An example will illustrate this: If there had been no written memorials of the life of John Wesley, there can be no doubt that the society which he founded would have handed down to the present day an account of the chief events of his life, which would have been accurate in its main outlines. Thousands of persons are now living who have conversed with those who have heard him preach; I myself have done so. It would therefore be impossible to impose upon them a wholly mythic account in place of that which would have been handed down by the Wesleyan body. Yet this society is founded on a set of dogmas, not on the historical facts of its founder’s life. The Christian Church therefore was in a far superior position for preserving a substantially accurate account of the chief events in the life of Jesus Christ, yet the interval which separates us from the death of Wesley is greater than that which elapsed between the death of Christ, and the publication of the latest of the Synoptic Gospels, even if we accept the dates which are assigned to them by our opponents. 6. When the knowledge of past events has perished, it is impossible to re‐ construct them by the aid of conjecture, except within the limits to which I have previously alluded. These limits must be strictly defined, otherwise that which is propounded as history becomes nothing else than a statement of our subjective impressions. Conjectures which cannot stand the test of historical verification cannot be accepted as facts of history. Nothing is easier than, when facts are wanting, to invent them, and thus bridge over the intervals which lie between others, the connecting links of which have perished. But how are we to know that such conjectural events were real facts, and not mere creations of the imagination? Clearly this can be determined in no other way than by subjecting them to a rigid verification. If they will not endure this, they must be rejected. Historical conjectures have no higher claims for acceptance than scientific ones. Both must be subject to the same tests, and must share the same fate. I do not deny that many such conjectures may have a considerable degree of plausibility; but, unless we rigidly reject from the rank of historic facts those that break down under the test of verification, histories will be converted into novels or poems. If our knowledge of the connecting links between events in the history of the past has perished, we shall not improve it by imagining facts, and calling the result by the name of history.