108 



J. W. SHIVE AND B. E. LIVINGSTON 



experiment was begun August 14, 8 a.m. The first permanent 

 wilting occurred in the chamber in the open, on August 14. On 

 August 17, permanent wilting was attained in the shelter and 

 in the glass box, and on August 18, the same stage was reached 

 in the chamber in the shelter. 



The numerical results of series III are set forth in table IV, 

 which conforms with the preceding tables. 



In consequence of the large volume of soil in the containers, the 

 time required for the plants to attain the permanently wilted 



7.0 . 



SERIES III 

 ( table: IV.) 



Fig. 3 



condition was comparatively long in this case, and yet the gen- 

 eral results of this series do not differ essentially from those of 

 the two preceding series. There is no such discrepancy here, in 

 the relative order of magnitudes of the five main averages, as 

 was found in the case of series II. It appears, however, that 

 although it is strictly true that the main averages do increase 

 with higher intensities of evaporation, yet averages (2) and (3) 

 are somewhat too low to allow all five to lie on a consistent curve. 



