ACANTHOPTERYGII. 97 



there are two ; sometimes instead of a first dorsal, tliere are 

 only a few free spines. The first rays of their anal are also 

 spines, and there is generally one to each ventral. 



The relations between the Acanthopterygii are so multi- 

 plied, and their different natural families present so much va- 

 riety in the apparent characters which we might suppose 

 would indicate orders or other subdivisions, that it has been 

 found impossible to divide them otherwise than by these same 

 natural families, which we are compelled to leave together. 



FAMILY I. 



PERCOIDES.(l) 



This family is so called because its type is the Common 

 l^erch. It comprehends fishes with oblong bodies, covered 

 with scales that are generally hard or rough, and whose oper- 

 culum or preoperculum, and frequently both, have dentated 

 or spinous edges, and whose jaws, the fore-part of the vomer, 

 and generally the palatine bones, are furnished with teeth. 



The species are extremely numerous, particularly in the 

 seas of hot climates ; their flesh is generally wholesome and 

 \agreeable. 



In a vast proportion of these Perches, the ventral fins are 

 inserted under the pectorals : they form a first division which 

 may be called Percoides Thoracici. 



They were nearly all comprised by Linnteus in his genus 

 Perca, but we have been compelled to divide them as fol- 

 lows, from the number of the branchial rays, that of tlie dorsal 

 fin and the nature of the teeth. 



In the first subdivision we find seven rays in the branchiae, 

 two fins on the back, and all the teeth small and crovvded.(2) 



(1) In my first edition this family also comprehended the Buccse Loricatae, the 

 Scieno'ides and the Sparo'ides. It was necessary to detach these three new families 

 from it, and I think I have been fortunate enough to discover sufficient charactei's 

 for that purpose. 



(2) The original expression en velours is one of the many instances in which the 



Vol. II. N 



