PAPERS ON GEOLOGY AND GEOGRAPHY 217 
openings were used by the ancient inhabitants of the region as 
homes, who built walls across the front, appropriating the up- 
per massive rack as a roof. This is true of all the canyon ruins, 
the best example being Cliff Palace. 
The other characteristic method of erosion in the plateau 
country is by cliff recession through stripping. The general 
degradation of the Plateau Province is accomplished predom- 
inantly by the erosion forces acting upon the almost vertical 
edges of the strata where weathering, and wash, and wind is 
especially effective. This sapping has gone on to an unusual 
degree on the northern escarpment of the mesa where condi- 
tions are very favorable. The almost vertical escarpment of 
about five hundred feet of Mancos shale, capped by a heavy 
sandstone, gives almost ideal conditions for cliff-recession. 
As rapidly as the softer shale is carried away below, huge frag- 
ments of sandstone slump off, making immense talus heaps 
Which in the aggregate resemble, somewhat, landslide topog- 
raphy. These heaps are comparatively short-lived because there 
is so little vegetation to protect them. The present mesa is 
only a skeleton of what it was originally. It has been cut up 
into many narrow strips by deep canyons which roughly par- 
allel each other. These canyons starting at the Mancos in the 
south have gradually worked their way headward until they 
have cut through the entire present length of the mesa. This 
cutting has gone on to such an extent that the canyon areas are 
about equal to that of the remaining plateau surface. From the 
standpoint of travel the region is as rough as most mountain 
areas of the west. The canyon walls are the dominating ele- 
ments of the relief, a relief produced by downward departures 
from the general level, instead of an upward departure as is 
common in mountainous areas. This gives an uncommon type 
of topography. None of the canyons at the present time have 
running water except during very severe showers. They are all 
aggrading in the main and show considerable fill. Wells sunk 
to the bottom of this loose material usually have sufficient 
water to supply the stock foraging in the upper canyons. These 
wells show from thirty to seventy feet of loose wash material, 
according to Mr. Prater, who had the wells sunk. If his report 
may be relied upon, and we believe it may, the canyons for- 
merly were much deeper than now and must have carried much 
more water. This points either to a much more arid condition 
