XIV 



been cut arrow the bedding of any 

 In the Utter case, we should probably meet with in- 

 stances of a conformity Ixrtwccn the external slopes 

 of the mounds and the inclination of the layers of sand 

 of gravel inside \ and this conclusion as to the IKUHUMS 

 retaining the original shape assumed by the loose mate- 

 rials at of their deposit would be amply confirmed 

 if we saw little basins filled with water or with peat, lying 

 between the mounds, for it is evident that had the slopes 

 been due to atmospheric denudation these hollows must 

 necessarily have been filled up, or rather could not have 

 been formed. Mounds of this kind, deriving their pecu- 

 liar forms from the circumstances of their formation, are 

 abundant in the north of Britain and in Scandinavia. 

 Some of them lie on open moors or hill -sides, and on 

 watersheds. These are known as fsktrs, kamts, or <wr. 

 A second kind occurs in valleys among the hills, with 

 the forms of rude crescent-shaped ridges curving from 

 side to side of their valleys, sometimes inclosing small 

 lakes: these are unmistakably the moraine-heaps of local 

 glaciers. A third kind consists of a stiff, stony boulder- 

 clay or earth. The mounds of this type are arranged in 

 one general direction across wide tracts of low ground in 

 Ireland and Scotland, where they are known as drums or 

 drumlins. They belong to the older glacial formations 

 of the country. 



Bouidfrt and Travtlltd Stottts.\n most countries 

 where hard rocks of any kind protrude above the soil, 

 scattered blocks of stone may be observed. When these 

 are of large size, and have no visible rock near them, 

 they are commonly assnmrd to be erratic masses which 



