CHAP. Ill 



VARIATION OF PURPURA LAPILLUS 



91 



is an exceedingly variable species, and in many cases the varia- 

 tions may be shown to bear a direct relation to the manner of 

 life (Fig. 35). Forms occurring in very exposed situations, 

 e.^. Land's End, outer rocks of the Scilly Is., coasts of N. Devon 

 and Yorkshire, are stunted, with a short spire and relatively 

 large mouth, the latter being developed in order to increase the 

 power of adherence to the rock and consequently of resistance 

 to wave force. On the other hand, shells occurring in sheltered 

 situations, estuaries, narrow straits, or even on open coasts where 

 there is plenty of shelter from the waves, are comparatively of 

 great size, with a well-developed, sometimes produced spire, and 

 a mouth small in proportion to the area of shell surface. In the 

 accompanying figure, the specimens from the Conway estuary and 

 the Solent (12, 5) well illustrate this latter form of shell, while 

 that from exposed rocks is illustrated by the specimens from Robin 

 Hood's Bay (13, 14). Had these specimens occurred alone, or 

 had they been brought from some distant and unexplored region, 

 they must inevitably have been described as two distinct species. 

 Mr. W. Bateson has made^ some observations on the shells 

 of Cardium edule taken from a series of terraces on the border 



1 2 3 4 



Fig. 36. — Valves of CartLuia edule from the four upper terraces of Shumisli Kul, a 

 dry salt lake adjacent to the Aral Sea. (After Bateson.) 



of certain salt lakes which once formed a portion of the Sea of 

 Aral. As these lakes gradually became dry, the water they 

 contained became Salter, and thus the successive layers of dead 

 shells deposited on their borders form an interesting record of 

 the progressive variation of this species under conditions which, 

 in one respect at least, can be clearly appreciated. At the 

 same time the diminishing volume of water, and the increasing 

 average temperature, would not be without their effect. It was 



^ Phil. Trans. 1880, vol. 180 B, p. 207. A somewhat similar case (the cele- 

 brated Steinlieim sori' s of PJmuirhis) is dealt with by Hilgendorf, JIB. Akad. 

 Berl. 1866, p, 474 ; and Hyatt, rroc. Amer. Ass. Sc. xxix. p. 527. 



