DISCUSSION : FORESTS, RESERVOIRS, AND STREAM FLOW 447 
vation, the available data for run-off are generally not in much better mr. Kuichling. 
condition, due to difficulties of measurement and observation. It is 
rarely the case that the gaugings of a stream at the same stage of 
flow agree with each other in the course of several years, even when 
the channel remains unchanged, and in but few instances are accurate 
measurements made of high floods. It also seems to be necessary to 
change the rating tables of many river stations from time to time 
in consequence of altered channel or more reliable gaugings, so that 
comparisons of discharge at some particular stage in different years 
cannot be made without much tedious study. Such work is usually 
impracticable in the case of private enterprises and is seldom done for 
public purposes, but it is indispensable when endeavoring to ascertain 
the cause of variations in discharge. 
It seems to the writer, therefore, that, with the scanty data now 
available in the United States, a sound conclusion cannot be reached 
in regard to the influence of forests on either rainfall or run-off; and 
that, until a much larger volume of accurate observations has been 
acquired, we must content ourselves with the more or less imperfect 
studies of the subject that have been made abroad. There is, however, 
no question about its importance from the point of view of national 
economics, and the writer hopes that the public interest taken in it 
will quickly lead to the needed enlargement of our Weather and 
Hydrographic Bureaus. 
Rosert Fietcuer, Assoc. Am. Soc. C. E. (by letter).—The follow- mr. Fletcher. 
ing facts have come within the range of the writer’s personal ex- 
perience and observation: Region: the upper Connecticut Valley at 
a point where there is a tributary area of about 3 250 sq. miles, north 
of Dartmouth College. Altitude of the country: from 380 ft. above 
the sea at the river level to 6 000 ft. in the White Mountains. Annual 
rainfall: between 30 and 40 in. Winter conditions: ground frozen 
from late November to the middle of April. Surface of the country: 
river bottoms, narrow valleys, generally steep hillsides; probably one- 
quarter or more in forest which is mostly recent growth excepting 
about the head-waters of streams in the extreme north. Furthermore, 
the writer is interested in the work being done by Dartmouth College, 
under the supervision of its forester, in systematic forestry applied to 
the College Grant which comprises about 40 sq. miles near the head- 
waters of the Connecticut River. Also, as an official of the Hanover 
Water-Works Company (which is the College and Village precinct 
in a joint corporation), he has supervision of about 2 sq. miles, the. 
entire drainage area tributary to the reservoir which furnishes the 
water supply. His period of personal observation extends over thirty- 
six years. 
As to forestry conditions in this region, the following facts are 
undisputed: The primeval forest has long since disappeared, excepting 
