04 STUDIES FOR STUDENTS 
sides are to be worn down; it is with respect to the graded lines 
of a river system that its whole basin is to be worn down. This 
conception, as announced by Powell, is of fundamental impor- 
tance; but it does not seem to gain in clearness or strength by 
expressing the control of erosion in terms of a warped surface, 
guided by the branching lines of the graded river system, instead 
of expressing it distinctly in terms of the branching stream lines 
themselves. Reference will be made again to this aspect of the 
problem further on. 
Not only do graded streams vary in slope in different parts ofa 
river system; the slopes may vary greatly in two neighboring 
river systems at the time of the general establishment of their 
grade. This may be illustrated by considering the unlike con- 
ditions obtaining in two rivers, alike in volume, but one flowing 
through an upland of resistant rocks, the other through a similar 
upland of weak rocks. The first would have to cut down a deep 
valley to a gentle slope before grade was reached ; because its 
load would be slowly delivered from the resistant rocks of its 
valley walls, and high walls would have to be produced by deep 
valley cutting before a balance could be struck between the 
increasing load from the walls and headwaters and the decreas- 
ing capacity of the river. The second river could not cut so 
deep a valley, however weak the rocks of its bed, because it 
would have an abundant load supplied by the rapid wasting of 
its valley sides, even when they were of moderate height, and a 
strong slope down the valley would be required in order to 
maintain a velocity with which the graded stream could bear 
the abundant load away. Only as the whole upland is worn 
down in the later stages of the cycle could the second stream 
wear down its valley to a gentle slope; and then the valley 
would be still shallower than when grade was first attained. 
The principle here considered is clearly recognized by Gil- 
bert, who instances the Platte as a river of the second kind, and 
states that Powell also had so described it (1876, 100). The 
difference between the two kinds of rivers is not satisfactorily 
indicated in terms of baselevel; but it is clearly presented by 
