Although in Arabia, in the time of the Prophet, descent was traced in the male line, the evidence is almost unlimited, going to show that it was not always so, but, on the contrary, that at an earlier age, relationships were reckoned through women, mothers being the recognized heads of families and tribal groups. In his work on _Kinship and Marriage in Early Arabia_, Prof. W. R. Smith says: If a kinship tribe derives its origin from a great father, we may argue with confidence that it had the rule that children were of their father’s tribe and kin; while on the other hand if we find, in a nation organized on the principle of unity of tribal blood, tribes which trace their origin to a great mother instead of a great father, we can feel sure that at some time the tribe followed the rule that the children belong to the mother and are of her kin. Now among the Arabs the doctrine of the unity of tribal blood is universal, as appears from the universal prevalence of the blood-feud. And yet among the Arab tribes we find no small number that refer their origin to a female eponym. Hence it follows that in many parts of Arabia kinship was once reckoned not in the male but in the female line. In reply to the suggestion that the several families of polygamous fathers might be designated by the names of their several mothers, Professor Smith observes: The point before us, however, is not the use of the mother’s name by individuals for purposes of distinction, but the existence of kindred groups whose members conceive that the tie of blood which unites them into a tribe is derived from and limited by descent from a common ancestress. That the existence of such a group proves kinship through women to have been once the rule is as certain as that the existence of patronymic groups is evidence of male kinship. In most cases of the kind the female eponym is mythical, no doubt, and the belief in her existence is a mere inference from the rule of female kinship within the tribe, just as mythical male ancestors are inferred from a rule of male kinship. But even if we suppose the ancestress to be historical, the argument is much the same, for where the bond of maternity is so strong that it binds together the children of the same mother as a distinct kindred group against the other children of their father, there also we may be sure that the children of one mother by different fathers will hold together and not follow their father. And this is the principle of female kinship.[86] [86] _Kinship and Marriage in Early Arabia_, pp. 26, 27. It is stated that the designation of tribal unity by a feminine appellation “is not an arbitrary fiction of later facts,” but that it is “one of the old standing figures of Semitic speech.” In Hebrew, _em_, which means mother, means also stock, race, or community.