DISCUSSION : FORESTS, RESERVOIRS, AND STREAM FLOW 511 
gated lands as opposed to the writer’s statement that the aggregate mr. Chitten- 
n. 
forest area of our country cannot be increased. As soon as Mr. Le 
Baron demonstrates the practicability of a large development of 
genuine forest in the arid regions the writer will modify his state- 
ment proportionately. As a general rule, lands under irrigation are 
devoted to intensive farming, and are too valuable to be converted into 
forest; though Mr. Johnston intimates the practicability of forest ex- 
tension by this means. 
Mr. Le Baron is very much astray in his inference regarding the 
saving of cost of installation of water-powers by the use of Govern- 
ment dams. In the first place, this would not apply to the great ma- 
jority of water-power developments, if such developments: are to be 
made on a really extensive scale. In the second place, the sites of 
locks and dams are not favorable on account of the low head and the 
great relative fluctuation. Compared with the power attainable, the 
installation is necessarily much more costly than where the same 
power-house and machinery would yield several times more power. 
The writer notes the interesting point brought out by Mr. Todd in 
regard to the percentage of run-off as affected by slope. That a greater 
percentage of rainfall runs off from flat than from steep slopes is at 
first not easy to admit, but where the soil is swampy and ground stor- 
age is not available, it is quite probable that this is true. The exam- 
ple is another proof of the necessity of care in making generalizations 
upon so uncertain and variable a subject. 
When Professor Swain says that “the author seems to be attempt- 
ing to convince us that, not only forests, but reservoirs, increase the 
violence of freshets,” he says something that has no substantial war- 
rant anywhere in the paper. The whole tenor of the argument is 
that the difference between the past and present is not great, al- 
though the records show that we actually had greater floods before 
the forests were cut than we now have. Only in the effect of forests 
on snow-melting, as described by the writer, can Professor Swain find 
any support for his remark, and only in rare combinations do reser- 
voirs operate to increase freshets. 
Professor Swain does not understand the writer’s reference to the 
suggested reservoir system for the River Rhéne in France. If he 
will examine page 39, et seq., House Document No. 141, 55th Congress, 
9d Session, or the French reports upon this subject, he will find the 
matter fully explained. With characteristic detail, the French engi- 
neers worked out a system of reservoirs for that stream which was 
ample to protect the valleys of the tributaries and the upper courses 
of the main stream; but they found, on applying it to a great flood 
e) 
