16 Col. E. H. Beddome on the 



scale yellowish ; 3 or 4 dark triangular blotches on the ante- 

 rior portion of the trunk and 1 or 2 similar ones about the 

 anal region ; belly yellowish, blotched with seaweed-like 

 purplish markings. (Ventrals 139 to 141, subcaudals 7 pairs 

 in the two specimens known.) 



Hob. South Travancore, on the Myhendra Mountain. 



A very beautiful variety, but not differing from the type in 

 any way but coloration. 



Var. e. arcticeps. 



Sill/bum arcticeps, Giintuer, Proc. Zool. Soc. March 16, 1875, p. 229. 

 Siiybura madwreiisis, Bedd. Proc. Zoo!. Soc. Nov. 5, 1878, p. 802. 



Brown or blackish pnrple above ; belly and sides of the 

 same colour, with irregular yellowish cross bars. (Ventrals 

 128 to 149, subcaudals 7 to 9 pairs.) 



Ilah. Mountains in the Madura district above the Cumbiira 

 valley ; mountains of North Tinnevelly, 4000 to 6000 feet 

 elevation. 



The typical arcticejjs of Giinther (two specimens) has 128 

 to 130 ventrals ; madurensis (Bedd.) 142 to 149 ; they are, 

 how^ever, I believe identical. 



Var. ^. incta. 



Back blotched with orange-yellow and black, the two 

 colours nearly equally divided, some scales being black and 

 some yellow, others half black and half yellow ; belly and 

 sides (2 or 3 rows of scales) jet-black, iridescent, with very 

 irregular broad orange-coloured blotches or cross bars. 



Hah, North Travancore near Peermede. 



A single example only of this most beautifully coloured 

 variety was captured on Mr. Maltby's coffee-estate at an 

 elevation between 3000 and 4000 feet ; it has 150 ventrals 

 and 8 pairs of subcaudals (the two central ones being single). 



Some herpetologists will probably consider these five 

 varieties distinct species ; they, however, I believe differ only 

 in coloration, and are, I think, only geographical varieties, 

 though probably quite permanent as such in their respective 

 localities. 



Dr, Giinther described B. hrevis as a distinct species on ac- 

 count of its shorter form and fewer ventrals, 122 to 131 against 

 139 to 155 in what he considered typical of this species; Ifound 

 out in India, however, that it was quite impossible to draw a 

 line between them or to distinguish S. hrevis in any way: the 

 ventrals vary from 122 to 140, the coloration is exactly the 



