80 
Research Bulletin _\ o. 3 
considerable difference in the rooting habits of some crop plants 
between the eastern and the western portions of the loess in Ne- 
braska, and to determine whether there was a gradual change 
from west to east this experiment was carried out in 1911. The 
location of the places where the soils were secured are indicated 
on the map in figure 20. At each of the five places an excavation 
was made a ad about 150 pounds of soil taken from each of the first 
6 feet. These bulk samples were shipped to the Experiment Sta- 
tion where each soil, in dry condition, was reduced to a fine state 
of division and mixed by passing it several times thru a one- 
eighth-inch screen as previously described (page 17). This screen- 
ing at the same time served to remove nearly all the roots. Three 
cylinders were filled with the soils from each of these places. 
Evaporation 5 ft. 4^ ft 3 ft. 
Fig. 20. Map showing the loess region of Nebraska (unshaded) and the 
sources of the samples used, as well as the normal annual precipita- 
tion and evaporation. 
Besides the 15 cylinders mentioned, 8 others w r ere used in the 
experiment. Six were filled with the same H O soils used in 1910, 
but instead of keeping the upper and the lower half of the surface 
foot separate the previously unused portions of these were mixed 
and used as one lot of soil. The hygroscopic coefficient was found 
to be practically the same as the average of the two parts used 
in 1910. These 6 are referred to as the H O cylinders. The sur- 
face 6 inches of soil from a field from the Experiment Station 
Farm was used to fill two cylinders which are referred to as those 
of Surface Soil. 
The hygroscopic coefficients of the soils used are given in Table 
19. The numbers given for the H O soils and for the surface soil 
are the averages of a large number of determinations, and those 
for the others are the averages of three determinations made on 
different days. 
