igi2.] N. Annandale : Batrachia. 29 



form of irregular spots and. veinings. I have found every stage in 

 the metamorphosis of this tadpole into M. parva, Boulenger. 



A second tadpole of identical structure but slightly different 

 colouration, larger size and more slender form also occurs in 

 the E. Himala^'as and was taken in considerable numbers in the 

 Abor foot-hills. It differs from that of M. parva in lacking the 

 dark markings on the sides or at any rate in having them much 

 less conspicuous and consisting only of minute spots. I believe 

 that this larva, which I have seen from the N. Shan States, is that 

 of M. major, Boulenger, but have only circumstantial evidence in 

 support of this belief. 



A fourth Megalophrys tadpole was taken by Mr. Kemp in the 

 Abor foot-hills. It differs from the one I have assigned pro- 

 visionally to M. major in being of an almost uniform dark brown 

 colour both above and below. There is no evidence that this is 

 the tadpole of M. kempii^ but it ma^' be stated that a very similar 

 larva was taken b}' Mr. F. H. Gravely on the Siamese frontier of 

 Tenasserim near the Dawn a Hills in company with a 3'oung frog 

 that apparentl}' represents M. heteropus (Boulenger), a species 

 hitherto known from the hills of the Malay Peninsula. 



It is a remarkable fact that whereas the tadpoles of M. has- 

 seltii (Tshudi), a Malayan species, are of typical Pelobatid struc- 

 ture, those of other members of the genus from the same region 

 are not only extremely different from that form but also extremely 

 close to one another in structure. Were it not proved that the 

 common Darjiling tadpole is the larva of M. parva, it would be 

 difficult to believe that it was specificallj' distinct from that of M. 

 montana, although there is no difficult}'^ in distinguishing the 

 adults. Seeing that these two larvae belong to different species, it 

 is evident that the larva of other species must only be identified 

 with great caution. 



Part II.— BIOIvOGlCAI.. 



Of all the frogs and toads taken in the Abor foot-hills onh^ 

 two {Rana limnocharis and Bujo himalayanus) entirel}^ lack ad- 

 hesive digital disks, and we may take it that the majorit3^ of the 

 other species are normallj^ arboreal in habits. Some few of them^ 

 how-ever, seem rather to be rupicolous and to live among stones at 

 the edge of streams. It is probable that Micrixalus horealis be- 

 longs to the latter category, but Ixalus tuherculatus , although 

 individuals w-ere actually found at the edge of a stream, is know^n 

 also to frequent tree- trunks. Phrynoderma moloch apparently 

 affects dead tree-trunks and lays its eggs in or near small masses 

 of water that accumulate in them. In the dense jungles of the 

 ^Malayo-Himalayan tract most of the Batrachia are as a rule 

 arboreal and Mr. Kemp's collection in this respect is no exception. 

 It may be noted, further, that the great majority of his specimens 

 are protectively coloured and resemble either the lichen-clad bark of 

 jungle trees or the green leaves of their foliage. 



