FORESTS, RESERVOIRS, AND STREAM FLOW 299 
it can be stated with certainty that the diminution of discharge at 
any considerable distance below the reservoirs for a given time would 
never be as great as the amount held back by the reservoirs in the 
same length of time, and that, the quicker and the higher the flood, 
the smaller the relative effect at all points below. It is only when 
such wave elimination merges into a constant quantity, continuing 
for a considerable time, that the full effect of a reservoir would be 
experienced at any point below. This, in fact, is what would actually 
happen in the contrary case of the low-water season when the reservoir 
discharge is kept up for a long time. 
Still another feature in the high-water effect of such reservoirs is 
the demand for water for power at all times. If there should ever 
result any really general use for all this water, as is predicted, then 
the consumption for power would make a considerable river in itself. 
Now this much cannot be shut off in any case. Street cars and shops 
must run and houses must be lighted whether the flood is ruining the 
lowlands or not. An example of this occurred in 1905 on the Upper 
Mississippi where the outflow from the upper dams was cut down to 
a minimum to reduce the flood in the valley at Aitkin, which was 
then being overflowed by the river. The mill at Grand Rapids, just 
below the reservoirs, made a strenuous protest, and even threatened 
legal proceedings to compel the release of the full normal flow of the 
river. 
Considering all the foregoing features of the operation of the pro- 
posed system, even if every reservoir were built with the full, estimated 
capacity, it would be extremely fortunate if 75% of the predicted 
results, either in flood protection or in aid of navigation, could be 
realized. 
It is in the matter of cost, however, that the weak point of Mr. 
Leighton’s system appears. Judged by any reasonable standard, his 
estimates are hopelessly wide of the mark. The method itself of 
getting at a basis of cost is inadmissible. For example, in determin- 
ing a unit of cost for that class of reservoirs which embrace the greater 
portion of the total storage, the figures for nine reservoirs are taken, 
counting as one the whole Upper Mississippi system. Only the Mis- 
sissippi system has been built; two others are under construction 
and six are merely projected. In accordance with almost universal 
experience, and especially in view of the great advance in prices of 
