FIFTH NATIONAI, CONSERVATION CONGRESS 199 



Although in a level country a spruce forest consumes more water than do culti- 

 vated fields, at high elevations, where the precipitation is from 43 to 47 inches, 

 it consumes only 9.2 inches, or 21.5 per cent of the total precipitation, less than 

 open fields, and nearly 15.7 inches or 34 per cent less than bare surfaces. 



The ability of the forest to check surface run-off is greatest when the ground 

 beneath is covered with an unbroken leaf litter. A forest without leaf litter, on 

 slopes at moderate altitudes, has little effect in checking run-off. The entire loss 

 of precipitation from such a forest was found to be 26.9 inches, while that from 

 bare surface in the same situation was 27.2 inches. Hence, for a forest to exer- 

 cise its most beneficial effect upon run-off, it must not be burned over, grazed, or 

 otherwise interfered with in its normal function. 



That a normal forest in the mountains saves more water for streamflow than 

 any other vegetable cover or any bare surface is shown also by the abundance 

 of springs in mountain forests. 



Reduction of surface run-oif means both an increase of underground seepage 

 and prevention of erosion, two important factors in the regulation of streamflow. 

 The action of mountain forests in protecting the soil against erosion and in in- 

 creasing underground seepage at the expense of surface run-off is the result of 

 their ability to lessen the severity of rainfall, to retard the melting of snow, to 

 offer mechanical obstacles to surface run-off, to hold the soil together, to keep it 

 in a permeable state, to increase its volume by constantly adding new soil, and to 

 absorb large quantities of water by its leaf litter. 



The forest modifies both the severity and the duration of the rainfall. By 

 its foliage and branches it breaks the force of the rain, so that the water reaches 

 the soil without violence and at the same time prolongs its duration. Alter a 

 storm water continues to drip from the leaves and twigs for one or two hours. 

 The water in the forest, therefore, falls more quietly and for a longer time and 

 has thus a better chance to be absorbed by the soil. 



The rapid melting of the snow in the spring, especially when the ground is 

 frozen or saturated with water, favors surface run-off and lessens. seepage. 



The influence of forests in retarding the melting of snow has been demon- 

 strated with especial precision in a ten years' series of observations carried on 

 at the Imperial Agronomic Institute at Moscow. These show that the period of 

 snow melting lasts within the forests from 26 (1904) to 57 (1902) days, while 

 snow in the open disappears within 6 or 7 days. Thus, in 1908, the melting of 

 snow, which began April 12, lasted in the forest until May 15 (34 days), but in 

 the fields, pastures, and all other open places surrounding the institute, only until 

 April 22 (11 days), while in the more exposed fields the snow had all disap- 

 peared as early as April 18, 7 days after it had begun to melt. The retention of 

 snow in the forest until May 15 was in spite of the fact that after April 22 there 

 were frequent warm rains. 



In an ordinary forest region the water in the streams in the spring is de- 

 rived from three sources: (1) The snow water that runs off from fields and 

 clearings; (2) the surface run-off from forest soil, however slight; and (3) one, 

 one and one-half, or two months later, after all the snow is melted, the under- 



