


316 STRUCTURAL AND FIELD GEOLOGY 
by which the great mers de glace were tunnelled, especially 
during the period of their final dissolution. In the low- 
lying parts of the country, many wide sheets of sand and 
gravel seem also to have been accumulated underneath the 
melting ice-flows, for they are often closely associated with 
eskers. In other cases, however, they may have been dis- 
tributed over the exposed surface of the low lands by water 
escaping from the gradually disappearing snow-fields and 
decaying glaciers of adjacent high grounds. 
In hilly and mountainous tracts, narrow and_ broad 
terraces and considerable plateaus of gravel, sand, and clay 
obviously mark the sites of ancient glacier-lakes. Such are 
the Parallel Roads of Glenroy, the wide sand-plains covering 
the watershed between the rivers Avon and Irvine in the 
neighbourhood of Loudoun Hill, and many similar terraces 
and flats occurring in the Northern Highlands and Southern 
Uplands. Even the hill-slopes overlooking the broad low- 
land tracts of Scotland now and again show strong evidence 
of the former presence of temporary glacial lakes, which 
appear to have come into existence after the hills in ques- 
tion had been divested of their icy covering, and while the 
adjacent lowlands were still thickly mantled by the gradually 
decaying mer de glace. 
There seems also to be little doubt that those tumultuous 
assemblages of hummocks, cones, and ridges known as kames, 
are of the nature of gravelly moraines, deposited in front of 
giant glaciers or district ice-sheets. Often associated with 
them are wide sheets of sand, loam, and clay, which spread 
out over the low-lying tracts, upon the borders of which the 
gravelly moraines have been accumulated. Perhaps one of 
the best areas for the study of these phenomena is the great 
valley of Strathmore. 
Although the external form of glacial and fluvio-glacial 
deposits is often original, yet occasionally widespread sheets 
of sand, gravel, etc., have been so cut up by subsequent 
epigene action as to present a more or less rapidly undulating 
or corrugated surface. When this is the case, such a denuded 
plain now and then simulates the appearance of “drums” 
and “kames.” Usually, however, the observer is not likely 
to mistake a surface of erosion for one of accumulation. If 

