" 298 FORESTS, RESERVOIRS, AND STREAM FLOW 
this is done there will be many years when the reservoirs will not 
fill. 
An important consideration in the use of the reservoirs for flood 
control is that of a proper combination of their outflow. To any 
one who will try to figure out how this can be accomplished over a 
water-shed of such vast extent, with storms arriving at different 
times in the various portions, with no way of telling when, where, 
or with what intensity they will arrive, with the varying distances of 
the different reservoirs from those points where flood control is par- 
ticularly important, the problem seems almost impossible—that is, 
impossible to realize the full effect based upon the aggregate capacity 
of the system. It is understood that Mr. Leighton has endeavored 
to do this, but it would be interesting to see the application to some 
of the great floods that might be designated. For example, in the 
flood of 1907, which reached its maximum at Cincinnati and Pitts- 
burg about the same time, no amount of holding back of the storm 
water on the Upper Ohio at that time would have helped the situa- 
tion at Cincinnati at all. 
Another important consideration in the effect of these reservoirs, 
as they would have to be operated to prevent floods, is the great change 
that takes place in a flood wave as it propagates itself down stream. 
The writer is unable to tell from Mr. Leighton’s paper to what extent 
he has considered it. The paper itself seems to indicate that the 
discharge held back by a particular reservoir produces a correspond- 
ing volumetric effect (not gauge effect, of course) at all points below, 
after making a due allowance of time for the transmission of the 
wave. This would be an erroneous conclusion. For example, a wave 
that might rise at Pittsburg from 100000 cu. ft. per sec. to 150000 
a day later and to 200000 the next day, and then fall at a corre- 
sponding rate, would not at any point below produce a maximum 
increase of 100000 sec-ft.; and the farther away the point considered, 
the less would be the increase. At Cairo, nearly 1000 miles below, 
the same wave would take a much longer time in passing, probably 
not less than a week, and the maximum increase would probably not 
be more than 25000 sec-ft. This is merely a general illustration, 
for exact data on the subject are not available. The problem is of 
such complexity that nothing but the results of long experience could 
establish a rule as to what might be expected in any given case; but 
