R. M. Deeley — Erosion hy Rivers and Glaciers. 389 



for the supremacy of fire and water respectively, geologists have 

 come to recognize the fact that the rocky crust of the earth has 

 been caused to reach its present form and condition by the operation 

 of quite a number of agents of change acting both separately and 

 conjointly. But it must not be assumed that any particular agent 

 everywhere and under all conditions produces similar effects, for 

 its work may really be both of a formative and destructive nature. 

 Thus, the sea washes down the cliffs, and then assorts and spreads 

 the debris over the sea-floor. Rivers also, although they erode deep 

 and broad valleys, spread over the lowlands wide deposits of gravel, 

 sand, and brick-earth. 



That there is still much to be settled concerning the erosive action 

 of both rivers and glaciers is clear, for papers treating of these sub- 

 jects frequently appear, the authors of which contend, with some 

 show of reason, that, from a theoretical point of view, the work 

 attributed to these agents is, sometimes, apparently inconsistent 

 with our views concerning their methods of working. More 

 especially is this the case in all matters relating to ice and its 

 work. But ice-work and river-work are so intimately bound up 

 the one with the other, that it is necessary to deal somewhat in 

 outline with the latter if the former is to be made clear. 



Rivers, like other elements of change, are not wholly erosive 

 agents, for we have seen that their valley bottoms, even at points 

 distant from the sea, are often covered by widespread deposits 

 of gravel, sand, and brick-earth, having thicknesses closely related 

 to the depths and volumes of water passing down the rivers by 

 which they were formed. 



In valleys, the falls of which are rapid, the streams cut vertically 

 into the rocks below, forming gorges or ravines having V-shaped 

 sections. In the lower grounds, where the rivers are less rapid, 

 vertical erosion proceeds slowly, the streams impinging against 

 their banks and more rapidly eroding their valleys horizontally 



than vertically. The lowland valleys therefore have the typical i 1 



section. Along the bottom the river winds, changing its position 

 constantly, and here and there cutting crescent-shaped masses out 

 of the hillsides. But the stream rebounds from side to side at 

 every turn, and constantly tends to accentuate the sinuousness of 

 its course by cutting into its banks. The river, however, does 

 not become broader, for the material eroded from the outside of 

 a curve, or brought down by tributary streams, is piled upon the 

 inside of the curve, and the channel, as it were, moves bodily 

 sideways. This filling up of the shallows on the inside of every 

 curve with gravel, sand, and brick-earth, results from the fact that 

 the centi'ifugal force of the water at the surface as it sweeps 

 round a curve is greater than that of the slower current at the 

 river bottom. The flow is consequently diagonally across the 

 bottom, and carries stones, etc., with it, much as the leaves in 

 a saucer of tea are piled up in a central heap when the liquid 

 is stirred. In fact, the mass of water rotates as well as moves 

 forward, sometimes in one direction and sometimes in another, 



