Mr. Willis, 
388 DISCUSSION : FORESTS, RESERVOIRS, AND STREAM FLOW 
snow is invariably to increase its intensity. This results from two 
causes, one affecting the falling of the snow and the other its melting.” 
Following Colonel Chittenden, snow which falls in the open may 
be divided into that which remains evenly spread and that which is 
gathered into drifts; and snow which falls in the forest may be 
divided also into two parts, namely, that which reaches the ground 
and that which hangs in the branches. The snow which is held back 
in drifts and that which is hung up in the branches of the trees and 
returned to the air by evaporation may be said to be withdrawn from 
immediate run-off. The statement is not quite correct in either case, 
but the qualifying considerations are not such as to modify the balance 
which may be roughly struck between the two quantities, when each 
approaches a maximum. There remains in the open an uncertain but 
large proportion which is melted, as all agree, early in the spring, and 
there remains in the forest a considerable proportion which is held 
in the protection of the trees until later in the spring and early sum- 
mer. From any such snow-covered region of moderate altitude, we 
commonly have spring floods, which are produced chiefly by the melting 
of the snow in the open, both from the areas that are evenly covered 
and from the surfaces of the drifts, and subordinately by a contribu- 
tion from the snow that is melting more slowly under the trees. Were 
all the snow in the open, the proportion which would be contributed 
as run-off to the early spring floods would be much larger, and these 
floods would be much more disastrous. Later in the season, the forested 
parts of such regions sometimes occasion severe floods in consequence 
of a large body of snow lingering until the warm winds strike it. 
This effect is undoubtedly unfortunate, but the question may well be 
asked whether it were better to have the June flood on top of that of 
April, rather than divided, as they are likely to be in consequence of 
the presence of the forest. The condition is one in which excessive 
precipitation results in run-off which is beyond the control of any 
natural agency, and high waters will ensue whether or not there be any 
forest. 
In the Northwest there is a special condition prevailing on the 
higher ranges which rise above the tree-line and even into the region 
of perpetual snow, a condition which is described by Colonel Chitten- 
den, and has the effect, that he regards as general. The writer is 
familiar with it in the high mountains of Western Montana, Idaho, 
Washington, and Oregon. The deeply sculptured crags and peaks of 
the mountains above the tree-line produce such eddies in the strong 
winds, that a very large part of the heavy snowfall is gathered into 
drifts. At these altitudes even “chinook” winds are cooled by expan- 
sion as the air currents push upward along the mountain sides, and 
they melt less snow than on lower slopes. Therefore the drifted snow 
banks linger far into or through the summer, continuously feeding 
