Mr. Willis. 
392 DISCUSSION: FORESTS, RESERVOIRS, AND STREAM FLOW 
effect of deforestation is to increase conspicuously the irregularity of 
precipitation. 
The principle which underlies the effect of forests on precipitation 
is very simple. Moist air currents rising into the upper strata of the 
atmosphere are dilated and cooled. Their vapor is condensed and falls 
as rain. Any effect which increases the profortion of moisture and 
lowers. the temperature promotes precipitation; any effect which 
decreases the proportion of moisture or raises the temperature lessens 
precipitation. 
It has been shown that forests transpire more water than is 
evaporated from the bare ground, and the experiences of balloonists, 
particularly in France, show that the air is cooler over a large forest, 
even at a height of 1500 m. Sometimes, over hot plains, a shower 
may be seen to begin to fall, but to evaporate before it reaches the 
ground. While these effects are certainly subsidiary to the great move- 
ments of moisture along the cyclonic tracts from the west and south 
toward the northeast across North America, they cannot be said to 
be negligible. 
Man cannot readily change a great plain from a forested region to 
one bare of soil, and when. he substitutes for the trees a vast stretch 
of grain, as has been done in tlhe Upper Mississippi Valley, he does 
not change the conditions of radiation so seriously as to affect the 
powerful rain-bearing currents which sweep up from Texas and the 
Gulf. The case is entirely different, however, in regard to mountain 
ranges. It is easy, with the aid of the rills and torrents, to remove 
their covering of vegetation, and they can be denuded so completely 
that the radiation from bare rock replaces that of forest for an entire 
mountain range. The precipitation over mountains is always less 
regular than that over plains, for the irregularities of the mountain 
slopes cause the air-currents to rise and expand and cool very un- 
equally; and when these conditions are exaggerated by removing the 
regulating effect of the tree surface and by substituting the aggravat- 
ing effect of the bare surface, conditions are produced that result 
in more violent floods and more excessive drought than existed before 
Man did his irrevocable work. 
“The commonly accepted opinion is that forests have a beneficial 
influence on stream flow: 
“(4) By preventing erosion of the soil on steep slopes * * *.” 
It is difficult for the writer to understand how any observer can 
say, as Colonel Chittenden does,* that there is a deficiency of evidence 
to support this view. As a geologist, the writer’s attention has been 
constantly drawn to the evidences of erosion under the varied condi- 
tions of relief and climate in nearly all parts of the United States 
* Page 270. 
