household, at once polygamic and polyandric, sprang sometimes the polyandric family, when several brothers had a single wife in common, and sometimes the polygynic family, when a single man married or bought several women, who might, or might not, be sisters to each other.
But has the familial group evolved in the same manner all over the earth and among all races? Except for the countries previously enumerated, precise and detailed information is wanting, and we are reduced to conjectures which are more or less probable. With rare exceptions, the races which it remains for us to examine have definitely emerged from primitive familial confusion, and they have adopted either maternal or paternal filiation. Have they first passed through the familial clan with classes of fictitious or real relations ? We cannot certainly affirm it. The existence of a totem and the custom of exogamy seem to bear witness in favour of this hypothesis; but these are insufficient proofs. The totem does not necessarily imply consanguinity; and exogamy may be dictated by very diverse reasons, for we often find exogamic tribes living side by side with endogamic tribes.
What is still more general than the clan, is the institution of the maternal family, or uterine filiation; but this familial type is not invariably deduced from a previous familial clan. Among many animal species the maternal family exists without there ever having been either clan or gens. As a matter of fact, in humanity as well as in animality, the uterine family establishes itself spontaneously, whenever the male abandons the female and her progeny. This familial type will therefore necessarily appear in every horde where there is no durable pairing of males and females, of men and women. In every ethnic group living in promiscuity, for example, uterine filiation shows itself, and it will be the same under a polyandric régime, unless fictitious paternity is established. In short, for the adoption of the paternal family, it is imperative that the wives should be appropriated by a particular man, though it is of no importance whether the marriage be monogamic or polygamic. But this possession of one or more women by one man to the exclusion of all others, presupposes already a complex social condition, which has necessarily been preceded by a
