398 PREHISTORIC EUROPE. 



It is in keeping with the evidence furnished by the Carse- 

 clays that the ancient river- terraces into which they pass as 

 they are followed inland, should betoken a much larger volume 

 for the rivers than these now attain. Not only did they formerly 

 overflow a broader area, but, despite the fact that their course 

 was at that time considerably shorter than now, they yet often 

 had quite a torrential character. This is seen in those broad 

 stretches of high-level sand, gravel, and shingle, with their fre- 

 quent tumultuous bedding, which extend over wide areas in the 

 upper reaches of the Tay above Stanley, and in the Earn above 

 Dalreoch. 



These appearances, taken in connection with the character 

 of the true Carse-deposits, seem to me to point to a period of 

 greater rainfall than the present, and also to a lower winter- 

 temperature. The Carse-clays I consider to be in large measure 

 made up of the fine " flour of rocks " derived from the grinding 

 action of glaciers which then occupied the mountain-valleys of 

 the Highlands, and from which muddy water escaped in large 

 quantities, especially during summer. The melting of the snow 

 and ice, and the more or less sudden disengagement in the warm 

 season of great bodies of water, account very well for the wide- 

 spread and thick deposits of sand, gravel, and shingle which 

 occur in the upper reaches of the valleys. Receiving the tribute 

 of so many swollen glacial streams and rivers, the estuary of the 

 Tay, we may well suppose, would be considerably freshened in 

 its upper reaches ; and the conditions at and for some distance 

 below Perth might thus be quite unfavourable to marine life. 

 It is perhaps for this reason that sea-shells appear to be absent 

 from that part of the old sea-bottom. 



I am thus led to believe that the accumulation of our Carse- 

 clays coincided with a period of local glaciation in our mountain- 

 valleys — a view which we shall presently see is supported by 

 another line of evidence. 



6. The deposits next in succession consist of river-gravel, 

 sand, and silt, which form terraces along one or both sides of 

 the present rivers. They point to a time when the sea gradually 



