eigenmann: the cheirodontin^. 



11 



than would the completeness or incompleteness of the lateral hne justify the col- 

 location of the genera having these characters in subfamihes. 



In the peculiar armature of the cheeks Prionobrama shows such similarity to 

 Aphyocharax that Cope placed it in the latter genus. In spite of its general appear- 

 ance and the backward position of the dorsal, Prionohrama is probably more nearly 

 related to Aphyocharax than to any member of the Glandulocaudinae. 



The pecuhar scaling of the caudal in two of the species recalls the similar 

 character in the Glandulocaudinae and in Argopleura of the Tetragonopterinse. 

 It is possible that this is also an independently acquired character in a number of 

 remotely related genera. 



While it is very probable that we are dealing with a natural group, it is certain 

 that different members point to three distinct subfamilies from which they may 

 have been derived, or to which they may have given rise. The Cheirodontinse 

 are certainly near the generahzed type of all the Characins. 



Secondary Sexual Differences. 



There are no conspicuous secondary sexual differences in the group under con- 

 sideration, unless there should be a striking difference in the lipochromes, which 

 have been dissolved in alcohol. 



The differences when present consist in the length of the fin-rays, in the fila- 

 mentous termination of the fins, in the development of hooks on the anal, and less 

 frequently on the caudals and ventrals of the male, in the development of scaly 

 pockets on the caudal of the males, in the color of the dorsal, and, what is unique 

 for this group, in the high development of the interhsemals in the male in Cheirodon. 

 The details are given under the respective species. 



Distribution. 



The Cheirodontinae apparently reach their maximum development in the 

 middle Amazon and the upper La Plata basins. It is quite possible that this 

 greater abundance in the middle Amazon and in the Paraguay basin is apparent 

 rather than real. Little collecting has been done with fine meshed nets in the 

 Orinoco and in the upper waters of the Amazon. 



The group as a whole has a very wide distribution. The genus Cheirodon has 

 a range all but coextensive with that of the subfamily. The species, with the 

 exception of Cheirodon piaba Liitken, are confined to rather limited ranges. Only 

 two species are found both in the Paraguay and in the Rio Guapore. A number of 

 species in the Madeira and Paraguay are evidently very closely related. 



