158 
Journal of Agricultural Research 
Vol. V, No. 4 
The results in Table III show the most surprising fact that the 
amount of moisture moved from the moist and warm column of soil to 
the dry and cold column of soil by vapor is very insignificant. It will 
be seen that at the temperature amplitude of 40° the quantity of mois¬ 
ture moved is only about 0.25 per cent, and at the amplitude of 20° the 
value is only about 0.035 P er cent - In comparison with the results of 
Table II, where it is shown that the maximum thermal movement of 
water at the thermal critical moisture content, when the soil mass is 
continuous, runs as high as 3.68 per cent in some cases, the above values, 
due. only to vapor movement and condensation, are extremely insig¬ 
nificant. 
From these results then it is safe to conclude that the thermal move¬ 
ment of moisture due to distillation is practically negligible, even at 
such high amplitudes of temperature as 20° and 40° C., which never 
exist during the night at the different depths on the soil, nor during 
such a long, continuous period as eight hours. This conclusion is indi¬ 
rectly substantiated by the studies of Buckingham (5) on the loss of 
soil moisture by direct evaporation from points below the surface. By 
exposing a surface of water or moist soil to evaporation into a confined 
space which was in communication with the outside air through a column 
of soil, Buckingham found that the actual mean rate of loss of water 
through diffusion of water vapor through soils in still air w r as very small. 
Another noteworthy fact in the foregoing experimental data is that 
the amount of distillation from moist and warm to the dry and cold 
column of soil is the same for all moisture contents. This might have 
been anticipated, since the amount of water vaporized depends prin¬ 
cipally upon the temperature and is not governed by the amount of 
water present. On the other hand, if the amount of water present in 
the soil is extremely small, the water is held by the soil grains with an 
attraction of great magnitude, causing a lowering of the vapor pressure 
of the absorbed water film and thereby producing a diminution in the 
rate of evaporation. Perhaps the water contained in the soil with the 
lowest moisture content was above the point where this lowering of 
vapor pressure occurs; and consequently the partial pressure of the vapor 
in the air space in this soil was the same as in the air space of the soil 
with the greater moisture contents. Furthermore, the values are so 
small as to lie within the experimental error, and the method of mois¬ 
ture determination may not be sufficiently sensitive and accurate to 
show any decreased evaporation by the soils with the lowest moisture 
content. 
In undertaking and performing the foregoing series of experiments 
it was taken for granted that there really is an upward movement of 
moist air during the night from the warmer soil below to the colder soil 
at the surface, where its vapor is condensed. This theory seems to be 
now very widely accepted, as already stated. The formation of the dew 
