investigations. Certain facts are observed. In what manner are they to be accounted for? that is, what must have been the nature and the order of the changes by which these appearances were produced? When we have conceived of a cause, or succession of causes, which will account for all the facts, and which alone can account for them, we may consider such cause or causes as matter of established truth. Thus, a geologist observes that a river has cut its way through banks a hundred feet high. Some thirty feet below the surface of the soil a layer of vegetable matter is discovered, the stumps of trees, standing upright, imbedded in the soil where they grew, and the trees broken off lying upon and by the side of them. Some thirty feet lower, another stratum of a similar character is observed. From the position of these trees it is evident that they also must have grown on the spot where they are found, and, of course, that each of these layers must have been, at the time of its growth, on the surface of the earth. There is but one way in which these facts can be accounted for. After the lower layer of trees had grown to its present size, the surface of the earth must have subsided until they were covered with drift for thirty or more feet. The subsidence was then arrested until another forest grew up. Another subsidence must have occurred until the drift covered the timber again to a similar depth. Then the whole surface must have been upheaved to its present position, and, afterwards, the river must have cut its way through the mass, thus laying bare the mode of its formation. As no other cause can be assigned. for these effects, we are warranted in believing that such events as these actually existed. It will be seen that direct and circumstantial evidence may frequently be found corroborating each other, and they then present the strongest possible ground of belief. If any marked event occur, not only will it be seen by witnesses, but it will be preceded by its appropriate causes, and followed by its appropriate effects. Thus, the death of Cæsar is proved by the testimony of eye-witnesses, and contemporary writers. But, besides this, the civil wars in the Roman empire, and the character of the parties that were formed immediately after that event is said to have taken place, can be accounted for on no other supposition than that of his violent death. So the invasion and occupation of Britain by the Romans is proved by the testimony of historians. But if such an event had occurred, we should naturally expect that some traces of their occupation would be observed in that island. Hence, we examine, and find there the remains of Roman encampments, walls, roads, Roman coins of that age, and inscriptions which could have been made by no other people. These facts can be accounted for on no other supposition than that of the conquest and permanent occupation of Britain by the former conquerors of the world. This coïncidence of direct and indirect evidence furnishes the most perfect ground of belief which we can conceive to any matter of fact. REFERENCES. Evidence of testimony - Reid, Essay 7, sec. 3; Stewart, vol. ii., chap. 2, sec. 4; Abercrombie, Part 2, sec. 3. Different kinds of evidence - Reid, Essay 2, chap. 20. Testimony of others a source of knowledge — Locke, Book 4, chap. 16, secs. 6-8; Abercrombie, Part 2, sec. 3. Laws of testimony Abercrombie, Part 2, sec. 3. Natural bias to truth-Abercrombie, Part 2, sec. 3; Reid's Inquiry, chap. 6, sec. 24. Hume's argument against miracles - Abercrombie, Part 2, sec. 3. SECTION IV. OTHER FORMS OF REASONING. I. Of probable evidence. Thus far I have treated of those modes of reasoning in which our premises are acknowledged to be true, and our conclusion is equally, that is, absolutely true. But all of our reasoning is not of this character. It frequently happens that our premises rise no higher than probability, and our conclusions can only reach the same level. Our process is, however, precisely the same, the only difference consists in the degree of certainty to which we arrive. When the reasons for believing a proposition to be true are not such as to establish belief, but only to show that it is more likely to happen than not, we say that such a proposition is probable. Thus, if the wind is in a certain quarter, I say that it probably will rain. I examine the evidence that may be adduced in favor of the proposition that the planets are inhabited, and I say that it is or is not probable. It may require the coöperation of several causes to render an event certain. If, however, only a part of these causes unite in a particular case, the event may occur, though we cannot expect it with confidence. So, if an intelligent being has several times, under given circumstances, acted in a particular manner, we form a distinct anticipation that he will act in the same manner under similar circumstances. But here our anticipation only amounts to a probability, for we know not what changes may have taken place in his character since we last observed him; and there may have arisen circumstances which affect him of which we are ignorant. When, in this manner, we attain to mere probability, we call our state of mind opinion; that is, we judge a proposition more likely to be true than false. We take such opinions as the grounds of our reasonings in a large number of cases in practical life. Thus, we say, It is probable that the character of a human being will be improved by affliction. A. B. has suffered affliction; therefore, A. B. is probably improved in character. Or, again: If there be war in Europe, the price of breadstuffs will rise. There will probably be a war in Europe; therefore, It is probable that the price of breadstuffs will rise. When only one of our premises is a doubtful and the other a certain proposition, the probability of our conclusion is equal to that of our doubtful premise. Thus, it being granted that if there be war in Europe prices will rise, the probability of our conclusion is precisely as great as the probability of a war. When, however, both of our premises are mere probabilities, the probability of our conclusion is greatly reduced, and can rarely furnish a ground for an opinion. Thus, If the south wind blow to-morrow, it will probably rain. The south wind will probably blow to-morrow; therefore, It is (very slightly) probable that it will rain. When so slight an indication of an event is given, it is manifestly of very little use in forming a judgment. From the fact that we reason from probabilities, very commonly, in the practical business of life, it has happened that this mode of reasoning has sometimes been confounded with that by which we arrive at practical certainty. It has sometimes been said that moral reasoning, or reasoning concerning matters of fact, is nothing else than a succession of probable arguments, each one reducing the liabilities of error, until they become so small as to be inappreciable. The cases, however, are dissimilar. In the one case, we proceed from an approximation to truth so near that neither we nor other men can discover any error, and the result is of the same character. In the other case, we proceed from an approximation to truth, but so distant that we can appreciate our liability to error; we know the uncertainty of our premises, and the result is a mere approximation similar to them, producing not belief, but merely opinion. For instance, suppose we endeavor to ascertain whether the battle of Waterloo was fought on the 18th of June, 1815. We proceed according to the laws of evidence as before stated. We apply the rule of perception, and the rule of human motive. We can discover no error, and no other man can discover any. I rely upon the result at which I have arrived with perfect confidence, and the state of mind of which I am conscious is belief, full, entire, and unquestionable. Again; the question is asked, when did the battle commence? I find that here the accounts vary. The best authorities differ, some placing it as early as ten o'clock, and others as late as one. I form an opinion, by comparing the accounts, and balancing the probable motives which would lead men into error. I form an opinion as to the time, but it is not belief. I am conscious of a state of mind very dissimilar to that in the preceding case. Or, again; from the data established by observation as accurate as the faculties of men will permit, we determine the distance and magnitude of the planet Jupiter. No error can be discovered either in our data or our reasoning. We know that there may be error, but that it cannot exceed a certain amount, and we rely on the result under this condition with absolute certainty. But when it is said the planet Jupiter is inhabited, we collect our data, and they give us nothing but a probability to reason from, and we arrive at nothing but an opinion. The states of mind differ not in degree but in kind. The one proceeds from data in which no error can be discovered by the faculties which |