520 DISCUSSION: FORESTS, RESERVOIRS, AND STREAM FLOW 
Mr. Guitten: about 8 in. The trees are possibly from 15 to 30 ft. high. When it 
is considered that in the native forest the trees are often 200 ft. high, 
and so thick that their tops everywhere interlace, and that there are 
trees of all heights and sizes, with dense undergrowth extending from 
the ground up to heights of 20 ft. or more, he will readily admit that 
the snow will find difficulty in reaching the ground, and that such 
extensive shelter will readily catch and hold temporarily, at least, a 
foot of snow. 
But does Mr. Pinchot think it quite fair to pass from this simple 
illustration of a single storm to a general proposition of “two-thirds 
of the snow on the branches and one-third on the ground”? Does he 
really fancy that the writer believes that, of a winter snowfall in these 
mountains of, say 9 ft., 6 ft. remains on the trees? 
To go into the elementary principles, the ability of snow to cling 
to any object depends very much upon its quantity, its condition, and 
the state of the wind. Snow in large flakes, and damp and sticky, 
such as is common in many of the storms of this region, will cling tc 
almost anything, and the degree to which it will pile up on slender 
supports in the absence of wind is extraordinary. On the other hand, 
when the temperature is well below freezing and the snow is dry and 
fine it will slip from any inclined surface or through a mesh of under- 
brush as fine, dry sand slips through the fingers. If such snow falls, 
as is nearly always the case, under the influence of the wind, which 
shakes such movable objects as tree tops, almost all of it will sift 
through to the ground. Moreover, as to proportions on the trees and 
ground, a great deal depends upon quantity. A much larger per- 
centage of a 2-in. snowfall will cling to the trees than of a 20-in. fall, 
other things being equal. 
Both Mr. Pinchot and Mr. Leighton have apparently derived much 
comfort from the writer’s comparison of the American River and 
Puta Creek water-sheds, in California, and both have referred to it 
as an “unfortunate” illustration. Let us see on which side of the 
question the misfortune of that illustration lies. The writer presents 
a map, Fig. 17, of the two water-sheds, prepared after he had read the 
report of Mr. Pinchot’s remarks and before he had seen his map, Fig. 
16. <A careful comparison of these two maps is invited, with particu- 
lar attention to the location of the gauging stations, the outline of the 
Sacramento flood plain, and the direction of the wind as shown on the 
writer’s map. Mr. Pinchot places Guinda (71) on the Puta Creek 
water-shed, but, unless all the maps which the writer has consulted 
are wrong, it is on Cache Oreek, quite outside the Puta Creek water- 
shed. The writer supposed that Calistoga (76) was within the water- 
shed, but it appears to be just outside, though practically on the bor- 
der. There is a discrepancy in the position of Tamarack on the 
American River water-shed, but it is immaterial. 
