In 1955, the late Dr. O. Schubart, Director of the Estaçao Esperimental de Biologia at Pirassununga (São Paulo, Brazil) kindly put at my disposal a small collection of fishes which he had assembled in the Mogi-Guassu (river) below the Cachoeira de Emas (Emas falls). Only a series of eight apparently conspecific specimens is discussed here. At first sight, the present specimens appear to closely resemble Pimelodella griffini Eigenmann, especially six examples with a prolonged first soft dorsal ray. However, excepting several minor differences, P. griffini has the dorsal spine prolonged instead of the first ray, while it has the adipose dorsal fin much shorter. Having only two specimens of P. griffini, moreover of dubious identity, available for comparison, extensive use was made of Eigenmann's (1917) monograph of the genus. As already indicated, of the eight apparently conspecific specimens only six show the filamentous prolongation of the first soft dorsal ray, which seemed to point to two possibilities: either the present material might consist of two separate new species, or the difference in length of the first soft dorsal ray has to be considered as a mere sexual character. Though sexual differences of this kind never seem to have been reported from Pimelodella species, the fact that no other differences were found in the usual diagnostic characters made the second possibility much more likely. This opinion was confirmed by a close examination of the morphological features of the abdomen and by dissecting part of the specimens. The examples with prolonged rays proved to have a wel developed conical sexual papilla, rather remote from the vent, while dissecting showed an obvious milt, making it clear that all were male. The two examples without