dence against this view: the allegation being that, from Greek and Roman times downward, these peoples, though militant, have been monogamic. It may, however, be replied that ancient European societies, though often engaged in wars, had large parts of their populations otherwise engaged, and had industrial systems characterized by considerable division of labor and commercial intercourse. Further, there must be remembered the fact that in Northern Europe, during and after Roman times, while warfare was constant, monogamy was not universal. Tacitus admits the occurrence of polygyny among the German chiefs. Already we have seen, too, that the Merovingian kings were polygamists. Even in the Carlovingian period we read that—
"The confidence of Conan 11. was kept up by the incredible number of men-at-arms which his kingdom furnished; for you must know that here, besides that the kingdom is extensive as well, each warrior will beget fifty, since, bound by the laws neither of decency nor of religion, each has ten wives or more even." (Ermold. Nigellus, iii., ap. Scr. R. Fr., vi., 52.)
And Kœnigswarter says that "such was the persistence of legal concubinage in the customs af the people that traces of it are found at Toulouse even in the thirteenth century."
Thus considering the many factors that have coöperated in modifying marital arrangements—considering also that some societies, becoming relatively peaceful, have long retained in large measure the structures acquired during previous greater militancy, while other societies which have considerably developed their industrial structures have again become predominantly militant, causing mixtures of traits—the alleged relations are, I think, as clear as can be expected. That advance from the primitive predatory type to the highest industrial type has gone along with advance from prevalent polygny to exclusive monogamy, is unquestionable; and that decrease of militancy and increase of industrialness have been the essential cause of this change in the type of family, is shown by the fact that this change has occurred where such other supposable causes as culture, religious creed, etc., have not come into play.
The domestic relations, thus far dealt with mainly under their private aspects, have now to be dealt with under their public aspects. For, on the structure of the family, considered as a component of a society, depend various social phenomena.
The multitudinous facts grouped in foregoing chapters show that no true conception of the higher types of family, in their relations to the higher social types, can be obtained without previous study of the lower types of family in their relations to the lower social types. In this case, as in all other cases, error results when conclusions are drawn from the more complex products of evolution, in ignorance
