18 FRESH-WATER FISHES OF SOUTH AMERICA— EIGENMANN. 



XVIII. 



Genera peculiar to Central America and Mexico. Those marked with 

 an asterisk (*) are immigrants from North America, where they are 

 still abundant: 



1. Lepidosteus* 1 sp. 



2. Aminrus* 1 sp. 



3. Ictalurus* 1 sp. 



4. Dorosoma' 1 sp. 



T). Characodon 2 sp. 



6. Girardiiiichtliys 1 sp. 



7. Pseudoxipliophorus 2 sp. 



8. Belonesos 1 sp 



9. Mollienesia 



10. Xiphoi)liorus 1 sp. 



11. Platyptecilns 1 sp. 



12. Agonostomus 3 sp, 



13. Chirostoma 2 sp. 



14. Neotioplus 1 sp 



The foregoing lists explain themselves in part, but a few remarks will 

 not be altogether out of place. It will be seen that genera of many 

 species usually have a wide distribution, and, conversely, genera of 

 wide distribution usually have many species. A comparison of Lists ii, 

 III, IV, and XVI on the one hand, with Lists vii, viii, ix, etc., and even 

 VI, on the other, will show this in a striking manner. In List ii, for 

 instance, of the genera found distributed over the whole of the Bra- 

 zilian subregion, each genus has at an average 15 species. In List 

 III, whose genera have but a slightly more restricted distribution, each 

 genus is composed of 9;^ species, at an average. In List vi, whose gen- 

 era, while they have a wide distribution, are yet much more restricted 

 than in the others mentioned, each genus has, on an average, but 3i 

 species. The genera of List vii have on an average, but If species, 

 and those of List viii4)ut 1. The number of species of each genus, 

 therefore, varies directly as the extent of its distribution, and, con- 

 versely, the extent of the distribution of any genus varies directly as 

 the number of species comi)osing it. 



There are a few genera which do not come under this general propo- 

 sition. Callichthys has but two species, and Moplosternum only three, 

 but the limits of variations of the species of these two genera are so 

 wide that the two species of Callichthys have received eleven different 

 names, and the three of Hoplosternum thirteen. The most noted ex- 

 cei^tion to the first half of the proposition is Hcmidoras (List vi), with 

 thirteen species. 



At a first glance it might appear that a genus with a narrow distri- 

 bution must necessarily, on account of its restriction, have few species, 

 but a closer inspection will show that this is not the case. Taking, for 

 instance. Lists ii and vil: The genus Pscudopimelodus has four repre- 

 sentatives in the region covered byvii; the genus Bhmn di a twelve ^ 

 Fimelodella five; Fimelodus seven; Trachycorystes five; Fseudouche- 

 ni2)ter us two; Ageneiosiis six; ioricaria nineteen, etc. This shows that 

 the genera of wide distribution have, on average, several times as many 

 species in a given system, even if it be as large as that of the Amazons, 

 as a genus restricted to this system; and that a genus of narrow dis- 

 tribution has not a small number of species simply because there is room 

 for no more. 



