about twenty genera. Nevertheless, a great gap still exists between these mammals and those of the Tertiary strata, since no mammal of any kind has been found in any part of the Cretaceous formation, although in several of its subdivisions abundance of land plants, freshwater shells, and air-breathing reptiles have been discovered. So with fishes. In the last century none had been obtained lower than the Carboniferous formation; thirty years later they were found to be very abundant in the Devonian rocks, and later still they were discovered in the Upper Ludlow and Lower Ludlow beds of the Silurian formation.
We thus see that such sudden appearances are deceptive, and are, in fact, only what we ought to expect from the known imperfection of the geological record. The conditions favourable to the fossilisation of any group of animals occur comparatively rarely, and only in very limited areas; while the conditions essential for their permanent preservation in the rocks, amid all the destruction caused by denudation or metamorphism, are still more exceptional. And when they are thus preserved to our day, the particular part of the rocks in which they lie hidden may not be on the surface but buried down deep under other strata, and may thus, except in the case of mineral-bearing deposits, be altogether out of our reach. Then, again, how large a proportion of the earth consists of wild and uncivilised regions in which no exploration of the rocks has been yet made, so that whether we shall find the fossilised remains of any particular group of animals which lived during a limited period of the earth's history, and in a limited area, depends upon at least a fivefold combination of chances. Now, if we take each of these chances separately as only ten to one against us (and some are certainly more than this), then the actual chance against our finding the fossil remains, say of any one order of mammalia, or of land plants, at any particular geological horizon, will be about a hundred thousand to one.
It may be said, if the chances are so great, how is it that we find such immense numbers of fossil species exceeding in number, in some groups, all those that are now living? But this is exactly what we should expect, because the number of species of organisms that have ever lived upon the earth, since
