I can now consider briefly some problems of general interest towards the solution of which the conclusions of this paper may contribute.
My own interest in the subject has come directly out of my attempt to carry out an ethnological analysis of Oceanic culture. If movements of people have carried cultures over wide regions of the globe, it is inevitable that some of the elements of these cultures must disappear and thus will there be lost links in the chain of evidence. In any attempt to analyse a cultural complex, it will be often necessary to assume such disappearance; and the probability and stability of any analytic scheme will be greatly promoted if one is able to assign motives for the disappearance, either from physical features of the environment, or from social or magico-religious features of the culture.
I hope that I have made it clear that in studying the history of culture we must be prepared for changes not to be accounted for by the likes and dislikes of the civilized and almost incredible from the utilitarian point of view. We must be very cautious in assuming that elements of culture are so useful or so important that they would never be allowed to disappear. If islanders can lose the canoe, of what features of culture can we safely say that they can never be lost?
A second most important aspect of my subject is one in which the loss of the canoe is especially concerned. In many regions of Oceania and in other parts of the world, islands are now to be found inhabited by people whose present means of transportation are wholly insufficient to have brought them from the nearest land. In dealing with such problems it has sometimes been assumed that under no circumstances is it credible that people could ever lose the art of navigation and it has therefore been concluded that the islands must have been peopled when they were connected with some continent by a connecting bridge of land. On