134 BARBOUR: ZOOGEOGRAPHY. 



common than B. candidus, it is nevertheless abundant enough to be much dreaded 

 by the natives. Bryant had two examples from Depok. 



In color my specimen is perhaps noteworthy, in that the forward prolonga- 

 tion of black from the first annulus reaches the internasals. Between this black 

 and the brown on the nose is a very narrow line of yellow. 



Bungarus candidus (Linn^;). 

 Linn£ Syst. nat., ed. 10, 1758, 1, p. 223. Boulenger, Cat. snakes Brit, mus., 1896, 3, p. 368-369. 



Type locality: — "Habitat in Indiis." 



Boulenger has grouped under the name B. candidus three forms of serpents 

 which, as he has shown, come from well-defined geographic areas. We have 

 no proof that these forms merge into one another where their respective ranges 

 meet. It seems, then, necessary, owing to their well-marked color-characters 

 and their distribution, to call them separate species. Thus we have: — 



B. candidus (Linne) from Java and Celebes. 



B. multicinctus Blyth (Stejneger, Bull. 58 U. S. nat. mus., 1907, p. 

 397-399) from Upper Burma, China, Hainan, and Formosa. 



B. caeruleus Daudin, from India to the Straits Settlements. 

 This last form is variable in coloration, but has not been shown to be com- 

 pletely connected with either of the others. 



Two specimens of B. candidus taken at Buitenzorg, January, 1907, agree 

 perfectly with the descriptions. When compared with a specimen of B. caeru- 

 leus taken at Lucknow, November, 1907, they show the following differences in 

 addition to the well-known divergence in coloration : — the width of the head of 

 the Javan specimens is about one fourth greater than in the Indian; the frontal 

 in the Javan examples is very nearly as broad as long, in the Indian the frontal 

 is one and one third times as long as broad; in the Javan individuals the length 

 of the suture between the internasals is two thirds that of the suture between the 

 prefrontals; in the Indian snake the first-mentioned is only one third as long as 

 the second. In the Javan specimens the superior preocular is much larger than 

 the inferior, while in the Indian they are of the same size. The scale-counts do 

 not aid in separating any of the three forms. Bryant's series, ten from Buiten- 

 zorg, and two from Depok, bear out these observations. One of his specimens, 

 however, has an almost square first temporal, a curious anomaly. 



