It leaves untouched all that can be inferred from the existence of the conditions which make organic evolution possible: matter which lives, multiplies, and varies; an environment which possesses the marvellously complex constitution required to make these processes possible. Selection may modify these conditions, but it cannot start them. It may modify the manner in which multiplication is secured; it may modify the lines which variations follow; it may enable organic species to adapt their powers to their environment, and (within narrow limits) their environment to their powers. But it cannot produce either the original environment or the original living matter. These must be due either to luck or to contrivance; and, if they be due to luck, the luck (we must own) is great. How great we cannot say. We cannot measure the improbability of a fortuitous arrangement of molecules producing not merely living matter, but living matter of the right kind, living matter on which selection can act. Here, indeed, Laplace’s calculator might conceivably help us. But suppose him to have done so, suppose him to have measured the odds against the accidental emergence of the desired brand of protoplasm, how are we to compare _this_ probability with its assumed alternative—intelligent design? Here, I think, even Laplace’s calculator would fail us; for he is only at home in a material world governed by mechanical and physical laws. He has no principles which would enable him to make exhaustive inferences about a world in which other elements are included: and such a world is ours. For a Greek philosopher to assert that the world is material was legitimate enough. He was in search of a universal principle; and if he found it in matter we need neither wonder nor criticise. After all, matter lies round us on every side; we are immersed in it; we are largely dependent on it. It may well seem but a small step further, and a very natural one, to treat it as the essence of all that is.