16 BULLETIN 122, U. S. DEPARTMENT OF AGRICULTURE. 
Gloucester stony loam, Nos. 11 and 12, shows a greater amount of 
silica in the subsoil and the iron and aluminum does not follow the 
general rule of being greater in amount in the subsoil. It would 
seem that the samples used in analysis might have been transposed, 
but the analysis of a second sample showed the same relation. This 
type is an exception to the rule, though the difference in the compo- 
sition of the surface soil and the subsoil is not great. 
Potash and magnesia are noticeably higher in the subsoil. In the 
Durham sandy loam, Nos. 19 and 20, the sands, or larger soil parti- 
cles, are mostly potash feldspars. This fact accounts for the greater 
amount of potash in the surface soil. Barium and lime do not seem 
to have been concentrated in either layer of the soil. Zirconium 
appears in the surface soil in larger amounts. The particles of mineral 
containing this element are comparatively coarse and resist surface 
wash. 
The average amount of sulphur trioxide present in these soils is 
0.13 per cent, whereas phosphorus pentoxide averages 0.14 per cent. 
In one case there is one-fourth as much sulphur trioxide as phos- 
phorus pentoxide. These facts support the evidence brought out by 
Hart and Peterson * and further by Shedd 2 that the sulphur content 
of soils is lower than that of phosphorus. The average of 16 Wiscon- 
sin soils was 0.08 per cent S0 3 and 131 Kentucky soils averaged 0.06 
per cent S0 3 . Hart and Peterson point out that average crops of 
cereals remove in the grain and straw two-thirds as much sulphur as 
phosphorus, grasses about the same, alfalfa somewhat more, and 
cabbage and turnips 2 to 3 times as much. They advise that sulphur 
be considered in fertilizer practices. 
MINERALOGICAL COMPOSITION. 
The micro-petrographical determinations were made by W. J. 
McCaughey, now of the University of Ohio, and William H. Fry, of 
this bureau. The examinations were made on separates made by 
mechanical analysis. 
Quartz was universally present in all the soils examined and its 
presence is not noted in the individual reports. 
(1) NORFOLK SANDY LOAM, SOIL. 
No. 4 sand. — Practically pure quartz sand with apatite and rutile 
inclusions. 
No. 5 sand. — Nearly pure quartz sand, with a. few subangular, 
though mostly clear, well-rounded grains. Minerals other than 
quartz, 3.5 per cent. Rutile, zircon, ilmenite, and tourmaline 
present. 
i Wis. Agr. Expt. Sta., Research Bui. No. 14. 2 Ky. Agr. Expt. Sta. Bui. No. 174. 
