124 BARBOUR: ZOOGEOGRAPHY. 



granulatus. It is found generally along the coasts of India, southeastern Asia; 

 in the Indonesian area it occurs generally from Sumatra to Ceram and Goram, 

 also in the Philippines and the Pelew Islands. Schlegel (Essai phys. Serp., 

 1837, 1, p. 171) reports its occurrence on the coasts of New Guinea. It is evi- 

 dently rare there, as I was unable to find it; and recent collections do not seem 

 to ha^'e contained it. It is unknown as yet from the coasts of Australia. On 

 January 25, 1907, a single specimen was taken at Buleleng, Bali Island, in salt- 

 water at the mouth of an estuary. Observations on its habits only confirm those 

 noted by Cantor and more recently by Flower (Proc. Zool. soc. London, 1899, 

 p. 679). This is apparently the first record of its capture on Bali, though its 

 occurrence there was certainly to have been expected. 



Fordonia leucobalea (Schlegel). 



Schlegel, Essai phys. Serp., 1837, 2, p. 345, pi. 13, fig. 8-9. Bodlenger, Cat. snakes Brit, mus., 

 1896, 3, p. 21-22. 



Type locality: — Timor. 



A very variable and widely ranging form. An example here is from Penang, 

 M. C. Z. No. 941. 



Known now from nearly all coasts and rivers from Bengal to Queensland. 



Boiga multimaculata (Boie). 

 BoiE, Isis, 1827, p. 549. Boulengek, Cat. snakes Brit, mus., 1896, 3, p. 63-64. 



Type locality: — Java. WTiile Boie does not make a definite statement as 

 to whence his specimens came, it is quite certain that they were from Java. For 

 he credits the name to Reinwardt, and we have evidence from Schlegel (Essai 

 phys. Serp., 1837, 2, p. 265) that this material was collected in Java. Schlegel 

 writes: "Cette espece inedite, egalement due aux recherches du professeur 

 Reinwardt, habite I'lle de Java, ou elle est aussi abondante que la prec6dente 

 [Boiga dendrophila] MM. Boie, Macklot, Kuhl et van Hasselt nous en ont fait 

 parvenir un grand nombre d'indiAadus." 



This arboreal serpent is not uncommon in the vicinity of Buitenzorg. It 

 also seems probable that a large series of specimens would show the examples 

 from Malaya to be subspecifically distinct from those from Burma. Bole's 

 specimens ranged in scale-counts as follows: — V. 202-213, C. 83-93. From 

 Boulenger's data {loc. cit.) and two specimens taken in March, 1907, at Buiten- 

 zorg, it seems that Javan examples average about 206 in ventrals and 85 in 

 subcaudals. Bryant got two at Depok, V. 217-218, C. 86-90. The range is 

 not quite within that mentioned above for Bole's specimens. For the five 

 specimens in the British museum from Burma and Indo-China, we have a much 



