DISCUSSION: FORESTS, RESERVOIRS, AND STREAM FLOW 463 
hearing on House Resolution 208, before the Committee on the mr. Pinchot. 
Judiciary, 60th Congress, Ist Session, it is shown that the Ohio, 
Cumberland, Wateree, Congaree, Savannah, and Alabama Rivers, 
disclose a marked increase of floods and flood duration. A more 
recent examination of the records of flow for the Allegheny, Mononga- 
hela, Muskingum, and Potomac Rivers shows precisely the same trend. 
These streams do not show general increases in the extreme height of 
floods, but rather a marked tendency of the waters to rush away in many 
sudden violent floods of short duration, which is to be expected in the 
case of barren slopes. Most streams of which we have records show 
a longer low-water period now than formerly, though the tendency is 
not so marked as the tendency toward increased floods. Where such 
a tendency is not in evidence, it may possibly be due to the geological 
formation of the water-shed, or to the filling of the stream channels, 
so that the general water level is raised. 
In contradistinction to these streams, which have mountainous or 
hilly water-sheds, the Red River, whose upper basin is entirely in the 
plains and in a region which has largely been brought under the 
plow in the last 15 years, shows both diminished floods and greatly 
improved low-water conditions during the past 16 years. This proves 
the case over again, only from another point of view. Surface condi- 
tions over a water-shed clearly have great influence on the flow of the 
stream which drains it. 
Fundamental errors appear to underlie the author’s discussion of 
influence of forests on snow-melting. The first is that the ground 
does not take up the water from the first melting of snow. The author 
says: “The water from the first melting of the snow blanket does not 
sink into the ground, but into itself. * * * The author has seen 
an 8-ft. covering of snow dwindle to 2 ft., with the ground beneath 
it still comparatively dry.” Does not evaporation and rapid drainage 
into the ground account for this? In a forest, especially in a mountain 
forest, the snow usually falls on unfrozen ground, and the natural 
warmth of the ground hastens the melting of the snow at the bottom. 
Whatever free water appears soaks at once into the earth. 
It is true that, because the sun is excluded in a dense forest, melting 
begins later than in the open, and Colonel Chittenden correctly states 
a universally known fact when he says that, “even after the ground 
in the open is entirely bare, except under the drifts, the forest areas 
may still be covered with an unbroken layer of snow.” Why? Simply 
because it is colder in the forest. It is colder when the process of 
melting begins, it is colder until the snow is gone in the open. The 
forest consequently acts in exactly the same way as the gulch on the 
‘north mountain exposure. It catches the average amount, or in some 
cases more than the average amount, of snow, and holds it till the 
melting time in the open is past. If the melting went on as fast in the 
