HERPETOLOGY OF JAPAN. 421 



DISTEIRA MELANOCEPHALA" (Gray). 

 Plate XX 1 11. 



184y. Hydrophis snhlxvis, var. vielanocephald (tRAY, ('at. Sn. lirit. Mus. . p. 53 

 (part: type-locality, ''Indian Ocean;" type in Brit. Mus.; Sir E. Belcher, 

 collector). 



18(i4 '^Hydro-phis atriceps Guenther, Kept. Brit. India, p. 371. pi. xxv, figs. /-/' 

 (type-locality, Siam; types in Brit. Mus.). 



1895. Hydrophis faseiatus Boettger, Offenbach. Ver. Naturk. 33-30 Ber., 1895, 



p. Ill (Miyakoshima. Riu Kiu); Kat. Schl. Mus. Senckenberg.. 1898. 

 p. 113 (Miyakoshima) (not of Schneider). 



1896. Hydrophis melanocephaliis Boulenger, Cat. Sn. Brit. Mus., Ill, p. 283, 



pi. XV b (Indian Ocean; type). — Wall, Proc. Zool. Soc. London, 1903, 

 pp. 95, 101 (" Loo Chocs"); 1905, II, p. 516 (Ishigaki; Irioniote).— Mcro- 

 cephalophis melanocephalus Stejneger, Journ. Sci. Coll. Tokyo. XII, 

 Pt. 3, 1898, p. 224 (Pescadores lis.). 



1901. Disteira orientalis Stejneger, Proc. Biol. Soc. Washington, XIV, Dec. 12, 

 1901, p. 191 (type-locality, Okinawa shima; type No. 29, Imper. Mus., 

 • Tokyo). 



1903. Distira subdnda Wall, Proc. Zool. Soc. London. 1903, p. 95 (part: "Japan, 

 Loo Choos;" not of Gray). 



The line drawn between this and the following species, D. fasciata, 

 viz, at the ;370-399th scale rows, is possibly an arbitrary one, but I am 

 inclined to believe that in this way we obtain better results and come 

 nearer the truth. It is not certain, however, that other characters 

 coincide with such a division, and possibly it might be better to sepa- 

 rate the two forms according to the scale rows around the body, but 

 the records are very defective on this point. On the other hand, there 

 may even be three forms, one more or less confined to the coasts of 

 India and the Bay of Bengal, with the maximum average of ventrals 

 and body scales; another intermediate form from the Malay island 

 world and atljacent coasts of the mainland, and a third, with a more 

 eastern habitat, including the China Seas, having the mininmm num- 

 ber of ventrals and scales. '' 



a From uiXai, black; KscpaXi'j, head. 



^ Reproduced in this work on Plate XXIII. 



c There is a very serious obstacle to the defining of these forms from the recoi'ds, viz, 

 the difficulty in counting the scales and scutes of these snakes exactly. In the records 

 we find the most divergent figures for the same specimens. Thus among the snakes 

 included by Boulenger in this species, British Museum Nos. a, d, and e seem to be three 

 of the four specimens mentioned by Guenther (Kept. Brit. India, p. 371) as H. lindsayi. 

 These four he gives as having, respectively, 345, 390, 424, and 449 ventrals, while 

 Boulenger's count is 402, 412, and 452. Guenther states that the number of ventrals oi 

 the type of his H. atriceps is 376; Boulenger counts 364. Guenther descriljes Gray's 

 H. ocellata as having 296-334 ventrals, mentioning three specimens, one of these the 

 type; according to Boulenger (p. 291), these three specimens (including the type) have 

 274, 275, and 290 ventrals; Guenther examined four H. ornata, among which the type 

 and that of H. inornata and gives the ventrals of the species as 252-260; according to 

 Boulenger the ventrals of the two types mentioned are 232 and 240. 



