Our own civilization is shot through with survivals, so that further illustrations are unnecessary. They suggest, however, another aspect of our general problem. Of course, in every culture different traits are linked together without there being any essential bond between them. An illustration of this type of association is that mentioned by Dr. Laufer for Asiatic tribes, _viz._, that all nations which use milk for their diet have epic poems, while those which abstain from milk have no epic literature. This type of chance association, due to historical causes, has been discussed by Dr. Wissler[11] and Professor Czekanowski.[12] But survivals show that there may be an _organic_ relation between phenomena that have become separated and are treated as distinct by the descriptive ethnologist. In such cases, one trait is the determinant of the other, possibly as the actually preceding cause, possibly as part of the same phenomenon in the sense in which the side of a triangle is correlated with an angle. A pair of illustrations will elucidate the matter. Primitive terms of relationship often reveal characteristic differences of connotation from their nearest equivalents in European languages. On the other hand, they are remarkably similar not only among many of the North American Indians but also in many other regions of the globe, such as Australia, Oceanica, Africa. The most striking peculiarity of this system of nomenclature lies in the inclusiveness of certain terms. For example, the word we translate as ‘father’ is applied indiscriminately to the father, all his brothers, and some of his male cousins; while the word for ‘mother’ is correspondingly used for the mother’s sisters and some of their female cousins. On the other hand, paternal and maternal uncle or aunt are rigidly distinguished by a difference in terminology. As Morgan divined and Tylor clearly recognized, this system is connected with the one-sided exogamous kin organization by which an individual is reckoned as belonging to the exogamous social group of one, and only one, of his parents. The terminology that appears so curious at first blush then resolves itself very simply into the method of calling those members of the tribe who belong to the father’s social group and generation by the same term as the father, while the maternal uncles, who must belong to another group because of the exogamous rule, are distinguished from the father. In short, the terminology simply expresses the existing social organization. In a world-wide survey of the field Tylor found that the number of peoples who use the type of nomenclature I have described and are divided into exogamous groups, is about three times that to be expected on the doctrine of chances: in other words, the two apparently distinct phenomena are causally connected.[13] This interpretation has recently been forcibly advocated by Dr. Rivers, and I have examined the North American data from this point of view.