Oct. 22, I917 
Direct Determination of the Hygroscopic Coefficient 149 
which has entered the small interstices of the soil from the rain or by watering.— 
Translation. 
Four years later, having completed his proof of the falsity of this view, 
he strongly discouraged determining the hygroscopicity of soils on the 
ground that it would cause confusion if very hygroscopic soils were 
indicated as retentive of an especially high percentage of water in a non- 
available form, when this was not an actual disadvantage, for the reason 
that the most hygroscopic soils were the very ones that were able to hold 
the largest amounts of total water (22, p. 248): 
By a remarkable coincidence the most strongly condensing soils usually are those 
with the highest water capacity and it would probably only cause confusion if we 
were to mark these soils with a blemish in regard to the supply of water on account of 
a not very productive correction.—Translation. 
However, he still clearly recognized that the amount of water remain¬ 
ing in a soil on the wilting of a plant was closely related to the maximum 
amount of moisture which the soil was able to absorb from a nearly satu¬ 
rated atmosphere. 
During the last quarter of the nineteenth century Hilgard and Lough- 
ridge appear to have been the only investigators who continued to attach 
any importance to the determination of the hygroscopic coefficient in 
connection with soil-moisture studies. The general attitude of soil inves¬ 
tigators during that period is well illustrated by the numerous publica¬ 
tions of one of the foremost, the late Dr. F. H. King, in which the deter¬ 
mination, the importance, and the use of the hygroscopic coefficient are 
ignored. 
At the beginning of the present century Mitscherlich began to call 
attention to the importance of the relative hygroscopicity of soils (23) 
to which he attached extreme importance. In Europe the importance 
of a knowledge of the hygroscopicity of soils is still practically ignored 
except by Mitscherlich and by Hall (rr, p. 84-88). 
Naturally, under such circumstances, the particular method developed 
by Hilgard received little critical consideration. It appears that no one 
except himself had investigated the method with the object of determin¬ 
ing the experimental errors involved, previous to our beginning the use 
of it on an extensive scale for the interpretation of field-moisture data. 
We had originally no intention of studying the method itself, and our 
observations have resulted from difficulties encountered in its use, part 
of them due to attempts to expedite the determination without lessening 
its accuracy. 
Lipman and Sharp (18), while associated with Hilgard, studied the 
effect of a rise or fall of the temperature and also of the thickness of the 
exposed layer of soil upon the amount of hygroscopic moisture absorbed. 
Briggs and Shantz (8, p. 64) studying the relationship between the 
moisture equivalent and the hygroscopic coefficient found that the deter- 
