1920.] G. A. BouLENGER : Frogs of the Genus Rana. 3 



the bones rugose and more or less confluent with the skin, on the 

 other those in which the nasals and the frontoparietals are reduced. 



Most species agree with the type of the genus, Rana tem- 

 poraria, in having a strong, horizontal clavicle (so-called precora- 

 coid), and a simple omosternum (episternum). But the omo- 

 sternum may be forked at the base, A-shaped, and this I regard 

 as a specialization; whilst the oblique direction of the clavicle, 

 diverging from the coracoid towards the median line, as in a 

 few African species, is certainly a*n approximation to the original 

 condition in the Cystignathidae. The African frogs of the subgenera 

 Hildehrandtia and Ptychadena show a high specialization in the 

 reduction of the clavicles. 



The species with simply pointed terminal phalanges I regard as 

 the most primitive ; a transverse distal expansion, which may be 

 carried so far as to give the bone the shape of a T or a Y, in con- 

 nexion with the scansorial discs, is a departure from the primitive 

 type, and so is the claw-shaped phalanx which is characteristic of 

 Ptvchadena. ' 



Basing our considerations on these theoretical conceptions, we 

 find that Rana hexadactyla is, among Asiatic species, the nearest 

 approach to the ideal prototype : Outer metatarsals separated, fully 

 webbed toes, pointed fingers and toes, distinct tympanum, absence 

 of dorsolateral glandular fold, condition of the skull. In the 

 shoulder girdle, with horizontal clavicles and forked omosternum, 

 it is, however, more advanced than some of its African congeners. 



I therefore start with R. hexadactyla^ from which the series 

 headed b}' R. tigrina and R. grunniens can be directly derived. How 

 other series, corresponding to the subgeneia into which the genus 

 is divided, may be derived from Rana, sensu striclo, is expressed in 

 the synopsis given further on. 



Having explained the principles by which I have been guided 

 in the classification of the species, I need hardly add that their 

 application is not without many difficulties, which I am still unable 

 to overcome. I trust, however, that the present arrangement con- 

 stitutes a marked progress on past attempts. 



A few words of explanation are necessary concerning the 

 method of taking measurements. 



The length of the head and of the snout are taken along 

 the axis, and the posterior extremity of the head corresponds to 

 the articulation of the skull with the vertebral column, which can 

 be approximately ascertained, in the flesh, by feeling with the 

 points of the compasses. The width of the head is the greatest 

 width, at the commissures of the jaws. 



The tympanum,* of which the greatest diameter is given, is 

 compared with the eye, measured along the side of the head. The 

 limbs are measured fully stretched out, the hind limb from the 



1 Cf. Boulenger, C.R. Ac. Sci. Paris, CLXV, 1917, p. 987-. 



2 In a recent paper published in the Records of the Indian Museum (XV, 

 1918, p. 97), Mr. Baini Prashad proposes to substitute the name ' tympanic area ' 



