14 



vents it being exclusively employed in survey work. Certain physical 

 constants, viz., the hygroscopic coefficient and the moisture equivalent, 

 are dependent upon the mechanical composition and express in a single- 

 valued term the relative fineness of texture and have the additional 

 advantage of being readily determined. Briggs and McLane (6, p. 

 20-21) and Briggs and Shantz (8, p. 73) have proposed the formulas — 

 Hygroscopic coefficient=r0.0007 sand -(- 0.082 silt -f- 0.39 clay and, 

 Moisture equivalent=:0.02 sands -)- 0.22 silt + 1-^5 clay. Alway and 

 Rost (4, p. 10), working on the relation of the hygroscopic co-efficient 

 of loessial soils to the mechanical composition, have pointed out that no 

 one formula for such a physical constant wnll do for all soils but that 

 as an approximation for the loess soils that they were dealing with 

 the above will serve. 



Similarly Alway and Russel (5, p. 842) dealing with the moisture 

 equivalents of the same soils have developed the formula : Moisture 

 equivalent=0.14 sands -J- 0.27 silt -}- 0.53 clay. The moisture equiva- 

 lents as calculated by these two formulas are shown in columns 9 and 

 10 of table 1. The results obtained by the two formulas are ver}' 

 similar. 



Previous Comparisons. 



A limited amount of data showing the relations of the chemical 

 composition of glacial soils to the age of the drift upon which they 

 occur is already available, but none of the studies were made under 

 conditions so favorable as those existing in Rice county. Hopkins and 

 Pettit (20) in Illinois, Brown (9) in Iowa, and McMiller (25) in 

 Minnesota have each reported a study. 



Hopkins and Pettit's study was based upon Leverett's drift survey 

 of Illinois in which six glaciations had been mapped. A considerable 

 number of samples were collected from three depths on each, 0-7, 7-20, 

 and 20-40 inches, and subjected to analysis for nitrogen, carbon, potas- 

 sium and phosphorus. The averages for the dififerent drift sheets are 

 summarized in table 2. 



As the glacial soils of Illinois are generally covered by a layer of 

 loess from 3 to 10 feet or more in depth (20 p. 193) the samples were 

 not of till but of the overlying mantle, and the age of the latter is not 

 necessarily similar to that of the former. Hopkins and Pettit appear 

 to have assumed that after each glaciation the freshly exposed till 

 became quickly covered with its mantle of loess, this being derived from 

 the till sheet, a portion of which it covered, and that the loess deposited 

 on the later till sheets was derived chiefly from these and did not affect 

 the soils on the earlier glaciations. There appears no satisfactory' evi- 

 dence that these assumptions are justified and hence it is doubtful just 

 what importance we should attach to their data in connection with the 

 question of the relation of the composition of a glacial soil to the age of 



