explanation of the functions of the father's sister in the matrilineal communities of Melanesia. There remains the very similar position of this relative in Tonga. The similarity is so great that there can be little doubt that whatever conditions explain the Melanesian facts will also explain those of Tonga, and it may be pointed out that there is no doubt that the cross-cousin marriage existed at one time, if it does not still exist, in Tonga, especially among the chiefs, the information given to me on this point confirming the account given by Mr. Basil Thomson.[1] Further, there would seem to be a close analogy between the functions of the aunt in the two places in taking the umbilical cord and the first menstrual blood respectively. We have in the Tongan practice an example of a custom, having its origin at a time when kinship with the father was beginning to be recognised, which has persisted long after this kinship has been fully established, and long after the change from matrilineal to patrilineal descent has taken place.
In conclusion, I should like to refer to the bearing of the facts I have related on certain questions of definition. Of all sociological terms there are none more important and at the same time used more indefinitely than "kin" and "kinship." In his book on the Melanesians Dr. Codrington has spoken of a child as not being of the same kin as his father.[2] Here Dr. Codrington has used the English word "kin" as the equivalent of the Mota word sogoi for those related to one another by common membership of a social group, in this case the veve or moiety of the whole population. Thus one of the meanings which has been ascribed to the word "kin" is membership of the same group, so that it excludes certain people related by consanguinity, and includes others with whom no genealogical connection can be traced. The same definition is implied, though
