and for our working hypothesis we must always be prepared to substitute one that works better. But the point to which the Science of Religion has been brought by Dr. Marett's article in the Encyclopædia of Religion and Ethics, and by Dr. Söderblom's book on the growth of the belief in God, is precisely whether Sir James Frazer's description of religion is absolute and final, or whether it is merely a working hypothesis which can now be improved—or rather a provisional definition which must now be amended and extended. If it be amended and extended, then we may be able to include under it the intichiuma ceremonies at one end; but the question which will probably arise in most minds is whether at the other end the belief in personal beings may not disappear from the definition. For if it does disappear, then it cannot possibly be a definition.
The difficulty thus raised might be serious, if it arose only when we were seeking for a definition of religion. But, so far from arising only then, it appears with just the same force whenever we attempt to define anything whatever that develops or evolves. The difference between the acorn and the full-grown oak tree is as great as that between the intichiuma ceremonies and a polytheistic or a monotheistic form of religion. But though no description we can give of the oak will describe the acorn, the fact remains that the oak grows out of the acorn, or that the acorn becomes an oak by a process of continuous growth. And the process is not only one of continuity but of change—of change in continuity and of continuity in change. No one imagines that the oak is preformed in the acorn—that if we take the acorn to pieces we shall find an oak inside. And it would be just as unreasonable to imagine that if we dissect one stage of religion we ought to find, preformed in it, the stages which later are to evolve from it. The fact that we do not find in an acorn an oak-tree preformed does not in the least shake the
