tablished. One may judge therefore of the weight of testimony necessary to prove a suspension of those laws, and how fallacious it is in such cases to apply the common rules of evidence."

It sometimes happens, however, that a prevailing opinion, or a prejudice, may so diminish the natural improbability of an event, that it shall appear easily overcome by the force of testimony.

"This has happened with men of the first abilities; and in the age of Lewis XIV., Racine and Pascal were two remarkable examples of it. It is humiliating to see with what complacency Racine, that admirable painter of the human heart, and the most perfect poet who has ever been, relates as a miraculous event, the cure of Mademoiselle Perrier, the niece of Pascal, and pensionnaire of the Abbey of Port-Royal: It is no less painful to read the reasonings by which Pascal endeavours to prove that this miracle had become necessary to the cause of religion, in order to justify the doctrine of the Nuns of that Abbey, at that time persecuted by the Jesuits. The young Mademoiselle Perrier, who was then about three years and a half old, was afflicted with a fistula lachrymalis; she touched her sore eye with a relique which professed to be one of the thorns of the crown placed by the Jews on the head of our Saviour, and she believed herself cured from that instant. Some days after, the physicians and surgeons attested the cure, and gave

it as their opinion (in which probably they were perfectly correct) that the medicines had had no effect in bringing it about. This event, which happened in 1656, made a great noise: All Paris," says Racine," flocked to Port-Royal. The crowd increased from day to day; and God seemed to take pleasure in authorizing the devotion of the people, by the number of miracles worked in that church."

The question here touched on, how far the evidence of testimony is able to overcome that which arises from our experience of the course of nature, is one of the most delicate and important which the Doctrine of Probability presents. That testimony itself derives all its force from experience, seems very certain. This, however, has sometimes been disputed; and it has been urged, that there is a natural tendency to believe in the testimony of others, independent of experience. That such a tendency really exists, we are willing to allow. A man who feels in himself a propensity to speak the truth readily supposes a like propensity in others; and therefore, previous to all experience, may be disposed to believe in their testimony. He soon learns, however, that he cannot trust safely to this principle; for he perceives, that though men have a tendency to speak the truth, they have often motives that lead them to do the contrary, that tempt them to conceal and even to pervert it; and how much

these opposite motives may counteract one another, is a matter only to be collected from experience and observation. Indeed, it is quite evident, that whatever propensity we naturally have to believe in testimony, it must be in itself extremely fallacious, as bearing no proportion to the probability of, the thing believed, or the likelihood that it will happen.

It is useless, therefore, in treating of probability, to talk of a tendency to believe, which, confessedly not being regulated by the experience of the past, cannot be depended on for its anticipation of the future. Such a tendency, whether natural or acquired, is evidently no better than a mere prejudice, and is as likely to lead to error as to truth. The evidence of testimony, then, is measured in the same way with other probabilities, and is expressed by the number of instances in which men, circumstanced in a particular way, have been known to speak true, divided by the number of cases in which they have given evidence whether true or false. It is true that the strict arithmetical value of this fraction is hardly possible, in any case, to be assigned. But a certain coarse and loose estimate of it may be formed, sufficient for directing the judgment and the conduct on ordinary occasions.

The first author, we believe, who stated fairly the connection between the evidence of testimony and the evidence of experience, was Hume, in his

Essay on Miracles, a work full of deep thought and enlarged views; and, if we do not stretch the principles so far as to interfere with the truths of religion, abounding in maxims of great use in the conduct of life, as well as in the speculations of philosophy.

Conformably to the principles contained in it, and also to those in the Essay now before us, if we would form some general rules for comparing the evidence derived from our experience of the course of nature with the evidence of testimony, we may consider physical phenomena as divided into two classes, the one comprehending all those of which the course is known from experience to be perfectly uniform; and the other comprehending those of which the course, though no doubt regulated by general laws, is not perfectly conformable to any law with which we are acquainted; so that the most general rule that we are enabled to give, admits of many exceptions. The violation of the order of events among the phenomena of the former class, the suspension of gravity for example,—the deviation of any of the stars from their places, or their courses in the heavens, &c.-these are facts of which the improbability is so strong, that no testimony can prevail against it. It will always be more wonderful that the violation of such order should have taken place, than that any number of

witnesses should be deceived themselves, or should be disposed to deceive others.

It is here very well worth attending to, how much the extension of our knowledge tends to give us confidence in the continuance of the general laws of nature, and to increase the improbability of their violation. Suppose a man not at all versed in astronomy, who considers the moon merely as a luminous circle that, with certain irregularities, goes round the earth from east to west nearly in 24 hours, rising once and setting once in that interval. Let this man be told, from some authority that he is accustomed to respect, that on a certain day it had been observed at London, that the moon did not set at all, but was visible above the horizon for 24 hours there is little doubt that, after making some difficulty about it, he would come at last to be convinced of the truth of the assertion. In this he could not be accused of any extraordinary and irrational credulity. The experience he had of the uniform setting and rising of the moon was but very limited; and, the fact alleged, might not appear to him more extraordinary, than many of the irregularities to which that luminary was subject. Let the same thing be told to an astronomer, in whose mind the rising and setting of the moon were necessarily connected with a vast number of other appearances; who knew, for example, that the supposed fact could not have happened, unless