EXPERIMENT STATION BULLETINS. 015 



ill the highest degree the power to form substances of hydration. Dry 

 hydrogels upon rewetting, are believed to enter into chemical combina- 

 tion with water in order to resume their colloidal state. 



If the water is physically adsorbed it is conceived that it is removed 

 from the liquid phase and becomes practically solid like water of crys- 

 tallization. It becomes then inactive or inoperative in the same man- 

 ner as when a solute is adsorbed. 



In which form of the above the inactive water exists, cannot be 

 stated definitely. It is very probable that it exists in both forms. But 

 in whichever form it might exist, the assumption that it is inactive or 

 unavailable to dissolve the salts tends to explain the results, at least 

 partially. Thus, for example, assuming that a clay soil adsorbs or 

 combines chemically with 15% of water, and that this clay at 36% 

 moisture causes a lowering of the freezing point of 0.034°C, and at 

 18%, .955°C, then at the former moisture content there is 21% of water 

 available to dissolve the salts in the soil, Avhile in the latter case there 

 is only 3% available for tlie same purpose. It would be natural, there- 

 fore, that the depression of the freezing point would be many times 

 greater at the low percentage of moisture content than at the high, 

 than would be expected from the difference of the total moisture con- 

 tent, just as the experimental data really indicate. 



The foregoing hypothesis, therefore, would tend to explain, at least 

 partially, why the lowering of the freezing point of the various types 

 of soil with the exception of quartz sand and possibly some extreme 

 types of sandy soil, should increase in a geometric progression while 

 the percentage of water decreases in an arithmetic progression. It 

 would also tend to explain why the lowering of the freezing point of 

 quartz sand should increase inversely proportional with the water con- 

 tent, which would tend to indicate that quartz sand causes very little, 

 if any, of the water to become inactive. That the latter fact might be 

 so is further shown by the solidification which can be induced when the 

 moisture content is only about 0.7%, practically at the air dry state. 



Mention has already been made that in conducting a preliminary 

 series of tesls to ascertain how close the lowering of the freezing point 

 of a soil sample could be duplicated it was found that at a high mois- 

 ture content the duplicate readings would agree almost perfectly with 

 one another for all types of soil but at a low moisture content the dupli- 

 cate readings would agree with one another very close only with some 

 soils, such as tlie quartz sand, kaolin, burned silicic acid, etc., but with 

 other; soils, such as the clay, loam, etc., the duplicate readings, would 

 vary appreciably, and indeed the first depression would be always greater 

 than the second. This latter phenomenon was made of further special 

 study by determining the lowering of the freezing point of a number 

 of tyi>es of soil at low moisture content at successive freezings. The 

 freezing process consisted of either allowing the soil sample to 

 freeze at its proper freezing point and then thawing it by holding the 

 tube containing the soil in the hand, or allowing the temperature of the 

 soil to fall 3 or 4°C after it had frozen and then thawing it as before. 

 Some of the results thus obtained are shown in Table S. The magni- 

 tude of the decrease of the depression due to the successive freezings 



