1918.] G. A. BouLENGER & N. Annandale : Rana tigrina. 65 



dorsal coloration is darker and more uniform. The mouth-disk and its 

 armature are closely similar except that the lowest tooth-row on the 

 lower lip is broader and the teeth larger, and that the fringe of papillae 

 is interrupted on the middle below. 



Geographical distribution. — The only point precisely ascertained 

 as to the general range of this species is that it occurs in South Siam, 

 including the provinces of Singgora and Patani in the Malay Peninsula, 

 as well as in Java. It is apparently synonymous with R. scJdueteri, 

 Werner, from North Borneo, but there is a possibility that the var. 

 angustopalmata of van Kampen from Celebes may be distinct, if it is not 

 synonymous with R. vittigera, Wiegmann, from the Philippines. 



III. POST-SCRIPTUM. 



By Gr. A. BoULENGER. 



Dr. Annan dale having most courteously communicated to me his 

 reply to the suggestions contained in the first paper, I will add a few 

 words rather than make any alteration to my original draft. 



As I say in the last paragraph, my opinion on the rank to be assigned 

 to R. cancrivora stands or falls on the question of the tadpole, and as 

 Dr. Annandale appears to have proved his case, I have no further reason 

 to disagree with him, except from the theoretical point of view. 



The old conception of the frog in its development climbing up its 

 own genealogical tree must be abandoned. As I pointed out twenty 

 years ago,^ " larval forms such as the tadpoles are outside the cycle of 

 recapitulation, the ontogeny being broken by the intercalation of the 

 larval phasis. The horny beak, the circular lip with its horny armature, 

 the spiraculum, the enclosure of the fore limbs in diverticula of the bran- 

 chial chambers, and such special adaptations as the ventral disc or sucker 

 of certain mountain forms, clearly point to tadpoles having had a develop- 

 mental history of their own. We need, therefore, not be surprised 

 at occasionally finding, wntliin the same genus, very different types of 

 tadpoles, or even a total suppression of the larval stages, as is actually 

 the case in the large and widely distributed genus Rana." That adap- 

 tational gyrinal polymorphism occurs has been pointed out by Camer- 

 ano,^ and I have myself drawn attention to a very remarkable 

 dimorphism, apparently non-adaptive, in Pelodytes punctatus.^ 



Our progress in the knowledge of the metamorphoses of Batrachians 

 has most decidedly invalidated the prediction of my late chief Dr. 

 GUnther who, in his Preface to my Catalogue of 1882, expressed the 

 opinion that probably the next step in perfecting the system of clas- 

 sification would be marked by a consideration of the larval stages. 



I conclude, from the close agreement of R. cancrivora with the other 

 forms grouped under R. tigrina, that the differentiation of the tadpole 

 has arisen independently from that of the adult, the cuspidate beak 

 and other buccal features of the R. tigrina tadpole being, of course, as 



1 Tailless Batrachians of Europe, p. 110. 



2 Aiti. Ace. Toriru, XXVI, 1890, p. 72. 



3 Proc. Zool. Soc, 1891, p. fil7, pi. xlvii, figs, 1, 2. 



