272 DR. MALCOLM SMITH ON 



Both in their natural habitat, and in captivity, thej' showed a 

 marked pi-edilefition for shallow water. The streams where I found them 

 were small and fairly swift, but it was in the quiet puddles here and 

 there that they usually congregated and could bo observed feeding. 

 Those that I kept spent most of their tima on the top of bricks placed 

 in their tank, where the water was never more than one centimetre 

 deep. Hidden there beneath the leaf of some aquatic plant, they would 

 poke their "funnel" round the edge and so feed. 1 saw the same thing 

 happen under natural conditions. Sometimes they lay quite still, with 

 the "funnel'' expanded on the surface of the water, but not feeding. 



Although when first caught and kept in a jar, they often asnm- 

 ed the vertical attitude as figured by Gadow, 1 seldom saw tlieni in 

 that position when placed in their permanent abode. Nor did I often 

 see them so in nature. Even when feeding in deeper water, their bodies 

 were usually kept in a more or less horizontal plane. 



That the funnel acts as a float, and is of assistance in that way 

 to the creature whilst feeding, is evident, but that its function ever 

 can be to enable its owner to float away upon flood water into safety, 

 as has been suggested, I doubt very strongly. Certainij', at the first 

 signs of disturbance in the water, mine in captivity invariably curled 

 up their floats and sank to the bottom. I never saw mine use the 

 " funnel " as a rasp, as Van Kampen has remarked, and they fed so 

 persistently at the surface, that I imagine this to be the usual method 

 of obtaining food. I quite agree with Annandale with regard to the 

 muscular action ot the structure. 



Of the si.x individuals which I succeeded in bringing home, only 

 two completed their development ; one at the end of ^^ay, and the other 

 a month later. The absorption of the "funnel" took place concurrently 

 with the absoi'ption of the tail, and occupied about ten days. When it 

 was practically completed, the creature left the water. 



Megalophrys pelodytoides Bouleng. 

 I obtained the tadpole of this species early in March on Doi 

 Nga Chang at about 1,000 metres elevation, together with a few adult 

 specimens. At this height the larvae were quite common, but as one 

 descended the hill, they became less numerous, and below 400 metres 

 were not seen at all. Some tan individuals that 1 brought down with 



JOrR.N. NAT. HIST. SOC. SUM 



