120 



BULLETIN 151, UNITED STATES NATIONAL MUSEUM 



the digital disks of the type he has reversed Nieden's identifications: 

 of C. petersii and utilized the material as types for two new species 

 and lists that there are no examples of C. petersii in the Berlin Museum. 

 The position is best seen by reproducing that portion of Ahl's key 

 which purports to separate the alleged species. 



D 1. Disks very small petersiu 



D 2. Disks rather large, or large, at least half as big as the tympanum — 



E ^ Outer finger one-third webbed macrops. 



E 2. Outer finger with only a rudiment of web at the base 



F '. Tibia contained twice in the bodj' length albescens. 



F 2. Tibia two and a half times in the head and body length 



G '. First finger scarcely smaller than the second; no tarsal fold fasciatus. 



Boulenger certainly stated "disks very small" for petersii^^ and so 

 they are when compared with those of many arboreal frogs and even 

 other members of the genus Chiromantis . In the above series they 

 are about half the size of their respective tympana; it appears that 

 the younger the frog the greater the area of the tympanum covered 

 by its digital disk, though doubtless some variation occurs, for in 

 No. 10376 those on the left are two-thirds the diameter of the 

 tympanum, while those on the right are only half the diameter of the 

 right tympanum. No. 9515, which was named for me at the British 

 Museum as petersii, does not differ from the others in the size of 

 its disks. The distinction between D ^ and D ^ of the key is an 

 imaginary one, based on the statement "disks very small" and the 

 figure of the type. 



One of the Dodoma frogs (No. 12175), which is the largest example 

 (65 mm.) which I have taken, might well be said to have its outer 

 fingers one-third webbed (that is, is macrops of the key); the rest 

 are undoubtedly less than one-third webbed. 



In the Singida frog (No. 10375) the tibia is contained twice in the 

 length from snout to vent (that is, is referable to albescens), while in 

 a Mwanza frog (No. 10376) it is two and a half times and would 

 together with the Mukwese specimen (No. 12175), which agrees 

 with it in lacking a tarsal fold (as do all the series), have to be re- 

 ferred to josciatus. The variation in the tibial length of the series 

 is as follows: 



" In this connect ion Mr. IT. W. Parker writes me (May 25, 1929): "As you suspected the disks of the type 

 are shriveled; I softened those of one hand in KOII and then spread them with needles under the binocular. 

 The transverse diameter of the disk of the third finger measured about 5.5 divisions of my micrometer 

 scale (10 divisions approximately equals 3 mm.) and the tympanum of the same specimen measured 9 divi- 

 sions of the same scale. For comparison the same figures fora larger frog from Dodoma were 8 and 12 mm. 



