PECTORAL FINS. 17 



Cirrhites, again, the last five rays are not only spinous, 

 but much thickened and prolonged beyond the mem- 

 brane ; a structure which excites a strong suspicion of 

 this genus being analagous, in its own circle, to Cheilo- 

 dactylus ; and this seems the intermediate state of deve- 

 lopment leading to the next. 2. Digitated pectorals 

 are exclusively confined to the typical Canthileptes, or 

 gernards, and spine cheeks (Scorpcenidce). Among the 

 first, and particularly in the genus Trigla, there are 

 three detached, finger-like processes, unconnected by 

 any external membrane, which are situated just before 

 the lower base of the pectorals, and which almost appear 

 to be detached rays of this fin, much thickened, and 

 somewhat removed from the connected rays. In the 

 Mediterranean and Atlantic Dactylopteri, these ex- 

 traneous rays are united by a membrane, so as to form 

 a spurious or supplementary pectoral ; while in those 

 from India, of which there now appears to be several 

 species*, this supplementary fin is united to the true 

 pectoral so as to form but one. The pectorals of nearly 

 all the remaining families of this tribe have the lower 

 rays, or those nearest the throat, thickened ; much in 

 the same way as in Cirrhites, but with this difference, 

 that the fins are so broad at their base, on account of 

 the number of rays, that they are often carried half-way 

 under the throat ; a character so very peculiar, that we 

 look upon it as a distinct modification. It seems pro- 

 bable that this unusual strength is given to such fish as 

 have very large heads, for the purpose of additional 

 support ; for it will be observed that the heads of all 

 these genera are not only large, but particularly heavy, 

 on account of the bony armature with which this part 

 is covered ; and it may be further remarked, that it is 

 among small headed fishes we find the most delicate 

 pectorals. In truth, however, nothing can be affirmed 

 with any degree of confidence on the reasons of these 

 variations from the ordinary structure. We cannot, 

 as in land animals, watch the habits and explore the 



* See the Appendix, wherein these are described. 

 VOL. I. C 



