1899.] FROM THE GOLD COAST. 721 



head without snout or as the spine of the pectoral, very slightly- 

 serrated along its posterior, and nearly smooth along its anterior 

 edge. Anal fin not reaching the caudal, when laid backward, with 

 11 or 12 rays, 7 or S of which are branched. Caudal fin deeply 

 cleft, with the upper lobe as long as the head. Upper parts 

 greyish brown, lower silvery. 



Three specimens from the Eiver Prah, 91 and 139 milliai. 

 long. _ 



This species represents in the Gaboon rivers the Nilotic Chfijs- 

 iclithys macrops, to which it is closely allied. In that species, 

 however, the anterior dorsal ray is greatly prolonged, even in 

 specimens which exceed the Prah fishes only slightly in length. 



Chrysichthys BtJTTiKOFEJii Stelnd. (Plates XLI. & XLII. fig. A.) 



Chrysichthi/s huttikoferl Steindaohner, Notes Leyden Mus. xvi. 

 p. 60 (1891). 



The examination of a small number of (chiefly young) specimens 

 of Ghrysichthi/s from various localities in the Gaboon country 

 has been attended with much difficulty and uncertainty. A j:)«j-^ 

 of them seemed to be identical with, or closely allied to, Ch. 

 bilttikoferi (Steindachner). Although they show certain slight 

 differences in the number of anal rays, extent of the tooth-patches 

 on the palate, form and comparative length of the snout, size of the 

 eye, and length of the dorsal and caudal rays, Steindachner's descrip- 

 tion applied more or less perfectly to all. However, the series of 

 specimens of any species from the same locality is still so incom- 

 plete that we are much in the dark as to individual variations, the 

 changes these fishes undergo with age, or as to any secondary 

 sexual characters. Some years ago I should not have hesitated to 

 refer all these specimens to the same species Ch. biUtikoferi, and 

 I am not i)y any means certain tliat this will not prove to be the 

 proper course to pursue, when sufficient materials are brought 

 together; but since more receut investigations of the West African 

 Fauna have shown the wide distribution and great specific develop- 

 ment of this genus, I am induced, after long hesitation, to 

 distinguish among the forms allied to Oh. buttilvfen several under 

 distinct names. 



The question, then, arises for which of the forms, distinguished 

 here, the name given by Steindachner should be retained. 

 Steindachner's type came from Liberia, is a unicum, and young, 

 being 20^ centim. long. I am indebted to Dr. Jentink' for a 

 sketch of this type as well as of its dentition. Unfortunately the 

 specimen presents those elements of uncertainty which render the 

 study of these fishes so difficult. As will be seen from the 

 accompanying sketch, the two patches of larger vomerine teeth are 

 connected witli each other by, and are in fact only a portion of, a 

 larger horseshoe-shaped band of minute rudimentary vomerine 

 teeth, extending backward on the palatine bones. In the Eiver Prah 

 specimens referred by me to Ch. hilttiJcoferi only the two patches 

 of larger teeth are visible, but none of the rudimentary ones. 



