Finally, we may learn much, even adopt much, from our enemies. The most dangerous antagonists of Christianity were, after all, the worship of Isis and Mithraism. It may be possible to trace the influence of these on their conqueror: the great work of M. Cumont on Mithras cult suggests at least many interesting religious parallels; and even the older Zoroastrian {25} literature must be considered within the range of necessary and legitimate comparison. As I am here concerned with illustration of method rather than with positive proof, I can only offer a very brief summary of the results which the anthropological study of Christianity has hitherto achieved, and may yet achieve. The religious affinities discoverable between the earlier and later “Mediterranean” systems may be classified according as they appear in the legends, in nomenclature and terminology, in external symbols and liturgical objects, in hieratic institutions, and finally in the ideas, aspirations, and concepts of faith. As regards legend and mythology, a great historic religion may of course claim to be free from all mythology; nevertheless it is a matter of experience that popular legends are sure at some period earlier or later to creep in, for the people insist on telling the old stories under changed names. I myself have heard the immemorial story of Odysseus walking inland with his oar, which the rustic mistakes {26} for a winnowing-fan, told about St Peter, St Paul, and St John on the coast of the Peloponnese; just as an old Norse legend about Odin and Baldur is retold of Christ and St Peter.[26.1] And students of mediæval hagiology will discover more and more clearly various fragments of pre-Christian mythology embedded in the legends of the saints. Such facts are the material of comparative folk-lore, which plays a useful but quite subordinate part in the work of comparative religion. Legends have indeed their own independent interest, poetical, ethical, and other; but the importance of mere mythology in the study of religion has been often much overrated; St Augustine, mistaking Greek legends for Greek religion, could discover no morality in it at all,[26.2] and modern scholars have inherited the fallacy. Myths are often irresponsible, capricious, volatile, and flit like a vapour round the solid structure of real {27} belief and ritual. A high religion may attract low myths: some of our own are not spotless, but Christianity can ignore them. The myth that is an essential fact for the student of religion is that which enshrines some living religious idea or institution, or one which proves the survival of some ritual or faith that belonged to an older system. I may note a few of this kind which illustrate the affinities of Judaic, Christian, and pagan legend. The cessation of human-sacrifice in the Mediterranean area, the awakening of the conviction that the practice was abhorrent to a merciful God, implied