says? Why but because we have learnt by experience that, when people have no apparent motive for lying, they commonly do speak the truth? But the same experience which has taught us this has taught us likewise that people do now and then lie without apparent motive. At best, therefore, there is never more than a probability that people are speaking the truth, while in this instance, the supposition that they might be speaking the truth would imply that there may be a truth against which there is proof amounting to certainty. For what they affirm is that something has happened the very reverse of what has invariably happened before in the same circumstances. Is it not infinitely more likely that people should be lying as they have often done before, than that the invariable course of nature should have undergone a variation? With evidence on the one side that has never yet deceived, with no evidence on the other save what has often proved deceptive, how can we hesitate which to accept? Even though the unanimity of testimony be such as might otherwise be deemed complete proof, it is here met by absolutely complete proof in the shape of a law of nature. The greater probability overwhelms the lesser. The stronger proof annihilates the weaker, leaving none of it behind, so that whoever still persists in believing that a law of nature has been violated, must be content to do so without one particle of proof. No quantity of testimony can furnish the smallest proof of a miracle unless the falsehood of the testimony would be a greater departure from antecedent uniformity--in other words, would be a greater miracle--than the miracle which it attests. Unfortunately it is but too notorious that there is not, and never has been, such a thing as uniform truthfulness of testimony to depart from.