Mr. Pinchot. 
460 DISCUSSION : FORESTS, RESERVOIRS, AND STREAM FLOW 
The theoretical argument is overwhelmingly in favor of the con- 
tention that mountain forests do cause rain; and a reasonably correct 
view of that contention may be reached by taking an extreme case. 
Take a heavily wooded water-shed, rising from the sea, as for 
instance in California, with prevailing on-shore winds during certain 
seasons. There are two things which lead one to suppose theoretically 
that the rain which does fall, under these conditions, should fall. 
One is the progressive cooling due to the air being forced into a higher 
altitude, resulting in a change in the amount of water vapor which 
the air will hold. The other is the fact that over the forest there 
exists a comparatively cool stratum of air, for some distance up, as 
has been ascertained by definite observations. We should, therefore, 
expect that a rain-cloud, passing from the sea over such a mountain 
range, would be inclined to drop its rain; first, because it was forced 
to cool itself by rising in height, and, second, because it met cool air 
above the forest. What would be the case if that same slope were 
deforested? The same influence of height would be observed, but, 
instead of the cool air above the forest, there would be a body of heated 
air rising from barren rocks and slopes, the whole effect of which 
naturally would be to prevent the cooling, and so to discourage 
precipitation. : 
This extreme case gives the argument as to the effect of mountain 
forests in causing rain. Many observations, some favorable, some un- 
favorable, and some neutral, to this theory, have been taken all over 
the world; but the best prevailing opinion of the men who have 
studied the matter most carefully seems to be that an effect of the 
forest on rainfall does exist, and that it increases rapidly with the 
altitude at which the forest grows. However that may be, the in- 
fluence of the forest in causing rain, whatever it may be, is a matter 
of comparatively small importance compared with its influence upon 
the distribution of water after it reaches the surface. 
Colonel Chittenden has made an unfortunate choice in comparing 
the floods of March 17th to 26th, 1907, in Puta Creek and the 
American River, California. His figures of precipitation for the 
American water-shed on the east are approximately a correct average 
of the various gauging stations of the water-shed. But his average 
precipitation of 22.7 in. for the water-shed of Puta Creek was taken 
apparently from only two stations, both on the extreme summit of the 
Coast Range. One station is not on the water-shed of Puta Creek 
at all, but across the main divide of the Coast Range on the west or 
rainy side, and the other is just inside the water-shed, but at its highest 
point. The larger part of the water-shed which lies below is wholly 
neglected. Fig. 16 is a diagram of these two water-sheds. 
Those who are familiar with California understand that the rain 
clouds from the sea come to the crest of the Coast Range with a con- 
