The agrobiologic test for normality in fertilizer experiments and variety comparisons. I. Varieties

A variety X nitrogen fertilization experiment with sugar beet seed by Bion Tolman is selected as an example showing the characteristics of a perfectly normal environment for crop increase, wherein the quantitative response of the crops to nitrogen fertilization and the relative seed-yielding abilities of three varieties of sugar beets may be accurately evaluated. The characteristics of normality are that experimental yield curves must show full homology with the curves of the standard yield diagram, which is based on the Mitscherlich-Baule yield equation y=A (1-10(-0.301x). The experimental curves obtained in Tolman's experiment coincide fully with normal curves of the diagram, and fulfill the condition that the spread between two values of y for any two values of A must increase as x increases, while the ratios between the yields of the varieties corresponding to given amounts of a plant nutrient shall be a constant. Conformity with this law is held to be a distinguishing mark of a normal soil for crop increase as regards varieties simultaneously grown thereon. Agronomists and plant physiologists are thus provided with means for distinguishing and evaluating normal and abnormal conditions for crop increase in field and pot tests. Conditions for making a most conclusive fertilizer test or variety comparison are, in the one case, to use graded amounts of one plant nutrient against two or more levels of a second nutrient in which the soil is also deficient, using a standard variety as the test crop, and in the second case to plant two or more varieties with graded amounts of one nutrient in which the soil is deficient. The resulting yield curves are then examined for the described characteristics of normality. In this agrobiologic method the diagram automatically reduces all variables, normal and abnormal, to a common denominator, which is not accomplished in any agrobiologic sense by the ordinary Fisherian analysis of variance.