562 STATE BOARD OF AGRICJLTURE. 



of working the soil between the fingers and passing it through a sieve 

 several times. When the sample appeared quite uniformly moistened 

 it was placed in a freezing tube, the latter closed with a rubber stopper, 

 and the soil allowed to remain in the tube at room temperature for about 

 24 hours. At the end of this period the lowering of the freezing point 

 of each sample of soil was measured in the tube in which it was originally 

 placed. 



For very high moisture content somewhat larger amounts of soil 

 were taken. These soil samples were put in a large evaporating dish, 

 the desired quantity of water added to them, and then mixed and worked 

 with the hands. When sufficiently mixed they were placed in the tubes, 

 and their freezing-point depression was determined in the same manner 

 as above. Those soils which had a tendency to be sticky at the high 

 moisture content were placed in the tube most conveniently by making 

 the wet sample into a round elongated stick and dropping it into the 

 tube. 



Whenever it was desired to ascertain the freezing-point lowering of a 

 soil just as it came from the field, it received no other treatments except 

 thorough mixing. 



The object of thoroughly mixing the samples, especially at the low 

 moisture content, was to moisten all the particles of the soil. If some 

 particles of a soil sample were not moistened, the freezing point lower- 

 ing would not be accurately determined. But if they were all sufficiently 

 moistened so that their film would freeze upon supercooling the freezing- 

 point depression attained would be an average of the depression of all 

 the various particles. 



The freezing-point depression of all soils could be measured from any 

 maximum moisture content down to a very low water content. The mini- 

 mum percentage of moisture is probably slightly below the wilting co- 

 efficient of soils. Below this minimum water content solidification could 

 not be induced, at least very readily, and the results could not be de- 

 pended upon. One of the principal reasons for this is that there is not 

 sufficient free water present to separate upon freezing to bring the tem- 

 perature of the system to its proper freezing point. 



At the high moisture content the freezing point lowering of a soil 

 could be duplicated almost exactly at successive freezings, but at the low 

 moisture content the duplicate determination would vary but the varia- 

 tion was always consistent. The explanation for this phenomenon will 

 be offered subsequently. 



The difficulty previously experienced in the sudden change of the cali- 

 bration of tlie thermometer was met also in the present research. This 

 change did not always happen but it did occur occasionally. The amount 

 of change was usually insignificant and in the case of soils short of 

 saturation, could be ignored. To guard, however, against any error aris- 

 ing from this source, the thermometer was calibrated very often during 

 a series of freezing-point low^ering determinations. 



It should be emphasized that the foregoing description of the apparatus 

 and procedure contains mainly the important improvements and modifi- 

 cations made during the course of the second investigation. For fuller 

 and more complete details of the method and procedure the reader is re- 

 ferred to the first paper already mentioned. 



Having given a brief description of the method, a consideration of 

 the experimental results is next in order. 



