20 CLASSIFICATION OF FISHES. 



the rays are always soft ; but those in the spine-rayed 

 fishes are strengthened by an external spine^ which of 

 course is never branched. Linnaeus employed the situ- 

 ation of the ventral fins to construct some of his primary 

 divisions; classing together those which had the ventral 

 placed before the pectoral, those in which it was imme- 

 diately beneath^ and those where it was placed behind. 

 This plan did very well in the infancy of our science; 

 but it was soon discovered that this artificial arrange- 

 ment separated the most natural and connected genera 

 into different orders, and that even, if rigorously acted 

 upon, individual species would be similarly dissevered. 

 M. Cuvier has therefore, with much propriety, rejected 

 these divisions, and yet not so thoroughly but that some 

 of his great groups are formed nearly on the very same 

 artificial principles as those of Linnaeus.* The situation 

 of these fins, however, is by no means unimportant^ 

 when used for subordinate characters : in some, as in 

 Pteracles and Uranoscopus, they are placed immediately 

 under the throat ; in others, as the sharks, they are 

 nearly half way between the pectoral and the caudal j 

 while in that extraordinary genus, Polyptej'us, it is 

 close upon the base of the caudal fin. 



(2-i.) The shape is no less diversified : in the great 

 majority of fishes it is symmetrical with the pectoral fin; 

 both being either round, as in the LabrincE, or pointed, 

 as in the SparincE. Several instances occur, however, 

 where this uniformity is disturbed : in some of the 

 C'hc£todonid<jB, the pectorals are obsoletely rounded (as in 

 Platax teira Cm\.), but the ventrals are particularly long 

 and pointed; while in Tauricthys imrius (if the figures 

 of these two singular fishes are correct) the pectorals are 

 acutely pointed, while the ventrals are decidedly rounded. 

 'We cannot but entertain some suspicions, however, on 

 the correctness of these figures; and, indeed, the diffi- 

 culty of making accurate drawings from preserved fish, 

 whether dried or in spirits, is frequently so great, that 



* Such, for instance, as the divisions of the order Malacopteryges, and 

 the insertion of Trachiniis and Uranoscopus among the perches, because 

 they have jugular ventrals. 



