In ages when any dispassionate comparative study was impossible, religious systems appear to have been considerably modified by the influence of those of conquered peoples on those of their conquerors, and vice versâ. Peoples who while at arm's length would insult and affect to despise each other's Gods, and would deride each other's myths, [121] appear frequently to have altered their attitude when one had conquered the other; and this not because of any special growth of sympathy, but by force of the old motive of fear. In the stage of natural polytheism no nation really doubted the existence of the Gods of another; at most, like the Hebrews of the early historic period, it would set its own God above the others, calling him "Lord of Lords." But, every community having its own God, he remained a local power even when his own worshippers were conquered, and his cult and lore were respected accordingly. This procedure, which has been sometimes attributed to the Romans in particular as a stroke of political sagacity, was the normal and natural course of polytheism. Thus in the Hebrew books the Assyrian conqueror is represented as admitting that it is necessary to leave a priest who knows "the manner of the God of the land" among the new inhabitants he has planted there. See 2 Kings xvii, 26. Cp. Ruth i, 16, and Judges xvii, 13. The account by Herodotos (ii, 171) of the preservation of the Pelasgic rites of Dêmêtêr by the women of Arcadia points to the same principle. See also hereinafter, ch. vi, § 1; K. O. Müller, Introd. to a Sci. Study of Mythol., Eng. trans., p. 193; Adolf Bastian, Der Mensch in der Geschichte, 1860, i, 189; Rhys, Celtic Britain, 2nd ed., p. 69; Max Müller, Anthropological Religion, p. 164; Gibbon, ch. xxxiv--Bohn ed., iii, 554, note; Tylor, Primitive Culture, i, 113-15; and Dr. F. B. Jevons's Introd. to the Hist. of Relig., 1896, pp. 36-40, where the fear felt by conquering races for the occult powers of the conquered is limited to the sphere of "magic." But when Dr. Jevons so defines magic as to admit of his proposition (p. 38) that "the hostility from the beginning between religion and magic is universally admitted," he throws into confusion the whole phenomena of the early official-religious practice of magic, of which sacrifice and prayer are the type-forms that have best survived. And in the end he upsets his definition by noting (p. 40) how magic, "even where its relation to religion is one of avowed hostility," will imitate religion. Obviously magic is a function or aspect or element of primitive religion (cp. Roskoff, Das Religionswesen der rohesten Naturvölker, 1880, p. 144; Sayce, pp. 315, 319, 327, and passim; and Tiele, Egyptian Rel., pp. 22, 32); and any "hostility," far from being universal, is either a social or a philosophical differentiation. On the whole question compare the author's Pagan