194 STRUCTURAL GEOLOGY. 



course is through indurated and homogeneous rocks, its channel is from 20 

 to 50 feet in width, and from 1,500 to 3,000 feet in depth; it runs through 

 a narrow but profound gorge. But when it passes a well defined geologi- 

 cal horizon from these coherent into extremely incoherent beds, its channel 

 abruptly widens, and the stream is a broad sheet of water many hundreds 

 of yards in width and but a few inches in depth. Again this condition is 

 well illustrated in the Platte River where it crosses the Plains. Here the 

 beds through which the river runs are incoherent, and although the river 

 has as great a fall as the Colorado through the plateaus, and although the 

 climatic conditions are essentially the same, yet the former runs in a broad 

 sheet scarcely below the level of the plain, while the latter runs in a narrow 

 groove at profound depths below the general surface. Thus it is that the 

 streams, though they may have great fall, and though the stream beds may 

 be of material that can be rapidly transported, yet they do not succeed in 

 excavating deep channels, for every particle taken from the bottom is re- 

 placed. Nevertheless general degradation is carried on by such streams at 

 a rapid rate, but not at a maximum rate, for the water permeating the sands 

 on either side is steadily and rapidly evaporated, and is thus lost as an agent 

 of degradation. There are many streams in the arid region of America, 

 running alternately through harder and softer beds, which are continuously 

 degrading the coherent rocks and intermittently degrading the incoherent 

 rocks ; that is, the streams are ever living where the beds are coherent, but 

 when they reach the sands the waters sink and re-appear where the beds 

 below are harder. Through these harder beds the streams caflon, and 

 through the softer beds low plains stretch either way from the course of 

 the stream ; ' and it is only during flood time that the channels are cut across 

 these plains from canon to cailon. 



But dropping these exceptions, all of them interesting cases, let us return 

 to the main argument. We have seen that ever are the effects of declivity 

 in degradation multiplied directly and indirectly. Wherever the rate of 

 degradation by any one method is increased it is due chiefly to increased 

 declivity, and wherever the rate of one method of degradation is increased, 

 the rate of all other methods is increased. 



We may not be able to give mathematic expression to rate of degrada- 



