1

together. And Morgan makes a similar remark with respect to the Fijians. The Kurnai of New Holland also defend the custom in question on these grounds.2 Since in the joint family group the brother succeeds to the headship of the community, and its interests and general protection are committed to his care, so also the widow and her young children are committed to him, and under primitive conditions these relations take the form of marriage. In proportion to the importance of the property of the deceased is the difficulty of allowing the widow to return to her own family. It is altogether irrational to seek for the causes of the connection between the widow and her brother-in-law in polyandry. That the brother-in-law had exerted marital rights in the husband's lifetime would only become a necessary condition if carnal considerations formed the corner-stone of the development of the family; but all we know of the life and habits of primitive men clearly shows that this was not the case. Carnal pleasures certainly took the most prominent place in primitive life, but they were also the most easily obtained, and therefore customs were not formed under the influence of considerations with respect to the means of sensual enjoyment.

Before concluding this series of researches, we must briefly mention McLennan's attempt to trace the existence of polyandry in the ancient Aryan race. We do not think that our conclusions would be invalidated, even if our forefathers lived in polyandry; this practice is, indeed, so often found in connection with the development of the family group, that it is quite possible that it was also practised by the Aryans. It thus becomes simply an historical question.

McLennan directs our attention to the following

1 Morgan, Systems, p. 583. See, above, the encroachments of uncles among the Sioux and Columbiaus.

2 Fison and Howitt, p. 204.

passage from Manu:-" If among several brothers of the whole blood, one have a son born, Menu pronounces them all fathers of a male child by means of that son, so that if such nephew would be the heir the uncles have no power to adopt sons." 1 McLennan does not, however, depend upon this text, since Draupadi's narrative makes it unnecessary by directly mentioning the fact of polyandry. But in his latest work he again refers to the passage, in connection with the conditions which exist in Thibet. He does not admit that the constitution of the father's brother into the father was founded on a fiction. He asserts that it was altogether contrary to the Hindu custom to hinder a man from obtaining sons of his own by fictitious and artificial regulations, and we have seen above that adoption by the uncle was prohibited. It was the chief and most persistent effort of Hindu government to promote the increase of heirs, that is, of separate households, and with these of the centres of religious worship. The prohibition of adoption contained in the text must, he thinks, have had its origin in some very ancient institution.

This view must be rejected as altogether unfounded. J. D. Mayne has shown that religious interests aimed at the dissolution of the joint family groups, and it was on religious grounds that the Brahmans advocated such a dissolution. The head of the family was himself its high priest. As long as the Brahman influence was still slight, the claim of the family group to maintain an heir who might offer sacrifice for them only reached the stage at which we find it in Manu's text, and subsequently, when this influence was fully developed, each brother desired to have a son of his own. The prohibition contained in Manu is merely that which has in all times and all places been the fundamental thought of men, namely, that only in cases in which the existence of the McLennan, "The Levirate and Polyandry," Fortnightly Rev., 1877, 2 McLennan, Patr. Theor., p. 336.

p. 698.

group was in danger of perishing for lack of legitimate heirs, whether that group consisted of the simple family, the joint family group, or the clan, it was permitted to have recourse to adoption. If, however, one of the brothers in a joint family group had a son, a legitimate heir was not wanted.1 McLennan also quotes the following passage from Apastamba:-"A husband shall not make over his wife, who occupies the position of a gentilis to others than to his gentiles, in order to cause children to be begot for himself. For they declare that a bride is given to the family of her husband, and not to the husband alone. That is at present forbidden on account of the weakness of men's senses.' "2 Here again McLennan finds a custom similar to that of Thibetan polyandry, based upon a polyandrous theory. J. D. Mayne, however, seems to us to be nearer the truth when he regards this text as a limitation of the Niyoga, which originally included any man whatsoever, and which was now restricted to the Sapinda or Samanodoco. The expression, "A bride is given to the family,' does not imply that she is given as a wife common to the family; the statement merely corresponds with Manu's previous statement, which revealed the ideas entertained in the joint family group.

If, in conclusion, we turn to the Draupadi myth, it is evident that the Pandava brothers, distributed into Draupadi, do not justify us in forming any general conclusion. They are ancestors of the Kshatry tribe, which came into India from one of the Kashmir valleys of the Himalayas; and the account only allows us to infer that polyandry occurred locally in an isolated Aryan tribe. Even this conclusion does not appear to me to be completely established, since the myth affords scanty in1 J. D. Mayne, p. 211.

See also H. Zimmer, p..325, who

2 McLennan, Patr. Theor., p. 304. does not admit the existence of polyandry. 3 McLennan, Patr. Theor., p. 305. J. D. Mayne, pp. 48, 63.

formation with respect to the reasons which led to Draupadi's connection with the five Pandavas, and also with respect to the ideas which justified this marriage.1

1 Lenormant, vol. iii. p. 497. McLennan, Fortnightly Rev., 1877, p. 699. Bachofen, Das Mutterrecht, §§ 94, 195. Wilkes, vol. i. p. 54. J. D. Mayne, p. 57.

CHAPTER V.

NOMENCLATURES.

Modern and primitive ideas-Descriptive and classifying nomenclatures— Their persistence-Descriptive nomenclature and monogamy-Classifying nomenclature and clan organization-Polygamy and polyandry -Malayan nomenclature-Punalua family-Turanian nomenclature -Distinction between the Turanian and the Ganowanian-Patriarchal family-Descriptive nomenclature-Nomenclature and the clanImportance of ceremonial-Paternal and maternal kinship-Punalua kinsfolk of Malayan nomenclaturc-Priority of kinship of cousins.Distinction between Turanian and Ganowanian nomenclatures-Ganowanian mother's brother-Development of special names-Son and nephew-Turanian marriage of cousins-Fijian nomenclature-Simpler nomenclature of Ganowanian women-Cayugan peculiarities -Tonganese nomenclature—Descriptive nomenclature-Karen and Eskimo nomenclature Chinese nomenclature Their general meaning.

THE readiness with which a husband in the primitive community permitted other men to hold sexual intercourse with his wife, was closely connected with the slight importance which he ascribed to the actual descent of his children. The idea is historically untenable that, even in the case of the primitive man, the chastity of his wife was an indispensable condition of marriage, and that lascivious customs consequently pointed to a time when promiscuous intercourse prevailed, and there was no tie of marriage. McLennan seeks to confirm the existence of such conditions by means of polyandry, an attempt which we have shown to be wholly unsuccessful. We must now turn to other writers, such as Morgan,