604 THE AMERICAN JOURNAL OF SOCIOLOGY
the reader who does not cast this article aside pardon me for the unavoidable imperfection of this attempt, an imperfection which is inevitable because of the newness and difficulty of the under- taking. Others will smooth and enlarge the way which I have thought necessary to open to the researches of sociology.
The equilibrium of any aggregate is never absolute. It is a rela- tive equilibrium, resulting from certain relationships among the elements of the internal structure and from certain relationships between these elements and the external environment. The structure and the life of a society are a more or less complete harmony of the elements of the society with the whole, and of this whole with the external world. The point is to explain how societies maintain themselves as structures, in a state of equi- librium of all their parts, either in'the presence of other societies or merely in the presence of their external environment, which is itself not social, although likely to become so. Thus the sea and the desert, which at first may be isolating external envi- ronments, may be transformed into internal or external social environments as soon as they become thoroughfares and are integrated into a social combination by fusion with the human factor. What are the natural limits of the structure of societies considered in their ensemble, and what are the equally natural limits of the many organs, groups of organs, and systems of groups which are united together in this structure of the whole?
Auguste Comte, following the method (chiefly subjective and deductive) adopted in the systeme de politique positive, approached the theory of the general limits of variation characteristic of the human order directly, without a preliminary study of special sociological limits. The general law of the latter can, however, be only the conclusion and the expression of those special con- ditions relating to the economic, genetic, aesthetic, scientific, moral, juridic, and political order. On the other hand, we shall persist in the course followed up to this time of proceeding from the particular to the general, from the simple to the complex, from the concrete to the abstract. We go further: according to us, the problem of sociological limits is comprehensible and
