Key points are not available for this paper at this time.
Abstract Thirty soil characteristics were used to calculate numerical indices of similarity for 29 Kansas soils. The indices varied from 90 for Ladysmith and Irwin soils (100 = perfect similarity) to 34 for Bates and Colby. In general, the numerical classification substantiated the present classification and the concepts of these soils and their interrelationships. This experiment suggests that the numerical classification of soils, like that of plants, animals, and insects, is possible. The necessary information for this kind of analysis is available for many of our soils. The major obstacle appears to be the selection and weighting, if any, of the characteristics to be considered. Once the characteristics are selected, the determination of numerical relationships is objective and repetitive.
Bidwell et al. (Sun,) studied this question.