EXPERIMENT STATION BULLETINS. 531 



The extensive studies on tlie subject of colloids have shown that soils 

 contain substance of this nature, and that some classes of soil possess 

 them more abundantly than others. In what structural form they exist 

 in the soil is not as yet definitely known ; but evidences seem to indicate 

 very strongly that a part of them at least exists as a gel coating 

 around the solid particles, and another part as individual gel aggre- 

 gates. Now, it is also known that colloidal substances swell when they 

 absorb water, and that this property of swelling increases with tempera- 

 ture. Their increase in volume at high temperature is probably due, (1) 

 to their inherent constitution, and (2) to the expansion of water they 

 contain. 



If soils, therefore, containing colloidal substances are placed in firm 

 and inflexible tubes, such as those used in the present research, in order 

 to oppose any horizontal thrust, and the temperature of these soils is 

 varied, the rate of flow of water through them, will be governed by the 

 viscosity of water, the structure of the soil mass, and the colloidal con- 

 tent of the soil. At about 4° C. the last two factors, and especially the 

 last, favor rapid percolation, but the first factor tends to reduce it. As 

 the temperature rises, the viscosity of water diminishes rapidly and fa- 

 vors the flow of water, but the colloidal material swells and tends to re- 

 duce the rate, by filling up the pores through which the water flows. 

 This swelling may not be considerable at the lower temperatures, and 

 allows the reduced viscosity to predominate, so that the velocity of perco- 

 lation increases with rise in temperature. When a certain temperature 

 is reached, however, the increase in volume, due to swelling of the colloids 

 and expansion of water attains a considerable magnitude, and since the 

 volume of the vessel remains constant, the mass of soil is subjected to 

 a strain: some rearrangement of the particles takes place; many of the 

 pores are closed up, and the rate of percolation commences to decrease. 

 At this temperature, therefore, the factor of swelling predominates over 

 the decreased internal friction of the liquid. With a further increase in 

 temperature this factor of swelling and the consequent closing up of the 

 soil interstices, is augmented. The total' result, will be, therefore, just 

 what the experimental data really show. On the other hand, as the soil 

 mass is maintained under a constant strain at the higher temperature, 

 there may occur some radical change in the structure of the soil mass 

 which will favor a high rate of percolation. Such changes were observed 

 in some of the experiments. 



In a sandy soil, however, the conditions are different. Here, there are 

 no colloidal substances to swell and close up the pores, and consequently 

 the rate of percolation follows more closely the law of viscosity. This 

 rate, however, is modified by a change in the structure that takes place 

 as the temperature rises. This change in structure in sand, seems to 

 favor a more rapid flow of water. The abnormal velocity of percolation 

 noted in this soil is due to the internal alteration in structure. That a 

 change occurs in structure by varying the temperature is further con- 

 firmed by the following observation : When the rate of passage of water 

 was determined say at 20° C. and then at 50° C. and the temperature 

 was brought to 20° C. again, the two readings at 20° C. were not at all 

 the same. 



It thus appears that the unexpected trend of the results is explainable 

 by the swelling of the colloidal material, and by the change in structure 

 in the soil mass, with rise in temperature. 



