DISCUSSION : FORESTS, RESERVOIRS, AND STREAM FLOW 397 
The paragraph on pages 250 and 251 contains another contradictory Mr. Leighton. 
statement, and presents certain new concepts, unsupported by supple- 
mentary evidence. The first sentence is an admission that, in periods 
of extreme drought, springs and little streams dry up more completely 
under deforested than under forest conditions. It is then stated that 
the difference is exaggerated in the popular mind. On the contrary, it is 
very doubtful if the difference between the two conditions could be 
exaggerated. One could hardly exaggerate a difference between flowing 
water and a dry spring or rivulet. The margin must be 100% in all 
eases. The exception noted by the author in the foot-note,* with 
reference to the drainage of low swamp lands, while true, has not even 
a remote relation to the conditions under discussion. The drainage 
of a swamp area is essentially the tapping of a reservoir. It bears the 
same relation to spring and rivulet flow as would an artificial storage 
reservoir. The water released during dry seasons must prolong the 
flow, of course, whether the release be effected by raising a gate or 
digging a trench. The author then states that, because springs and 
streams are dried up, there is no reason why there should be any 
appreciable diminution in the flow of the great river. This idea is 
strange and new, and brings up a question as to the sources of supply 
in a great river during a period of extreme drought. No one can 
well deny that rivers are supplied during such periods from ground 
storage, which collects almost entirely in the small streams and rivu- 
lets, and unites ultimately to form the great river. The immediate 
drainage on any great river, which contributes seepage directly to the 
main artery, is always comparatively small and sometimes negligible. 
Why, for example, has there been any water whatsoever in the Ohio 
between Pittsburg and Cairo during the summer and fall of 1908? 
No one believes that, had the immediate drainage area contributing to 
this main artery been depended on, the Ohio channel would have been 
anything but a dry bed. The supply that has been contributed to this river 
has come from the millions of small rivulets, each of which has drawn 
on its immediate ground storage, and, by uniting, has formed the 
brook which, also drawing upon its immediate storage, has formed the 
larger tributary, and these, supplemented to some extent by immediate 
ground storage, unite to form the river which, in spite of the lack 
of rain, has persisted. In other words, the great ultimate sources of 
supply for any river lie in these small rivulets, and the discharge of 
any river is but the result of their union. This is a matter of ele- 
mentary physiography, long ago enunciated, repeatedly confirmed, and, 
as far as the speaker knows, undenied up to the present instance. 
Admitting, therefore, the truth of the proposition and, in view of 
the author’s admission that springs and rivulets dry up more com- 
pletely under deforested than under forest conditions, what becomes 
* Page 250, 
