218 



SCIENCE 



[Vol. LVI, No. 1443 



per cent. This gives a ratio of potash to silica, 

 alumina and iron of 1 : 26.7 or very nearly 

 the ratio in the igneous rocks. The "terrigenous 

 clay" of the deep marine sediments has a rather 

 larger content of SiO^ and a somewhat smaller 

 content of KjO, giving a ratio of 1 : 36.4. 



The average composition of the soil colloids 

 shows somewhat less SiO^ than is contained in 

 igneous rocks but considerably more Al^Oj and 

 Fe^Og, indicating probably that much of the 

 quartz from the disintegrating rock remains as 

 such in the soil. On the other hand, it is seen 

 that the marine sediments have gained in SiOg 

 over the soil colloid for which we can offer at 

 present no explanation. 



An examination of the Miami silty clay loam 

 soil shows the following very interesting figures. 

 If we assume a weight of 3,500,000 pounds per 

 acre foot, the composition on a water-free basis 

 with 81.11 per cent, of SiOj shows 2,838,850 

 pounds of SiO, per acre; ALO, at 10.46 per 

 cent, gives 366,100 pounds per acre; Fe,03 at 

 2.80 per cent, gives 98,000 pounds per acre, 

 while K^O at 2.27 per cent, gives 79',450 pounds 

 per acre. The amount of soil colloid at 19.3 per 

 cent, gives 675,500 pounds of colloid per acre 

 foot. The SiO, content of this colloid amounts 

 to 365,780 pounds, the Al^Og 189,340 pounds, 

 and the Fe^O, 90,449 pounds, while the K„0 

 content of the colloid is 19,522 pounds per acre. 



The total soluble salts of the Miami silty 

 clay loam soil collected after passing through 

 the Pasteur-Chamberlain filter, after treatment 

 of one part of soil to five parts of water, 

 amounts to 60 parts per million of the soil. 

 The SiO. in soluble form in this soil collected 

 under these conditions is therefore 22.43 pounds 

 per acre and the A\fi^ and Fe^Og 6.16 pounds, 

 while the K,0 is 27.16 pounds per acre. 



This bureau has determined from observation 

 of river waters the presence of colloidal mate- 

 rial but has made no analysis to detei-mine the 

 amount or composition of ithis colloid. Unfor- 

 tunately, in all of the river analyses that have 

 been made the amount of silica, alumina and 

 iron in colloidal solution owing to the lack of 

 methods has not iieen determined and has not 

 been included ii the evidence presented of 

 chemical denudati^.>fl.. It seems likely that the 

 colloids carried by the rivers approximate in 

 composition the soil colloid and arguing from 



the figures above given pertaining to the Miami 

 silty clay loam soil the 2.13 per cent, of K^O 

 and the 9.24 per cent, of silica, alumina and 

 iron as reported for the average river water 

 gives only one part of the story. If the ratio 

 of 1 : 1.6 of potash to silica, alumina and iron 

 of the soluble salts is changed to the ratio of 

 1 : 33 as in the colloidal solution, then in addi- 

 tion to the soluble potash, silica, alumina and 

 iron as reported in the river water analyses, we 

 will have a colloidal solution carrying, in its 

 dry parts, 2.89 per cent, of potash, 54.15 per 

 cent. SiO,, 28.03 per cent. AI.O3 and 13.39 

 per cent. Fe^Og. 



The total additional amount of potash in the 

 colloidal matter carried by the rivers, as com- 

 pared with the total amount of potash in true 

 solution, and the total amounts of SiOj, AUOg 

 and Fe^Oj carried in this colloidal matter, as 

 compared with the amounts carried in true 

 solution, can not be determined without further 

 investigations and further determinations of 

 the material carried by the river water. If we 

 knew the amount of soluble material in the 

 river water, the amount of colloidally dissolved 

 material and the amount of unaltered rock 

 fragments there would still be one item which 

 would make comparison difficult and would be 

 seemingly difficult to evaluate, and that is the 

 amount of CO., that entei-s from the air and 

 fi'om decomposing organic matter, which would 

 combine with some of the constituents notice- 

 ably with lime, either in the soil or in the river 

 water. It seems likely, however, from the con- 

 sideration of the facts given that this would be 

 a small factor. 



Until we evaluate all these factors and par- 

 ticularly until we determine the actual loss 

 through chemical denudation of silica, alumina, 

 iron, potash and other electrolytes in the col- 

 loidal state carried by the rivers, we are in no 

 position to even speculate as to Avhether chem- 

 ical erosion is a selective process which might 

 change the chemical composition of the soil or 

 whether materials leave the soil in aboiit their 

 original ratios which would not change mate- 

 rially the chemical composition of the soil 

 upon which water has acted. 



Milton Whitney 



Bureau of Soils, 

 U. S. Dept. of Agriculture, 

 Washington, D. C. 



