Reviews. 355
has perhaps risen to a prominent position with some rapidity by virtue of unusual fertility ; how far this was due to its foreign origin does not appear. Evidence was collected showing that there was a dual grouping of the clans ; in Sabai the single village was formed of one long street, in which there were two clans on one side and three on the other ; and the bulk of the marriages took place across the street. As to Mabuiag the information was somewhat hazy; our authors suggest that the two groups occupied separate localities, and that the ancient separation was obscured by local movements of population, This view at first sight does not fit in with Dr. Rivers' suggestion that female descent went out only a little more than a hundred years ago ; for clearly there can be no local grouping unless the determinant spouse (/'.^., the one from whom the children take their totems) retains her or his residence, and the husband or wife, as the case may be, removes from among his own people. Of such a custom we might expect to find distinct traces in the law of inheritance, but the rule appears to be equal division of property between sons and daughters ; this and the large testatory powers may of course be due to the change from female to male descent, but a discussion of this point would have been useful. The custom postulated by Dr. Rivers existed, however, and as a matter of fact it is still usual for the husband to remove and live with his wife. This is mentioned on p. 230, but it is by no means clear from the expressions used there that it was ever a universal practice ; there cannot well have been a local distribution, however, unless the exceptions were very few. Statistics on this point might readily have been prepared from the genealogies, and a fuller discussion of the question, together with that of the prior existence of matriliny, would have been very desirable. If, as appears to be the case, the clans were also localised (p. 173), it is not difficult to understand why a polygamist took his other wives from the same clan as the first (p. 244). This, and still more the marriage of sisters, would tend to secure that their landed property was less scattered than would otherwise have been the case.
An important question, on which no conclusion is reached, is that of subsidiary totems. By this is meant, the authors say (p. 180), not anything analogous to Mr. Howitt's sub-totems, which he re- gards as nascent kin-totems, but an arrangement by which a man has as secondary totem an object which may be the chief totem of
2 A 2
