252 
THE GRAND CAlfON DISTRICT. 
Conceive now the stream cutting nearly but not quite through the 
second hard group C. The case here is not quite the same as when it was 
cut near to the base of A ; for in the earlier or iiigher stage of the river all 
the debris came from A alone ; while in the present case A, B, and C are 
all furnishing fragmental matter. Hence the corrasion is slackened and a 
small amount of talus may accumulate at the edge of the stream. Still the 
edges of C will be nearly vertical except very near the bottom. 
Finally, conceive the stream to have cut through the soft group D, and 
suppose that at the bottom of this group it remains for a considerable period 
at a base-level The edges of D are steadily sapped and C is undermined. 
But the undermining of C cuts off the foot of the slope in B, increasing 
the declivity in that group and facilitating the descent of fragments. And 
this, in turn, accelerates the rate of weathering in B and the rate of under- 
mining of A. Thus, curiously enough, the state of aflFairs at the bottom of 
the canon influences the rate of recession at the summit of its wall. 
We must now note the fact that a talus acts as a protecting mantle to 
the rocks it covers, screening them partially from dissolution by weather- 
ing. The heavier the talus the greater is the protection. But the amount 
of talus which can remain at any given level is dependent inversely upon the 
slope ; and the talus is always descending by the action of the rains. Hence 
the accumulation is greatest at the bottom of the canon. When, therefore, 
the recession of the wall has gone so far that the descending fragments do 
not fall at once into the stream, the lowest beds of all receive the most pro- 
tection. The rate of recession of these lowest beds, therefore, becomes 
retarded. The protection diminishes as we go higher and the rate of reces- 
sion increases correspondingly. 
It follows at once that the talus is the regulator of the cliff-profile ; for 
it checks the rate of recession in the softer beds, keeping their recession 
down to the mean rate, while, by undermining, the recession of the hard 
beds is brought uf to the mean rate. As soon as the talus is established on 
the lowest slope (D), the cliff may be said to have attained its normal pro- 
file, and in all subsequent recession that profile undergoes little change. 
The only modification it receives is a decrease in the slope D, which 
becomes longer and also takes the form of a curve, concave upwards. The 
