EQUALIZING INFLUENCE OF FORESTS 



493 



the year and over the area considered, 

 the slope of the ground, the area of 

 forest, cultivated land, etc. ; the num- 

 ber of lakes and reservoirs, the tem- 

 perature, and other elements. The 

 chief of all of these is undoubtedly the 

 rainfall and its distribution. A great 

 fall of rain, long continued, will prob- 

 ably cause a great flood whether there 

 are forests or not, although, as before 

 explained, there is abundant evidence 

 for the contention that the action of 

 the forest is to diminish the flood. 

 Meteorological phenomena are admit- 

 tedly variable and uncertain, and, of 

 course, they are entirely incapable of 

 control. The rainfall varies from year 

 to year in long cycles, the extent of the 

 variation being such that in the United 

 States it has generally proved impos- 

 sible to determine with certainty 

 whether the rainfall over a given terri- 

 tory which has remained in essentially 

 the same physical condition is increas- 

 ing or not. The rainfall at a given place 

 may va'ry from thirty inches in one year 

 to fifty or sixty inches in the following 

 year, and its distribution is subject to 

 similar variations. These variable ele- 

 ments, therefore, may mask the influ- 

 ence of forests or of reservoirs, but the 

 important point is that these two arc 

 the only clftnciits subject to man's con- 

 trol. It is admittedly physically pos- 

 sible, by reforesting and by the con- 

 struction of storage reservoirs, to make 

 the flow of a given stream practically 

 uniform throughout the year, although 

 to do so would in most cases involve a 

 prohibitive cost ; and, moreover, it 

 would be physically impossible to regu- 

 late a reservoir and allow the water to 

 flow out of it in such a way as to pro- 

 duce this effect, because the future can- 

 not be foreseen. Observations of gauge 

 readings on rivers, therefore, are in- 

 conclusive in themselves. Fortunately, 

 however, we are not without valuable 

 evidence on this point. Mr. M. O. 

 Leighton, Chief Hydrographer of the 

 United States Geological Survey, has, 

 during the past summer, made an elab- 

 orate study of the floods of the Tennes- 



see River, in which he has endeavored 

 to eliminate the effect of the rainfall 

 and its distribution by comparing the 

 number of days of flood with the num- 

 ber of individual rainstorms of suffi- 

 cient magnitude to produce floods. The 

 record shows that during the last half 

 of the period studied the number of 

 days of flood was actually less than in 

 the earlier part of the period, notwith- 

 standing the deforestation which has 

 recently taken place. The rainfall, how- 

 ever, has also been less in the latter 

 period, and the results of Mr. Leigh- 

 ton's study are that the diminution of 

 the rainfall has been much more than 

 sufficient to account for the diminu- 

 tion of the floods, so that the actual 

 result is that the floods have been in- 

 creasing, the percentage of increase be- 

 ing 18.75 i n the l as t seventeen years, as 

 compared with the seventeen years pre- 

 vious. This study is the best contri- 

 bution to the subject which has come 

 to the writer's knowledge, and it seems 

 conclusive. The experience in France 

 also furnishes valuable evidence in this 

 matter. In 1857, M. F. Valles, a French 

 government engineer, published a work 

 in which, and in some subsequent pa- 

 pers, he gave almost the identical argu- 

 ments advanced by Colonel Chittenden, 

 maintaining that forests diminished the 

 rainfall, increased the floods, and dimin- 

 ished the supply of grain by withdraw- 

 ing lands from cultivation. He also 

 maintained that floods were beneficial, 

 by bringing silt from the mountain 

 sides to the plains. His work, how- 

 ever, seems to have been entirely with- 

 out influence, for immediately after its 

 publication the French government en- 

 tered upon a policy of forest protection 

 and reforestation, particularly in the 

 mountain regions, which has been con- 

 tinued up to the present time. Up to 

 January i, 1900, the state had acquired 

 over 400,000 acres,, or 629 square miles, 

 for the purpose of controlling torrents. 

 Of this area, 440 square miles are in 

 the Alps, 145 square miles in the central 

 plateau and the Cevennes, and forty- 



