A NEW EVAPORATION FORMULA 115 
for winds below 
(^-2aJ] E>, 
La 100 
that it contained the additional wind-term, 
10 m.p.h. 
The weighted-mean values of E h E 2 , and E, derived from the two six-months 
solutions, U 4 and T t) were as follows, the method of weighing being the same as 
that shown in Table 26, page 82: 
#!= +0.561 ±0.065 E 2 = +0.690 ±0.125 E 3 = +1.167 ±0.485 
These values were studied in the light of the following criteria: 
(1) That a decrease in the wind velocity can not be accompanied by an in- 
crease in the evaporation. 
(2) That the evaporation can not be negative for any positive value of e. 
The above values are grossly inconsistent with the first criterion, and no 
assumed value of E 3 near that quoted would be consistent with both criteria, Ei 
and E 2 remaining unchanged. From the studies which had been made of the 
evaporation computations up to this time it appeared that the values of E x and E, 
were much more reliable than the values of E t . It appeared also that the reliability 
of the evaporation curve was high for wind velocities of 10 m.p.h. and above, and 
that this reliability decreased very rapidly as the wind velocity drops below 10 
m.p.h. At this time it appeared, moreover, that the evaporation might become 
independent of the wind velocity for such low wind velocities that there is but very 
slight ventilation at the evaporation surface — water, ground or vegetation. 
In view of the best values of E h E 2 , and E 3 quoted above as having been ob- 
tained up to this time, of the two criteria mentioned, and of the three considerations 
mentioned in the three preceding sentences, it was concluded at this time that the 
relation between evaporation and wind velocity (a) is a straight line relation from 
10 m.p.h. upward with increasing evaporation corresponding to increasing wind 
velocity, (6) is a curved line between 8.8 m.p.h. and 10 m.p.h. with increasing 
evaporation corresponding to increasing wind velocity, (c) is independent of the 
wind velocity below 8.8 m.p.h., (d) that the evaporation from a free-water surface 
is best expressed in units of 0.001 foot of depth per day by the following three 
formulas : 
If the wind velocity is 10 m.p.h. or above +0.56 e+0.69 e [ r^r— 2.4 J 
If the wind velocity is 8.8 to 10 m.p.h. +0.56 e+O.69/— -2.4^ + 1.17 e( — -2a\ 
If the wind velocity is less than 8.8 m.p.h., +0.46 e. 
The estimated evaporation from land is taken as 0.62 times the evaporation 
from water, as in Solution V,. 
The derived values of E h E 2 , and E s from Solution BB X are grossly inconsis- 
tent with both criteria above, as they were in Solution BB 2 — strong indication 
that the ei jqq — 2.4 j i^-term does not belong in the equation. This evi- 
dence is strengthened by the internal evidence of the solution. 
This evidence shows that the very large difference between the assumed and 
the derived value of 2?,— the largest of all the differences — is due mainly to the 
influence of the term el r^ — 2.4 ] E>. 
