Hume's sudden conversion, so amazing at first sight, from flattest denial to positivest assertion of causal power, becomes intelligible when he is seen immediately afterwards using his newly adopted creed as a fulcrum whereon to rest his argumentative lever in his assault upon Miracles. About that celebrated piece of reasoning, startling as the avowal may sound, there is, to my mind, nothing more remarkable than its celebrity, for, on close inspection, it will be found to be entirely made up of (1) the demonstration of a truism, and (2) the inculcation of a confessedly misleading rule. Not far from its commencement will be found a definition which, if correct, would leave nothing to dispute about. A miracle, we are told, is 'a violation of the laws of nature,' of laws which a firm and unalterable experience has established. But if so, _cadit quaestio_. Of course, there can be no alteration of the unalterable. No need, of course, of further words to prove that a miracle thus defined is an impossibility. Let us suppose, however, the word _unalterable_ to have been used here by a slip of the pen instead of _unaltered_, and that Hume really meant by a miracle any alteration of what had previously appeared to be the constant course of nature. Even so, we shall have him contending that no amount, however great, of testimony however unimpeachable, ought to be accepted as adequate proof of such an alteration. Of what he urges in support of this position much may be at once dismissed as altogether irrelevant. That the most honest witness may be the dupe of optical or auricular illusion, or of a distorting or magnifying imagination; that there is in many minds a natural predisposition to believe in the marvellous, and that the love of astonishing often gives exaggerated expression to the exaggerations of the fancy; that self-interest and religious zeal often furnish additional motives for mendacity, and that testimony, even when sincere at first, is apt to become corrupted at every stage as it passes from hand to hand, or is committed to paper--all this, together with any further enumeration of circumstances calculated to invalidate testimony, is quite beside the real question. It merely proves what no one needs to have proved, the propriety, viz., of weighing evidence and balancing adverse probabilities; and even though it proved in addition that of all the so-styled miracles on record, there is not a single one the evidence for which is sufficient, it would still prove nothing to the purpose. For Hume is arguing against the credibility, not of any miracles in particular, but of all miracles in general, those included the witnesses for which are of indisputable intelligence and undisputed veracity. Be the quality of the testimony what it may, no quantity of it, according to him, can be sufficient. This is the essence of his thesis, the only part of it in which there is any novelty, and in behalf of this part all that he has to say may be