THE DIFFERENTIAL CHAEACTEES OF CHAKAOINOII) AND 

 ERYTHRINOID FISHES. 



r>y Theodore Gill, LL. D. 



In my list of Families and Subfamilies of Fishes (131)3) I have 

 ndmitted two families of Ileteroguaths, Characida' (oi- Chaiaciiiida-) 

 and Erythrinid;e. As the limits and concepts of Miiicli they are the 

 exjiressions are quite different from those hitherto current, it is a duty 

 to no longer defer the reasons which have influenced me. 



The two families in question have been adujitted byotlier naturalists, 

 but have only been differentiated by the development of an adipose fin 

 in one (Characiuids) and the absence of it in the other (Erythrinids). 

 The mere x)resence or absence of a bag of adipose tissue is, however, of 

 too little importance to justify distinction as a family character, although 

 in most cases it happens to be coordinate with other features, and hence 

 available as a diagnostic mark.' Nevertheless, in at least the entire 

 subfamily Stevardiiuu' it fails, for the small fishes in (piestion appear 

 to be more nearly related to Tetragonopterines than to Erytiirinines. 

 A character of more importan(!e, apparently coordinate with other 

 structural modifications, and which has been the cause of my acce[»t- 

 ing the two families, is to be found in the structure of the posterior 

 part of the skull. The differences observable in due examination are 

 expressible in the following diagnoses: 



Family CHARACINID^. 



( I'rimarij Si/noiiyiny.) 



<^D€rmoptifes, Dumkril, ZooL Aualytique, p. 146, 180(i. 

 <6rtZ;«o)iidi, Rafinesque, ludice d'lttiolog. 8iciliaua, p. .32, 18U). 

 <C^Dvrmo])teria, Rafinesque, Analyse cle la Nature, p. <S7, 815. 

 <^Characini, Mulleu, Arcliiv Natnrgescb., 9. Jabrg., I, p. 323, 1843. 

 <^Characins {Characidce), Agassiz, Rep. Brit. Assn. Adv. Sci., 1844, p. 293. 



'The developipent of an adipose fin may occasionally fail as a generic and even 

 specific character, as among the Nannostouii. " Xur Lei eiuer einzigen dieser Arten 

 IVhlt.-die Fettflosse vollstiindig; bei einer zweiten Art besitzen von vier Exeni- 

 lil;>ren drei einc Fettfiosse, wiibreud sie dem vierten Individnum fehlt '" (SteiudacL- 

 uer, Icb. Beit., V, p. 74, 1876). 



Pruccedings (if flic I'liitid States X.-ilinnal Miisomn, \c.l. XVI I I. Nd. 1056. 



205 



