First: The positions taken by unbelievers are theories, which rest on the smallest basis of historical evidence. I readily admit that where there is a known fact, but the recollection of the events which would give an account of its origin has perished, if a theory can be propounded which fully accounts for the fact, then it has a right to take its place as an historical event which rests on evidence of the highest probability. An example derived from the mode in which the study of comparative philology discloses the history of the past will explain my meaning. We have before us the facts of language. The history of those who formerly used it has perished; the accounts of their migrations have nowhere been preserved. But certain facts of comparative philology justify the assumption that certain primitive races of men must have migrated in particular directions. These assumed migrations are really a theory, but one which is exactly adequate to account for the facts which language unquestionably presents. Thus the facts of the Indo‐Germanic languages justify the assumption that in the pre‐historic ages, migrations westward must have taken place, of which history contains no record. Still the theory affords so perfect an explanation of the facts, that the occurrence of the migrations is as certain as if they had been recorded by contemporaneous writers. On similar grounds it has been inferred with a degree of probability so high as to be equal to certainty, that a language earlier than the Sanskrit, and from which both it and the Indo‐Germanic family of languages have been derived, was spoken by a previous race. Investigations of this kind are largely adding to our historical knowledge. Let us observe the basis on which such arguments rest. In all these cases we have before us not mere conjectures, but a distinct and positive fact, or set of facts. The connecting links are missing. By the aid of conjecture we propound a theory; or in other words, we suppose a set of events to have occurred, which, if they really happened, would be adequate to account for the facts in question. When they thus account for them, and for them alone, and no other conjectural occurrence will do so, the assumed fact is fully entitled to take its place in history as an event which has actually happened. The reason of this is, that it can stand the test of historical verification.