in which things seemed natural and rational that now appear unnatural and devoid of reason, and in which, therefore, if myths were evolved, they would, if they survived into civilisation, be such as civilised men find strange and perplexing.
Our first question will be, Is there a stage of human society and of the human intellect in which facts that appear to us to be monstrous and irrational—facts corresponding to the wilder incidents of myth—are accepted as ordinary occurrences of everyday life? In the region of romantic rather than of mythical invention we know that there is such a state. Mr. Lane, in his preface to the Arabian Nights, says that the Arabs have an advantage over us as story-tellers. They can introduce such incidents as the change of a man into a horse, or of a woman into a dog, or the intervention of an Afreet without any more scruple than our own novelists feel in describing a duel or the concealment of a will. Among the Arabs the agencies of magic and of spirits are regarded as at least as probable and common as duels and concealments of wills seem to be thought by European novelists. It is obvious that we need look no farther for the explanation of the supernatural events in Arab romances. Now let us apply this system to mythology. It is admitted that Greeks, Romans, Aryans of India in the age of the Sanskrit commentators, and Egyptians of the Ptolemaic and earlier ages, were as much puzzled as we are by the mythical adventures of their gods. But is there any known stage of the human intellect in which similar adventures, and the metamorphoses of men into