112 Annals of the Carnegie Museum. 



deep; nucleus nearly central; apical margin rounded, without 



teeth; radial system very strong, both basal and apical, the radii 



meeting at the nucleus; large discal polygonal areas developed in some 



scales of Hoplerythrimis; lateral and basal circuli fine (much coarser 



in Hoplerythrimis; finest in Hoplias) ; apical circuli differentiated, 



longitudinal, slightly oblique near the middle line. 



Hoplias malabaricus (Bloch). Apical radii five or six, counting the 



sublateral ones; basal about six, the outermost, when complete, 



arched at upper end. 



Hoplias macrophthalmus (Pellegrin) (Plate XXV'III, fig. 2). Large 



scales, over ten mm. across; apical radii, counting sublateral, about 



sixteen to eighteen, basal about eight to eleven; outer apical and 



basal both arched neat base when complete. The figure sufficiently 



shows the arrangement. 



Hoplerythrinus unitceniatus (Spix) (Plate XXVIII, fig. 4). Scales 



about five mm. long and six broad; about four to eight apical radii, 



about the same basal, and usually some lateral; large pentagonal 



areas developed in some scales; apical circuli very widely spaced. 



Erythrinus erythrinus (Bloch & Schneider). Scales about three and 



one-fourth to three and one-half mm. long and broad; sculpture 



and pattern as in Hoplerythrinus, but fewer radii. 



I give a figure (Plate XXVIII, fig. 3) of the scale of the Asiatic 



Barbus chola, a Cyprinid type which shows in its squamation a strong 



resemblance to Erythrinus. Here, so far as the scales go, the Chara- 



cinids and Cyprinids meet, and it is at least significant that this 



occurs in the Erythrinine group on the one hand, and the Barbus- 



group on the other. 



Appendix. 



Bryconccthiops microstoma Giinther (Plate XXVIII, fig. 5). Ksibi 

 River {Bates). British Museum. This African genus was omitted 

 from my paper on the Characinldae of that continent. The scale 

 is about five and one-half mm. broad with weak sculpture, the 

 broad nuclear region finely pustulose, and the apical field with 

 nothing to represent the circuli, except a sparser pustulose ornamen- 

 tation, even this failing toward the margin. The radial system is 

 represented by a few irregular polygonal areas, from which arise 

 imperfect and very asymmetrical radii. This is a weak type related 

 to Alestes. 



Hydrocyon forskalii Cuvier. (Plate XX\TII, fig. 6.) I give a figure 



