Mr. Smith, 
3864 DISCUSSION: FORESTS, RESERVOIRS, AND STREAM FLOW 
contribution through the underground circulation will continue, just 
as long as the forest bed retains any portion of the rainfall. 
The author’s view, that delayed melting of snow is not beneficial, 
must also be challenged. While the writer too has found dry ground 
beneath a dwindling snow bank, on digging into the sand he has dis- 
covered the water beneath. Again, it seems probable that the bottom, 
as well as the top, of the snow cover is also a melting surface both 
in the open and in the forest. This is shown by the manner in which 
certain spring flowering plants penetrate the snow wherever the cover 
becomes thin. Thus it appears reasonable to believe that the time of 
snow-melting is a period of contribution to the underground flow, 
and the total contribution from snow is necessarily largest where condi- 
tions favor the longest period of melting. 
On the subject of the influence of forests on snow-melting, obser- 
vations will differ somewhat. As contrasted with Colonel Chittenden’s 
excellent and detailed observations in the Yellowstone Park, the writer 
might mention the fact that even later in the same season, July Ist, 
1899, on Table Mountain, in Central Washington, there was snow in 
the forest at an elevation of 6000 ft., while, at the same elevation in 
the open, the snow had long disappeared, and: the flood conditions on 
the Yakima River had. passed. At one place in the forest on Table 
Mountain, the drifted snow was so deep on this date that the writer’s 
horse was stalled and had to be dug out, so that, as in Colonel 
Chittenden’s case, the difficulties attending travel forcibly drew atten- 
tion to the depth of snow. Furthermore, the abundant springs of ice- 
cold water below the escarpment of Table Mountain, which persist 
through the summer, suggest that the snow banks in the forest above 
were contributing to the underground supply rather than to direct 
run-off. 
The illustration, Fig. 1, Plate XL, is of a stream (White River, 
Washington), the water-shed of which is not aptly described as “the 
most densely wooded and perfectly protected water-shed in existence.” 
White River is a glacial stream, as its name indicates, and has a well- 
deserved reputation for devastating floods. It has its principal source 
in the largest glacier on Mount Rainier, and, as the writer knows from 
observation, 23 sq. miles of the Mount Rainier portion of its water-shed 
are not forested, but are covered with ice and snow twelve months of the 
year, not to mention other snow-covered areas near the crest of the Cas- 
cades and at least 50 sq. miles of burnt and cut-over land in the lower 
portions of its basin. This fact explains why a warm “chinook” will 
cause the White to rise much more quickly than its sister stream, the 
Green, which, as its name suggests, has no glacial tributaries. The 
ranchers in the Lower White River Valley recognize the advantage of 
the high-water stage of the Green and Cedar Rivers ‘reaching the valley 
after the maximum flow of Upper White River has passed. The Skagit, 
