Relation of X on a vail able ^Yater to Hygroscopic Coefficient 33 
heavy rains; this suggested a connection between the injury to 
the wheat and the presence of whitewash on the glass. On July 
10, 86 days after planting, the 4 plants were quite dead. They 
had each reached a height of about 21 inches and had each formed 
one spike, but only one had formed seed and this bore only two 
badly shrunken grains. The cylinders were split lengthwise and 
the soil examined for roots. These were found to reach to the 
bottom of the cylinders. The weight of dry matter in the two 
plants from the one cylinder was 3.5 grams and in those from 
the other 4.0 grams. 
As, on opening the cylinders, it was found that in each a 
crevice had developed between the soil column and the cylinder 
wall, extending from the surface almost to the bottom of the 
cylinder, thus allowing comparatively free circulation of air and 
consequent loss of water from the sides of the soil column by di- 
rect evaporation, the moisture data have been discarded as value- 
less. In all later experiments this source of error has been 
carefully guarded against by watching for the first appearance 
of such crevices and at once filling them with fine, dry soil. 
GROWTH OF WHEAT ON SEMIARID SOIL COMPARED WITH 
THAT ON HUMID SOIL. EXPERIMENT OF 1909. 
In this experiment Red Fife wheat was grown in 6 cylinders, 
3 filled with soil from humid eastern Nebraska and 3 with soil 
from semiarid western Nebraska, all having been saturated with 
water previous to the planting of the seed. As checks two other 
cylinders, the one filled with the humid soil and the other with 
the semiarid soil, were left unplanted but otherwise treated 
like the six planted cylinders. 
The cylinders used in the experiment were of galvanized iron, 
being 6 feet long, 6 inches in diameter, closed at the bottom and 
soldered to make the seams water-tight. In the bottom of each 
was a nipple carrying a one-hole rubber stopper to permit of 
drainage while the soil was being saturated; a glass plug placed 
in the hole in the stopper rendered the cylinder water-tight. Be- 
fore filling the cylinders with soil each was filled with water to 
test the tightness of the seams. The general construction of the 
cylinders is shown in figures 1 and 2. The latter illustrates one 
of the difficulties of working with such tall cylinders; thus, it was 
necessary to carry the cylinder into the open in order to photo- 
graph it, the roof of the greenhouse being too low. 
The semiarid soil was from the same prairie field on the H O 
Ranch as that used in the 1908 experiment, but taken to a depth 
of 6 feet, being removed from the excavation in eight portions — 
2 
