120 J. W. SHIVE AND B. E. LIVINGSTON 



the Washington greenhouse. There appears no reason at all 

 to doubt the experimental accuracy and reliability of the ex- 

 tensive study of Briggs and Shantz. 



If the supposition of Caldwell be true, that there exists a range 

 of evaporation intensities within which this factor does not effect 

 the soil moisture residue here dealt with, this range must lie in the 

 region of high intensities, not in that of low ones as supposed by 

 Caldwell. But to lie within this range the intensities must 

 be extraordinarily high for natural conditions; they must be as 

 high as, or higher than, the intensities experienced in the open 

 during the Tucson summer, perhaps above 3 or 4 cc. per hour 

 from the standard atmometer. 



Turning now to the details of the quantitative relation which 

 is found to hold between soil moisture residue and atmospheric 

 evaporating power, this relation appears, from the present 

 experimeixtation, to be a consistent one within the limits em- 

 ployed in these studies. It is very satisfactorily expressed as a 



y 

 logarithmic function of the form x = k", where x denotes the 



evaporation intensity as measured by the porous cup atmometer 

 and y denotes the soil moisture content at permanent wilting, 

 measured in terms of the dry weight of the soil. If the constant 

 ratio k of this series be taken as 2, then the constant difference 

 (a) is shown by our experiments to lie within the range of values 

 from 0.32 to 0.55. The five values of this difference, for the 

 five different series above described are 0.50, 0.32, 0.50, 0.55 and 

 0.40, and the average of these is 0.45. The first three of these 

 values were obtained from experiments with a soil mixtm^e having 

 a water holding power of 32.0 per cent., and with three different 

 plants Zea, Phaseolus and Capsicum. The fourth value (0.55) re- 

 sulted from tests with plants of Phaseolus in a soil having a water 

 holding power of 38.2; the last value (0.40) was derived from plants 

 of Zea in a soil with a water holding power of 57.3 per cent. From 

 these statements it appears that the variations just given, in the 

 value of the constant difference for the logarithmic functions 

 here in question, are not to be related to the water holding power 

 of the soil used. Neither do these variations appear to be related 



