different explanation is given of the change which led to
regulation, and the establishment of the maternal family.

According to Mr. McLennan, the primitive group or horde, though
originally without explicit consciousness of relationships, were yet
held together by a _feeling_ of kin. Such feeling would become
conscious first between the mother and her children, and, in this way,
mother-kin must have been realised at a very early period. Mr.
McLennan then shows the stages by which the savage would gradually, by
reflection, reach a knowledge of the other relationships through the
mother, sister and brother relationships, mother's brother and
mother's sister, and all the degrees of mother-kin, at a time before
the father's relation to his children had been established. The
children, though belonging at first to the group, would remain
attached to the mothers, and the blood-tie established between them
would, as promiscuity gave place to more regulated sexual
relationships, become developed into a system. All inheritance would
pass through women only, and, in this way, mother-right would tend to
be more or less strongly developed. The mother would live alone with
her children, the only permanent male members of the family being the
sons, who would be subordinate to her. The husband would visit the
wife, as is the custom under polyandry, which form of the sexual
relationship Mr. McLennan believes was developed from promiscuity--a
first step towards individual marriage. Even after the next step was
taken, and the husband came to live with his wife, his position was
that of a visitor in her home, where she would have the protection of
her own kindred. She would still be the owner of her children, who
would bear her name, and not the father's; and the inheritance of all
property would still be in the female line.[25]

[25] _Studies in Ancient History_, pp. 83, _et seq._

We have here what appears to be a much more reasonable explanation of
mother-kin and mother-right than that of Bachofen. Yet many have
argued powerfully against it. Westermarck especially, has shown that
belief in an early stage of promiscuous relationship is altogether
untenable.[26] It is needless here to enter into proof of this.[27]
What matters now is that with the giving up of promiscuity the whole
structure of McLennan's theory falls to pieces. He takes it for
granted that at one period paternity was unrecognised; but this is
very far from bein