1915.] NATURAL SCIENCES OF PHILADELPHIA. 263 



dachneri. In the earlier accounts of P. nigricans little note is taken 

 of the color. Prof. Starks states that Madeira River specimens 

 have the lengthwise lines and cross-bars more distinct than those 

 from Para. 



PoBoiluriolitliys bimaculatus (Linnseus). 



Head 3 to 3^; depth 2\ to 2f ; scales 32 to 35 in 1. 1. to caudal base 

 and 2 or 3 more on latter; 7 scales above 1. 1., rarely 6; 7 scales below 

 1. 1., rarely 6; 12 to 15 predorsal scales; snout 3| to 4 in head; eye 

 2f to3; maxillary 2| to 2f; interorbital 2| to 3; length 45 to 58 mm. 

 Six examples. 



Among my earlier material belonging to this species, the specimens 

 from the Tocantins headwaters each have a cluster of dusky dots 

 at base of each lateral scale, no other dots on outer portions of 

 scales, where apparently none were ever present, and thus lengthwise 

 series of inconspicuous spots are evident. In these the predorsal 

 scales are interrupted on the anterior median line, though closel}^ and 

 irregularly approximated, several being saddled over the ridge 

 of the posterior half. In the larger of my examples of Astijanax 

 hartleUii the predorsal line is more or less interrupted, though in the 

 larger the squamation is mostly destroyed. The Paramaribo A. 

 orientalis is pale in color, and in agreement with the Ceara material, 

 though without pigmented dots on the sides at present, and the 

 fallen predorsal scales have left pockets showing they were probably 

 more or less completely placed as saddles. A. lacustris shows the 

 scales of the predorsal closely approximated, though or.ly those of 

 the posterior , half formed saddles. A. jacuhiensis shows the pre- 

 dorsal scales with a nearly complete naked strip in front and without 

 the dark pigment dots, or only very faint sparse ones, at the bases 

 of the scale exposures. The body is also much deeper. Prof. 

 Starks mentions that Lake Extremos, Lake Papary and Ceara Mirim 

 examples have fewer anal rays (25-27), while in the Para material 

 they were more numerous (31-32).-* Of the first he says: ''These are 

 perhaps referable to Astyanax bimaculatus novcB Eigenmann, though 

 the lateral band is not so definite as in the picture of the original 

 specimen." A. bimaculatus novce I have been unable to find noticed 

 elsewhere. Finally I have described A. rwpununi^ from British 

 Guiana, which in no way differs from Ceard material. It shows the 

 predorsal scales nearly completely forming saddles over the median 



4 Stanford Univ. Publ. (Fishes Stanford Exp. Brazil), March 17, 1913, p. 16. 

 * Proc. Acad. Nat. Sci., Phila., 1914, p. 242, fig. 6. 



