I am not here concerned to argue generally for the shortness or longness of the periods of geological time; let us, for the purposes of argument, admit a very wide margin of centuries and ages; but _some_ limit there must be. The sun's light and heat, for one thing, are necessary, and though the bulk of combustible material in the sun is enormous, there must be some end to it. Sir William Thomson has calculated (and his calculations have never been answered) that on purely physical grounds, the existence of life on the earth must be limited to some such period as 100 millions of years; and this is far too short for uncontrolled evolution. We know from fossils, that species have remained entirely unaltered since the glacial epochs began, and how many generations are included even in that! If no change is visible in all that time, how many more ages must have elapsed before a primitive _Amoeba_ could have developed into a bird or a Mammal? In Florida Mr. Agassiz has shown that coral insects exist unchanged, and must have been so for 30,000 years. When we remember also the enormous destruction of life that takes place, supposing that in a given form a few creatures underwent accidental changes which were beneficial and likely to aid them--still what chances were there that the creatures which began to exhibit the right sort of change should have died before they left offspring! the chances against them are enormous: and the chances have to be repeated at every successive change before the finally perfected or advanced creature took its place in the polity of nature. Moreover, there is the chance of small changes being lost by intercrossing: our own cattle-breeders have most carefully to select the parents, or else the favourable variety soon disappears. How then, seeing the power of stability which at least some forms are found to exhibit--seeing too the enormous chances against the survival of the particular specimens that begin to vary, and the further chances of the loss of variety by intercrossing; how can we get the millions of millions of years necessary to produce the present extreme divergence of species? The fact is that the force of this objection is likely to be undervalued, from the mere difficulty of bringing home to the mind the immeasurable time really demanded by uncontrolled evolution. Nor is the question of time left absolutely to be matter of belief or speculation. For here and there in the geological records of the rocks, we _have_ certain intermediate forms--or forms which we may fairly argue to be such. But looking at the very considerable differences between the earlier and the later of these forms--differences greater than those which now separate well-defined species, it seems questionable whether any of the divisions of Tertiary time, taking all the circumstances into consideration, could be lengthened out sufficiently to accomplish the change.