To attempt to define magic further than has been done in our description of the notions of primitive man is like trying to embrace a phantom. Magic rested upon man’s conjecture of the characteristics and processes of nature, not on a knowledge of nature correctly deduced from observation and experiment. As one would expect, there went with these mistaken notions a fantasticalness both in reasoning and in practical procedure. The follower of magic is apt to be on the watch not for facts or laws, but for hidden mysteries; he is fond of ceremonial and symbols; he enjoins upon himself and his fellows the necessity of secrecy in their operations and mysticism in their writings. Again, magic is, as has been said, præternatural; its outcome is to be marvelous. It assumes the existence of wonderful properties in various objects and of wonderful bonds of sympathy between different things. Finally, we should remember that man always is a factor in magic. His knowledge, skill or power is always essential to the performance of a feat of magic. Even when demons do the deed, they must be invoked. A miracle may be contrary to natural law but it is not magic, for man is not the cause of it. Even if wrought in answer to his prayer, the miracle is not magic, for the gods answer only if they choose. But the magic formula compels the desired marvel; by it man coerces nature or even deity. Such are some of the chief characteristics of magic. Yet with these granted, it remains, like superstition or religion, a vague term at best. The reader may disagree with me as to exactly what beliefs and practices should be included under it, and it is indeed a nice question just where magic begins and ends. Much of alchemy, for example, was nothing but chemistry of a rude sort, and perhaps even its theories that metals may be transmuted and life greatly prolonged will some day prove to have had much truth in them. On the other hand, alchemy was based to a considerable extent on a belief that plants, animals and minerals have properties and powers which they cannot have: and if we ever do succeed in making gold or putting off old age, it is quite certain that such a consummation will never be accomplished by the fantastic methods which alchemists usually employed. Similarly we shall see that the practice of allegorical interpretation of past writings and the Pythagorean doctrine of numbers, which perhaps at first thought one would not regard as magic at all, nevertheless bear at least some resemblance to it. But after all our thesis is not to establish a certain definition for the word “magic,” or to prove that such and such ideas and acts are magical. A name signifies little, and the word magic has had too many different meanings in different periods and for different men to allow any one to assert with confidence that he has found an absolutely correct definition. I employ the word simply because it seems the most