ANALOGIES OF THE SILURIDiE. 333 



to recognise in these characters the genuine indication of 

 the broad-headed or cartilaginous type of animals ; for 

 they are not only the primary external marks by which 

 that type is distinguished, but are almost peculiar to this 

 division of the Siluridcs. The same form, indeed, but 

 differently modified, is seen in the Sturisoma rostrata*; 

 so that we once more enter among the Loricarince, with 

 which we began our survey, and the circular succession 

 of all the sub-families becomes sufficiently evident. 



(277.) Let us now compare the preceding divisions 

 with the primary orders of fish, placing each in the 

 natural series of their succession. This comparison, 

 also, will elicit several other characteristics of the Si- 

 lurian groups which have not been yet touched upon. 



Sub-families of ^ , ■ „i rn.^ *^ „ Orders of 



SiLURio^. Analogical Characters. p^^^^^ 



Loricarince. ^The two most typical groups in 7 Malacopteryges. 

 Pimelodince. (_ their respective circles. j Acanthopteryges. 



SUurirup i'^^^^ excessively long ; the caudal,] . 



buurirue. ^ ^^.^^^ ^^^ ^j^j.^^! gj^^ ^^^^^ united. ] ^podes. 



Asp-edince. pyeymall, vertical ; gills immove- jp^^^^^^^^^^^^ 

 Sorum.. {''llf^.tl'JS'' P-^-'^^'^j Carx.Agx.es. 



We need not again advert to the difficulty that arises, 

 in some cases, of determining the more particular ana- 

 logies of the two typical groups of one circle, with those 

 of another circle ; which, as we have had frequent oc- 

 casion to observe, appears to be sometimes reversed. 

 In such cases we must fall back upon the strength of 

 our affinities, and leave the analogies to be explained 

 in a more advanced state of the science. No one, how- 

 ever, can deny that \he Loricarince oxidi Pimelodince follow 

 each other, and this in as unquestionable a manner as 

 the soft-finned fishes are succeeded by the spined order. 

 These points, therefore, being determined by absolute 

 affinity, — an affinity acknowledged by all writers, — it 

 follows, from the collateral evidence derived from the 

 other groups, that these four are analogous to each 



* Loricaria rostrata of Spix. 



