Now, when we find Mr. Codrington taking the same precautions in Melanesia, and when his account of Melanesian myths reads like a close copy of Mr. Sproat's account of Aht legends, and when both are corroborated by the collections of Bleek, and Hahn, and Gill, and Castren, and Eink, in far distant corners of the world, while the modern testimony of these scholarly men is in harmony with that of the old Jesuit missionaries, and of untaught adventurers who have lived for many years with savages, surely it will be admitted that the difficulty of ascertaining savage opinion has been, to a great extent, overcome. If all the evidence be wrong, the coincidences of the witnesses with each other and of the savage myths they report with the myths of Greeks and Aryans of India will be no less than a miracle. We have now examined the objections urged against a system founded on the comparative study of savage myths. It cannot be said of us (as it has been said of De Brosses), that "whatever we find in the voyages of sailors and traders is welcome to us; that we have a theory to defend, and whatever seems to support it is sure to be true". Our evidence is based, to a very great extent, on the communications of missionaries who are acknowledged to be scholarly and sober men. It is confirmed by other evidence, Catholic, Dissenting, pagan, scientific, and by the reports of illiterate men, unbiassed by science, and little biassed by religion. But we have not yet exhausted our evidence, nor had recourse to our ultimate criterion. That evidence, that criterion, is derived from the study of comparative institutions, of comparative ritual, of comparative law, and of comparative customs. In the widely diffused rites and institutions which express themselves in actual practice we have sure evidence for the ideas on which the customs are founded. For example, if a man pays away his wampum, or his yams, or his arrow-heads to a magician for professional services, it follows that he _does_ believe in magic. If he puts to death a tribesman for the sin of marrying a woman to whom he was only akin by virtue of common descent from the same beast or plant, it seems to follow that he _does_ believe in descent from and kinship with plants and beasts. If he buries food and valuable weapons with his dead, it follows that he _does_, or that his fathers did, believe in the continued life of the dead. At the very least, in all three cases the man is acting on what must once have been actual beliefs, even if the consequent practices be still in force only through custom, after the real faith has dwindled away. Thus the belief, past or present, in certain opinions can be deduced from actual practices, just as we may deduce from our own Coronation Service the fact that oil, anointed on a man's head by a priest, was once believed to have a mysterious efficacy, or the fact that a certain rough block of red sand-stone was once supposed to have some kind of sacredness. Of all