DISCUSSION: FORESTS, RESERVOIRS, AND STREAM FLOW 493 
of the river just before the storm began was about 1000 cu. ft. per sec. mr. Chitten- 
It should be said that the figures for rainfall are made up from 9% 
two stations, the only ones available, viz., Cedar Lake, west of the 
center of the water-shed, and Lester, just south of the eastern por- 
tion. Both these stations are in valleys, and probably do not represent 
accurately what took place in the hills. But the explanation of so large 
a freshet compared with the rainfall is in any event obvious. The 
month had been very rainy and the storage capacity of the water-shed 
was exhausted. There was a considerable quantity of snow in the hills, 
both on the ground and on the trees. The temperature was high, and 
the water from the snow came down with that from the rain; so that, 
in spite of the heavy layer of humus over-spreading the whole water- 
shed, and in spite of the presence of a’ considerable lake which has a 
moderating effect on freshets, the discharge rose and fell quite as 
rapidly in response to precipitation as on an eastern stream in the 
hilly country where the forests have been cleared away. Plate XLVI 
is a view of the water-shed. 
Cedar River is the stream which Mr. Pinchot, in the January, 1908, 
number of the Annals of the American Academy of Political and 
Social Science (devoted to the subject of American Waterways), has 
compared with Owen Creek, Arizona. Both streams are given as hav- 
ing a water-shed of about 173 sq. miles; that of Owen Creek is barren 
and treeless, with much rock exposed, and the water flows off as from 
a roof. Little rain falls, and there is no record of the precipitation that 
caused the heavy freshet of 1896 which Mr. Pinchot cites; but, from all 
experience in that country, the storm was evidently in the nature of 
a cloudburst. There was a record of only about 15 in. of rainfall for 
the year at the nearest Weather Bureau station. The freshet cited 
was estimated at 9000 cu. ft. per sec., or a little over 50 cu. ft. per sec. 
per sq. mile. 
The data given for Cedar River in the flood of 1897 were as follows: 
Rainfall, 93 in. for the year, estimated at 150 in. in the mountains; 
maximum flood flow, 3 601 cu. ft. per sec., or 21 cu. ft. per sec. per sq. 
mile. 
Of course any such comparison, as illustrating the effect of de- 
forestation, is wholly inapplicable. It would even be as fair a com- 
parison to take the City of Seattle, with its roofs and pavements. A 
legitimate method would be to compare a forested area with a de- 
forested area in the same kind of country. The comparison as given 
omits all reference to distribution of rainfall, yet in Phenix, in 1896, 
rain fell on only 51 days while in Seattle and vicinity it falls on the 
average 160 days in a year. Moreover, the flood flow of Cedar River, 
as stated, is evidently wholly in error, for the flood of 1897 is locally 
considered to have been greater than that of 1906 when it reached 
10 800 cu. ft. per sec., or 75 cu. ft. per sec. per sq. mile. The area of 
