We have now to consider the principle of causality so far as it is implied in the theistic inference, and the theistic inference so far as it is conditioned by the principle of causality. It is not necessary to discuss the nature of the principle of causality in itself or for its own sake; it is even expedient, I believe, not to attempt to penetrate farther into its metaphysics and psychology than the work on hand imperatively requires. We must of course go as far as those have gone who have maintained on metaphysical or psychological grounds that the principle of causality warrants no theistic inference; we must show that their metaphysics and psychology are irrelevant when true, and false when relevant; but we may be content to stop when we have reached this result. The truth of theism has been very generally represented, both by those who admit and by those who deny the validity of the theistic inference, as much more dependent than it really is on the truth or falsity of some one or other of the many views which have been entertained as to the nature of causation, and the origin of the causal judgment. We are constantly being warned by theists that unless we accept this or that particular notion of causation, and account for it in this or that particular manner, we cannot reasonably believe in the existence of God; we are constantly being assured by anti-theists that belief in God is irrational, because it assumes some erroneous view of causation, or some erroneous explanation of the process by which causation is apprehended. But it will be found that representations of this kind seldom prove more than one-sidedness and immaturity of thought in those who make them. An accurate and comprehensive view of the nature of causation, and of our apprehension of it, will, it is true, have here as elsewhere great advantages over an erroneous and narrow one, but hardly any of the theories which have been held on these points can be consistently argued by those who hold them to invalidate theistic belief. Even utterly inadequate statements and explanations of the principle of causality--as, for example, those of Hume and J. S. Mill--are not more incompatible with the theistic inference than they are with any other inference which is a real extension of knowledge. Unless they are understood and applied more rigidly than by those who propound them, they allow us to draw the theistic inference; if understood and applied so as to forbid our drawing it, they logically disallow all scientific inference except such as is purely formal and deductive. In a word, if compatible with science they are compatible with theism, and if incompatible with theism they are incompatible with science.