土壌試料の調整と分析誤差に関する2, 3の考察

Abstract

This report deals with the preparation of sample and analytical errors in quantitative determination of carbon in soils. Some forest soils were used as test material for this purpose. In proportion to grinding by which the soil granule will be smaller, the value of carbon tends to increase, while it is reverse concerning analytical error represented by the coefficient of variation. Such relation, however, depends upon the sort of parent material, and effect of treatment is not always the same. Although the personal errors in chemical analysis are unnegligible, most of them would be eliminated by improvement of analytical method.