26 CLASSIFICATION' OF FISHES. 



is attached to the base of the anterior ray of the second 

 dorsal ; so that^ in effect^ although there are two divi- 

 sionsj or fins, yet they are^ in fact^ connected, and, con- 

 sequently, become one. The genus Gadus, as now 

 restricted, has obviously three dorsal fins, each separated 

 by an interval from each other {fig. 2.a) ; while in the 

 genus Blepsias there are also three (6), nearly of the same 

 form ; and yet, because they are very sHghtly united by 

 a membrane, in the manner above described, this genus 

 is said to have but one dorsal. Now the transition 

 from two approximating dorsal fins to one, cleft be- 

 tween the spiny and branched rays, is so gradual, that 

 it is impossible to define every stage of tbe progression ; 

 and much ambiguity will always attend this part of the 

 definition of the subordinate groups. Nevertheless, as 

 characters taken from the fins will be eventually found 

 to be of much more real and practical value than has 

 hitherto been supposed, we should propose the adoption 

 of the following terms, as calculated, in some degree, 

 to express the modifications just mentioned, "Wliere, 

 for instance, a portion, however small, of the naked 

 back intervenes between one or more of these divisions, 

 as in the common cod (a), they may be considered, as at 

 present, three distinct fins. When the last membrane 

 of the first dorsal is in any way united to the anterior 

 ray of the second dorsal^ we might consider them as 

 "two dorsal fins united;" and when this union is so 

 close, as that the membrane in question ascends up 

 the side of the next ray, instead of descending in a 

 direction to the base, we would then t^rm the dorsal fin 

 "'single, but emarginate,'' deeply, or slightly, as the 

 case may be. The annexed cuts will more effectually 

 explain our meaning ; and we may now consider their 

 other pecuHarities. 



(SO.) The rays of the dorsal are either simple or 

 branched. Simple rays, again, are of two sorts : 

 sometimes they are slender and flexible, although with- 

 out any joints ; in which case they are generally ter- 

 minated by a fleshy or membranaceous filament; and 



