HERPETOLOGY OF JAPAN. 387 



Genus HEMIBUNGARUS« Peters. 



1843. Bmchyrhynchus Fitzixgeh, Syst. Rt'pt., p. 28 (type, Elups callujaslvr) (not 



of Laporte, 1833). 

 1862. Hemibungarus Peters, Mon. Ber. Berlin Akad. Wiss., 1862, p. 037 (type, 



Elaps calligaster) . 



The only difference between the genera Ilemlhungarus and Cal- 

 liophis appears to be the presence of several (1-3) small solid teeth 

 on the maxillary bone behind the grooved fang in the former, and it 

 is ([uestionable whether these so-called genera really represent genetic 

 groups. The two species occurring within our limits, namely, Cal- 

 liophis macclellandii in Formosa and Hemibungarus jayonicus in the 

 Riu Kill Islands seem to be so closely related to each other that the 

 latter appears less nearly allied to any of the other species of Ileml- 

 hungarus. The only difference, besides that of dentition, consists in 

 the greater length of the parietals in H. japonicus and in a slightly 

 different pattern of coloration. Thus far no Hemibungarus has been 

 recorded from Formosa, but this island is as yet so imperfectly known 

 that no conclusion can be based on this fact. 



The relationship of these snakes (through Calliopliis) is undoubtedly 

 Ilimalayo-Chinese, but beyond this statement it would be hazardous 

 to generalize for the present. 



The two forms described in the following pages under the names 

 of H. japonicus and H. boettgeri are evidently very closely allied in 

 spite of their very different coloration. The scutellation is almost 

 identical, though it must be pointed out that in the former (9 s])ecimens) 

 the number of ventrals averages higher, namely, 208.5 (maximum 2 16, 

 minimum 196), the corresponding figures for the latter (11 specimens) 

 being 193.5 (204 and 183). While the scale formulas thus show some 

 overlapping, the difference in coloration is so striking and apparently 

 so constant, without any intermediary specimens, that I do not feel 

 compelled to use trinominals for these forms, especially in view of 

 their geographical isolation, the former being confined to the O-shima 

 subgroup, the latter to the Okinawa subgroup of the Riu Kiu 

 Archipelago. 



The}'^ may be distinguished as follows: 



0* Red (white in alcohol) with one, or three, narrow dark-brown longitudinal striijes, 

 and distant rings of the same color; ventrals 196 to 216 H. japonicus, p. 387. 



H- Black, with four narrow longitudinal red stripes (whitish in alcohol) on back, and 

 distant white half-rings; ventrals 184 to 200 //. boctUjcri, p. 389. 



HEMIBUNGARUS JAPONICUS & (Guenther). 

 HAI (according to Dr. Ijima). 



1868. Callopliis japonicus Guenther, Ann. Mag. Nat. Hist. (4), 1, 1868, p. 428, pi. 

 XVII, fig. c (type-locality. Nagasaki [?J; type in Brit Mus.; Whitely, 

 collector). — Heviihungarus japonicus Boulenger, Cat. Snakes Brit. Mus., 

 Ill, 1896, p. 395 (part: Oo-shima). — Wall, Proc. Zool. Soc. London, 

 1905, II, p. 51^ (part: Amami-o-shima). 



a From 77///-, half-; and Bungarus, a nearly related genus of snakes. 

 b Signifying Japanese. 



