Effects of Forests Upon Run-off in the Rockies 69 



pect would not be desirable since the exposure would not be 

 typical. The mean or average watershed then, should drain 

 to the east or west, so that one slope will be exposed to the 

 south and the other to the north. In this way the effect of 

 insolation on the two slopes will be fused in the behavior of 

 the stream which drains both. 



To obtain average surface conditions the soil on the water- 

 sheds must be consistent with the steep character of the slopes 

 and the geologically young state of the Rockies, i.e., a fine loam 

 lying in a thin layer intermixed with much rocky debris and 

 numerous outcrops. The amount of humus in the soil is of 

 course an important agent in retarding percolation of water 

 and should be considered quite carefully in order that an aver- 

 age may be obtained. 



To avoid any changes in geological structure or precipita- 

 tion the areas for study should be contiguous or nearly so. 

 They should lie on the same general slope of a simple, narrow 

 mountain range or on a spur well removed from the higher 

 ground. The presence of high mountains behind the water- 

 sheds, even though the drainage is apparently in an 'opposite 

 direction, may complicate matters since such a drainage area 

 might be furnishing a great deal of underground water to the 

 streams being measured. The experiment should be located 

 where the precipitation is moderate and yet large enough to 

 give the drainage stream a minimum flow of 0.5 cubic feet 

 per second. Here a.gain the middle altitude seems most de- 

 sirable for fulfilling the above requirement. 



With the points just enumerated in mind, two watersheds 

 should be selected with the utmost care and consideration in 

 order that the results obtained will be convincing in every 

 respect and a provision made for every contingency. In the 

 inauguration of the experiment, one of two plans may be used ; 

 (1) the experiment may be started with the two areas having 

 a forest cover about equal or, (2) with one of the areas bare 

 and the other forested. In the first plan observations of cli- 

 mate and stream flow should be carried on for a period of two 

 years, then one watershed should be denuded ; in the second, 

 the area now bare should be brought into a forested condition 

 during the period of observation. The experiment cannot be 

 considered absolutely reliable or complete unless simultaneous 



