INTRODUCTION TO SOCIOLOGY 697
analysis and this description can be furnished only by statistics, and especially by economic statistics. From these alone can we advance to the study of the functions and organs, the apparatus and systems, and later to the study of the societies considered in their ememble at first from the concrete and descriptive standpoint, and finally from the abstract and qualitative.
Having indicated the four classes of possible modifying influ- ences on societies, Comte, entering into their details, sets forth how they operate.
1. Influence of the modifications of the inorganic environ- ment: (a] through the common longevity or interval between generations the dead governing the living; () through the diminution of the population; (^) through the more or less rapid multiplication of the population.
2. Influence of modifications of the biological environment, through the races. This influence is obscure and is badly eluci- dated. According to him, it is sometimes confounded with that exercised by the different physiques arising from the differ- ences of the inorganic environment and transmitted by heredity. He seeks to explain by the races that which he formerly explained by climates. At all events, this influence is weakened more and more by the continued mixture of races.
3. Influence of the direct modifications of society, resulting spontaneously from its own play. Comte understood by this the modifications exercised by societies upon each other. They result from the concrete multiplicity of the social centers, although in abstract sociology there may be only one people. Even without conquest, these modifications would have arisen spontaneously sooner or later, "with only slight variations." Yet he adds that this action may become systematic, so as to abridge and diminish the transitions.
These considerations are, indeed, very important and very sound, but in placing itself from the very first in the abstract point of view of a single humanity, the spontaneity of the devel- opment becomes incomprehensible, unless we adopt as the point of departure the constant multiplicity of the social factors of which I have proposed a hierarchical classification, and whose
