THE TRIBE OF BLENNIDES. 71 



one half the typical length. In Chiroleptus they are 

 long^ but with the rays all united. Trachipterus seems 

 a prototype of Trichosoma by its low and narrow dor- 

 sal fiUj while in Brachirus the pectorals are again short- 

 ened^ and the rays connected and branched. We may 

 perhaps be thought to have carried our analysis^ or at 

 least our patronymic divisions^ rather too far in the two 

 lasr groups ; but it is clear that if we had not done this 

 in the case of Apistes, the analogies on which we have 

 so lately expatiated would not have come to light. 

 There is a wide difference between such sub-genera as 

 merely state minute differences, and those which indi- 

 cate the stages of circular groups^ like Aputes and Ma- 

 crochirus. We have nothing particular to observe on 

 the remaining genera of the Scorpcence, since their sci- 

 entific characters will be subsequently detailed. Whether 

 Trichodon or Sehastes is the intervening form between 

 Tcenianotus and Scorpana is a secondary question : both 

 are perfect chironectiform types j and we believe the 

 firsi is the true one of this group. 



CHAP. VI. 



ON THE TaiBE OF BLENNIDES, OR GOBIES AND BLENNIES. 



(74.) This tribe, which we have named from the typi- 

 cal family, is composed of those acanthopterygious 

 fishes which have the ventral fins differently construc- 

 ted, as before observed, from all others of the order. 

 They are either of two or three small rays, or enveloped 

 in a thick skin, or they are united together into a funnel, 

 or finally they are totally wanting. They are besides 

 very remarkable from the rays of their fins being more 

 slender than any of the spiniferous tribes ; so that in 

 fact there are few, if any, which possess the strong 

 robust rigid spines, so common in the group we have just 

 left ; the ventral fins, when they exist, are also generally 

 placed considerably before the pectorals. All these 



F 4 



