GEOLOGY: C. KEYES 
35 
consequence of the general lowering of the highlands by stream-action, 
while the intermont lowlands are being filled up; because some of the 
best examples of terracing border broad plains having rock-floors. For 
the same reason it does not appear possible that there ever occurs 
during so-called topographic maturity an adjustment by water action 
between one bolson and an adjacent lower one which results in the 
terracing of the higher. There is little or no actual evidence to show 
that bajadas were all formed during periods of glaciation; since some of 
the most typical expressions of these sloping plains are found surround- 
ing low knolls near sea-level, and far below all possible altitudes of 
glacial action in the region. Neither does it appear likely that baja- 
das were built up during interglacial epochs of materials which accumu- 
lated in the mountains when the latter were covered with ice; for this 
does not explain the many cases in which rock-floors are present. Nor 
is it any better to postulate a recent increase of temperature and a 
different distribution and amount of rain-fall abetted by the advance- 
ment of the area in the geographic cycle; for terracing is now going on 
before our very eyes at an astonishingly rapid rate, and as quickly is it 
also completely obUterated. In many localities over-grazing is mani- 
festly a potent and direct cause of the tremendous recent trenching 
and corrading by sporadic storm-waters of the soft temporary soil 
accumulations in the desert. 
In recently setting forth reasons for believing that the gradational 
effects displayed by the intermont plains of arid regions are mainly 
accomplished by means of the winds, I have attempted to point out the 
fact that the action takes place chiefly uphill instead of gravitationally 
down-stream as in the case of running waters. I have also endeavored 
to emphasize the point that the relatively steep slope of the bajada- 
belt represents the highest possible wind-gradients, just as the river- 
bed in a peneplain approaches the lowest possible water-gradient. This 
statement appears to be amply supported by the results arising from 
the artificial diversion of arroyo-courses over the smooth bajadas. 
Were the leveling tendencies of the winds wholly absent from the 
desert regions it is quite possible that the corrasive effects of what 
desert waters there are would be much the same as they are in humid 
lands, differing only in degree. This is well shown in the cases of wing- 
dams that have been constructed to protect lines of railway from the 
disasters of the flood-sheet and the latter has come before the earth- 
works have had time to be leveled by the winds. In one instance in 
particular the culvert and track were washed out in less than an hour's 
time, and a canyon, 75 feet deep, 50 feet wide and several miles long, 
