that in every account of Melanesian affairs given to the world tribes are spoken of; but a belief that every savage people is made up of tribes is part of the mental equipment of a civilized visitor; when one reads of the 'coast tribes' or the 'bush tribes,' nothing more is meant than the people who inhabit the coast or the inland part of some island.
There is, however, one very remarkable exception to this general rule of division in the Solomon Islands; it is not to be found in Ulawa, Ugi, and parts of San Cristoval, Malanta, and Guadalcanar, a district in which the languages also form a group by themselves, and in which a difference in the decorative art of the people, and in the appearance of the people themselves, thoroughly Melanesian as they are, can hardly escape notice. In this region, the boundaries of which are at present unknown, there is no division of the people into kindreds as elsewhere, and descent follows the father. This is so strange that to myself it seemed for a time incredible, and nothing but the repeated declarations of a native who is well acquainted with the division which prevails in other groups of islands, was sufficient to fix it with me as an ascertained fact. The particular or local causes which have brought about this exceptional state of things are unknown; the fact of the exception is a valuable one to note[1].
Speaking generally, it may be said that to a Melanesian man all women, of his own generation at least, are either sisters or wives, to the Melanesian woman all men are either brothers or husbands. An excellent illustration of this is given in the story of Taso from Aurora in the New Hebrides, in which Qatu discovers and brings to his wife twin boys, children of his dead sister: his wife asks, 'Are these my children or my husbands?' and Qatu answers, 'Your husbands to be sure, they are my sister's children.' In that island there are two divisions of the people; Qatu and his wife could not be of the same, Qatu and his sister and her children must be of the same; the boys therefore were possible husbands of
- ↑ 'Descent is still uterine in some parts of Fiji; most of the tribes, however, have advanced to agnatic descent.'—Rev. L. Fison.
