DISCUSSION: FORESTS, RESERVOIRS, AND STREAM FLOW 453 
Soil wash may be due to gullying the surface, or bank erosion by mr. snow. 
streams. The first, such.as is common in the Piedmont Section of 
the Southern States, is almost unknown in New England. The latter 
occurs both in forest and open land. Torrential streams generally keep 
their banks intact in both forest and open, except when jams occur 
in the timber. On quiet reaches the bottom land in heavy timber is 
usually a network of sloughs, sand-bars, snags, and swamps due to 
drift. In the open, banks are frequently cut at the bends, and 
disastrous cut-offs sometimes occur. A fringe of small-growth trees 
and bushes along the banks of the streams, through level land, is a 
great protection against this class of damage. 
The greatest damage from erosion is due to sediment deposited on 
arable land or in bars which tend to divert the stream and spread it 
out into shallows. Without doubt the sudden removal of forest growth 
from a water-shed may result in erosion that will be pernicious to the 
land below, as will any disturbance of the existing equilibrium. Saw- 
dust from lumber mills, poisonous sediment from paper and woolen 
mills, waste from tanneries, drainage from cities, wash from highways 
and streets, and the various encroachments of grasping abutters will 
destroy the regimen of a stream unless it is properly controlled by 
competent authority. Once a stream in the open is choked and spread 
out into shallows, its waters are exposed to the pumping effect of the 
summer sun, and a small dry-season run-off will follow. 
Serious erosion sometimes occurs from the undercutting by sub- 
soil water from adjoining land when timber fringes are removed from 
stream banks. Such removal can hardly be called deforestation, but 
some of the minor features of the relation of wood growth to streams 
have more actual influence on the question at issue than the sponge 
theories of our friends, the lecturers. The sponginess of the thin- 
soiled rocky hillsides of New England is a negligible quantity. 
As the author states, the amount of stream flow depends on the 
amount of precipitation, and forests are not the cause, but the natural 
effect, of abundant precipitation. This effect, if correctly spelled out, 
will read that mighty forests are produced by heavy precipitation 
because they are the most rapid dissipators of the superabundant 
moisture. Replace the tree growth of the Coast Range in Washington 
and British Columbia with close-cropped lawn grass, and the streams 
and rivers will carry much more water than they do now. 
In a half century the streams, springs, and wells of New England 
have not been as low as they have been in 1908, although the area 
in tree growth is much greater now than 50 years ago. The reason 
of the low water, however, is simply the lack of rain. 
The proper objects of forest conservation are well set forth by the 
author. His views accord with those of the writer except, perhaps, 
in the matter of devoting convenient lowlands to timber growth. 
