32 



CLASSIFICATION OF FISHES. 



being rounded. The Balistidce, the most "perfect of 

 the cheloniform fishes, and which represent the spine- 

 rayed, order (Acanthopteryges), consequently present 

 us with nearly all the modifications of fin observable 

 among their prototypes ; but in the whole of the remain- 

 ing families., and the entire order Apodes, where we have 

 the most sluggish of all fishes, as the Chironectidce , the 

 Lophidcs, CyclopteridcE, Murcenida, &c, not a solitary 

 example occurs of a forked tail, much less of those ad- 

 ditional helps for speed just noticed, which have been 

 given to the Scomberidce. If we carry our inquiry 

 into the minor groups or families, we shall find the 

 same determinate prevalence of one set of characters in 

 the fins, running through each particular group. Nu- 

 merous instances of this will be brought before the 

 reader, for the first time, in the progress of our work. 

 Having now adduced sufficient reasons, as we imagine, 

 for the opinions above expressed, we may at once pro- 

 ceed to notice the different forms observable in this fin. 

 (34.) The caudal fin presents every modifica.tion 

 between a perfectly lanceolate shape, where the largest 

 rays are in the centre (fig. 3. a), to that of a deeply 



forked one, wh ere 



the central rays 



: are so short as 



almost 



come 



giving 



to be- 

 obsolete, 

 the tail 

 an appearance of 

 being divided 

 into two parts. 

 The first of these 

 forms is shown in the genus Cepola, and its repre- 

 sentatives the Indian gobies ; the second runs through 

 the whole of the mackerel, tunny, sword-fish, and a large 

 proportion of the Zeidce, or dories. Besides these, there 

 is also a third, peculiar only to two or three genera, 

 where the tail may be said to be doubly forked ; a few 

 of the central rays being lengthened nearly as much as 



