36 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 119 



of pelvic fin and distal portion of its anterior 4 to 5 rays bright red. 

 Distal tips of second to about fourth pelvic rays black. Belly silvery; 

 area below primary horizontal stripe bright gold. Entire length above 

 primary horizontal stripe with narrow iridescent gold or sometimes 

 silvery-red stripe. Above this mth a greenish dark-brown area, same 

 color as top of head and dorsum. Primary, secondary, and horizontal 

 stripes black. Iridescent golden to pale reddish spot near anterior 

 nostril. 



Remarks. — Nannostomus marginatus has been reported only from 

 the Guianas. It now appears to be a widespread polymorphic species 

 found in the Guianas and in at least several widely separated areas in 

 the Amazon Basin. Hoedeman's subspecies picturatus from Sm'inam 

 is described very inadequately and some of the description appears to 

 be in error. For example he reported a difference in caudal fin ray 

 counts between the British Guiana and Sm^inam populations. He 

 stated the caudal rays to be 10 + 10 in specimens from British Guiana 

 and 8 to 9 + 9 to 8 in specimens from Surinam. All specimens of nan- 

 nostominans and indeed all characids that I have examined, except for 

 a few obviously abnormal specimens of the genera Hyphessobrycon and 

 Moenkausia, have had a principal caudal fin ray count of 10/9 (equals 

 10+9). I am not sure whether Hoedeman counted principal caudal 

 fin rays or only unbranched rays. However, he reported the wrong 

 count for either method. It should be noted that in counting the prin- 

 cipal caudal rays of two of Hoedeman's types of A^. bifasciatus, the 

 typical count of 10/9 was found. Hoedeman (1954a) reported 8(7) + 

 8(7). Despite the probable errors in some of Hoedeman's counts, and 

 possibly also his measurements, probably it will be found eventually 

 that his subspecies picturatus is valid. Also the populations noted 

 here from lower and upper Amazon Basin very probably will, when 

 adequate material becomes available, prove to be sufficiently different 

 to merit subspecific recognition. An example of the difference in 

 teeth of different populations can be seen in figiu-e 8. The two 

 opposing teeth to the right are from British Guiana specimens [CAS 

 (lUM) 11700] while those to the left are from Lagoa Grande, Brazil 

 (SU 50219). 



Specimens examined. — Four, paratypes, CNHM 52776, SL 17.0-18.5 mm., 

 Brutish Guiana, Maduni Creek, Mahaica River, 1908, Carl H. Eigenmann. — 

 Two, CNHM 54932, SL 19.4-20.3 mm., British Guiana, Hubabu Creek, Demerara 

 River, Oct. 1, 1910, Max EUis.— One, paratype, SU 21941, SL 17.6 mm., British 

 Guiana, Cane Grove Corner, Mahaica River, 1908, Carl H. Eigenmann. — Two, 

 paratypes, CAS(IUM) 11697, SL 16.0-16.4 mm., British Guiana, Lama Stop-Off, 

 Mahaica River, 1908, Carl H. Eigenmann.— One, paratype, CAS(IUM) 11698, 

 SL (damaged), British Guiana, Crab Falls, Essequibo River, 1908, Carl H. 

 Eigenmann. — One, paratype, CAS(IUM) 11699, SL 19.5 mm., British Guiana, 

 Rockstone sandbank, Essequibo River, 1908, Carl H. Eigenmann. — Two, para- 



