Paley's argument, either supply its alleged deficiencies, or remove the belief in their existence. The great questions agitated in that work have been much complicated of late by skeptical quibblings and metaphysical difficulties. The legitimacy of the whole reasoning has been called in doubt, and the points to be proved have been varied and distorted by the makers of philosophical systems. Some complaints might be done away, and much obscurity be dispelled, if the nature of the evidence were once fairly considered, and the relation fully determined which this subject bears to other sciences. This was the scheme of Lord Brougham's first Discourse, but the execution was imperfect, and these volumes do not complete the design. As the subject is of great interest in both a religious and a philosophical view, some desultory remarks upon it may be acceptable to a portion of our readers.

The great problem of Natural Theology is to prove the existence of a God, all the other questions being subsidiary to this, and in great measure dependent upon its solution. Two modes or classes of proof are presented, called the argument a priori and a posteriori. These appellations are unhappily chosen, for in such a case reasoning a priori is impossible, without assuming the very point at issue; we cannot argue from cause to effect in order to prove the existence of a First Cause. And if the meaning of the term be restricted to original and intuitive perceptions, which are independent of experience, the distinction implied by the two phrases does not exist. These first principles of belief are implied in every act of ratiocination; they are taken for granted in the argument from experience, and in every other proof. Besides, we cannot go behind the Divine existence in order to find a basis of proof; we cannot assume a more comprehensive proposition, from which the fact itself can be deduced. We must reason upward to the first principle of all things; and every argument urged with this design must be a posteriori.

But the implied distinction really exists, though improperly designated by two such phrases. In the one case we proceed by moral evidence, and the conclusion is termed in logic only probable, though it may amount to the highest degree of certainty, of which any argument based on experience is susceptible. In the other, the steps are linked together by demonstrative evidence, and the conclusion follows with mathVOL. LIV.NO. 114. 15

ematical certainty. We take no account of those, who assume the Divine existence as an intuitive truth, because their opinions admit of no argument, and to them Natural Theology does not exist as a distinct science. The question between the two modes of proof may appear to be one of pure curiosity, for the inquirer will surely ask, why they cannot be placed side by side, since neither excludes or limits the other, but only offers it fresh support. It is not enough to answer, that a position is improved in strength by removing every rotten or useless prop, which gives at least the appearance of insecurity to the fabric. The very existence of the dispute shows that neither of the proofs is wholly unnecessary, for there are some minds which rest with greater assurance on one argument, and some on another. Neither can remove what is useless to himself, without doing injury to his neighbour. To justify the rejection of either mode of reasoning, it must be shown, that our idea of the point to be proved is affected by the nature of the argumentative process. If the method a posteriori leads to an imperfect or grovelling conception of the Divine Existence, if it abandons the inquirer when he has advanced only half-way, forcing upon him a contingent truth, in place of that absolute and necessary conviction, which, on such a subject, his nature imperatively requires; or if the argument a priori conducts only to a confused and pantheistic notion of a God; if it destroys his personality, and identifies him with an abstract principle, then it becomes a duty not only to prefer one mode of proof, but to expose the fallacy of the other. Here lies, we apprehend, the real ground of dispute. Not only are the two methods unlike; the ultimate theories are contradictory. The question of preference between them ceases to be merely speculative. It exerts a direct and practical influence on our whole scheme of religious belief.

One preliminary remark is necessary, before entering upon the main question. The process by which belief is formed, often differs widely from the manner in which it is substantiated. Our opinions are often imbibed from education, or instinct, or casual circumstances. When attacked, they are often defended by arguments, which had no share in their formation, and in fact never occurred to us, before we had occasion to use them. Such is the case with the elements of religious truth. They were taught to us in infancy, or our minds were predisposed to receive them. "Man," says

Benjamin Constant, "is by nature a religious being, just as he is endowed with the use of language, and a disposition for society. He does not reason out his first creed; he adopts it in a great measure from impulse." All this may be true, but such a disposition does not, in itself, constitute an argument for the truth of his belief. It may be made the basis of such an argument, and he may reason up from it till he arrives at entire conviction. Other proofs may go along with it of equal, or even superior force, and it is no valid objection to them, that they had no influence in creating the original disposition to believe. Very few persons, probably, have been convinced for the first time by the proofs which theologians adduce; their assent may be modified or confirmed by such considerations; but it proceeded originally from another source, and was supported by different influences.

