INDUCTION has greatly predominated over deduction throughout the foregoing chapters; and readers who have borne in mind that Part II. closes with a proposal to interpret social phenomena deductively, may infer either that this intention has been lost sight of, or that it has proved impracticable to deal with the facts of domestic life otherwise than by empirical generalization. On gathering together the threads of the argument, however, we shall find that the chief conclusions forced on us by the evidence are those which Evolution implies.
We have first the fact that, little as it might have been expected, the genesis of the family fulfills the law of Evolution under its leading aspects. In the rudest social groups nothing to be called marriage exists: the unions of the sexes are extremely incoherent. Family groups, consisting of mothers and such few children as can be reared without permanent paternal assistance, are necessarily small and soon dissolve: integration is slight. Within each group the relationships are less definite; since the children are mostly half-brothers or half-sisters, and the paternity is often uncertain. From such primitive families, thus small, incoherent, and indefinite, there arise, in conformity with the law of Evolution, divergent and redivergent types of families—some characterized by a mixed polyandry and polygyny; some that are polyandrous, differentiating into the fraternal and non-fraternal; some that are polygynous, differentiating into those composed of wives and those composed of a legitimate wife and concubines; some that are monogamous, among which, besides the ordinary form, there is the aberrant form distinguished by a wife
- Conclusion to the chapters on "The Domestic Relations," which complete vol. i. of the "Principles of Sociology."