22 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 119 



adipose fin of bifasciatus is placed more like that of members of 

 Nannostomus than that of Poecilobrycon harrisoni, which has the 

 adipose fin posterior to the anal fin, not over it as in Nannostomus. 

 Hoedeman remarked that bifasciatus has a vestigal adipose fin; 

 however, I find no trace of one in his 3 specimens. In addition I 

 faU to find any generic significance in the placement of the adipose 

 fin in nannostominans. It is somewhat more posterior in Nanno- 

 stomus trijasciatus and Poecilobrycon harrisoni but more anteriorly- 

 placed in all other species of nannostominans which possess this 

 fin. Concerning the swimming position, I can make little comment 

 because I have not seen living specimens of bifasciatus. However, 

 it should be remarked that Poecilobrycon harrisoni normally swims 

 horizontally like all living species of Nannostomus that I have seen 

 and that all species of Nannostomus will occasionally swim slightly 

 obliquely when feeding. The color claimed for the caudal fin may 

 be like that of Poecilobrycon egues, but I cannot determine this from 

 the specimens at hand. The precise significance of this color pattern 

 if present would be difiicult to determine. From the key in the 

 present paper it would seem that bifasciatus is closest to beckfordi 

 in several characters and in fact may be rather closely related to it. 

 I faU to find that Hoedeman presented sufficient evidence to indicate 

 either that bifasciatus is a relatively primitive nannostominan or 

 that it is intermediate between Nannostomus and Nannobrycon. The 

 osteology of bifasciatus is typically that of Nannostomus. Weitzman 

 (1964) noted that pyrrhulininans and nannostominans undoubtedly 

 had a common ancestor. Presumably a relatively primitive nanno- 

 stominan would have some characters suggesting pyrrhulininans. 

 The only nannostominan that does this is Nannostomus espei, and 

 even in this species the few such characters are difficult to evaluate. 



Specimens examined. — Three, paratypes, AM 100513, SL 27.2-28.3 mm. 

 (only 2 specimens could be measured as snout of 1 damaged), Surinam, Berg en Dal 

 at Surinam River. 



Nannostomus digrammus Fowler 



Figures 5, 6 



Nannostomus digrammus Fowler, 1913, p. 525 (original description; type locality: 

 Rio Madeira, about 200 miles east of long. 62°20' W., Brazil; ^ holotype in 

 Academy of Natural Sciences of Philadelphia), 1950, p. 261 (listed). 



Poecilobrycon diagrammus. — Hoedeman, 1954a, p. 84 (copied description). 



Poecilobrycon digrammus. — Hoedeman, 1954b, p. 71 (listed). 



Nannostomus beckfordi beckfordi. — Sterba and Tucker (not Giinther), 1962, p. 209, 

 (aquarium description of "gold anomalus" but counts from Giinther's 

 original description of beckfordi; see remarks below). 



2 Bohlke (1955, pp. 8-12) doubts the quoted locality and presents reasons why 

 it may be considered as being 20 miles east of long. 62°20' W. 



