FOREST TYPES IN CENTRAL ROCKY MOUNTAINS. 75 



(2) It indicates the marked differences that are possible between 

 succeeding years and between growing seasons, the latter explaining 

 far more completely than the records of precipitation do, the varia- 

 tions in growth rates of trees from year to year. The growing seasons 

 of 1917 and 1920 are especially to be contrasted. 



(3) It gives a basis on which, in the absence of a long-term record, 

 the evaporation may be computed for other years, in order to show 

 more clearly the cyclic character of growing conditions. 



If the indicated annual evaporation is reduced to depths, a figure 

 of 28.96 centimeters or 11.4 inches, is found to be the approximate 

 yearly loss. As the normal precipitation is 21.1 inches, the evapora- 

 tion represents 54 per cent of the precipitation. At a corresponding 

 elevation at Wagon Wheel Gap, streamflow measurements indicate 

 that about 74 per cent of the total precipitation escapes as evapora- 

 tion from snow, soil, and vegetation. For the 20 months of observa- 

 tion, the measured evaporation is only 25.9 centimeters per } T ear, or 

 only about 47 per cent of the normal precipitation when entirely 

 unaffected by trees. The instrumental measurements, therefore, are 

 evidently conservative. Evaporation records from open pans usu- 

 ally show more evaporation than precipitation in arid climates. 



Evaporation record extended by formula. — In a general way the rate 

 of evaporation from any type of evaporimeter or vessel of water is 

 determined by the capacity of the atmosphere for additional vapor, 

 which affects the rate of diffusion from the evaporating surface, and 

 by the rate of wind movement, which not only assists diffusion but 

 may be the principal source of heat for continuing evaporation. If 

 the evaporating surface is well heated by sunshine, however, it may 

 be warmer than the air; the wind will then be not a source of heat 

 but a means for loss. As evaporation is further complicated by 

 other factors, precise formulas for computing it, even from free-water 

 surfaces, have never been devised. However, for approximate pur- 

 poses in this study, to obtain a better idea of the relative evapora- 

 tion for different years, it seems advisable to determine the ratio 

 between the actual evaporation of Type 4 instruments during 1917 

 and 1918 and the evaporation which may be computed from the 

 wind, humidity, and sunshine data combined in a manner to give the 

 most consistent results. The coefficient so obtained will then be 

 applied to the corresponding atmospheric conditions for other years, 

 to elaborate the evaporation record for the control station. 



Admittedly, because the sunshine record does not indicate the 

 intensity of sunlight as determined by its angle of incidence on a 

 horizontal surface, the coefficient should vary according to the eleva- 

 tion of the sun. It has been considered sufficient to calculate C for 

 each month, assuming that the variable atmospheric conditions 

 affecting intensity would probably be fairly well compensated in the 

 records of twenty months. 



The formula used, after three other trials, to compute the coeffi- 

 cient for each decade of the 20 months is as follows : 



<7= E 



( W+ VD) SS 



