370 



Journal of Agricultural Research 



Vol. XII, No. 6 



error from this source, determinations on pure water were checked 

 several times each day. The general magnitude of error in technic is 

 indicated by the following typical instances of duplicate experiments: 



Soil No. 



loA 

 loA 

 iB. 

 iB. 



I5A 



Freezing- 

 point de- 

 pression. 



°c. 

 0.065 



. 062 



.066 



. 069 



.049 



Soil No. 



ISA 

 14A 

 14A 

 4B. 

 4B. 



Freezing- 

 point de- 

 pression. 



°c. 



o. 052 

 • 043 

 .050 

 •157 

 •15s 



DESCRIPTION OF SOILS USED 



In the article by Stewart {8) the reader will find a detailed description 

 of the soils under the same laboratory numbers. For the sake of con- 

 venient reference, a list of the soils used is included here (Table I). 

 Each soil was investigated under three general conditions: (i) Soil from 

 tanks in which barley crop was grown for two seasons ; (2) corresponding 

 soils kept under identical conditions but with no crop the second season ; 

 (3) soils air-dried, sifted, and kept two years in closed bins. All tank 

 soils were kept constantly at nearly optimum moisture contents with 

 distilled water. 



Table I. — Description of soils used in this investigation 



UNFREE WATER OF SOILS 



One of the most important factors affecting the concentration of the 

 soil solution is the moisture content. This has already been pointed 

 out by Bouyoucos and McCool, and the former has recently devised a 

 dilatometer method for more accurately measuring the unfree water 

 of a soil (7). In the present investigation an attempt was made to 

 approximate the unfree water in each soil by careful determinations of 

 the freezing-point depressions at different moisture contents. These 

 data could then be used in reducing the observed depressions to definite 

 and comparable moisture percentages, especially where the observed 



