The tendency towards a reform in the method of sociological treatment of kinship and family has been shown not only from the side of purely theoretical writers. Some of the modern field workers, who happily for our science are at the same time distinguished scholars, have achieved a considerable advance in the method of collecting evidence. This refers in the first place to the Cambridge School of Ethnology, whose members under the lead of Dr. Haddon have obtained such remarkable results from their work in the Torres Straits Islands. Dr. Rivers, who specially worked out the chapter on kinship in the joint publication of this Expedition, has, by the introduction of the genealogical method of inquiry as well as by the systematic study of the functions of kin, given perhaps the most useful instruments of inquiry into the social working of family and kinship organization. Thus both our theoretical conceptions and our methods of getting at the facts are certainly approaching more and more the first postulate of scientific study: the possibility of an adequate description of facts and their mutual dependences as they exist now in living primitive societies. Only on a basis of such knowledge are further speculations fruitful. As regards the general principles of sociological method much has been done in recent times by the French school of sociology, grouped round the editor of the _Année Sociologique_. The important question, how methodically to present evidence, has received its full attention in the excellent works of Dr. Steinmetz and his pupil, Dr. Nieboer, which are examples of a clear and conclusive way of utilizing ethnological sources. I am glad to acknowledge my intellectual indebtedness to both these schools. I have tried to collect sufficiently complete evidence, and in this endeavour have used some of the older sources whose trustworthiness might perhaps be disputed. But many of their observations are highly valuable if properly interpreted; and moreover it was necessary to bring their statements into line with the newer evidence for the sake of critical comparison, as much of what they say has been uncritically accepted and given without reference by some secondhand compilers (for instance, Waitz-Gerland, vol. vi.; Cunow) and hence found its way into the newer sociological literature. The statements I have taken from the different authors are quoted at length, and I do not think that I have thus uselessly increased the bulk of the volume. By an unprejudiced collection of evidence, which is, moreover, presented in a manner independent of, and accessible without reference to, the theoretical discussion, I hope to have given a useful compilation of observations which may serve for further theoretical purposes other than those of the present writer. In order to make short and yet clear references possible a list of the works quoted is given at the end. With its help the short indications in the footnotes will be perfectly plain.