Superstition, like science, has her hypotheses, and, like science, she reasons from experience. But her experience is usually fantastic, unreal, or if real capable of explanation by causes other than those alleged by superstition. A man comes in at night, and says he has seen a ghost in white. That is merely his hypothesis; the existence of ghosts in white is not demonstrated. You accompany him to the scene of the experience, and prove to him that he has seen a post, not a ghost. His experience was real, but was misinterpreted by dint of an hypothesis resting on no demonstrated fact of knowledge. The hypotheses of superstition are familiar. Thus, an event has happened: say you have lost your button-hook. You presently hear of a death in your family. Ever afterwards you go anxiously about when you have lost a button-hook. You are confusing a casual sequence of facts with a causal connection of facts. Sequence in time is mistaken for sequence of what we commonly style cause and effect. In the same way, superstition cherishes the hypothesis that like affects like. Thus, the sun is round, and a ball of clay is round. Therefore, if an Australian native wishes to delay the course of the round sun in the heavens, he fixes a round ball of clay on the bough of a tree; or so books on anthropology tell us. Acting on the hypothesis that like affects like, a man makes a clay or waxen image of an enemy, and sticks it full of pins or thorns. He expects his enemy to suffer agony in consequence, and so powerful is 'suggestion' that, if the enemy knows about the image, he sometimes falls ill and dies. This experience corroborates the superstitious hypothesis, and so the experiment with the image is of world-wide diffusion. Everything is done, or attempted, on these lines by superstition. Men imitate the killing of foes or game, and expect, as a result, to kill them in war or in the chase. They mimic the gathering of clouds and the fall of rain, and expect rain to fall in consequence. They imitate the evolution of an edible grub from the larva, and expect grubs to multiply; and so on. All this is quite rational, if you grant the hypotheses of superstition. Her practices are magic. We are later to discuss a theory that men had magic before they had religion, and only invented gods because they found that magic did not work. Still later they invented science, which is only magic with a legitimate hypothesis, using real, not fanciful, experience. In the long run magic and religion are to die out, perhaps, and science is to have the whole field to herself.