352 CLASSIFICATION OF FISHES. 



tail, existing between the Pimelodince and all the true 

 Silurince. In the typical examples of the two next 

 groups which stand opposite each other, namely, the 

 genus Silurus and the sub-family Loricarince, the 

 adipose dorsal does not exist ; and it is further to be 

 noted, that these two genera — Silurus and Loricaria 

 — are the pre-eminent types of their respective circles. 

 Plotosus and Sorubium represent each other most com- 

 pletely by their " very large, long, and depressed 

 head * ;" and by the snout projecting over ^the lower 

 jaw, so as to render the mouth inferior, as in the Squa- 

 lidce and the other typical cartilaginous fishes. Equally 

 beautiful is the analogy between Cetopsis and the minute- 

 eyed genera Astroblepus and Aspredo ; and we find a 

 perfect representation of all three in the genera Cyclo- 

 pium, as well as in the types of the cheloniform fishes. 

 We next find Ageniosus standing, in our table, opposite 

 to the Siluridce ; and this analogy at first sight appears 

 forced, because, as the silures represent the anguilliform 

 fishes, it would seem to follow that Plotosus, from 

 having all the hinder fins united, should be compared to 

 them, rather than to the Sorubince : but further con- 

 sideration on the nature of the anguilliform type will 

 show this reasoning to be fallacious. Great length of 

 tail is, indeed, one of the chief indications of the anguilli- 

 form structure ; but it is not the only one, for, through- 

 out the greater part of the representations of this form 

 among the acanthopterygious order, the body is exces- 

 sively thin ; and analysis proves that the highest de- 

 velopement of this particular character is, that the body 

 becomes semi-transparent or diaphanous. Now, by far 

 the greater part of the Silurince are remarkably thin 

 fishes, — that is, the body is very much compressed, 

 although the head is remarkably depressed. But it is 

 only in our genus Silojiia (one of the types of Agenio- 

 sus), that we find both head and body compressed, and 

 this to a much greater degree than in any other forms 

 of the sub-family we are now investigating. Hence it 



.* Russell thus describes the head of Plotosus. 



