g true. The idea of the father's relationship to the
child is certainly known among the peoples who trace descent through
the mother; the system is found frequently where strict monogamy is
practised. Again, Mr. McLennan connects polyandry with mother-descent,
regarding the custom of plurality of husbands as a development from
promiscuity. Here, too, he has been proved to be in error. Whatever
the causes of the origin of polyandry, it has no direct connection
with mother-kin, although it is sometimes practised by peoples who
observe that system.

[26] _History of Human Marriage_, pp. 51-133. It is on this
question that my own opinion has been changed, compare _The
Truth about Woman_, p. 120.

[27] See next chapter on the Patriarchal Theory.

For myself, I incline to the opinion that the system by which
inheritance passes through the mother needs no explanation. It was
necessarily (and, as I believe, is still) the _natural_ method of
tracing descent. Moreover, it was adopted as a matter of course by
primitive peoples among whom property considerations had not arisen.
Afterwards what had started as a habit was retained as a system. The
reasons for naming children after the mother did not rest on
relationship, the earliest question was not one of kinship, but of
association. Those were counted as related to one another who dwelt
together.[28] The children lived with the mother, and therefore, as a
matter of course, were called after her, and not the father, who did
not live in the same home.

[28] Starcke, _The Primitive Family in its Origin and
Development_, pp. 36, 37.

All these questions will be understood better as we proceed with our
inquiry. The important thing to fix in our minds is that mother-kin
and mother-right (contrary to the opinion of McLennan and others) may
very well have arisen quite independently of dubious fatherhood. It
thus becomes evident that the maternal system offers no evidence for
the hypothesis of promiscuity; we shall find, in point of fact, that
it arose out of the regulation of the sexual relations, and had no
connection with licence. It is necessary to understand this clearly.

Bachofen is much nearer to what is likely to have happened in the
first stage of the family than Mr. McLennan, though he also mistakenly
connects the maternal system with unregulated _hetairism_. Still he
suggests (though it would seem quite unconsciously) the patriarchal
hypothesis,