But ethnology then discovered other phenomena also that seemed to favour this view. _Two_ lines of argument, particularly, have here played an important rôle, and still retain a measure of influence. The first argument was again derived from the ethnology of Australia. This region possesses a remarkable institution, describable neither as monogamy nor as agamy, but appearing, at first glance, to be an intermediate form of association. This is the so-called _group marriage_; several men are united in common marriage with several women. Either a number of brothers marry a number of sisters, or a number of men belonging to one kinship group marry in common women of another. Group marriage, therefore, may seem to represent a sort of transitional stage between promiscuity and monogamy. At first, so we might picture it to ourselves, the union of all with all became restricted to more limited groups, and only later to the union of one man with one woman.
But had not a further argument been added, perhaps neither female descent alone nor group marriage would have attracted to this theory so many prominent ethnologists, including, besides Howitt, the two able investigators of Australia, Spencer and Gillen, the learned exponent of comparative ethnology, J. G. Frazer, and a number of others. This further argument was presented with particular thoroughness by the American ethnologist Lewes Morgan, in his history of primitive man, "Ancient Humanity" (1870). It is based upon what Morgan has termed the 'Malayan system of relationship.' We are not, of course, familiar with this as a system of actual relationship; it occurs only in the languages of certain peoples, as a system of names--in short, as a nomenclature--referring in part to relations of kinship, but chiefly to age-relations within one and the same kinship group. The name 'Malayan' is not entirely appropriate as applied to this system. The nomenclature is found particularly on the island of Hawaii, though it also occurs in Micronesian territory. Its essential characteristic may be very simply described. It consists, or consisted, in the fact that a native of Hawaii, for example, calls by the name of 'father,' not only his actual father but also every man of an age such that he could be his father--that is, every man in the kinship group of the next older generation. Similarly, he calls by the name 'mother,' not only his own mother but every woman who might possibly, as regards age, be his mother. He calls brother and sister the men and women of his own generation, son and daughter those of the next younger generation, and so on up to grandfather and grandmother, grandson and grand-daughter. The Hawaiian native does not concern himself about more distant generations; great-grandfather is for him the same as grandfather, and great-grandchild the same as grandchild. The terms, thus, are of the simplest sort. The brothers and sisters of a man, whom we designate in the accompanying diagram by M, are placed alongside of him in the same generation; above, as an older generation, are fathers and mothers; still