DISCUSSION : FORESTS, RESERVOIRS, AND STREAM FLOW 399 
there might be uniformity in the stream below, because, after one small Mr. Leighton. 
unit of the supply had exhausted itself, the next would supply the 
deficiency, and so on to the end; but, as the author forcibly states, 
natural run-off cannot be controlled with such mathematical accuracy. 
Uniformity in small rivulets begets uniformity in the main artery 
supplied by those rivulets. The cause and effect are so clear as to 
become axiomatic. If, as the author contends, deforestation assists 
in the drying up of springs and rivulets, then, as a necessary conse- 
quence, the flow of the main affluent must become less, and the effect 
will be more and more pronounced as the proportionate amount of 
deforested area decreases, until finally, when there is absolute de- 
forestation, there must be this tendency of small rivulets and springs 
to dry up over the entire area. If the author would give consideration to 
the all-important matter of ground storage and apply thereto some of 
the fundamentals of geology and physics, he could readily see the 
fallacy of his argument. 
With reference to the section concerning the influence of forests on 
snow-melting, it appears that the author’s observations have been 
confined to a comparatively limited area in a high altitude. Several 
of his observations are faulty for the areas which he is considering. 
He has apparently neglected all consideration of the physical principle 
of diathermancy. His premises will be considered in their order of 
presentation. 
The statement that forests prevent the formation of drifts, and 
distribute the snow in an even blanket over the ground is, of course, 
quite erroneous. Every country boy has floundered through drifts 
in the forest, and Dr. George Otis Smith, in his discussion on this 
paper, testifies that in the very region concerning which Colonel Chit- 
tenden appears to be most familiar, and in the same year in which 
the latter made some of his observations, his (Dr. Smith’s) horse 
became stalled in a snowdrift in a dense forest on or about July Ist. 
Further, the author states that, “in the open country, the snow is 
largely heaped into drifts,” which is quite as erroneous as the first 
statement. On the tops of peaks and in extremely high altitudes, con- 
stantly subject to great winds, this may be true at times; but it is 
certainly untrue for the greater part of the country, the quantity of 
snow in the drifts being but an exceedingly small part of the total accu- 
mulation. Even in high altitudes, the heavy and uniform blanket of 
snow is common, as can be shown by hundreds of photographs on file 
in the Geological Survey. 
Consider this, together with the author’s admission that the period 
of snow-melting in the open country begins much earlier than in the 
forest; and consider further the well-known physical fact that the 
melting of snow takes place at the bottom and not at the top, except, 
of course, in the event of warm melting rains. This is a point 
