In the Pueblo of Orayba the relatives of a married woman who dies take her property and her children, only leaving to the husband his horse, his clothes, and his weapons;[899] for by marrying the woman does not cease to belong to her original clan. Among the Pipiles of Salvador a genealogical tree with seven branches was painted on the wall of the common house, and save in the case of a great service rendered to the clan, a man could not intermarry with any persons related up to the degree indicated by the genealogical tree.[900] In reality, this people had got beyond familial confusion, or of purely _totemic_ relationship, but the principle regulating conjugal unions had not yet changed. In Yucatan marriage between persons of the same name--that is to say, of the same clan--entailed the penalty of being considered as a renegade.[901] The savage Abipones were also exogamous, according to Dobritzhoffer. This rule naturally gives way in proportion as civilisation develops. The Nahuas still prohibited marriage between consanguineous relatives; but at Nicaragua the prohibition only applied to relatives of the first degree.[902] We have previously seen, in describing the family amongst the animals, that it is habitually maternal; it is around the female that the young group themselves. As for the male, if he does not abandon the family, he exercises no other function but that of chief of the band. It must surely have been thus that the first human hordes were formed, and when man became intelligent enough to take note of filiation, it was uterine parenthood alone that he considered worthy of account. The primitive family was maternal, for in the confusion of sexual unions paternal filiation would have been difficult to determine; no importance was therefore attached to it in early times, and the father was not looked upon as the parent of his children. We shall find the maternal family, or at least traces of it, in many countries, but it is especially among the Indians of North America that it has been the best preserved and the best studied. In the eighteenth century it was already remarked by Charlevoix, Lafitau, and Lahontan,[903] that the Redskins always bear the name of their mother, and that it is through a man’s sister that his name is transmitted to descendants. The American clan is based on uterine filiation; it comprehends all the descendants, in the female line of an ancestral mother, real or hypothetical. It is therefore exactly the contrary of the agnatic _gens_ of the Greco-Roman world.