22 



BULLETIN 700, U. S. DEPARTMENT OF AGRICULTURE. 



so that for this, purpose the atmometer is superior to the free water 

 surface. 



Figure 7 shows the relation of evaporation, as obtained by the two 

 methods, to wind movements. Here, again, agreement between the 

 two graphs consists in a slope in the same direction, since high wind 

 velocity accelerates evaporation rate and low wind velocity retards 

 it. This figure shows an agreement of 11 periods in the case of 

 evaporation from the free water surface with the average daily wind 

 movement, or a percentage agreement of 37. The evaporation from 

 the atmometer, on the other hand, shows for the same time an agree- 

 ment of 18 periods, or a percentage agreement of 60. 



From the above comparisons it is evident that the evaporation rec- 

 ord obtained by means of the spherical atmometer agrees more closely 

 with the relative humidity, the temperature, and the wind velocity, 

 and is a more reliable index when short periods are to be considered 

 than that obtained from the free water surface. When periods ex- 

 tending over several days or longer are considered, however, either 

 instrument may be used with good results. 



Because of the advantage of the rain-correcting device used in con- 

 nection with the porous cup atmometer, this instrument was espe- 

 cially suited to the experiments here presented, and accordingly has 

 been used throughout. Of the three types of porous cup atmometers 

 available for field use, namely, the white sphere, .the black sphere, 

 and the white cylinder, all were operated in each type station through- 

 out the season. The sum of the daily means of each set of cups and 

 the difference in evaporation between the black and white sphere are 

 presented in Table 7. 



TABLE 7. Summary of evaporation from white cylindrical, wliite spherical, and 

 black spherical atmometers, and of difference between white and black 

 spherical atmometers during the period of experimentation. 



The figures given in Table 7 are platted in figure 8. It is interest- 

 ing to find that the graphs representing the three evaporating instru- 

 ments here used are all practically parallel. This parallelism also 

 holds for shorter periods, as is shown in figure 9. It would appear, 

 therefore, that one might select any one of these instruments to ascer- 

 tain the evaporation. However, because of the desire to use evap- 



