EXPERIMENT STATION BULLETINS. 569 



corresponding depressions between tlie liglit and lieavj types of soil wiil 

 be revealed subsequently. In the meantime the reader is warned not to 

 conclude that the variation in the depression between the different types 

 of soil and especially between the closely related types signifies absolute 

 difference in total salt content. This is on account of the fact that the 

 free moisture content of the different soils is probably not exactly equiv- 

 alent, and free water, as will be subsequently seen, is one of the con- 

 trolling factors in the freezing point lowering. 



The degree of concentration in parts per million of solution and of 

 osmotic pressure in atmosphere is also very different at the two moisture 

 contents and in all soils, which, of course, is quite natural. At the high 

 moisture content the concentration varies from 3GG p. p. m. in the case 

 of sand (Soil No. 12) to 1829 p. p. m. in clay (Soil No. 33). The osmotic 

 pressure in the same soils range from .181 to .994 atmospheres in the 

 sand and clay respectivel3^ At the low percentage of moisture the con- 

 centration ranges from 92G8 p. p. m. in the case of sand (soil No. 12) to 

 33,415 p. p. m. in the heavy silt loam (soil No. 1). The osmotic pressure 

 for the same soils is 4.58 and 1G.48 atmospheres for the sand and silt 

 loam respectively. 



The data in Table 3 also reveal the remarkable fact which was observed 

 ii! the original investigation, namely, the lowering of the freezing point 

 increases in all soils with the exception of quartzi sand and possibly of 

 some of the extreme types of sand, at a far greater rate than the per- 

 centage of water decreases. In other words the ratio of the freezing- 

 point depression and of the percentage of water are not inversely pro- 

 portional (approximately) as might be expected, save only in the quartz 

 sand and some of the sands. Thus, Minnesota clay (Soil No. 33) at 

 39.837o of moisture gave a depression of .075°O and at 22.39%, .987°C, 

 the ratio of the percentage of water at the low and high moisture con- 

 tents is only 1.78 while that of the depression is 13.1G. Kentucky Car- 

 rington clay loam (Soil No. 42) which at 92.7G% of moisture produced 

 a lowering of the freezing point of .039°C while at 39.28%, 1.075, shows 

 a ratio of 2.37 in percentage of water and 27.56 in the lowering of the 

 freezing point. In the case of the quartz sand, however, the ratio of the 

 depression is directly inversely proportional to the water content. Thus, 

 at 1.5% of water the depression is .070°C and at 15%, .007°C, the inverse 

 ratio, therefore, being 10 to 10 respectively. Apparently the quartz sand 

 yields entirely different results from those of the soils, especially the 

 heavier types. 



MATHEMATICAL LAW WHICH THE FRBEZING-POINT LOWERING OF SOILS AT 

 DIFFERENT MOISTURE CONTENTS FOLLOWS. 



In the first investigation it was discovered that the results of both the 

 quartz sand and of the soils followed distinct and definite mathematical 

 laAvs. The results of the quartz sand and sands could be mathematically 

 expressed by the simple equation M -D = K, wdiere K is a resultant 

 contant, M the percentage of moisture content and D, the observed de- 

 pression of the freezing point. While the results of the other soils fol- 

 lowed an entirely different mathematical law and indeed the geometric 

 progression law. That is, the lowering of the freezing point of the soils 

 increased in a geometric progression while the percentage of water de- 



