494 DISCUSSION: FORESTS, RESERVOIRS, AND STREAM FLOW 
Mr. Chitten- the Cedar River water-shed is overstated by 20%, and no mention of 
den. the lake is made. It appears, however, that, notwithstanding the unfit- 
ness of the parallel, and the defective data for Cedar River, the illus- 
tration stands decidedly against the argument which it was intended 
to enforce. 
A recent freshet in the Skagit River, in which snow played very 
little part, and in which the rainfall was by no means excessive, still 
further confirms the view that the moderating effect of these forests 
on floods is much less than is popularly supposed. The rainfall at 
Sedro Wooley (the only station on the water-shed), which, for a week 
prior to November 16th, 1908, had been nothing, was 0.46 in. on the 
16th, 1.11 in. on the 17th, 0.01 in. on the 18th, and 0.26 in. on the 19th. 
The corresponding gauge readings were 37, 48, 52 and 47.6 ft.—a total 
rise of 15 ft., the river reaching bank-full stage. The rainfall was 
doubtless heavier in the mountains, though records at three points 
outside the water-shed do not indicate that it was so. Mr. H. L. De- 
vine, weather observer at Sedro, and for nineteen years a resident of 
that place, stated that the absence of snow alone prevented a most 
disastrous flood. The writer cannot account for the fact that such a 
rainfall could produce such a result in such a country, except upon 
the supposition that previous rains had left the forest bed well soaked, 
and also that, through long periods of time, the water had developed 
little channels under the débris through which it flowed away much 
more easily than would seem possible from superficial appearances. 
The Skagit River at Sedro has a rapid fall, and a 15-ft. rise in gauge 
height means a great increase in discharge. 
How well the inhabitants of this region understand the danger in- 
volved in heavy snowfalls in the lower altitudes is well shown by what 
is transpiring at this writing. A week of frequently recurring heavy 
snow storms with light wind had loaded the trees and brushwood 
heavily with snow. (Plate XLVII.) This was followed by rain and 
warm winds, though very fortunately with intermittent cold snaps. But 
the farmers, understanding fully what would happen if the warm spell 
should be prolonged, immediately began strengthening the weak 
places in the river dikes and otherwise preparing themselves for a 
probable flood. 
Proposition 4 is hardly comprehensive enough to cover the entire 
subject of snow-melting as treated in this paper, because it applies 
specifically to the Rocky Mountain region. A more complete summary 
of the general influence of forests upon the run-off from snow-melting 
is given in the following statement: 
Snow has the effect of postponing the run-off from precipitation to 
a period subsequent to that in which the precipitation occurred. If the 
melting is accompanied by rain, the result is to add the run-off of two 
ot more storms together. 
