216 ILLINOIS ACADEMY OF SCIENCE 
ample. In its upper course it has cut through the entire Mesa 
Verde series and is developing a wide valley in the softer 
Mancos shale. Similarly, Morfield and Prater canyons have 
cut through the sandstone in the upper courses and have de- 
veloped comparatively broad and open valleys. 
The situation of the mesa on the border of the Pla- 
teau Province, and at the foot of the La Plata moun- 
tains is also important. The mesa on the north is bounded by 
a sharp escarpment averaging about 1,000 feet in height, over- 
looking lower country from which the Mesa Verde sandstone 
and the Mancos shale have been removed. This has been ac- 
complished by stripping, a gradual recession of the escarpment 
due to sapping. This stripping is comparatively rapid because 
of the much greater rainfall in the mountain area than in the 
plateau country. The La Platas are regularly mapped with 
the areas having the heaviest precipitation in the state. Their 
drainage is predominantly to the southwest and erosion, there- 
fore, is unusually active on the soft shales at the base because 
of the abundance and clearness of the waters. 
THE CHARACTER OF THE EROSION 
The erosion is predominantly by canyon cutting and by cliff 
recession through stripping. Erosion on the flat plateaus, ex- 
cept by wind, is almost negligible in comparison. Where water 
runs over an impervious surface the result is sheet wash. 
But where the soil is very porous, as on the greater part of the 
plateau, the absorption is rapid enough to prevent even much 
sheet wash. Only where the water drops over the rim rock 
and in the bottoms of the canyons can there be much ero- 
sion. The canyons, therefore, are deep and narrow, the upper 
end usually box-like. The erosion is headward, up slope, 
toward the source of the supply. Since the slope 
is all in one general direction, the growth of the canyons 
is in one direction, also, producing the parallel arrange- 
ment of canyons so conspicuous on the mesa. 
The box-like endings of canyons is especially noticeable in 
the small canyons where the headward cutting now is extremely 
slow. In these small canyons the massive layer is well sup- 
ported on the sides, and the undercutting into the softer layers 
is sometimes several hundred feet. Many of these cavern-like 
