IX. REMARKS ON SOME FORMS OF 

 DIPSADOMORPHUS. 



By F. Wall, Major, IMS., C.M.Z.S. 



Many of the forms now recognised as species in the genus 

 Dipsadomorphits exhibit extremely close affinities. A close study 

 of the head shields of many of the species (I have examined no 

 less than thirteen of the twenty-three known) shows a number and 

 disposition so similar, that, with the single exception of the rostral 

 shield in some few species, I can find no means of differentiating 

 between them. The only points made special use of by Mr. 

 Boulenger, viz., the height of the prseocular and the size of the 

 posterior sublinguals, with the separation of the fellows of this 

 pair, I find too inconstant to place an}' reliance upon. 



The close similarity' of these shields in the different species 

 probably accounts for the frequent confusion among them by 

 various observers. Thus trigonata has been mistaken for gokool 

 by Ferguson {ReptU. Fauna Ceylon, 1877, p. 21), Phipson 

 {Journ. Bom. Nat. Hist. Sac, vol. ii', p. 247) and Traill (Journ. 

 Bom. Nat. Hist. Soc, vol. ix, p. 499). Gokool was considered the 

 young of cynodon by Cantor {Cat. Mai. Rept., 1847, p. jy). 



The multifasciata of Giinther was confused for a long time with 

 ceylonensis. Stoliczka {Jourii. Asiat. Soc. Bengal, vol. xxxix, 

 p. 199) could not see the justification for considering it a species 

 apart, though Blyth and most herpetologists since his time wholly 

 support Giinther's views. Boulenger {Faun. Bri.. Ind., Rept. and 

 Batrach., 1890, p. 359) did not separate it from ceylonensis, though 

 later {Cat., iii, 1896, p. 69) he too has accepted Giinther's opinion. 



I have for a long time thought that the species ceylonensis and 

 liexagonotus , as regarded by Mr. Boulenger in his Catalogue (1896), 

 comprise more than one form fit to rank as a species, and 1 have 

 been accumulating observations for some years which now enable 

 me to speak with conviction. 



The separation of the species in this genus is mainly, if not 

 wholly, dependent upon the difference in the number of the 

 scale rows, the degree of enlargement of the vertebrals, and the 

 differences in the ranges of the ventrals and subcaudals. 1 think 

 more use may be derived from the scale rows by counting them 

 in two situations instead of in midbody alone. I find that at a 

 point two heads-lengths before the anus the rows are fewer than 

 in midbody, and at least in one instance the counts are useful, for in 

 the two species trigonata and gokool, which are extremely alike, the 

 rows come to 15 in the former and only to 17 in the latter. 



