456 
Journal of Agricultural Research 
Vol. VII, No. 10 
having the least evaporation. The sands show somewhat the same 
results as were found in the experiments mentioned above, except that 
there is a tendency for the losses to increase with increasing size of soil 
particles when the particles are more than 2 mm. in diameter. 
COMPACTING THE) SOIL 
A set of six galvanized-iron cans 13 inches deep and 11 inches in diameter 
with an opening at the bottom’through which water could be added 
from below to maintain the soil at a constant moisture content were 
filled with soil. Two sections of the soil at various depths were com¬ 
pacted in order to determine the effect of compacting on evaporation. 
The cans contained 12 inches of Greenville loam made up to about 
15 per cent moisture. 
In can 1 the surface 2 
inches were compacted; 
in can 2 the second 2 
inches; and so on,until 
in can 6 the bottom 2 
inches were compacted. 
The packed layers con¬ 
tained 20 per cent more 
soil in a 2-inch layer 
than a similar volume 
of the loose soil. These 
cans were weighed 
weekly for seven weeks 
and the loss made up 
through the side tubes 
at the bottom, which «vere kept closely stoppered except while water 
was being added. 
Table VI shows that cans 1 and 2, compacted in the first and second 
2-inch layers, respectively, lost much more heavily than the cans in 
which the packed layers were farther from the surface. 
Table VI .—Loss of moisture from cans of soil containing a 2-inch section compacted 
at various depths 
Can No. 
2-inch section 
compacted. 
Total loss. 
Top 
Gm. 
2. 
Second.... 
Third 
I, 205 
1,045 
885 
870 
880 
Fourth 
tfifth 
6. 
Bottom.... 
505 
Fig. 14.—Loss of water in 180 days from glasses having dry mulches 
of various kinds suspended above free water. 
