196 
Journal of Agricultural Research 
Vol. VIII, No. 6 
15 per cent of the water to become inactive or unfree either physically 
or chemically, or both, and at 39 per cent of moisture this clay gives a 
depression of 0.075 0 and at 22 per cent, 0.987° C., then in the first case 
there is 24 per cent of moisture free or active to dissolve the salts, while 
in the second case there is only 7 per cent of free or active water for the 
same purpose. The depression of the freezing point at the low moisture 
content therefore would be many times greater than that at the high 
than would be expected from the total percentage of water content. 
It was also found that the magnitude of the lowering of the freezing 
point of soils at the low moisture content decreased with successive 
freezings. This was true, however, only with complex and colloidal 
types of soil, such as the silts, loams, and clays, and not with the simple 
and noncolloidal soils, such as the sands and light sandy loams. 
In explanation of this phenomenon the hypothesis was offered that a 
large portion of the water which was made inactive or unfree and thus 
removed from the field of action as far as the freezing-point lowering is 
concerned, was due to the colloids which the soils contain. This inactive 
or unfree water existed in the colloids both as physically adsorbed and 
loosely chemically combined water. Upon freezing, these colloids coag¬ 
ulated, the bonds uniting them with the water broke, and the combined 
or unfree water became liberated. This liberated and free water went to 
dilute the original solution and thus decreased the lowering of the freezing 
point. Thus, for instance, if there were 5 per cent of moisture at the 
beginning of the first freezing, there were probably 7 per cent at the end 
of the first freezing, 7.5 per cent at the end of the second freezing, and so 
on, until all the colloids were coagulated. 
It was further discovered in these researches that solidification could 
not be induced to take place below a certain minimum moisture content, 
but it could be started extremely easily above this critical point. This 
minimum water content varied with the type of soil and appeared to 
correspond quite closely to the moisture content at which plants begin to 
wilt, or the wilting coefficient of soils. Since the percentage of moisture 
in many of the soils was quite high, amounting in some clays to 20 per 
cent and in some humus clays to 40 per cent and yet solidification could 
not be started, it appeared logical to conclude that the remaining water 
in the soil did not exist in a free state but in an inactive or unfree 
condition. 
All evidences therefore seemed overwhelmingly in favor of the view 
that soils cause some of their moisture to become inactive or unfree and 
thus lose its solvent action and be removed from the field of action, so 
far as the freezing-point lowering is concerned. 
The desire now arose to measure quantitatively the amount of this 
water which the different soils cause to become inactive or unfree and 
lose its solvent action. Such a determination, it was thought, would be 
of considerable scientific and practical importance. 
