396 H. M. Beelejj — Erosion hy Rivers and Glaciers. 



takes place and the temperature is low, but if the rock be dense 

 and impervious the water cannot readily escape. However, on the 

 top and lee side of the boss the pressure is relieved, the water 

 escapes through grooves, joints, etc., and the ice freezes to the 

 rocks with which it is in contact, tearing up, if the joints allow it 

 to do so, large masses which are then carried forwards. The rate 

 at which erosion is effected in this manner must depend largely 

 upon the direction of the dip and the jointing. In many valleys, 

 however, the rocks are so massive and impervious that they have 

 succeeded in largely resisting the erosive power of the ice. But 

 even in the absence of large hummocks in the path of the glacier 

 the rocky floor must always be so irregular that the local variations 

 of pressure must give rise to local freezing. In this way the whole 

 forward stress of large masses of ice is concentrated locally, and the 

 hard floor is dealt with in detail. When we remember that a glacier 

 may at one time be moving over a bed of sand and slowly drag 

 along with it an inch or two of its surface without in any way 

 disturbing the lower portions, or glide easily over a dense tough 

 clay, and in another may meet some obstruction, be compelled to 

 move or more or less change its course, and displace the obstruction 

 to a greater or lesser extent, it is clear that we must expect to meet 

 with all sorts of conflicting evidence concerning the real nature 

 of the work glaciers accomplish. Indeed, in the study of glaciated 

 districts we find every kind of action : in one place rocks are 

 polished and scratched, in another deeply grooved, or they may be 

 broken and bent over at a sharp angle at their upper surfaces. 

 Soft clays are bent, crumpled, and contorted, striated and polished 

 stones kneaded eight or more feet deep in marls and clays, and 

 masses of older rock drawn up and buried in the beds above. 

 The Cromer cliffs exhibit splendid examples of this action. 



Indeed, just as we find that the deposits formed by rivers vary 

 with the rate of flow, the nature of the rocks they erode, and many 

 other circumstances, and that they are sometimes acting as denuding 

 and at other times as depositing agents, so it is reasonable to argue 

 that the effects wrought by glaciers must also depend upon their 

 rate of flow, their thickness, the nature and surface contour of the 

 rocks over which they move, whether they actually enter the sea 

 or lakes, or terminate on land, etc., etc. 



Towards its extremity, where it enters the warm air of the low- 

 lands, the water on the surface of a glacier finds its way through 

 crevasses, or melts its way down the glacier-side to the bottom of 

 the ice-filled valley, and then flows through a wide tunnel until it 

 emerges into the open. The tunnel-sides are of course constantly 

 closing in, and the stream by melting its retaining-walls is con- 

 tinually moving its position. In this way boulders find their way 

 into the stream, are well rounded, and finally ejected along with 

 sand and fine mud. 



Outside the tunnel at the end of the glacier the volume of the 

 stream is seldom sufficient to carry away all the coarse debris, 

 which, therefore, collects at the end of the glacier as an enormous 



