not the family, but the clan. We lay the greatest stress upon this circumstance, and it is only because this has not been done before that descent through the female line has been explained by ignorance of the paternity of the child. A clan is an exclusive group, and it is absolutely impossible for one person to belong to two distinct clans. In Australia, intermarriage between members of the same clan was never allowed, and hence it was necessary to define the clan to which the child belonged on one side only; either the father or the mother must be ignored. In Australia, it is usual to ignore the father, yet the male line is maintained by the Kurnai, Gournditch-Mara, Turra, Moncalon, and Torndirrup tribes, and by others visited by Eyre.1 The line of descent has an important effect on the localization of the groups. If the child follows the parent on whom the family place of abode depends, the local isolation of the clan can be carried out; if otherwise, a fresh confusion of the clans will constantly occur. As a rule, the place of abode depends upon the will and humour of the husband, and the female line of descent, therefore, breaks down any fixed boundary between the territories of clans.

The female line of descent does not, however, affect the family. Among the Australians, indeed, the family has no fixed organization; the rape of married women, and divorces are of daily occurrence, and the bond which unites a father to his adult son is therefore unstable. Yet the tie of marriage cannot be regarded as so loose as to make it impossible for a father to feel any certainty with respect to the child, for all reports agree in declaring that the men are extremely jealous. Only the custom of hospitality, which leads a man to offer his wife to his friend or guest, can give rise to doubt; but the arrival of travellers is not of frequent occurrence. We shall presently consider the meaning of such customs, but we must pass over it now, as well as the strong affection which, accord1 Fison and Howitt, pp. 276, 285. Nind, p. 44. Eyre, vol. ii. p. 328.

ing to all witnesses, Australians feel for their children, since, as we shall presently see, such affection affords no proof of the assurance of fatherhood. As an instance of the sharp distinction between that which concerns the clan and the family, we need only note that the obligations of the blood-feud will sometimes place a father and son in hostile opposition to each other, since they belong to different clans, and yet the father's land is divided among his sons.

Since, therefore, we find that the clan alone is affected by the one-sided line of descent, the question arises whether the circumstances of the Australians afford data from which we can decide whether descent was originally traced through the male or female line, and what were the reasons for the transition from one to the other. This may be premised: that either the reasons for adopting the female line were of universal application, or that it may be adopted under all sorts of conditions; one or other of these assumptions is warranted by its wide diffusion. We find the male line in use among the Kurnai, which are divided into clans in accordance with their districts, not with their kobongs. The same line is in use among the Moncalons and Torndirrups, who are no longer quite distinct tribes, yet show traces of earlier division. Their fusion seems to have been produced by a subsequent kobong division. I have no idea how the formation of groups dominated by the female line could have arisen, since this line of descent interferes with the permanence of the groups already formed. Hence I can easily understand how groups which were formed and maintained under the male line were confounded together under the female line, without, however, altogether disappearing. I am therefore inclined to regard the female line as a later development; and this surmise seems the more probable, since usually the female line only appears in conjunction with the kobong groups. Among the Kurnai, the only tribe which is without kobongs, the female line is not observed.

If we venture to decide on the stage of civilization to which the Australian tribes have attained, we should assign those who maintain the male line of descent to the lowest place. The tribes visited by Eyre were on the whole more degraded than those visited by Grey; the former maintained the male, the latter the female line, and among the former, as we have already said, the kobong organization was only dawning. A single motive cannot be assigned for the adoption of the male line of descent by these tribes, and, indeed, we may infer the contrary, for the rights established by marriage appear to be less decided in the tribes observed by Eyre than in those described by Grey. Grey states that a widow becomes the property of her dead husband's brother, while Eyre reports that she returns to her own. family after her husband's death.1 Only among the Kurnai we find slight indications of her becoming the brother's property. It is, on the other hand, possible to discover that the transition from the male to the female line was due to an urgent motive. As the family instinct became more developed, the importance of the distinction between brothers and sisters of full and half blood would become greater. When the husband was regarded as the only important person, it mattered little what became of the wife after the husband's death. Brothers and sisters born of one mother became an inconceivable idea. Yet polygamy led to the conception of such an idea. The daughters were subject to the father; on his death his brother took his place, and it was necessary to distinguish between the whole and the half brother, since the former must become the guardian. Among primitive men, as a rule, everything depends upon the first shock which their theories receive; the imagination is set to work in another direction, and cannot be restrained within due proportions and limits. In this way the

1 Grey, vol. ii. p. 230. Eyre, vol. ii. p. 319. Waitz, vol. vi. p. 776. 2 Fison and Howitt, p. 201.

unrestricted descent through the female line may have spontaneously arisen, although in the first instance it was only a feeling of convenience which induced the race to distinguish the children of polygamous families by their descent through their mothers.

We do not claim unconditional assent to these suggestions. We only assert that the conditions which exist among the Australians do not entitle us to regard the female line of descent as the primitive one, nor to explain it by the fact of promiscuous intercourse. Their circumstances do not afford proof of either opinion; they rather indicate that the female line was a later development, to which they were led for other reasons than those which referred to the question of paternity.

CHAPTER II.

AMERICA.

Exogenous totem-groups-General diffusion of female line-As to the primitive existence of the clan-The tamanuus of the peoples of Columbia-Tylor, Lubbock, and Spencer - Medicine and totemSystem of names in clan-The Columbians the primitive tribesPosition of husband with respect to members of wife's family-Effect of this position on line of kinship-Influence of locality and household-Caribs-Non-existence of clans among them, and vague definition of kinship-Labour performed by son-in-law-Paternal authority and household among Brazilians-Tattooing a bond of union-Likewise a distinctive mark-Obligation of blood-feud-Slaves and freemen-Undeveloped clans-Forbidden marriages-Tribes of La Plata -No patria potestas-Hereditary dignity of chief-Araucanians -Coherence of the family-Female line not primitive-Origin of duties and privileges-Germ of development of groups-Nomenclature-Mexico-Paternal line and localized clans--The same in Peru -The couvade-Lubbock's hypothesis-Diffusion and meaning of the couvade.

WE find in North America very marked forms of clans which are distinguished by their totems or kobongs, and in that country as well as in Australia the clans are exogamous; that is, marriage between members of the same clan is not permitted. The formation of tribes is also more permanent than in Australia, when persons belonging to different tribes, but possessing the same totem, stand in close relations to each other; in America this is not the case. In some instances the clan organization is absent, and these exceptions will help us to understand its history-a history which reveals