[408] Polyandry is favored by poverty and scarcity of women; but it is essentially the outgrowth of ancient sexual relations: HELLWALD, _op. cit._, 258-61; agreeing with LIPPERT, _Kulturgeschichte_, II, 10. MARSHALL, _A Phrenologist amongst the Todas_, 223 ff., follows Lubbock and McLennan in regarding polyandry as a survival of communism. On the other hand, FRITSCH, _Die Eingeborenen Süd-Afrikas_, 227, is decidedly of the opinion that polyandry among the Kafir Herero is the direct result of poverty and low condition (_niedrige Gesinnung_); it is, he says, "keine Sitte, sondern eine Unsitte," harmonizing with the laxity of their moral ideas. [409] SMITH, _Kinship and Marriage_, 125 ff., 128. [410] WAKE, _Marriage and Kinship_, 172, 134-78. [411] STARCKE, _op. cit._, 135, 139, 128-70. More satisfactory, from a scientific point of view, is the result of Westermarck's inquiry. This is so, not only because we feel that he is probably right in his conclusion, but because his argument affords an excellent illustration of the success with which the statistical method may be applied to social questions. The way for a solution of the problem had been prepared by McLennan and his critics. They had established a strong probability that poverty and scarcity of women are in some intimate way connected with polyandry. Westermarck shows that there is, in fact, a close relation, but that relation is a consequence of natural selection. The ultimate causes of polyandry, he demonstrates, are identical with the forces which have produced a numerical disparity between the sexes.[412] First of all the assumption[413] that "monogamy is the natural form of human marriage because there is an almost equal number of men and women," is proved to be untenable by an appeal to the statistics of population, which reveal a considerable variation in the numerical proportion of the sexes. Among many peoples the men are greatly in majority; among others there is a corresponding surplus of women. This disparity is in part easily explainable by referring to the varying conditions of life among different peoples. The "preponderance of women," for instance, "depends to a great extent upon the higher mortality of men" due chiefly to the "destructive influence of war" and the other dangers and hardships to which primitive men are exposed. On the other hand, the surplus of men may, in some degree, be ascribed to female infanticide and, still more, to the severe labor and harsh treatment which usually fall to the lot of women among low races.[414] [412] WESTERMARCK, _op. cit._, chap. xxi, in connection with chaps. xx and xxii.