig. 5), 

 with Lumpus (PL II. fig. 6), and Ctenolabrus (PL II. fig. 7). In the 

 young Poronotus (PL II. fig. 8), the embryonic caudal lobe is more 

 prominent. 



In Lophius, the termination of the notochord remains unchanged 

 quite late in life; the tail of the young Lophius (PL IL fig. 10) show- 

 ing no trace of any ossification of the vertebral column, or degenera- 

 tion of the extremity of the notochord, at a time when the young fish 

 can readily be recognized as a young Lophius, from the presence of 

 the peculiar appendages of the pectorals and of the anterior dorsal. 

 In Gasterosteus (PL II. fig. 13), the embryonic caudal is again very 

 prominent: it can readily be traced in PL II. figs. 14, 15, until the 

 tail has assumed the shape it finally takes in the adult. In all the 

 genera thus far described, the tail gradually jjasses from a strictly ven- 

 tral appendage, placed below the dorsal column, to that of a terminal 

 tail placed in the conthiuation of the vertebral column. 



In Phycis and in the Cod, the structure of the tail is somewhat dif- 

 ferent ; the accessory fin-rays, both of the dorsal and ventral side, are 

 very numerous (PL II. figs. 19, 20), far outnumbering what are usu- 

 ally called the principal rays of the tail. These accessory rays early 

 make their appearance (PL II. fig. 18) ; so that, although the terminal 



