6 BULLETIN 122, U. S. DEPARTMENT OF AGRICULTURE. 
MANGANESE. 
Contrary to the experience of Gortner and Rost 1 the method for 
the determination of manganese outlined by Hillebrand 2 has given 
excellent results. If the soil was finely ground and ignited prior to 
the treatment with hydrofluoric and sulphuric acids, the insoluble 
residue 3 contained only the merest traces of manganese, and in 
many cases gave no color whatever. The residue was fused with 
sodium carbonate with a very little niter, and in case the color indi- 
cated manganese the melt was dissolved in acid and oxidized to per- 
manganate. In all cases the manganese in the residue was below 
0.01 milligram of manganous oxide, MnO. 
FERROUS IRON. 
Ferrous iron is undoubtedly present in soils, for ferrous minerals 
are common constituents, as shown by microscopic examination. 4 
But since organic matter is invariably present, no attempt was made 
to estimate ferrous iron, for there is a certainty that some ferric iron 
would be reduced during the analytical operations, and, further, if 
titration with permanganate solution were attempted some per- 
manganate might be reduced by persisting organic matter. The 
iron is therefore calculated as ferric oxide. 
LOSS ON IGNITION. 
The loss on ignition is a figure difficult to duplicate, especially in 
soils high in iron and aluminum. Although the crucibles were sub- 
jected to a comparatively low temperature and an oxidizing flame, 
there was probably some reduction by the organic matter. Since 
the sum of the constituents is in most cases well over 100, it is prob- 
able that errors due to this cause were greater than those due to the 
imperfect dehydration of hydrated minerals. 
SULPHUR. 
The ignitions with sodium carbonate and nitrate for the determina- 
tion of sulphur were made in an electric furnace, thus avoiding any 
contamination by the sulphur of a gas flame. 
RARE EARTHS. 
The results obtained for the rare earths are to be taken as qualita- 
tive in nature, that is, as showing the presence of the rare earths rather 
than the absolute amounts. However, work by different analysts 
i J. Ind. Eng. Chem., 4, 522 (1912). 
2 Bui. 422, U. S. Geol. Survey, p. 116. 
3 This residue contained zircons, fluorides of calcium and the rare earths, barium sulphate, and, strange 
as it may seem, fragments of quartz. 
* McCaughey and Fry, Bui. 91, Bureau of Soils, U. S. Dept. Agr. (1913). 
