Unfortunately, it does not go very far--or, we had better say, it is of restricted application. By the time we find well-preserved fossils in the rocks, the main groups of the animal kingdom and their chief subdivisions had been already differentiated, with the one important exception of the vertebrates; while time, heat, and pressure have so modified the earlier strata as to destroy the fossil forefathers of insects, molluscs, crustacea, and the rest, which they must have contained. Within the vertebrate stock, then, we can learn a great deal from the semi-direct methods of paleontology: but for the history of the other groups and for their origin and interrelations, we are driven back upon comparative anatomy and embryology, into another field of more circumstantial evidence. When, for instance, we find that the fore-limbs of bat, bird, whale, horse, and man, although so different in function and in detail of structure, are yet built upon the same general plan, and upon a plan wholly different from that of the limbs, say, of a spider or an insect, we must either deny reason and say that this similarity means nothing; or assume that its cause is supernatural, outside the province of science, that it is the expression of some eternal Idea, or some plan of a personal creator (in which case, be it noted, the idea or the plan often appears to our intellect as unreasonable and indeed stupid); or finally that it implies community of origin with later divergence of development. When we are dealing with the smaller sub-divisions of some larger group, this method too gives us information of the same order of accuracy as does paleontology: but when we try to understand the relationships of these larger groups, then we are forced to renounce any claim to detailed knowledge. In broad outline, however, a great deal still remains, and this broad outline we can employ for our valuation of the whole sweep of biological progress, just as we can use the greater accuracy of vertebrate paleontology and comparative morphology to fill in the detail within a restricted field of its operation. From these various evidences, direct and indirect, we can paint for ourselves a picture of the evolution of life which, in spite of inevitable gaps and errors, is in its main features adequate and true. Let us not be misled by the fact that disputes can and justifiably do arise over details: as Professor Bateson put it recently[4]:-- “If the broad lines do not hold, then we must sink into irrationality or turn to flagrant supernaturalism.” Let us then remind ourselves of some of these broad lines.