DISCUSSION: FORESTS, RESERVOIRS, AND STREAM FLOW 4651 
Colonel Chittenden’s position. Some features of the paper,. however, Mr. Snow. 
may, perhaps, be made to mean more than was intended. 
Conditions vary for each locality, and a statement that is true 
for one place may not hold in another. The author well says that 
no fixed rule can be upheld for all localities and conditions. What 
follows must be considered as pertaining only to that portion of New. 
England between the latitude of Boston and the Canadian line. 
Certain vital elements of the problem have not been mentioned 
in recent discussions, as far as the writer has observed. That forests 
evaporate an enormous quantity of water from foliage is proved by 
the water level in swamps becoming higher after a heavy growth has 
been cut off. On ordinary upland forest this evaporation tends to 
‘lower the subsoil water more than the growth on open land. Surface 
springs and running streamlets, on the other hand, are’ protected from 
evaporation by the shade of the forest, and stand a better chance to 
persist through a hot summer than if in the open. It is not surface 
water, however, that principally feeds our streams during ordinary 
or low-water periods. It is subsoil water reaching the streams below 
the water surface. Many times wells and pits have been sunk 
near streams with the idea of getting the advantage of - filtration 
through a few feet of earth, and, behold, the supply was all from the 
side farthest from the stream. The greater part of New England is 
covered with a blanket, from 5 to 15 ft. thick, of interglacial gravel 
resting on subglacial till or ledge. Deep'bank cuttings all over the 
_region show water oozing out at the top of the impervious stratum 
below the gravel blanket. This natural reservoir is practically inde- 
pendent of surface cover. = 
In long-continued rains, when the ground is not frozen, the run-off 
-is practically the same from open as from forest-covered land, with 
‘the following exception, in connection with freshet effects: Forest 
-Streams are generally choked with driftwood, and sometimes very 
formidable jams occur, which spread devastation when they break. 
A stream may even be turned from its course when one of these 
‘jams holds permanently, as is indicated on the Ammonoosuc River, 
in New Hampshire, a few miles above the Village of Littleton. These 
obstructions do not increase or-decrease the total run-off, but they do 
‘have a marked effect on the possible amount of damage that high 
water may do. 
-In the case of moderate rains, the absorbent action of open and 
- wooded land is probably very nearly equal, as a whole, with the 
balance in favor of the open land, if the ground is not frozen. New 
England grass-land, with its thick mat close to the surface, and with 
- stalks and roots leading down into the soil, is certainly almost ideal 
as an absorbent. Deciduous forest, on the other hand, covers the 
ground with broad leaves that act like shingles to keep the water 
