resolved into a sophism, followed by a repetition of the same begging of the question, as is involved in his afore-cited definition of a miracle. In substance it runs as follows. All testimony is at best but a description of the results of sensible experience--of observations of the senses--but the most faithful description must needs be a less vivid presentation of truth than the reality described. A single original is better evidence than any number of copies. Your own personal experience is more trustworthy than any number of mere records of the experience of other people, and where the two conflict, the former always deserves preference. Now the personal experience of each one of us assures us that many sets of natural phenomena take place in perfectly invariable sequences, in sequences so invariable as to appear to be, and to be familiarly spoken of as, manifestations or operations of certain inflexible laws of nature. Within our experience there has never been a single deviation from any such law. Wherefore, though all the rest of mankind should unite in asserting that they have observed such a deviation, we ought not to believe them. Even though, for example--the example, however, being not Hume's, but my own--we were, on leaving home some morning, to hear on all sides that, while we were yet in bed, the sun was seen to rise in the west instead of the east, and though we found the statement repeated in the 'Times' and 'Daily News,' and presently afterwards saw it posted up at the Exchange as having been flashed by electric wire from New York and Kurrachee, we are not for a moment to doubt that these reiterated and mutually corroborative statements are utterly false. For, numerous and consistent as they may be, they are but copies of the experience of other people, while, although we may have to oppose to them only our own single experience, still that single experience is original, and therefore of more worth. The value, moreover, of any experience is, irrespective of originality, determined by the difference in number between the results of opposite kinds which it has discovered. The smaller number is deducted from the larger, and the balance represents the probability that the results which have most frequently occurred hitherto will continue to occur henceforward. The larger the deduction thus to be made, the smaller the probability, and _vice versa_; and when the deduction is _nil_, or when there has hitherto been complete uniformity, the probability becomes virtual certainty. When two original experiences are opposed to each other, their respective values, ascertained in the same manner, are compared, and trust is reposed in one or the other accordingly. Now, according to these principles, difficult as it might be in the case supposed for any one to conceive what motive all the rest of the world could have for lying, it would be still more difficult to believe them to be speaking truth. For why do we ever believe anything that anyone