490 DISCUSSION: FORESTS, RESERVOIRS, AND STREAM FLOW 
Mr. Chitten- from 5 to 10 miles wide, extending along the northern border of the 
den. Yellowstone Park, and constituting the greater portion of the grazing 
ground for game. A road extends east and west for 50 miles through 
this section, and 4 miles of it are in a heavy forest. The drifts form 
so heavily in the open section that in many places it is impossible to 
maintain travel except by leaving the road and going around the drifts. 
This trouble is never experienced in the forest. 
Fig. 1, Plate XX XVII, accurately depicts the condition of the 
snow in a dense forest. Someone has hinted that this is a fresh snow- 
fall. <A glance at the left of the picture, where the ground at the foot 
of the trees is entirely bare from sun melting, would have saved the 
necessity of that suggestion. The writer took this picture himself, 
and at the date on which it was taken the snow all over this country 
lay in a solid, nearly uniform blanket, fully 4 ft. thick. Within ten 
days it was nearly all gone. The two pictures in Plate XX XVII will 
bear study. They tell the whole story, and the writer can only say, of 
those who belittle the difference between forest and open country in 
the matter of snow distribution, that they are not familiar with the 
subject. Surely Mr. Pinchot has never “entered into the treasures of 
the snow” in his study of our western water supply, or he would not 
speak lightly of the great reservoirs formed by the drifts of our moun- 
tains. They do more to equalize and prolong the flow of such streams 
than any other cause, except, possibly, the-deep subterranean springs. 
Mr. Leighton is also in error in what he says about the melting of 
snow from underneath. This does not take place to any appreciable 
extent in the Rocky Mountains. The heavy snows—the real accumu- 
lation of winter—come in February and March. Long before this the 
ground under the thin covering of the earlier snows is frozen solid to 
a considerable depth. It does not thaw out during the winter enough 
to cause any appreciable melting of the layer of snow above it. Even 
such melting as does occur soaks into the snow rather than into the 
ground, for the snow is the better sponge. The writer has examined 
this matter too many times to admit of any question upon it. The 
sub-surface melting is practically a negligible quantity, and the win- 
ter’s supply of snow, except such as escapes in evaporation, gradually 
compacts until it reaches a stage of saturation at the time when the 
genuine warmth of summer arrives. 
This condition also prevails, though to a less degree, in the eastern 
deciduous forest, as confirmed by Messrs. Le Baron and Snow. The 
difference in temperature between the forest and open country is not 
great, and while the ground is more protected in the woods than under 
the thin covering of the open, it is not as well protected as under the 
open country drifts. But there is one feature of these forests that ef- 
fectually hinders the assumed soaking into the ground from sub- 
surface melting, and that is the flattened out layer of leaves to which 
