The second sort of magic acts by spells which constrain spirits or gods to do the will of the magician. This magic involves itself in religion when the magical ceremonies are, so to speak, only symbolic prayers expressed in a kind of sign-language. But if the idea is to put constraint by spells on a god or spirit, then the intention is magical and rebellious. Though the official priest of a savage god may use magic in his appeal to that deity, he is not a wizard. It is the unofficial practitioner who is a witch, just as the unqualified medical practitioner is a quack. In the same way if a minister of the kirk was clairvoyant or second-sighted that was a proof of godliness and inspiration. But if a lay parishioner was second-sighted, he or she was in danger of the stake as a witch or wizard. These, briefly stated, are the points of contrast and points of contact between magic and religion. The question has recently been raised by Mr. Frazer, in the new edition of his 'Golden Bough,' whether magic has not everywhere preceded religion. Have men not attempted to secure weather and everything else to their desire by magic, before they invented gods, and prayed to them for what magic, as they learned by experience, failed to provide'? This question cannot be historically determined. If we find a race which has magic but no religion, we cannot be certain that it did not once possess a religion of which it has despaired. I once knew a man who, as a child, suffered from toothache. He prayed for relief: it did not come. He at once, about the age of eight, abandoned religion. What a child may do, in the way of despair of religion, a childlike race may do. Therefore, if we find a race with magic but without religion, we cannot scientifically say that the race has never possessed a religion. Thus the relative priority of religion or magic cannot be ascertained historically.