142 Journal of Agricultural Research voi. xviii. no. 3 



Potassium. — ^There is a trend upward in quantity of this solute in 

 soil No. I from the lowest to the highest moisture content; in soil No. 2 

 no significant effect is produced in the first two sets of determinations; 

 however in the two latter it is seen that there are increasing quantities 

 of potassium from soils of 5 per cent to 15 per cent moisture, and a 

 decrease in the soil of 20 per cent moisture. 



Calcium and magnesium. — The striking feature is the depressing 

 action of the excess water in soil No. 2 and the almost identical per- 

 centages of depression for the two solutes at the corresponding sampling 



date. 



GENERAL DISCUSSION 



In preparing the water extracts it was observed that the extract of 

 the soil at the lowest moisture content filtered much faster than that at 

 the highest, with a regular gradation between the two extremes, and that 

 the difference in filtering speed was greater between the silty clay loams 

 of highest and lowest moisture content than between corresponding 

 samples of the sandy loam. A reasonable explanation may be found 

 in the physical structure of the soil at the different moisture contents. 

 It has been pointed out by Fippin^ that the continued wet condition or 

 dry condition does not produce any change in structure, but that alter- 

 nate wetting and drying produce granulation, caused by expansion and 

 contraction, which are directly proportional to the degree of wetting 

 or drying. Thus the dryer silty clay loam may approach in structure 

 that of a sandy loam. Klein ^ suggests that the increased granulation and 

 hence greater access of water to the soil particle by drying may be over- 

 come by the greater amounts of material held soluble in the soil at the 

 higher moisture contents. 



When a condition of saturation is reached there is a marked depression, 

 as has been noted. The sandy loam soil at 20 per cent moisture was 

 saturated, and water was standing on its surface a quarter of an inch in 

 depth. This in itself is an indication of the results which might have 

 been expected, and which were in fact obtained. This soil contained 

 5 per cent more water than its optimum, while soil No. i with 25 per cent 

 moisture was 3 per cent above its optimum content and showed no 

 apparent depression of solutes. Silty clay loam soils have a greater range 

 between their optimum moisture contents and their saturation points 

 than sandy loam soils because of the difference in soil texture. 



The range of variation in moisture content covered in the present 

 study was very much greater than the variations occurring in the investi- 

 gations^ referred to in the early part of this paper. In a season's work 

 with these two soils it was observed that for soil No. i the moisture con- 



'FippiN, Elmer O. somb causes of soii. granulation. In Proc. Amer. Soc. Agron., v. 2, 1910, p. 

 106-121, fig. 11-18. 1911. 



' Klein, Millard A. studies in THE drying of soils. In Jour. Amer. Soc. Agron., v. 7. no. 2, p. 49-77. 

 3 fig. Bibliography, p. 75-77. 1915. 



* Stewart, Guy R. op. cit. 



