1865.] POSXEE AND TOPLEY DENUDATION OF THE WEALD, 471 



side of the transverse valley, and these amphitheatres will extend 

 themselves backwards along the strike, as shown by the dotted 

 lines, fig. 18. Soon we shall have sufficient area to support a 

 brook, and thus we shall get two brooks at right angles to the 

 transverse valley, fig. 19. Fig. 20 shows a section from r to s, and 

 fig. 21 a section from ^ to g. Of course, rain rimning down the 

 slope, h a, will gradually wear off the face of the clay, and under- 

 mine the sandstone. In time the end of the sandstone, 6, will suc- 

 cumb to the never-ceasing atmospheric agencies, and an escarpment 

 will begin to be formed. An escarpment will be formed, and not 

 an even slope, on account of the difference in hardness between 

 the clay and the sandstone ; and the latter will project, because it 

 will suffer less from the action of rain than the clay. In the case 

 we have assumed, there is another element to be taken into consi- 

 deration, besides hardness. The sandstone will soak in a great deal 

 of the rain that falls upon it, whilst every drop that falls upon the 

 clay will produce a certain amount of mechanical erosion. How- 

 ever, where there is a steep slope on the sandstone the rain may 

 produce considerable mechanical erosion, and the face of the 

 escarpment will gradually be worn back, as shown by figs. 22 and 23, 

 The sandstone-plain will also suffer to a certain extent, and its 

 general level will be lowered slightly ; but it wiU suffer much less 

 than the face of the escarpment, as its slope is but small. 



The rate at which the escarpment is worn back will depend on 

 the rate at which the river deepens its valley. It must not be in- 

 ferred from this that the escarpment would not go on wearing its 

 way back, if the stream merely performed the office of carrying the 

 rainwash down into the transverse valley. The escarpment would 

 continue to wear its way back, but the difference in level and, con- 

 sequently, the slope between the edge of the escarpment and the 

 bottom of the valley would constantly be getting less ; if the level 

 of the land remained stationary, the amount of rainwash would get 

 less and less, and in time the slope would get so small that rainwash 

 would not be carried down, and the formation of the escarpment 

 would cease. If, however, the stream at a has an excavating power, 

 which enables it to preserve a certain slope between itself and the 

 escarpment, then the wearing back will always go on. The ex- 

 cavating power of the stream in the longitudinal valley will depend 

 on that of the transverse valley ; and if the sea-level remains con- 

 stant, the transverse stream will go on deepening its bed and lessen- 

 ing its excavating power, until at last it ceases to have any at all. 

 A slight elevation of the land would once more give the transverse 

 stream an excavating power, which in time would be communicated 

 to the longitudinal streams. 



From what we have said it mil be seen that we consider escarp- 

 ments to be due to the difference of waste of hard and soft rocks 

 under atmospheric denudation. When once a transverse valley bas 

 been formed, longitudinal valleys will be formed along the strike of 

 the soft beds, and escarpments will be formed by the hard beds on 

 the side on which the beds dip away from the valley, as in fig. 23. 



