The Use of Mitscherlich's Regression Law in the Analysis of Experiments with Fertilizers

to the statistical analysis of experiments with fertilizers. The above equation was introduced by Eilh. Alfred Mitscherlich 2 in the study of fertilization of soils in Germany. Parameter A measures a maximum yield which could not be exceeded by the use of the fertilizer in consideration. Parameter c measures the efficiency of the fertilizer and b measures the soil content of the fertilizer in the control plots in a form assimilable by the plant. Mitscherlich carried out many experiments in Germany, most of them in pots kept in green-houses. The fitting of the curve was obtained with only two levels of the fertilizer, since a constant known value was attributed to c, always the same for each fertilizer. But Kletschkowsky and Shelesnow 1 proved that c is not constant, but varies with the amounts of other fertilizers. From that time on the equation has not been used much, partly because of the apparent difficulty of fitting it by satisfactory methods. Several authors have applied very crude processes of fitting, some of which have tended to bring the use of the curve into disrepute. But in the last few years advances have been made in the theory, so that the fitting of the curve can now be obtained quickly and accurately in most cases suitable for practical application, and efficient estimates may be computed for the variances of the estimates of the parameters.