admitted. The conclusion, therefore, by Col. Ellis cannot rest on this. But the conclusion itself is disputed by Professor Robertson Smith's well-known theory of the origin of religion in the tribe, and its gradual localisation.
I think I can see that a part of the difference between Col. Ellis and those who have argued for opposite views is perhaps due to a want of harmony in matters of terminology, so fruitful a source of error in our science; but it is clear that this does not account for all the difference. On p. 209, Col. Ellis, in discussing the law of debt, says it "seems to show that the community preceded the family, which one would certainly expect to be the case, when it is remembered that men must have dwelt together in groups long before any such notion as that of kinship had been formed." And in pages previous to this passage Col. Ellis has discussed the beliefs and examined the customs and ritual of the Ewe-speaking people, on the fundamental principle that the oldest cult consists of the worship of local deities, that these disappear, the tribes changing from place to place, and the local deities gradually developing into a tribal worship, and ceasing as local worship. Col. Ellis's theory seems well supported by his evidence, and I should hesitate to oppose it by the result of any researches of my own. But it is diametrically opposite to the central theory of Professor Robertson Smith's researches into Semitic belief, and the importance of the question to folk-lorists is of great magnitude. Local cults seem to me to be a very large element in folk-lore. If their decay commences in primitive times, as the result of their development into primitive tribal belief, how is it that we have them surviving in the latest times as an important element in folk-lore? If it is right to compare the folk-lore of modern civilisation with the beliefs and customs of savages—and, of course, our Society especially has strenuously advanced this right—we must be quite sure of the grounds of our comparison. If Col. Ellis's theory is correct, the local cults in folk-lore
