62 John Hemtt 



of the ground. Within the chamber, the eggs undergo the whole 

 of their development, moderately damp and cool yet without 

 water. No swinmiing tadpoles emerge therefrom, but the 

 embryos that wriggle their long tails within a capsule of trans- 

 parent jelly may fairly be regarded as incipient tadpoles. E>r. 

 Rattray's report is as follows: — *' We hunted for long days 

 for the tadpoles you wanted (Heleophr^ne) without success, but 

 I found the nests of a frog. They were spherical, about an inch 

 in diameter, just under the surface of the ground and contained 

 about 20 eggs. The whole cycle of development must be com- 

 pleted in the nest. The Hogsback hotel is situated on a narrow 

 plateau at the top of the Tyumie Valley, and my aneroid gave 

 the height at 3,600 ft. The southern top of this plateau is 

 bounded by the steep, forest-clad slope which descends rapidly 

 to the Tyumie valley. It was in the depths of this forest that 

 I found the nests. Three of them occurred on bare patches of 

 ground, but the rest were among a sparsely growing forest grass. 

 They were just as far away from water as was possible in a 

 locality abounding with shrams, for those I found were on a slight 

 watershed between two shrams.*' Writing again in January, 

 1919, at a time of widespread drought, Dr. Rattray reported: 

 '* Nests were not so easily found as last year and it is evident 

 that this toad required a certain degree of moisture, for even 

 the comparative drought which Hogsback has been undergoing 

 plainly destroyed many broods. I could find no newly deposited 

 eggs, but found one or two nests \vith frogs in almost the adult 

 form. I tried a few crude experiments of transferring the 

 embryos to water in the hope of upsetting the normal development 

 and making it revert to that of the order but only succeeded in 

 drowning the specimens." 



This little frog, at that time new to science, was described by 

 the writer under the name of Anh})drophT\;ne rattra^i Rattray's 

 frog belongs to the family RamdaCy including the common water- 

 frogs, and is not directly related to those frogs in tropical 

 America, of the family Cy^stignathidae, which omit their tadpole 

 stage in a similar way. Thus, it would seem that frogs belonging 

 to different families and living in different regions have in- 



