We

because it is kindred in principle to a philosophy which has had a good deal of acceptance in our day, that of the Positivists. Hume has objected that our argument employs the law of causation in an illicit manner. The inductive principle that like consequents have like antecedents, which is at the bottom of our reasoning, may hold throughout nature, but will it hold outside of nature? have no right, he contends, to stretch the analogies of the Universe beyond its limits, and thus assert the existence of a supramundane Being. We reason, we say (in popular language), by analogy. But has the Universe, as a whole, any analogue? Works of design may have intelligent makers within the limits of our world, but we have no right to apply the law to the world as a whole, and so rise above it.

I have already fully admitted that this argument from design could never establish the existence of an Eternal Being of infinite wisdom and power. The Maker of the Universe, as thus made known to us, is after all only a finite Being. We need some other line of reasoning, pointing to the same Being, to convince us that he is in any respect truly infinite. If this shortcoming of our argument were all that Hume' meant to assert, his criticism might at once be admitted. But I understand him

1 See Hume's Essay 'On a Particular Providence and a Future State;' and also on 'Necessary Connection.'

to mean something more, viz., that this reasoning cannot in reality disclose anything at all as to God. It becomes illicit as being carried beyond the limits of the Universe. Now, there seems to be here an arbitrary circumscribing of the space within which our generalisation will hold. We do but proceed by that principle of induction which, as I shall endeavour hereafter in another essay more fully to point out, is at the bottom of all our objective knowledge. Why must we stop at the limits of the visible creation? Let us review our proceeding. Like consequents, we say, have like antecedents. In the works of man we see that traces of design have arisen from an intelligent Maker. We examine a part of nature. We see traces of design. We say this part must have had an intelligent Maker. We examine more, we repeat the conclusion. We see a unity in nature which implies a common author. We accordingly extend our inference to the whole known Universe. Where does the reasoning become illicit? Certainly not because the Being whose existence is inferred is invisible, or, in more general language, is known only from the hypothesis of our argument. If so, we should have no right to attribute reasonable minds to one another. The natural philosopher has no such scruples. He freely imagines invisible agents to explain sensible phenomena on mechanical prin

ciples, as in the undulatory theory of light or the atomic theory of matter, and I think I may add the Newtonian theory of gravity. No doubt he is very careful to compare his hypothesis with facts, to test it by observation and experiment. But if it stands this test well and persistently, he will accept the hypothesis at last with confidence. And indeed in so doing he only carries further the conduct of everyday life. Much of sensation itself is subjective.1 What we popularly call seeing, for instance, is an interpretation of certain physical effects on our frames, which we learn to make by comparison of other sensations and other mental work. Dr. Whewell was right when he contended that there was no stable distinction between facts and theories, paradoxical as the saying may seem. And it would be a just exhibition of the same truth to say that our notion of the external world is in reality a hypothesis which our minds have framed to explain certain sensations. If so, why should it be illicit to frame, according to principles drawn from mental experience, a further hypothesis of a God to explain this first hypothesis of a world?

I have dwelt a little longer on this objection of Hume because, as I have said, the principles involved seem to me to be the same as those at the

1 See 'Popular Lectures on Scientific Subjects,' by H. Helmholtz, translated by E. Atkinson.

bottom of the controversy with the Positivists. So far as the teaching of M. Comte was a protest against purely à priori speculation, and an insisting upon the need of conforming our theories to the results of observation, it seems to me worthy of praise. The continued and wonderful success of the experimental philosophy must, I think, justify such teaching to everyone. The progress of that philosophy, indeed, gives hope that it will explain consistently those mental phenomena over which metaphysicians have wrangled in vain. But when M. Comte goes further, when he seems to restrict the sphere of knowledge to mere phenomena, and the work of science to their registering and arrangement, then he is to my thinking unphilosophical. As I have pointed out, we cannot observe without making hypotheses-without doing something of the very thing which M. Comte would forbid as characteristic of the metaphysical or theological stage of human thought. As is well known, he objected to the undulatory theory of light. His principles, thoroughly carried out, would leave us in universal ignorance or doubt, a self-destructive conclusion. One remark more before I leave this objection. I have said that we needed some other reasoning besides this causation argument to prove the eternity of God, or, indeed, his infinity in any respect. On that subject I will speak further by-andby. But let me at once notice one point. It may

be thought that if upon any grounds we pronounce the Deity to be an eternal, and therefore an uncaused, Being, we set aside the very principle of sequence upon which we have been reasoning. I should reply that we cannot upon any grounds assert that an eternal or uncaused being cannot be a cause to a finite being the starting point of such a series as we see in nature. It is true that experience does not suggest the idea of such a being, but at the same time it cannot justify us in refusing to accept a reason that may come from any other quarter for believing that such a being exists.

The objections which I have now described were felt, and were more or less distinctly brought forward, in past days, but they did not prevent the Design argument from having the popularity of which I have spoken. So recently as forty years ago, in the days of the Bridgewater Treatises, it was generally relied upon in this country. As an instance of its influence at a still later time, I may, I think, quote the tenacity with which the habitation of other worlds was maintained. This persuasion arose in great part from an inverse application of the argument. If worlds were not inhabited, why had they been made? But there has been a change. Let us now go on to examine how.

There has been of late years a prodigious advance of knowledge, and especially of the physical sciences.