we suppose them u> have (wen originally perfectly hori- 



xontal, they rau*t kuve 1* down against the 



lopes of a mountain about four and a half miles high ; 



sloped gently away I'nmi tin- underlying rocks, 



the height of the m have been still greater. 



Hut nut only must a stupendous mountain have been 



round so as to lie on its side ; the whole of the 



rocks must have been removed except a narrow 



cake cut across the bedding parallel with the original 



slope of the mountain. In reality there are not two 



. i p f 



sets of rocks in the line of section. The whole is one, 

 subject here and there to local crumpling and faulting, as 

 might have been seen by more careful and extended 

 observation. 



Where no satisfactory evidence can be obtained of the 

 stratigraphical relations of two groups of rocks, that is, 

 where neither t them exhibits its dip and strike in such 

 a way as to show whether or not they arc uncon- 

 formable, proofs of a discordance between them can 

 sometimes be found in the presence of conglomerates 

 in the one scries derived from rocks that occur in place 

 in the other. This kind of proof may undoubtedly in 

 many cases be taken to establish an unconformability. 

 Hut it is not always sufficient For example, a vc 

 conglomerate derived mainly from the detritus of an 

 underlying lava may occur in a perfectly conformable 



