SYSTEM OF INSECTS. 427 



fers a degree of pre-eminence upon the latter, symboli- 

 zing the feline race, which seems to throw no small weight 

 into their scale. 



There are two Classes of Vertebrate animals with 

 which insects may appear to claim kindred. Thejishes, 

 and the reptiles. Fishes in their fins exhibit no small 

 resemblance to insects ; the pectoral and ventral ones 

 representing their arms and legs, and the dorsal ones 

 their wings: Pegasus Draco in this last respect is not 

 unlike a butterfly a . In some genera (Ostracion, Pega- 

 sus, &c.), like insects the animal is covered with a hard 

 shell or crust, formed by the union of its scales. The 

 oral cirrlii of many fishes seem analogous to the palpi 

 of insects ; and in some a pair longer than the rest re- 

 present their antennae^ '. Another circumstance in which 

 insects and fishes correspond, is the wonderful variety of 

 forms, often in the greatest degree eccentric, that occurs 

 in both Classes. Some of the cyclostomous fishes, as 

 Ammoccetus, Gastrobranchus, are supposed to connect the 

 fishes with the Annulosa, by means of the Annelida as 

 an osculant Class c , which Mr. MacLeay regards as the 

 passage to the Chilopoda d : his Mandibulata he con- 

 siders as passing into the Anoplura by means of some 

 osculant Order as yet unknown e . But 1 must confess 

 I can see no good ground for this last transition : the 

 Anoplura appear much more nearly related to thePsoctdte, 

 especially by the apterous Atropos pulsatoria f than to any 



a N. Diet. d'Hist. Nat. xxv. 115. xxvii. t. M. 8./. 1. 



b Piso Hist. Nat. 63. Curui 1. Jundia v. 



c N. Diet. d'Hist. Nat. xxvii. 235. Hor. Entomolog. 203. 



Ibid. 281. <> Ibid. 354, 390, 397. 



1 This insect, except in its antennae, so nearly resembles a Nir- 

 mus, that it might be mistaken for one. See Coquebert lUusir. 

 Icon. i. t. ii./. 14. 



