784 JOURNAL, BOMBAY NATURAL HISTORY SOCIETY, Vol. XV III. 



prone to be thrown into one or more creases resembling sutures. 

 I think in these examples the shield is without doubt entire, except on 

 the right side in one specimen where it appears divided by a suture. 



Lachesis purpvreomaculatus. 

 Three specimens were included in Major Evans' collection 

 with the scales 25 in midbody, and which, according to 

 Mr. Boulenger's Catalogue, should be considered as belonging to 

 this species. I have the greatest doubt of the validity of this species, 

 for I can see no differences other than the number of the costal rows 

 between specimens called by him gramineus and purpvreomaculatus. 

 I have examined three and four times over all the available speci- 

 mens of both in the British Museum collection and again in other 

 Institutions and have taken particularly careful notes of the com- 

 moner varieties with 21 rows in midbody. Again now J have 

 contrasted the three specimens with 25 costal rows, with five 

 others from Burma where the costal rows are 23, and five 

 others with 21 rows. I can find no means of separating these 

 except by the costal rows, and consider that all these specimens 

 belong to one species which is remarkable for the variability 

 of the costal rows, viz., 19 to 27. Specimens in which the costal rows 

 number 25 to 27 should, I think, be considered as variety purpureo- 

 maculatus of the species gramweus, which title claims priority. 



Haplocercus ceylonensis. 

 1 have examined two specimens of this rather uncommon Ceylon 

 snake. I notice that the anterior nasal shield is confluent with 

 the 1st supralabial, and the same feature occurred in the only 

 specimen 1 saw in the Colombo Museum last year. The 4th 

 supralabial alone touches the eye in both specimens. The ventrals 

 and subcaudals are 190 + 42, and 217 + 44. These were both 

 obtained, I am almost certain, at Henaratgoda (Ceylon). 



