1862.] , EAMSAT GLACIAL OEIGIN OF LAKES. 191 



not stretched, and they occupy a smaller horizontal space than they 

 did previous to the formation of the chain. 



Let us suppose a set of strata of (say) 14,000 to 20,000 feet in 

 thickness, like the rocks of North Wales, and let these be spread 

 out horizontally over thousands of square miles. Let these strata, 

 from any cause, be compressed from the right and left so as 

 to be contorted, and occupy a smaller horizontal area than they 

 did before disturbance. Then, at a great depth, where the super- 

 incumbent strata pressed heavily on the lower beds, the latter would 

 be crumpled up, cleavage would often supervene, and gaping frac- 

 tures would be impossible ; for, where mere fractures occurred, the 

 walls of the cracks would be pressed more closely together. But 

 nearer the surface, where there was less weight, and at it, where 

 there was none, the beds would extend into larger curves than they 

 did lower down ; and where the limits of extensibility were passed, 

 shattering might take place, and yawning chasms might ensue. In 

 all violeDtly contorted countries, however, as in the cleaved rocks 

 of North Wales, for instance, the present surface shows those origi- 

 nally deep-seated contortions that since disturbance have been ex- 

 posed by denudation; otherwise the rocks would not be cleaved. 

 I therefore do not believe that in any country I have seen, such as 

 Wales or Switzerland, there are any lakes now occupying yawning 

 fractures, consequent in Switzerland on post- eocene or post-miocene 

 disturbances. On the contrary, they lie in hollows of denudation, 

 shortly to be explained, of later date than these disturbances. 



Fourthly, again, it may be supposed that the great lakes lie each 

 in an area of special subsidence; but, in reply to this, it is evident that 

 among the unnumbered lakes of Switzerland and the Italian Alps it 

 would be easy to show a gradation in size, from the smallest 

 tarn that lies in a rock-basin to the Lakes of Geneva and Constance. 

 Neither do I see any reason why mere size should be considered the 

 test of subsidence. Disallowing that test, we should require a great 

 number of special subsidences, each in the form of a rock-basin, in 

 contiguous areas. Between the Seidelhorn and Thun, for example, 

 we should require one for the Todten See, several on the plateau on 

 the north immediately under the Seidelhorn, one for the lake at the 

 Grimsel, another for the drained lake at the Kirchet*, and another 

 for the lakes of Brienz and Thun. In Sutherlandshire these areas of 

 special subsidence would be required by the hundred, and in North 

 America by the thousand. 



Signor Gastaldi, in a masterly memoir on the composition of the 

 Miocene conglomerates of Piedmontt, considers with reason that 

 the large angular blocks of these strata, many of them far-trans- 

 ported, and some of them foreign to the Alps and Apennines, have 

 been deposited from ice-rafts; and thence he infers the exist- 

 ence of glaciers during a part of the Miocene epoch. But, admitting 

 this, it is evident that the distribution of the post-pliocene glaciers of 



* See the " Old Glaciers of Switzerland and North Wales." 

 t "Sugli element! che compongono i conelomerati Mioceni del Piemonte." 

 Turin, 1861. 



VOL. XVIII. PAKT I. 0' 



