INORGANIC COMPOSITION OF SOME AMERICAN SOILS. 5 
When received at the laboratories the sample was spread on brown 
wrapping paper and allowed to air-dry. It was then passed through 
an iron sieve of 6 meshes to the linear inch. The soil clods not 
passing were crushed with a wooden rolhng-pin on brown paper and 
passed through the sieve. The stones, sticks, roots, etc., were 
rejected. 
The soil passing through the sieve was well mixed and quartered 
down to a subsample of 7 to 10 pounds. This was crushed in an 
iron mortar to break up the larger soil particles, such as iron and 
manganese concretions and the like. It was then quartered down 
to a sample of 50 grams, which was ground till it passed entirely 
through a silk bolting cloth of 100 meshes to the linear inch. 
In this work of subsampling and grinding great care was taken, 
for it was recognized that in some cases the rarer elements were 
segregated in comparatively large pieces of their mineral species. 
For instance, manganese is often found in large concretions and 
zircons occur in fairly large, hard crystals. If such samples were 
not carefully ground and mixed, that portion taken for analysis 
might contain an undue number of these crystals or concretions. 
NOTES ON THE METHODS OF ANALYSIS. 
For the determination of the major constituents the soil was first 
ignited to destroy organic matter, then fused with sodium carbonate, 
following the procedure outlined by Hillebrand. 1 Where there were 
choices of different methods or special difficulties encountered, varia- 
tions were adopted. Notes explaining these points are given below. 
IRON AND ALUMINUM. 
Generally soils are higher in iron and aluminum than rocks and 
contain but little manganese and comparatively small quantities of 
calcium and magnesium. The precipitate of the iron and aluminum 
group, obtained by the use of sodium acetate, is difficult to handle 
when large, and has a tendency to run through the filter. Although 
this part was filtered off when the filtrate from the iron group was 
concentrated, the precipitate of the calcium oxalate contained much 
more iron and aluminum than when the iron group was precipitated 
with ammonia. Accordingly precipitation of this group was made 
with ammonia. Manganese may have been present in the iron 
group precipitate, but in small quantity only, for nearly all was 
recovered in the ammonium sulphide precipitate, as shown by a 
comparison of the amount obtained in this way with the total amount 
determined by the use of hydrofluoric acid on another sample of the 
soil. 
i Bui. 422, U. S. Geol. Survey (1910). 
