DISCUSSION : FORESTS, RESERVOIRS, AND STREAM FLOW 449 
(1) The moderating and equalizing influence of forests on the Mr, Fletcher. 
run-off of ordinary rainstorms is a fact too obvious to be questioned. 
We: must not, however, belittle the absorbent and retentive power of 
tillage and grass land under proper conditions of cultivation and slope. 
(2) The author’s second proposition is quite confirmed by the writer, 
who has tramped for miles over the drainage area of the reservoir 
in the height of heavy rainstorms, and found the saturated woodland 
shedding the rainfall in swift and furious streams, even more rapidly 
than the meadows where grass was growing. 
(3) The author’s proposition concerning run-off in summer the 
writer would apply to woodland and open country alike. Throughout 
the summer and autumn, when the grass, and crops, and trees, alike 
absorb moisture from the soil, heavy rains in the reservoir district 
produce but little effect on the level of the pond. In this case, the 
area of the reservoir (34 acres) is about one thirty-eighth part of the 
drainage basin, and the whole district is underlaid by an impervious 
boulder clay, so that nearly all the seepage eventually reaches the 
pond. In winter, however, the pond level rises quickly, even with 
moderate thaws or small rainfalls. 
(4) The author’s proposition concerning the effect of forests on 
snowfall to aggravate the flood discharge of rivers is not true in this 
region. Snow always remains in the woods late in the spring (except 
among young deciduous growth on the south exposure), and disappears 
gradually during April and May. Furthermore, the ground in the 
woods freezes much less than that in the open, and probably in spots 
allows percolation of water from the snow more or less throughout the 
winter; for, in the reservoir under observation, the surface level has 
been known to rise throughout the winter notwithstanding the draft 
upon the reservoir. (A temperature of 20° below zero is commonly 
experienced during three months of the winter.) 
(5) Soil erosion depends so much upon circumstances, as already 
noted, that we may pass the question here. 
(6) As to climatic conditions in this region, there has been no 
marked change in the mean annual precipitation since 1835, when 
the keeping of systematic records began. Of course, after the past 
season of phenomenal drought, by which we are more than 10 in. short 
of the average during the past year, the most ardent faith in statistics 
is likely to waver a little, as though some change were impending; 
but the records show that our forefathers had similar experiences and 
probably similar misgivings. 
Mean annual 
‘Period of 21 years. precipitation. 
1835 to 1855 40.77 in. 
1866 to 1886 30.63 “ 
1887 to 1907 36.438 “ 
Apparently, there is a gradual decrease, in the long run(?), but we 
can judge better 20 years hence. 
