In any pluralistic metaphysic, the problems that evil presents are practical, not speculative. Not why evil should exist at all, but how we can lessen the actual amount of it, is the sole question we need there consider. 'God,' in the religious life of ordinary men, is the name not of the whole of things, heaven forbid, but only of the ideal tendency in things, believed in as a superhuman person who calls us to co-operate in his purposes, and who furthers ours if they are worthy. He works in an external environment, has limits, and has enemies. When John Mill said that the notion of God's omnipotence must be given up, if God is to be kept as a religious object, he was surely accurately right; yet so prevalent is the lazy monism that idly haunts the region of God's name, that so simple and truthful a saying was generally treated as a paradox: God, it was said, _could_ not be finite. I believe that the only God worthy of the name _must_ be finite, and I shall return to this point in a later lecture. If the absolute exist in addition--and the hypothesis must, in spite of its irrational features, still be left open--then the absolute is only the wider cosmic whole of which our God is but the most ideal portion, and which in the more usual human sense is hardly to be termed a religious hypothesis at all. 'Cosmic emotion' is the better name for the reaction it may awaken. Observe that all the irrationalities and puzzles which the absolute gives rise to, and from which the finite God remains free, are due to the fact that the absolute has nothing, absolutely nothing, outside of itself. The finite God whom I contrast with it may conceivably have _almost_ nothing outside of himself; he may already have triumphed over and absorbed all but the minutest fraction of the universe; but that fraction, however small, reduces him to the status of a relative being, and in principle the universe is saved from all the irrationalities incidental to absolutism. The only irrationality left would be the irrationality of which pluralism as such is accused, and of this I hope to say a word more later. I have tired you with so many subtleties in this lecture that I will add only two other counts to my indictment. First, then, let me remind you that _the absolute is useless for deductive purposes_. It gives us absolute safety if you will, but it is compatible with every relative danger. You cannot enter the phenomenal world with the notion of it in your grasp, and name beforehand any detail which you are likely to meet there. Whatever the details of experience may prove to be, _after the fact of them_ the absolute will adopt them. It is an hypothesis that functions retrospectively only, not prospectively. _That_, whatever it may be, will have been in point of fact the sort of world which the absolute was pleased to offer to itself as a spectacle.