44 
Research Bulletin No. 3 
tained an average of 17.8 per cent total water and had a hygro- 
scopic coefficient of 7.6. In the adjacent prairie a shallow pit, 12 
feet in diameter, was dug on November 9, 1907. During the fol- 
lowing four days 28 inches of water was added. Ten days later 
the first six feet contained 21.1 per cent total water and at the end 
of the following April, the surface having been protected from 
evaporation in the interval, it contained 17.3 per cent total water 
■ — much the same as the soil of the adjacent orchard. Thus the 
same depth of soil in the cylinders with a quite similar hygro- 
scopic coefficient, averaging 8.5, contained about 25 per cent free 
water while similar soil in the field, both when saturated by 
fallowing and when saturated by irrigation, contained only about 
10 per cent free water. 
On the Experiment Station farm, part of a young orchard had 
been kept in clean cultivation for ten years. In the most moist 
soil that was found in this field the total water in the first 6 feet 
was 26.1 per cent, with an average hygroscopic coefficient of 11.9, 
while the humid soil in the cylinders contained about 35 per cent 
total water and had an average hygroscopic coefficient of 12.8. 
Summary of experiment. — At the time the seeds were planted 
the cylinders of semiarid soil carried more than twice as much 
water as the same soil would have retained under field conditions, 
while the cylinders of humid soil carried at least one-fourth more 
than the same soil in the field would have retained. The loss of 
water — wholly by direct evaporation — from the two unplanted 
cylinders during the four months of the experiment was equiva- 
lent to about 5.0 inches of rainfall in the case of the semiarid 
soil and to 2.5 inches in that of the humid soil. Normal plants 
were produced on both soils and the yields of grain, calculated 
to an acre basis, exceeded 25 bushels on the semiarid soil but 
were less than 11 bushels on the humid soil, the former being 
much higher than could be expected from the same depth of sub- 
soil under field conditions, because of the excessive quantity of 
free water at the time of planting. In no case, however, was the 
proportion of grain to straw or the loss of soil water per unit 
weight of dry matter produced more favorable than might be 
expected in the field. 
The plants on the semiarid soil made a much greater growth 
and rooted much more deeply. They also made a more economi- 
cal use of the soil water than did those on the humid soil; how- 
ever, this economy may have been due entirely to a smaller 
evaporation from the cylinders of semiarid soil. After the plants 
had matured there was a marked difference between the semiarid 
and the humid soil in the moisture condition of the subsoil but 
not in that of the surface foot. The transpiration of water from 
