105 



than in the forest. Then, too, another point in favor of the forest is 

 the fact that in it the soil itself is more efficiently shaded than in the 

 field. From these observations alone one must conclude that when 

 the land is changed from a forest to a field condition there will follow 

 a large loss of water by evaporation. 



It is just as obvious, when one stops to examine, that during a heavy 

 rain the conditions existing in the forest are such as to retain a larger 

 portion of the water than is held by the field. In other words, much 

 of the water, w 7 hich soaks into the ground in the forest, flows off as 

 surface wafer from the field and is almost at once carried out of the 

 country. 



Let us make this statement as exact as our present information will 

 allow. On pages 17 and 18 of the report of Major C. W. Raymond, 

 United States Engineer, upon Flood Protection of the City of Wil- 

 liamsport, w r e find the following statement : "Colonel Torrelli affirmed 

 as the result of careful observation that four-fifths of the precipitation 

 in forests is absorbed by the soil or detained by the surface of the 

 ground to be gradually given up in springs and gentle rills. And only 

 one-fifth of the precipitation is delivered to the rivers rapidly enough 

 to create floods. Upon the same slopes and surfaces, denuded of their 

 forests, the proportions are reversed/' 



That is to say, in the forests four-fifths soak into the ground and in 

 the fields but one-fifth succeeds in doing so hence then there is a sav- 

 ing by the woods of three-fifths of the water-fall over that of the fields. 



It w r ill be interesting to note just what this means. From data col- 

 lected by this office it appears that in the year 1896, in the counties of 

 Clarion, Forest, Indiana, Jefferson and Warren, in the State of Penn- 

 sylvania, the areas cleared of timber aggregated about thirty-two 

 square miles. These counties drain almost wholly into the Allegheny 

 river. It is not possible as yet to obtain the exact annual rain fall 

 for the region. It is safe, how r ever, to place it at thirty-eight inches. 

 The quantity of water falling upon thirty -two square miles would be 

 4,881,619,353,600 gallons. In an United States gallon there are 231 cubic 

 inches. Dividing the number of cubic inches of rainfall by the num- 

 ber of cubic inches in a gallon, we w r ould ascertain that the rainfall ag- 

 gregated 21,132,551,314 2-7 gallons annually over the thirty-two square 

 miles cleared. If this rain fall were converted into a stream one foot 

 deep, one hundred and twenty-five yards wide, and which flowed at a 

 speed of four miles an hour, it would require fourteen days, twenty 

 hours, forty-one minutes and thirty-five seconds to pass a given point. 

 If it were converted into a solid cube it w r ould form a mass of water 

 471.21 yards in length, width and depth. 



If the statement of Colonel Torrelli is to be depended upon, the re- 

 moval of timber from those thirty-two square miles would mean that 



