Cockerell — Scales of Cohitid and Homalofterid Fishes. 207 



asymmetrical; length of scale about 1190 m, breadth about 850; 

 sculpture much as in the last two species ; basal and apical radii 

 well developed, latter variable and more or less broken; apical 

 about 18, the outer ones curving basally toward the nucleus ; circuli 

 widely spaced. Lateral line scales have the radii fewer and the 

 circuli closer. The luicleus is about 290 m from base. There is 

 nothing wliatever in these scales that aftVjrds any radical distinction 

 from the Cobitines, especially Lepidocephalichthys. 



(8) Homaloptera maculata. (Khassya; Jerdon; B. Mus. ) A very curious 



fish, the under side tlat. (By no means so specialized as Gasiro- 

 myzon, however.) Scales fairly large, nearly 2 mm. long, and 

 nearly as broad; apex broadly rounded, base flattened; radii all 

 round, strong, not very close (about 15 apical), very short, owing 

 to the very large nuclear area, over which are scattered spots arising 

 from the breaking up of the central parts of the radii, the width of 

 this nuclear area is about 850 m; circuli widely spaced. The sculp- 

 ture here is essentially as in Gaslromyzon, etc., except for the 

 remarkably modified central region. 



(9) Homnloptera brucei. (Meekalan, Tenasserim; Fea; B. Mus.) Scales 



smaller, about 1275 m long, and broader, l>ut of entirely the same 

 pattern. The apical radii are irregular and more or less broken up. 



The Cobitinft have their headquarters in the Indian region, where Day 

 recognized nine genera and 4f> species. (Botia, G species; Acanthopsis, 1 ; 

 Somileptes, 1; Lepidocephalichthys, ?); Acnnthophthnlmus , 1; Apua, 1; 

 Jerdonia, 1; Nemachilichthys, 1; Nemachilus,31.) From Japan, Jordan 

 and Fowler recognized (1903) live genera and six species. The only 

 African genus is Nemachilus {N. abyssinicus Blgr., Lake Tsana, Abys- 

 sinia). Misgurnus, Acanthopsis and Cobitis get as far west as France. 

 Not a single species has reached America. Three species of Cobitis occur 

 in the Upper Miocene beds at Oeningen ( Wangen), Baden; I have exam- 

 ined specimens of them in the British Museum. 



The Homalopterinse, with four genera, are exclusively Asiatic, occurring 

 in China and India, and the Malay Peninsula and Archipelago. 



It seems rather remarkable that a group so diverse, and in some of its 

 members so specialized, and tlierefore probably of great antiquity, should 

 not have spread more widely. No doubt the carnivorous Characinids 

 may have proved an obstacle in Africa, and it must also be remembered 

 that the past distribution may have been wider than the present. The 

 discovery of a fossil Cobitid in America may not be altogether out of the 

 question. 



The scales, all of essentially the same type, are more or less degenerate, 

 but hardly specialized. They appear to represent the earlier type of 

 Cyprinoid scale in a weak form, and that is why they remind one of cer- 

 tain scales of various genera of true Cyprinidae. 



