76 KEY. A. IRVING ON THE 



carried by mere solution within a few centuries only. It is some- 

 what incomprehensible to me that the significance of the fact that 

 so many of the valley-lakes of the northern Alps lie among strata 

 where conditions specially favourable to subsidence are found, has 

 been so much overlooked by writers in this country. The subject 

 will be found more fully treated in the excellent text-book of 

 Credner, who (p. 211) mentions the Konigsee as a special instance 

 of a lake most likely formed by subsidence. The evidence is indirect 

 but strong. 



Here we are led to the recognition of another agent which, so 

 far as I know, has not been before specified as promoting subsi- 

 dence. It has been pointed out by Credner that the conversion of 

 anhydrite into gypsum by taking up water of crystallization, breaks 

 up a valley-floor under which the more massive beds may have been 

 removed by solution, and thus promotes subsidence. But I am not 

 aware that it has ever been before pointed out that, where underground 

 erosion from either of the causes mentioned had taken place in pre- 

 glacial times beneath the floor of a valley (the natural operations 

 which go on now having gone on then), the dead weight of the enor- 

 mous glaciers * which filled vcdleys thus undermined mag have crushed 

 in their floors. For such reasons I think that the suggestion of 

 Playfair that the bed of the Lake of Geneva had sunk owing to 

 underground erosion, deserves more attention than it has received. 



The connexion between the frequent occurrence of lakes and 

 limestone regions becomes more important when we consider the 

 relative numbers of lakes in such regions and among the crystalline 

 rocks. Something like 100 lakes may be counted (large and small) 

 among the stratified deposits of the Alps, while I do not think a 

 dozen (if we exclude mere tarns) could be pointed out as lying 

 among the crystalline rocks. And if we extend this examination 

 to other parts of the earth's surface, we find a similar rule holds 

 good in Ireland (where tho lakes occur for the most part in the Car- 

 boniferous Limestone), in the Apennines, in the region of the Middle 

 Danube, in the Balkan peninsula, and in Asia Minor. 



Again,- it is possible, reasoning from what we know of solution 

 of the chalk strata at the surface in England by rain-water holding 

 C0 2 in solution, to see how, in some cases, lakes once formed among 

 limestone mountains may have been deepened by a process of che- 

 mical solution. The long exposure of the snow which, on melting, 

 feeds many of these lakes, and the mechanical division of the water 

 during its descent into them from the mountains, are both circum- 

 stances favourable to the saturation of the water which enters the 

 lakes with the gases of the atmosphere. So far as this cause 

 operates, it tends to deepen a lake, though it could only do so effec- 

 tually in a case where the quantity of detritus carried into the lake 

 was exceptionally small. 



Other Lake-forming Agencies. — (1.) It is clear that any cause which 

 leads to a change in the contour of a valley of erosion by disfcur- 



* Such as that of the Etsch-Thal, which filled the valley up to about 150O 

 metres (Credner, ibid. p. 663). 



