traditional conceptions of masculine and feminine employment and that differences in this regard cannot fail to affect the course of the family life. A polygynous Thonga becomes a parasite supported by his gardener-wives; a Kirgiz wife performs the household tasks, while the husband not merely tends the herds but also supplies firewood, tills the soil, and manufactures all household vessels: the Toda woman has hardly any duties besides pounding and sifting grain, cleaning the hut and decorating clothing. It is not a question of woman's theoretical status, for that is doubtless lowest among the Mohammedanized Kirgiz, and probably lower among the Toda than with the Thonga; it is simply a question of what labors are conventionally allotted to each sex. Dr. Laufer has forcibly pointed out how at a higher level of civilization "the forms of Chinese fam ily life and the psychical relations of the members" are radically different from our own because woman never superintends nor even approaches the kitchen, which is always far removed from the center of the house and thus never serves as a family rallying-point.*

SEGREGATION OF UNMARRIED

Among the customs often giving to primitive family life a very peculiar cast as compared with ours is the segregation of unmarried young men and women from the remainder of the community, frequently at adolescence, sometimes even at an earlier period. Thus, among the Dravidians of southern India the youths no longer sleep with their parents but in a separate club house, and the girls in a dormitory of their own, superintended by a matron. Every Kariera camp is divided into the married people's and the bachelors' camp, the latter including widowers, the former single women and widows. The Masai usage has already been described by which the bachelor braves reside in a special kraal with the immature young girls, while all married men have settlements of their own.

These customs introduce us to the principle of dividing a community on the basis of age, with or without a farther recognition of a sexual segregation. The relevant problems of distribution and interpretation will receive detailed discussion. Suffice it for the present to call attention to the inevitable rending asunder or at least serious weakening of the family ties where the adolescent children are separated from their parents by these fixed institutions."

SEXUAL SEGREGATION

Still more drastic in its effects on the family as a social unit is the separation of husband and wife either by the segregation of men in a club house of their own or by exclusion of the women from those forms of public activity which especially engross the attention of the men. This, too, is a subject for ampler treatment further on, but one or two characteristic illustrations must be cited here by way of anticipation.

Among the Hupa the women lived in the family house, where their husbands came to eat meals before and after their daily tasks. In the evening the men retired to the sweathouse, which served not only as a Turkish bath and club but as a dormitory as well. This, however, does not compare with the isolation of the women in parts of Melanesia, such as the Banks Islands, where the men and indeed the adolescent boys not merely sleep but eat apart from the women, membership into the men's club being early purchased to shorten the ignominy of having to feed with the women. Finally may be mentioned the well-nigh universal Australian custom of barring women from attendance at those sacred rites about which most of masculine thought revolves in its more serious moments."

ADOPTION

The very constitution of the family may be altered by the legal fiction through which parents rear as their own

the children of another couple. In many cases the children are related to their foster parents, but this is by no means prerequisite. A common motive for adoption is lack of issue. Thus, a Chukchi couple without offspring will adopt the child, preferably the son, of a relative, and the boy then becomes the principal heir. The sentimental relationship comes to approach very closely that based on the natural tie. With the Crow Indians it is common for men and women to adopt a sibling's child, and if anything there was exaggerated demonstration of affection as if to compensate for the subconscious feeling that after all the tie was factitious. But probably nowhere is adoption so prevalent as in Murray Island of the Eastern Torres Straits group, where children for no manifest reason are adopted even before birth and brought up entirely as members of the adoptive parent, often remaining in ignorance of their real parentage till adult life or even till death."

SUMMARY

Although the character of the primitive family is appreciably altered by the usages sketched above, these modifications do not invade the bilateral principle. A man may spend the major part of his working and sleeping hours away from his wife, but for all that he is linked to her by the common interest in the children of the household, really or putatively his own, and by their economic and industrial partnership; and similar considerations apply to the other conditions mentioned, which often strangely affect the dynamics of family life from a Caucasian point of view. In short, the bilateral family is none the less an absolutely universal unit of human society.

REFERENCES

1 Lowie, 1917 (a): 40 seq., 51. Junod: 1, 44, 212, 226, 253, 262. Reports, v: 144 seq. Keysser: 45, 85. Roth, 1906: 6. Spencer and Gillen, 1899: 18. Malinowski: 158-167. Brown: 147. Spieth: 191.

'Lowie, 1912: 223; id., 1913: 169. Keysser: 86. Whiffen: 165. Thalbitzer: 65, 72. Bogoras: 596.

'Goddard, 1903: 56-58. Freire-Marreco: 269-287. Lowie, 1917 (a): 46. Schinz: 304, 311. Gurdon: 78. Jochelson, 1908: 744; id., 1910: 92. Cranz: 1, 215 seq. Boas, 1888: 579. Thalbitzer: 59. Murdoch: 410. Kroeber, 1917 (a): 105.

4

Hahn. Rivers, 1906: 567. Laufer, 1917: 148; id., in Amer. Anth., 1918: 89. Radloff: 462.

'Baden-Powell: 172. Goddard, 1903: 57.

6

Brown: 147.

Rivers, 1914 (b): 1, 63.

'Bogoras: 556. Lowie, 1912: 219. Reports, vi: 64, 177.

CHAPTER V

KINSHIP USAGES

IN

IN PROVING the bilateral character of the family, I have called attention to the social relations that obtain between an individual and the relatives on both his father's and his mother's side. As a matter of fact, primitive law usually goes much further and establishes definite functions for every relationship not only by blood but by marriage as well. In our society no fixed conduct is prescribed towards a maternal uncle or a sister's son or the husband of a father's sister. In primitive communities, on the other hand, a specific mode of behavior may be rigidly determined for each and every possible form of relationship. From the point of view of any individual this means that his tribesmen are classified into certain categories, each one of which implies an altogether special set of social rules to be observed by him. He is bound to render services to an individual of one class; with a member of another he may jest and take liberties; with persons of a third category he must have nothing to do except through intermediaries; and so forth. Proximity of relationship may or may not count; usually, as Mr. Brown has explained for the Kariera, a savage owes the same type of conduct to a more remote as to a closer kinsman addressed by the same relationship term, but the intensity of the obligation is greater for the nearer relationship. As this author further remarks, a native may be at a complete loss how to treat a stranger who falls outside of the established rubrics. What most frequently happens is that by a legal fiction, or it may