26 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 119 



border of opercle. On body, primary horizontal stripe begins anteriorly 

 underneath fleshy opercular flap and extends through approximately 

 lower half of foiu-th scale row and upper third of fifth scale row. 

 Primary stripe widening posteriorly until, over anal fin, there covering 

 enth-e fourth, and one-half of fifth scale rows. Posterior to this, 

 primary horizontal stripe covering almost entu-e lower half of caudal 

 peduncle and continuing as an acute wedge posteriorly over caudal fin, 

 ending at base of caudal fork. Below primary horizontal stripe, 

 sides and belly silvery with few irregularly scattered melanophores. 

 Broad silvery stirpe extending from eye to caudal base above primary 

 horizontal stripe. Silvery stripe with oblique band, about 3 scales 

 long, located just anterior to dorsal fin. This area spotted with large, 

 dark melanophores. Few small melanophores scattered over entire 

 silvery stripe. Narrow, dark secondary horizontal stripe just dorsal 

 to silvery lateral stripe, beginning at upper posterior edge of operculum 

 and ending on upper part of caudal peduncle posterior to adipose fin. 

 Dorsum pale brown above secondary horizontal stripe. Areas of head 

 and eye not covered by dense black pigment, silvery with few scattered 

 melanophores. Dorsal, pelvic, pectoral, and anal fins colorless. 

 Caudal fin, in addition to having narrow wedge of primary horizontal 

 stripe, with fau'ly dense aggregation of melanophores along first 3 

 upper principal rays of upper lobe. Some melanophores scattered 

 over entire fin. Color in life not known with certainty. 



Remarks. — Nannostomus digrammus was known previously only 

 from the four specimens reported in the original description. However, 

 this species has been imported as an aquarium fish. Several years ago 

 a prominent aquarist, Mr. Frederick Stoye, sent a specimen of a 

 nannostominan to Dr. George S. Myers at Stanford University. 

 This specimen, labeled as the "gold anomalus," proved to be a large 

 male digrammus. 



All the specimens belonging to the original type lot of digrammus 

 were females, and the holotype illustrated by Fowler shows the anal 

 fin of a female. 



Sterba and Tucker (1963, p. 209) believed the so-called "gold 

 anomalus" to be Giinther's original Nannostomus beckfordi. However 

 Stoye's specimen mentioned above would seem to show that the "gold 

 anomalus" of aquarists is N. digrammus. Aquarists' records of this 

 fish cited by Sterba and Tucker (1963) are diflScult to compare. 

 Perhaps more than one species of fish is involved. However, live 

 specimens of Nannostomus beckfordi from British Guiana are compar- 

 atively pale in color, never exhibiting the colors described for the 

 "gold anomalus." 



Specimens examined.— Two, paratypes, ANSP 39190 and 39191 SL 18.0-19.0 

 mm., Brazil, State of Amazonas, Rio Madiera, about 200 miles east of long. 



