ON T THE DESTRUCTION OF ROCKS. 241 



marked ; in which cases they must be referred to the 

 effects of water alone. 



As some of these changes are, from their extent or 

 other circumstances, of a very interesting nature, and 

 may perhaps serve to explain some difficulties in 

 geology, it will be necessary to bestow a few words 

 on them. 



This event is so common in granite, that examples 

 of it must have occurred to every geologist. It is 

 frequent in Cornwall and on the opposite coast of 

 Britany ; and the result is a clay mixed with quartz 

 gravel, and, in some cases, where the mica has re- 

 sisted decomposition, with that mineral also. In 

 some kinds of gneiss, it is equally common ; and re- 

 markable examples of this nature may be observed in 

 Aberdeenshire, in Guernsey, and in the isle of Sky. 

 If less frequent in micaceous schist, it is still suffi- 

 ciently conspicuous in that rock, in many parts of 

 Scotland ; as, in Cornwall, it is of very noted occur- 

 rence in the soft argillaceous schists, or the " killas," 

 of that district. That the white, orpure, sandstones 

 undergo this change in the same manner, is proved 

 by the occurrence of a hill of considerable extent near 

 Kildrummie in Aberdeenshire. This hill consists, al- 

 most entirely, of loose sand ; and thus its original 

 rocky state might be disputed, did not the perma- 

 nence of numerous harder veins that intersect it, and 

 the remaining marks of the joints which divide the 

 beds, prove its true nature. It has often been 

 doubted whether the sand beds in England, which 

 contain sandstone and sand together, were in the pro- 

 gress of decomposition or induration ; and it has, 

 I believe, been generally decided in favour of the 

 latter. The present fact may throw some doubt on 

 the propriety of that decision, and will at least ren- 



VOL. i. R 



