26 



INDIANA UNIVERSITY 



Of particular interest are the following instances where the 

 complete lateral line is giving rise to an incomplete lateral line : 



1. Ctenobrycon hauxwellianus Cope. Of this species I was 

 able to examine nearly 1,400 specimens from the Amazon between 

 Montalegre and Tabatinga. In all of these, the lateral line is 

 complete. I also examined 19 specimens collected by Haseman 

 in the Lagoa de Parnagua, of the Paranahyba basin. In seven 

 of these, the lateral line is complete, six are mutiliated, so it is 

 not possible to say whether it is complete or not, and in six it stut- 

 ters; that is, while it extends quite to the caudal, many scales 

 along the line are without pores. A stuttering lateral line is the 

 character of the genus Psellogrammus, from the Paraguay basin. 

 While the species Ctenohrycon hauxweUiafms is stable in the 

 Amazon, and the process of being transformed into a distinct 

 species of the genus Psellogrammus has been completed in the 

 Paraguay, it is now indifferently a Ctenobrycon or a Psellogram- 

 mus in the Paranahyba basin. To make the statement general, 

 a widely distributed species may undergo modifications in a re- 

 stricted area while it is not changing over its general area. 



2. Phenacogaster, which is one of the most highly specialized 

 of the genera of Tetragonopterinae, is found from Guiana to 

 ]>olivia, and in the Rio San Francisco. In all the specimens of 

 the various species from Guiana to the Beni, except one, the 

 lateral line is complete. In one specimen from the Beni, it is 

 developed on but 26 of the 39 scales. In the closely allied form 

 of the upper Paraguay basin it has been permanently reduced 

 to but eight scales. It has been changed into the genus Vesica trus. 

 ]t is inconceivable that Vesicatrus should have been derived from 

 any other source than Phenacogaster. 



3. Astyanax inconstans, daguae, Moenkhausia cotinho and 

 others are also mutating or inconstant. It is not only evident 

 that the lateral line has changed in different genera, but it has 

 changed and is changing several species of one genus into species 

 of another genus. In Moenli-hausia cotinho both forms with a com- 

 plete and incomplete line are found throughout its known area of 

 distribution. To make this statement general, a widely distributed 

 species may offer material for the same evolution throughout its 

 area of distribution. 



It is not so clear in which direction evolution went, in regard 

 to the scaling of the caudal. If we assume again that the majority 

 indicate the primitive condition, the naked caudal is primitive, 



