There are always certain drawbacks, certain new liabilities to error, which follow the step of each fresh advance in science. The shadow of this kind which attends the comparative method which had been adopted with such splendid results, not only in many natural sciences, but in almost all branches of pre-historic study--the comparative study of laws, institutions, language, myths, and creeds--is a tendency to confound the condition of these things with which we are actually concerned with their condition at some previous time. As Mr. Tylor admirably says about language, that, interesting as it is to trace the history of words, our understanding of their actual meaning is not always facilitated by a misty sense that at some previous time they meant something else, so we may say of many other things--laws, for example, and customs, or, still more, myths and religions. It will be obvious, for instance, that our appreciation of the place in history of certain personages will be very little affected by tracing some of the stories told about them to quite different countries and periods in the history of the world. Suppose (for example) that we should find in New Zealand legends a story closely analogous to the story of Harold’s oath to William the Bastard. It would be by no means safe to affirm that, if we sifted the multitudinous legends of the world, we should not be able to find some pretty close analogy to William’s celebrated trick of concealing the venerated relics beneath the altar. How, it may be asked, would such a discovery affect our estimate of the parts which William and Harold played as the rival claimants for the English throne? If the reader can answer that question he can decide the influence which studies into the religion of the Maoris or Andaman Islanders are likely to have over his estimate of the _rational_ parts of an historic creed. Such a discovery as we have imagined would suggest the possibility that some remote channel of tradition had fathered an old myth upon Harold and William. But it would give us no clue as to how well it fitted upon their characters, how far it gained general currency at the time. Upon these questions alone depends our estimate of the position which the two historic personages occupied in the world of their day. For a story which is generally believed is almost the same as a story which is true. Or, if the reader prefers a story which is really a myth, take the history of Hasting at the siege of Luna, with which most readers will be acquainted, and how he gained an entry into the town by feigning death and obtaining that his body should be carried within the walls for Christian burial. _That_ is undoubtedly a myth; it is found to be sporadic among the histories of the Vikings and of the Normans, their descendants. Should we discover that a very similar story has been current among the Incas of Peru, how far could that discovery affect our estimate of the supposed character of Hasting?