Different ages differ enormously in the severity of proof which they exact, in the degree of accuracy which they attain. The credibility of a statement also depends not only on the amount of its evidence, but also on its own inherent probability. Everyone will feel that an amount of testimony that would be quite sufficient to persuade him that a butcher's boy had been seen driving along a highway is wholly different from that which would be required to persuade him that a ghost had been met there. The same rule applies to the history of the past, and it is complicated by the great difference in different ages of the measure of probability, or, in other words, by the strong predisposition in certain stages of knowledge to accept statements or explanations of facts which in later stages we know to be incredible or in a high degree improbable. Few subjects in history are more difficult than the laws of evidence in dealing with the supernatural and the extent to which the authority of historians in relating credible and probable facts is invalidated by the presence of a mythical element in their narratives. Connected with this subject is also the question how far it is possible by merely internal evidence to decompose an ancient document, resolving it into its separate elements, distinguishing its different dates and its different degrees of credibility. The reader is no doubt aware with what a rare skill this method of inquiry has been pursued in the present century, chiefly by great German and Dutch scholars, in dealing with the early Jewish writings. At the same time, without disputing the value of their work or the importance of many of the results at which they have arrived, I may be pardoned for expressing my belief that this kind of investigation is often pursued with an exaggerated confidence. Plausible conjecture is too frequently mistaken for positive proof. Undue significance is attached to what may be mere casual coincidences, and a minuteness of accuracy is professed in discriminating between the different elements in a narrative which cannot be attained by mere internal evidence. In all writings, but especially in the writings of an age when criticism was unknown, there will be repetitions, contradictions, inconsistencies and diversities of style which do not necessarily indicate different authorship or dates.