18 PROdEEDINGS OF THE NATIONAL MUSEUM. vol. xxxiii. 



identical, with Cope's Astyanax phoenicopterus. There seems no doubt 

 that the five specimens before us belong to the same species, although 

 they do not have the same generic characters. Of the species of 

 Ilemigr'cunmuH it approaches Jiltheni most closely. 



Tetragonopterus argenteus Cuvier. 



No. 39403. One specimen, Brazil, collected by H. H. Rusby. 

 No. 44831. One specimen, Bolivia, Gibbon collection. 

 No. 1631. Paraguay, Page collection. 



Genus ASTYANAX Baird and Girard. 



In the difiicult and highly interesting group of Characins there is 

 no more difficult nor more highly interesting genus than Astyanax. 

 Its divergence toward Ilemigramnms, which is like an Astyanax with 

 an incomplete lateral line, toward Iletn'Jynjco)}^ which is like an 

 Astyanax with a completely denticulated maxillary, toward Moeiik- 

 hausid., which is like an Astyamix with a scaled caudal; and especially 

 to\f 2i,rdi Petersliis (which appears to be its African counterpart), together 

 with its universal distribution in tropical and subtropical America, all 

 indicate its central position in the system of Characins. Some of its 

 species are well marked, but for the most part there are groups of 

 species within which the specific characters are not well fixed. The 

 most notable of these groups is that centering in Astyanax rutilm. 

 This is a widel}^ variable, univer.sally distributed species, with which 

 throughout its distribution there are allied closely related species. In 

 the south are iherlngli^fasciatus, cuvleri, and perhaps others. In the 

 north, notably in Central America, Mexico, and Peru, there is an espe- 

 cially trj'ing series of species, varieties, or local forms. It will require 

 much larger series representing a much wider array of localities than 

 are at my command to finally describe the status of these forms. Those 

 of the northernmost localities, north of central Mexico, represented 

 b}^ specimens with a small number of anal rays — from 20 to 25 — may 

 readily be set aside as viexlcanm. From southern Mexico there have 

 been vecovdiQA fasciatus { — rutilus) and aeneus (Oaxaca). Through the 

 courtesy of Prof. S. E. Meek, I have been able to examine many 

 specimens from Perez, all of which appear to be xneus, and others 

 from Montzorongo, some of which are seneus^ and others with a larger 

 eye and maxillary extending considerably beyond origin of e3^e may ( ?) 

 represent rutilus. 



From the Pacific slope of Guatemala have been recorded rutilus^ 

 micropldhalmuH^ and luiwiJh. I am not acquainted with either of the 

 two latter. From the Atlantic slope probably come rutilus, seneus^ 

 and hrevimmius ; the last, a species with few anal rays, is probably a 

 mexicanus. From Nicaragua have been recorded mrstedii, which is 

 not distinguishable from s'neus or rutilus., and nicaraguensis^ which is 

 distinguished by the increased number of maxillary teeth. 



