And now, I can say what I take to be the difficulty, and the escape. Mr. Balfour draws a kind of parallel between the scientific creed, which is, as he would put it, "based upon" a metaphysical doctrine, and the theological creed, which has a similar foundation. If the metaphysical foundation is so uncertain in both cases, must not the scientific be as uncertain as the theological? If we know nothing about atoms, or, on the other hand, about souls, we must be either sceptical in both cases, or credulous in both. There are the same underlying difficulties, and if we manage to overlook them in the case of science, why not overlook them in the case of theology? Conversely, if we elect to be sceptics in theology, how can we escape from scepticism in science? And, as a thorough-going scepticism is, doubtless, an impossible state of mind in practice, the conclusion of many people will be to accept belief in spite of certain gaps in our logical foundations. This, no doubt, is eminently convenient for the "constructive" process adumbrated by Mr. Balfour, which I certainly regard as extra-logical. But is any such dilemma really offered to us? The obvious answer is, that scientific truth, as Mr. Balfour admits, is not "based upon" metaphysical theory. The astronomical doctrine of a Newton remains equally valid, whatever is the ultimate nature of space or laws or atoms; whether we are materialists or empiricists or idealists. The philosophical "basis" is not really a set of truths which we must know before we can know the astronomical theory; but simply a set of hypotheses which have to conform to the truths given by experience. The unassailable truths are just the facts which we observe, and which science enables us to describe accurately and state systematically. If a metaphysical doctrine has any bearing upon these facts, which seems to be doubtful, it must conform to the facts, and not the facts to it. So long as no such theory is proved, we can afford to remain metaphysically sceptical without losing our hold upon the scientific truth. Now, I should say, what is true of the physical sciences is true of all our knowledge. We may study the moral sciences as we can study the physical sciences. We can observe and colligate the facts of emotion and volition, as we can observe the position of the stars and the laws of heat. Therefore, in so far as theology is an attempt to give a theory of the universe in general, we must accept or deny the doctrines just in so far as they serve to explain or fail to explain the facts. But, in any case, the facts will remain unaltered, and will not vanish because we may be unable to understand them. But theology corresponds, also, not to the scientific method, but to the ontological inquiries which are represented by Mr. Balfour's "naturalism". Both doctrines, as I should say, lead to incoherence, to contradictions covered by ambiguous language, and to hopeless difficulties, which, in theology, are described as inscrutable mysteries. I am, therefore, quite ready, with Mr. Balfour, to reject naturalism, but, on the same grounds, I also reject the transcendental theology. Attainable truth is equally independent of all such theories; and were it otherwise, we should be doomed to hopeless scepticism. Mr. Balfour's analogy, therefore, apparently upsets his conclusion. I believe in heat, and I believe in the conscience. I reject the atoms, and I reject the doctrine of atonement. I reject it, if it be meant for science, because, so far from explaining the facts, the facts explain how the false doctrine was generated. I reject it, if it is meant for philosophy, because, like other transcendental theories, it leads to hopeless controversies, and appears to me to be incredible as soon as any such theology as is tenable by a philosopher is substituted for the crude theology of a savage.