that such marriages were common before the system was instituted to check them; in short, it implies that exogamy was a deliberate prohibition of a former unrestricted practice of incest, which allowed the nearest relations to have sexual intercourse with each other. This implication is confirmed, as Messrs. Howitt, Spencer, and Gillen have shown for the tribes of Central Australia, by customs which can be reasonably interpreted only as a system of group marriage or as survivals of a still wider practice of sexual communism. And, as the custom of exogamy combined with the classificatory system of relationship is not confined to Australia, but is found among many races in many parts of the world, it becomes probable that a large part, if not the whole, of the human race have at one time, not necessarily the earliest, in their history permitted the practice of incest, that is, of the closest interbreeding, and that, having perceived or imagined the practice to be injurious, they deliberately forbade and took effective measures to prevent it.
That is the great generalisation reached by L. H. Morgan from his discovery of the classificatory system of relationship. It is perhaps the most remarkable achievement of Fison and Howitt first, and of their disciples Spencer and Gillen afterwards, that their researches among the Australian tribes have not only lent powerful support to the conclusions of the American ethnologist, but have given us an insight into the machinery by which the great social reform was effected. The machinery was, indeed, simple; it consisted merely in the bisection, whether single or repeated, of the whole community into two exogamous classes. In Australia the application of this machinery to effect this purpose is seen more clearly than in any other part of the world, because in many Australian tribes the bisection has been repeated oftener than anywhere else, or, rather, oftener than it is known to have been repeated elsewhere; for it is possible that among other races of
