112 



J. W. SHIVE AND E. E. LIVINGSTON 



TABLE V— Continued 



§5 

 -'■ t 



00 



■J S 

 « 2 



O ^ 



m 



^ 2 a o 



- « « 



W Ki a o 



a "^ a. Z 



« g Cj " 



s" 2 a ^ Q 



S o 



■ r " 



i- -H W _ 



H « 2 o 



S ^. i' S S 



5.8 



6.9 



w 



gtf M 



o ^ J o 



S 5 " ■< 



05 > 

 m D «i 

 ►J O H . 

 -< ai IZ > 



6 H w 5 

 ^ ►-( ^ ^ 



per cent, of dry 

 wt. 



8.13 1.37 



• 8.26 • 1.40 



Av. 8.37 (5) Av. 1.41 



^§ 

 D a 



E-i» s 

 y ^ Q 



li, •< W 

 o J C5 



g« 



O J « 

 E -! O 



5 O tH 



8.22 



9.10 



Av. 8.66 



1.39 



1.53 



Av. 1.46 



a 

 o 



o 

 a 



» 



to (K 



o 



a 



Open 

 Open 



Open 

 Open 



IS "^ 



§^ 

 is 



Q 



2 s o 



« « 



a a o 



i^ II. 2 



S 2 ■ 



PSMH 



o " ^ 



a K 



A s: ^ 



n "^ ->■ 



Sept. 11, 11 

 Sept. 11,11 



Sept. 11, 15 

 Sept. 11,15 



deg., C. 



29-33 

 29-33 



34 

 34 



>< K B 



H a 2 



" o S 



Q a g 



S o I 



« H " 



a o a 



i> :? o 



J - w 



K 



^ a 



a o 

 a. c 



per cent. 



20-19 

 20-19 



* Calculations made by means of the Briggs and Shantz formula, from the water 

 holding power of the soil used. 



t The word box here denotes the glass box; ch., sh. denotes chamber in shelter; 

 ch., open denotes chamber in open. 



X The hours of the day are simply numbered in a single series, from 1 to 24; 

 thus, the thirteenth hour is the first hour after midday, etc. 



In this series there is no irrationality in the relative arrange- 

 ment of the main averages with reference to the corresponding 

 evaporation intensities; here there is always an increase in the 

 moisture residue accompanying each increase in the atmos- 

 pheric evaporating power. The curve formed by these five 

 averages is clearly convex upward and approaches a vertical 

 line with low evaporation rates, quite as in the cases previously 

 discussed. Here average (1) falls almost low enough to coincide 

 with the calculated value of the moisture residue (5.92) derived 

 from the Briggs and Shantz formula, a feature not encountered 

 in any of the earlier series. This is probably related to the fact 

 that the soil mixture here used possessed a higher water holding 

 power than did the mixture employed in earlier series. A graph 

 of the results of series IV is given in figure 4, constructed as 

 were the preceding graphs. 



The generalized curve of figure 4 shows almost as satisfactory 

 agreement between its points and the corresponding ones rep- 



