The foregoing discussion has been mainly concerned with the collective ideas which define parental kinship, and the different sets of social facts in which these ideas find their expression have been enumerated. It also dealt with collective feelings, and the different facts in which these are to be looked for were surveyed. We must now emphasize the fact that just as we may say that the different ideas determining kinship converge towards one central concept, or rather flow out of one common central idea of kinship, so there is also an intimate connection between the ideas determining kinship and the feelings bound up with it. This becomes obvious if our own social conditions be considered. As mentioned above, a father in our society loves his child in a great measure because he knows that it is his own offspring. In societies in which the idea of consanguinity (in the social sense) does not exist, such a connection between feelings of paternal love and knowledge of a physiological procreation would be impossible. And it would be of the highest sociological interest to trace what form such connections assume. An attempt at such a study would be possible in our own society and in other higher societies, although there would be serious difficulties enough. But there would hardly be sufficient material to attempt it in any lower society, and there is absolutely no possibility of doing this for Australia. A brief summary of the foregoing argument may now be given. It was stated at the beginning that parental kinship corresponds to a very complex and manifold set of phenomena; moreover in various societies this relationship is determined by various elements. The problem is to find in all this complexity the structural features, the really essential facts, the knowledge of which in any given society would enable us to give a scientifically valid description of kinship. In other words, the problem is to give a general formula defining kinship, which would state its constant elements and give heed to the essential varying elements therein; that formula being on the one hand not too narrow for application to the various human societies, it would be on the other hand not too vague to afford quite definite results when applied to any special case. A final solution of this problem cannot be arrived at _a priori_, but only by way of induction, after the facts in the different human societies have been studied. And in order to attempt such a preliminary study of the Australian facts, the foregoing remarks have been given; they aim at a general definition of the kind just described in the form of a tentative or preliminary sketch. Consequently in the first place the attempt was made to ascertain what could be taken as the constant elements in individual parental kinship. What appeared to be nearly universal in this connection is the fact that infants and small children are always specially attached, and