1918.] N. Annandale: Molluscs of the hde hale. 151 



shallow- water L. auricularia, from which, according to Forel,i the 

 deep-water " species " are derived. 



The cases of L. shanensis and L. mimetica are not altogether parallel 

 to that of L. andersoniana , for we have not here a difference induced in 

 the shell by a change from still to running or running to still water, 

 but rather, at any rate in L. shanensis, an apparently gradual evolu- 

 tion, the physical changes in environment connected with which are 

 complex and therefore obscure. I have already pointed out that phase 

 D of L. shanensis and also L. mimetica have a considerable resemblance 

 in shell-form to the deep-water Swiss species or varieties L. foreli and 

 L. abyssicola. Swiss conchologists agree with Forel that these deep-water 

 Limnaeae are derived from the common shallow-water L. auricularia, a 

 species of which the geographical range extends as far south and east as 

 the great plateau just north of the Himalayas and even into Kashmir. 

 There is in the collection of the Indian Museum a fine series of shells of 

 this species from various localities on the Pamirs. I have found among 

 these shells numerous specimens of a phase that is abundant in small 

 streams at high altitudes in Central Asia. Like the typical form of the 

 species, it has an expanded mouth to the shell, but this feature is not 

 nearly so pronounced as in the forma typica. The shell is small compared 

 with that of the forma typica, although somewhat larger than that of 

 L. bowelli, with w^hich I propose to compare it as being also a running- 

 water form from high altitudes on an Asiatic plateau. L. auricularia 

 is a very plastic species and a large number of varieties have received 

 names ; I select one (var. andersoni, Clessin ^) which seems to bear much 

 the same superficial relationship to the Pamir form as phase A of Z. 

 shanensis does to L. bowelli. It was found in the Gulf of Bothnia in salt 

 or brackish water and differs from the Pamir phase in that the shell is 

 considerably narrower and its mouth less patent. I have arranged 

 outline figures of the Pamir race, of the variety andersoni and of //. 

 foreli side by side with those of L. bowelli and phases A and D of L. 

 shanensis. It is easy to see how closely parallel the differences and 

 resemblances are. In all cases the lowest figure is much more highly 

 magnified than the other two. 



L. mimetica, so far at any rate as the shell is concerned, seems to bear 

 much the same relationship to the forma typica of the common Indian 

 L. acuminata, Lamarck, an extremely plastic species, as phase D does 

 to L. bowelli, and among the varieties of L. acuyninata it would be 

 possible to select at least as complete a series of intermediate forms as 

 in the case of L. auricularia and L. foreli. Extreme types of shell in 



1 Le Leman, III, p. 102 (1900). See also Clessiu, MalaJcoz. Blait. XXIV, pp. 171-177. 

 pi. iii, figs. 1-4, 8-9. More recently Roszkowski {Zool. Anz. XL, p. 375) has demonstrated 

 that the genitaha of L. foreli agree with those of L. cvata rather than those of L. 

 auricularia. The bearing of this fact on the present inquiry is rendered loss evident 

 by the existence of many phases intermediate between L. ovata and L. auricularia so 

 far as their shells are concerned. Indeed, Bollinger apparently claims to have found a 

 complete series of shells linking the two together. It may be, therefore, that the true 

 specific distinctions are anatomical and not conchological. In any case the facts as 

 represented by Roszkowski — I have been unable to consult Bollinger's work — have Sj 

 distinct bearing on my remarks in the next paragraph. 



a Clpssjn, Malakoz. Blatt. XXV, p. 73, p!. iii, fig. 8 (1878). 



