It is not, then, in the name of this or that philosophy, but in the name of universal experience, that we banish miracle from history. We do not say, "Miracles are impossible." We say, "Up to this time a miracle has never been proved." If to-morrow a thaumaturgus present himself with credentials sufficiently important to be discussed, and announce himself as able, say, to raise the dead, what would be done? A commission, composed of physiologists, physicists, chemists, persons accustomed to historical criticism, would be named. This commission would choose a corpse, would assure itself that the death was real, would select the room in which the experiment should be made, would arrange the whole system of precautions, so as to leave no chance of doubt. If, under such conditions, the resurrection were effected, a probability almost equal to certainty would be established. As, however, it ought to be possible always to repeat an experiment--to do over again what has been done once; and as, in the order of miracle, there can be no question of ease or difficulty, the thaumaturgus would be invited to reproduce his marvellous act under other circumstances, upon other corpses, in another place. If the miracle succeeded each time, two things would be proved: First, that supernatural events happen in the world; second, that the power of producing them belongs, or is delegated to, certain persons. But who does not see that no miracle ever took place under these conditions? but that always hitherto the thaumaturgus has chosen the subject of the experiment, chosen the spot, chosen the public; that, besides, the people themselves--most commonly in consequence of the invincible want to see something divine in great events and great men--create the marvellous legends afterward? Until a new order of things prevails, we shall maintain then this principle of historical criticism--that a supernatural account cannot be admitted as such, that it always implies credulity or imposture, that the duty of the historian is to explain it, and seek to ascertain what share of truth or of error it may conceal.