EXPERIMENT STATION BULLETINS. 585 



stances, the bydrogels of alumina, iron, manganese, hydrated silica, etc.; 

 that the hydrosols are continually being produced in the soils by physical 

 action such as that of water on clay, or by chemical changes such as the 

 decomposition of silicates ; and, that there are in nature various agencies 

 such as changes in temperature, concentration of solution, etc., by which 

 hydrosols are changed to hydrogels and vice versa. Both heating and 

 freezing have been found to cause coagulation of the hydrosols and hy- 

 drogels and this coagulation is accompanied bj' a decrease in the ab- 

 sorptive power. It has been abundantly proven for example that when 

 clay is heated its absorptive power for water is greatly diminished. 



The foregoing hypothesis appears to be further confirmed by the fol- 

 lowing experimental data: (1) When some of the soils in Table 10 

 which show a large diminution in the depression upon refreezing were 

 heated to red heat and then their depression was determined at a very 

 low moisture content in successive freezings, it was found that the 

 magnitude of the depression remained practically constant at successive 

 freezings. The logical inference that may be drawn from these results is 

 that heat coagulated or destroyed all the colloids and probably no un- 

 coagulated colloids were left to be coagulated by freezing and thereby 

 the inactive water to become available. 



(2) When the same unheated soils as above were frozen in the di- 

 latometer it was found that the amount of expansion in the second freez- 

 ing was considerably greater than that in the first, indicating that there 

 was a larger amount of free water to freeze in the second freezing than 

 in the first. These results are considered by us to be an absolute proof 

 that the process of freezing causes inactive water to become free or 

 .available. 



The foregoing hypothesis, therefore, would seem to explain very well 

 the diminution of the lowering of the freezing point of the agricultural 

 soils at low moisture content with repeated freezings. It would apppear 

 also to suggest that such substances as quartz sand, kaolin, burned soils, 

 etc., either do not contain colloidal material or if they do it is not 

 coagulable, and consequently not of the same nature as that of the agri- 

 cultural soils. It might be. however, that these substances reabsorb the 

 water made available during freezing, upon thawing. If that is the 

 case then it w^ould mean that freezing did not cause a permanent change 

 in the physical condition of the substances, and consequently the above 

 conclusion appears valid. 



At the high moisture content coagulation of the colloids of the agri- 

 cultural soils should also take place but since the percentage of water 

 present is high and the soil solution is very dilute, the amount of water 

 that is released upon coagulation is too small to make any noticeable 

 effect upon the depression. Consequently, the duplicate deterniinations 

 agree in soils with high moisture content, but do not with low water 

 content. 



