Summary
The fundamental objective of soil classification is the achievement of an agreed world–wide system of maximum usefulness to the widest possible range of users. Such a classification should be natural rather than artificial because soils are natural bodies. Further, it should be hierarchical rather than co–ordinate because hierarchies are readily comprehensible and memorable, can be constructed on the basis of a few characters, and do not need to give equal weight to all criteria used in classifying soils. They also have the advantage of closely denned, mutually exclusive classes which can still make allowance for occasional aberrant profiles or pedons.
The historical development of soil classifications is discussed with particular reference to the Russian and French and especially the U.S. Dept. Agric. 7th Approximation. The latter is defended against the main criticisms that have been levelled against it. It is maintained that it has been desired by sound inductive as well as deductive reasoning, that it treats soils as polythetic individuals, that it considers pedons three–dimensionally and that it permits non–hierarchical sub–classifications. Attention is given to the more practical criticisms that there is an undue emphasis on soil moisture and temperature regimes, unknown genetic criteria and the shallower horizons, and that it handles ses‐quioxidic soils inadequately. Positive advocacy is made on the ground that the 7th Approximation will facilitate and speed soil research especially in developing countries.