1920.] Dr. Malcolm Smith : Sea Snakes. 25 



side in two more ; the posterior temporal is as large as the 

 anterior or larger ; a cuneiform scale is present in all after 

 the third infralabial. 



Colour. — None of my series (except one juvenile) is 

 completely banded. Above they are pale grey, with from 

 60 to 80 dark grey dorsal rhombs, which in some of them 

 are continued round the body as pale bands, but in the fore 

 part of the body only. The transition from the dark dorsal 

 rhomb to the paler ventral band occurs with a fairly clear 

 line of demarcation. Below creamy white ; head, neck 

 below, and anterior part of belly, black. One specimen. 

 No. 1150, is very pale, being almost white in the posterior 

 part of the body, with the dorsal marks only just visible. 



To H. fasciatiis typica I refer the H. gracilis recorded 

 by Hanitsch (Kept. Raffles Mus., 1897, p. 101) . For its scale 

 formula see the table. 



Hydrophis fasciatus atriceps. 



Hadrophis atriceps, Giinther, Rept. Brit. Ind., p. 371, pi. xxv, 



fig. 1 (1864). 



Disteira cincinnatii. Van Denburgh and Thompson, Proc. 

 California Acad. Science, (4), 111, p. 41, Dec. 1908. 



The scalation and distribution of this form have already 

 been dealt with. 



Number of specimens examined, 60 ; Gulf of Siam, 

 58 ; Cap St. Jacques, 2. 



The sudden and marked diminution in the number of 

 scale rows and ventrals, as well as the alteration in colour, 

 which characterizes nearly all my specimens from the Gulf 

 ot Siam as compared with those from the Straits, induced 

 me for a long time to believe that they were distinct. I 

 cannot however find any certain grounds upon which they 

 can be separated. Between an example from Pulau Angsa, 

 with 57 scales round the body, 504 ventrals, and the back 

 with dark rhomboidal marks, and another from Ban Yao 

 with 41 scales round the body, 327 ventrals, and the body 

 marked with complete bands, there is every degree of 

 gradation. 



Variation. — As one would expect, with fewer scales 

 round the body in proportion to those upon the neck, this 

 form is relatively stouter anteriorly than the Malaccan one. 



The rostral shield may be as broad as high ; the frontal 

 shield in six examples is longer than its distance to the 

 rostral, in two of these being in contact with the nasals. In 

 No. 1261 the frontal is very small, much shorter than its 

 distance to the rostral. The anterior temporal in one 

 instance (No. 1252) is divided on one side to form two 

 superposed shields ; the posterior temporal shield is seldom 

 as large as the anterior, and is often divided in two by a 

 vertical suture. Normally the supralabials are undivided, 

 but in one instance there is fragmentation of the fifth. In 

 five examples the temporal shield reaches the border of the 

 mouth to the exclusion of the sixth and seventh labials. A 



