270 RANIDZE. 



in Germany. It is, therefore, not serving the interest 

 of exact taxonomy and zoogeography to be satisfied 

 with the comprehensive notion of Rana esculenta. 

 Attempts should be made at a division of the specific 

 type into races or sub-species. With this object in 

 view, I have for the last fifteen years been amassing 

 material and information, and have subjected the 

 many hundreds of specimens which have passed 

 through my hands to a most minute examination and 

 comparison. 



The first attempt at subdividing Rana esculenta 

 into sub-species, published by Camerano in 1881, 

 proved on the whole a failure. In various papers 

 contributed since 1884 I have endeavoured to throw 

 some light on the matter; and have ultimately pro- 

 posed, in 1891, to divide the species into four principal 

 forms, viz. — 



1. Var. ridibunda, Pall, (cachinnans, Pall. ; caucasica, 

 Pall.; tigrina, Eichw. ; dentex, Kryn. ; ? maritima, 

 Risso ; hispanica, Fitz. ; latastii, Gamer. ; bedriagse, 

 Camer. ; fortls, Blgr. ; peres't, Seoane). 



The largest and most widely distributed form, 

 inhabiting the whole of Europe with the exception of 

 the north-western and central parts and Italy, 

 Western Asia as far east as North Baluchistan, 

 Afghanistan, and Eastern Turkestan, and North Africa. 

 As being on the whole the least specialised form, i. e. 

 departing less from the normal pattern of the genus 

 to which it belongs, it deserves to rank first in the list, 

 although the denomination of forma typica pertains to 

 the next form, as having been first described under the 

 name of Rana esculenta. 



2. Forma typica (esculenta, L. ; sylvatica, Koch). 

 Northern and Central Europe and Italy. 



3. Var. lessons, Gamer. 



The habitat of this frosr, which in its characters is 

 intermediate between the preceding and the next, is 



