Dec. 4 . *9*6 
453 
Evaporation of Moisture from the Soil 
twice as long to evaporate the last io per cent as it did the first. In 
each case the last water that was driven off was the hygroscopic moisture. 
This probably accounts for the greater time required. 
SIZE OF SOII, PARTICLES 
Since evaporation takes place almost entirely at the surface, the rate 
of capillary movement directly affects this form of water loss. Because 
the size of soil particles and the porosity of the soil influence capillary 
movement, they indirectly affect evaporation. Principi (14) says that 
materials having the greatest pore space permit greatest evaporation. 
Wollny (21) reports no 
capillarity with particles 
more than 2 mm. in 
diameter, and also an 
increase as the particles 
get finer, though in clay 
the movement is slow. 
Losses from below the 
surface take place, ac¬ 
cording to Buckingham 
(2) by diffusion and 
vary with the square of 
the porosity of the soil. 
On account of the 
great difficulty in in¬ 
terpreting the results 
connected with this 
phase of the evapora¬ 
tion problem when or¬ 
dinary soils are used, 
most of our work has 
been done with sands 
having grains of 
different diameters. The early experiments were conducted with river 
sands and gravels; but because of greater ease in obtaining uniform size 
of particles and because of lesser influence due to the composition of the 
materials, most of the later work was done with graded quartz sand. 
In all trials except those in saturated atmosphere beneath the soil, 
these tests were made in the copper evaporimeters with free water main¬ 
tained from about 1 to 3 cm. below the surface of the soil, the distance 
varying in different experiments. At weighings, which occurred about 
a week apart, the vessels were made up to their original weight. 
The results of these tests are presented in figures 10, 11, and 12. 
Fig. 9.—Time required at different temperatures to drive off half 
and all the water from sand containing 20 per cent moisture. 
