1862.] 



SMITH SPLIT BOULDERS. 



163 



The block is composed of trap, apparently the same as that of the 

 island, but at such a distance from any neighbouring height as to 

 preclude the supposition that it could have fallen from it. I see there- 

 fore no other hypothesis by which we can account for its present 

 position than that of supposing that it must have fallen from an 

 escarpment of ice. 



We have thus two independent glacial phenomena which belong 

 to a period subsequent to the formation of the forty-feet terrace, 

 showing that the lengthened period of its formation belongs to the 

 Glacial Epoch. 



Sketch of the Sjplit Boulder on Little Cumhra, Western Isles. 



[N.B. In the foreground the shore shows glacial striae.] 



There is yet one circumstance connected with this locality which 

 requires to be noticed. The scratched surface of the ancient terrace 

 passes under the sea ; and although it has been exposed to its wasting 

 action for a length of time equivalent in duration to that of the 

 present sea-level, the striae have not been obliterated. 



Here we have in juxtaposition two distinct cases of the effects 

 of the wasting action of the sea. In the most ancient of these, or that 

 when the cliff and terrace were formed, we have a removal of rock 

 amounting to at least a hundred feet ; in the second, or that of the 

 present sea-level, the amount of wearing away of the same rock 

 cannot exceed a small fraction of an inch. 



I am convinced that no decided change of level has taken place in the 

 West of Scotland during the historic period ; but there may have been 

 small changes : and it is no objection to such a supposition that they 

 have not been observed and recorded ; such changes of level either pass 



