1 72 ARirs. 



tions which have been laid down by Valenciennes 

 between Bagrus^ Arius, and Plmelodus, among 

 which they ought most properly to range. These 

 have been taken from the teeth, their position and 

 form, which it is difficult to ascertain from notes 

 and drawing only. In all, the form is remarkably 

 similar; so much so, that many of the Siluri of 

 Hamilton Buchanan run into the others, and have 

 been separated. This is, indeed, so much the case, 

 that we are not sure if the dental classification be 

 not too minute and artificial. In Bapms, the pala- 

 tine teeth are disposed in single transverse bands ; 

 in Arius, as characterised by Valenciennes, and 

 which is made to contain species from Africa, Asia, 

 and South America, the palatine teeth form two 

 distinct and separated patches, and are generally 

 inserted in the palatine bones, although they some- 

 times advance upon the angles of the vomer. In 

 Plmelodus, again, the palate is smooth, and entirely 

 without teeth. + Unfortunately, Valenciennes does 

 not give his characters so that they can be tran- 

 scribed. Pimelodus, as characterised by Spix and 

 Agassiz, would contain all those of which we have 

 drawings or any notes, but we have chosen to place 

 two of them, provisionally, along with Arius^ on 

 account of their approaching nearer in form to some 

 of those figured in the various works to which the 

 French ichthyologist is at present contributing, par- 

 ticularly in the strong armature of their fins, &c. 

 The first we shall for the present entitle 



f Valenciennes, Hist. Nat, des Poissons, xv. p. 21. 



