Apr. 9,1917 
Water-Retaining Capacity of Soil 
31 
water supply by capillarity is not an important factor in crop production on Nebraska 
upland soils (9, p. 10). 
The hygroscopic coefficients of the soil samples were not determined, 
but the mechanical analyses of eight sets of samples from the first 3 feet 
permit a calculation of these values by the Briggs and Shantz formula 
(8, p. 73); these vary from 6.1 to 8.5. From the extremes of moisture 
found he considers 16 to 18 per cent the maximum amount of water the 
soil can retain against gravity, and 7 to 8 per cent its minimum point of 
available water (9, p. 18-19). From this it appears that on that type of 
soil the hygroscopic coefficient is approximately the lower limit of avail¬ 
able moisture, and that the maximum water content when downward 
movement ceases lies between 1.8 and 2.6 times the hygroscopic coeffi¬ 
cient, if it is assumed that all the soil samples taken are sufficiently similar 
to justify such a comparison. This is in accord with our findings reported 
below. 
CHARACTER OF SOILS USED 
The soils were selected to represent some of the most important types 
of Nebraska, especially those of loessial origin, and not the whole range 
in texture from coarse sands with a hygroscopic coefficient less than 0.5 
to clays with one in excess of 20. They include (Table I) six silt loams 
derived from the loess, five loams of residual origin, and one dune sand. 
Soil D is surface soil, Marshall silt loam, from the Experiment Station 
farm at Eincoln, and A the corresponding subsoil, taken from the third 
to the fifth foot. E and H similarly represent the surface soil and subsoil 
of the substation farm at Culbertson, and C and G corresponding depths 
of a prairie near McCook, both on Colby silt loam. Soils I, J, K, L, and 
M are from areas of residual soil mapped by the United States Bureau of 
Soils as belonging to the Sidney series (22, p. 58). I and K are surface 
soil and subsoil from a loam near Imperial, and M and L from a sandy 
loam near the same place. Soil J is a subsoil from the silt loam near 
Madrid, part of a bulk sample on which various studies (2, p. 4653, p. 249) 
have previously been reported. K and L are from the same depths as 
A, H, and G—viz, 3 to 5 feet—while J was from the fourth to the sixth 
foot. 
Soil Q is a dune sand taken from a “blow-out” near Dunning, and is 
typical of the subsoil of the very extensive sand-hill region. 
Soil B, the only soil from outside Nebraska included in the study, is 
from the Sulphur Spring Valley Dry-Eand Experimental Farm, north of 
Douglas, Ariz. This was included because of its interesting conduct in 
the field. It was taken from the surface of a field which after two years 
of clean summer fallow, without bearing any crop at all, was found to 
contain no available water to a depth of 2 feet. 
