S32 CLASSIFICATION OF FISHES. 



organisation is essentially the same. The great deve- 

 lopement of the tail, however, from being the primary 

 characteristic of the whole group, may be traced in all 

 the aberrant types, notwithstanding the numerous vari- 

 ations they present in other parts of their structure. 

 Quitting these, however, by means of the eel-like genus 

 Plotosus, we enter at once into the fourth sub-family, re- 

 presented by the genera J.?/)rec?o of Artedi (^Asp, Gronovii 

 Sw., fig. 80.) Platystacus (in part) of Bloch, and two 



80 



/ 



/ 



or three others. The first of these is probably the type 

 of the whole : they are distinguished by the excessive 

 smallness of their eyes, which are vertical, or placed 

 close together on the top of their head ; and the aper- 

 ture of the gills is merely a simple cleft of the skin, as 

 in the Plectognathes, or cheloniform genera. Here, also, 

 we arrange the remarkable genus Astrohlepas of Hum- 

 boldt*, one of the most extraordinary fish in the whole 

 family, since its connection to certain Mysti of Grono- 

 vius is quite evident. In the Heterohranclius 5-tenta- 

 culatus of Spix, the head is protected or mailed by bony 

 plates, as in the Heterobmnchi of Egypt; while its length- 

 ened muzzle shows such an affinity to the Sorubinie, that 

 we consider its real affinities to be with these latter fishes. 

 The genus Sorubium of Spix we place after the Aspre- 

 dincp: in these the dorsal fins are two — the hinder one 

 being either adipose, or with imperfectly developed rays: 

 the head, however, has now become uncommonly long 

 and large, the muzzle obtuse and dilated, and the upper 

 jaw considerably longer than the under. We cannot fail 



* This is not alluded to by JL Cuvier, even in his notes. 