The distinction between moral and demonstrative evidence, relates not merely to the inherent difference between the two processes, but to the difference between the truths, which are substantiated by them. Historical facts rest upon one; abstract propositions upon the other. The creation of the world is a fact, just as much as the foundation of a particular city; it can be proved only by testimony, or from data collected by observation and experience. Abstract propositions can lead only to what is abstract, unless more is gathered in the conclusion than what was distributed in the premises. The existence of a creative Deity, then, can be proved only by what is called the argument a posteriori. In strictness, the present existence of external nature is a fact known only by experience; it is not a necessary truth, for we can conceive of its non-existence, and the idealist philosopher boldly denies its reality. It cannot be assumed as a datum in any species of demonstrative reasoning. Here lies the great defect of the argument adopted by Clarke. All activity, all manifestation of self, may be denied to the infinite Being, whose existence he endeavoured to prove. His argument must be eked out with facts drawn from experience, or the doctrine will coincide with that of Epicurus, who admitted the existence of a God, but denied that he had any agency in the affairs of this world. "Quæ natura primùm nulla esse potest; idque videns Epicurus, re tollit, oratione relinquit, Deos." *

* De Natura Deorum, I. 123.

But we go further. The great truth of Natural Theology is in itself a fact of momentous interest. The being of a God is a reality, an existence in concrete. As such, it is not an object of mathematical or abstract reasoning. All demonstration begins by arbitrary definitions, and ends in abstractions. We might as well think of applying it to prove the fact of a deluge, or of any other event in the world's history, or to show the present existence of an electric fluid, pervading all matter, as to attempt using it in this inquiry. We do not limit its application to mathematics, nor overlook its successful introduction into the mixed sciences. Many propositions in ethics may be established with the same certainty, that attends the conclusions of the geometer, and by a perfectly similar process. Still, they are abstract propositions, and their application to particular cases, to the conduct of individuals, must always be contingent. The reason is obvious. We can speak with certainty of a subject of reasoning, only when its properties are all known and fixed, and its relations are determinate. Particular substances, things existing in concrete, cannot be thus perfectly determined. We can never be sure, that all their qualities have been taken into view, that the conclusion, at which we have arrived, may not be vitiated by something omitted in the primary definition. Or the attributes may shift during the process, or attendant circumstances may modify them in some unforeseen way; and the possibility of such change, small though it be, still makes the result contingent. If a stone be propelled by hand, no mathematical skill, no acquaintance with the laws of motion, can mark out with precision the curve that it will describe, or the exact point at which it will reach the earth. There are a hundred attendant circumstances, which cannot be accurately appreciated, or stated with precision, but which must modify the result. But let the problem be stated hypothetically, let it be a stone of ideal, and therefore exact, measurement, let the propelling power be assumed of an exact force, let it be taken for granted, that no extraneous influences can operate, and the geometer will show the course that the missile must take, and the spot where it must fall to the ground. The result can never be verified by experiment, but it must be true.

Another instance may be taken from the very elements of mathematical science. The geometer does not, as his name

would imply, measure the real earth. No boundaries are marked out, no actual limits are fixed, with the perfect precision which his method requires. He measures an ideal extension; his figures are perfect by hypothesis; they are limited by supposition to given conditions. Even the diagram before him is not the true object of his reasoning, but only its symbol. He proceeds, therefore, with absolute certainty to a determinate result. The law or rule, which he has investigated and established, is applied, it is true, to actual measurements; yet only by approximation. The nicest instrument which the skill of a Troughton or a Ramsden ever framed, only approximates the ideal perfection that the mathematician requires. The abstract result is certain; its application to real things, to existences in concrete, is contingent. Such is the nature of demonstrative reasoning, that this law must always hold. The mathematician owes his success, the precision and certainty of his results, only to his quitting the real world, and dealing with pure abstractions and hypotheses, to which, in strictness, his conclusions are limited. He who would obtain results of the same character, must pursue the same method. The moment he leaves this ideal region, and comes down to real things and events, to the actual instead of the possible, the sphere of demonstration ceases.

The question whether demonstration is limited to quantity, or how far it is applicable in morals, is hardly worth discussing, for it cannot affect the conclusion which we have just established. We incline to believe, that no principle, out of pure mathematics, can be demonstrated, which is not in itself intuitively certain. There are moral as well as physical truths, which can be built up on others of a similar character, or deduced from them, there being a necessary connexion among them. But in every such case it will probably be found, that the reasoning is unnecessary, because the truth of the first proposition is intuitively perceived, and therefore it needs no support. But, however this may be, absolute certainty belongs to the proposition, only when couched in general terms. It can be applied to particular cases only by approximation. The moral judgments of men do not always coincide; some actions are considered as meritorious in a particular age, or among certain nations, which are justly censured by posterity, or by a neighbouring people. Such disagree