OP a NS 
DAVIS: THE PLATEAU PROVINCE OF UTAH AND ARIZONA. 37 
of the escarpment, B, as to stand there above the level of the weak 
strata which underlies the strong escarpment maker. Under such con- 
ditions, the future retreat of the escarpment will be relatively slow, 
and its aged form must therefore be less abrupt than at an earlier stage 
of the cycle when the graded floor, A B’, rose only midway in the mass 
of the weak strata; for at this earlier stage, the retreat of the escarp- 
ment must have been accelerated by the sapping of its base, and it must 
then in its vigorous maturity have had a bold and constantly re- 
freshed face. 
Now if the aged region enters a new cycle of erosion, in consequence 
of broad uplift, the main river quickly erodes a canyon and already in 
the youth of the new order of things will be found at a significant depth 
beneath its former channel. In due time, the side strearas will entrench 
themselves beneath the peneplain of the former cycle; a new graded 
floor A” B’' will be opened with respect to the streams, and the subdued 
cliff will be actively attacked as the weak strata beneath the cliff-maker 
are worn away ; the cliff will thus be again steepened or refreshed. 
Figure 13. 
Generalized section of the plateau from the Colorado canyon to the Vermilion cliffs near 
Pipe spring. Abbreviations: —R. W., Redwall limestones; A, Aubrey group; P, Per- 
mian; Sh, Shinarump sandstone; L 7, Lower Trias clays; T, Triassic sandstones. 
Inasmuch as the Colorado canyon is to be explained as the result of 
a new attack of erosive forces on an uplifted peneplain, there seems to 
be equally good reason for explaining also the bold cliffs in which the 
strong Mesozoic strata outcrop north of the canyon as the result of a 
revival of erosive forces on the subdued cliffs that once bordered the pene- 
plain. It is not possible at present to determine the amount of retreat 
that the cliffs have suffered in the renewed attack made upon them in 
the canyon cycle ; but in the case of the Vermilion cliffs at Pipe spring 
the retreat is likely, it seems to me, to have been several miles at least. 
The reason for assigning so considerable a measure to the retreat of 
these cliffs is found in the probability already stated that a large volume 
of weak Permian and lower Triassic strata have been worn off of the 
plateau in front of the cliffs during the canyon cycle. At the beginning 
of that cycle, when the Colorado river, C, Figure 13, flowed at or above 
