(M193) But even if we grant for the sake of argument that this theory offers a ready explanation of the widespread belief in human immortality, it is less easy to see how the theory accounts for the corresponding belief of so many races in the immortality of the lower animals. In his dreams the savage recognises the images of his departed friends by those familiar traits of feature, voice, and gesture which characterised them in life. But can we suppose that he recognises dead beasts, birds, and fishes in like manner? That their images come before him in sleep with all the particular features, the minute individual differences, which distinguished them in life from their fellows, so that when he sees them he can say to himself, for example, “This is the very tiger that I speared yesterday; his carcase is dead, but his spirit must be still alive”; or, “That is the very salmon I caught and ate this morning; I certainly killed his body but clearly I have not succeeded in destroying his soul”? No doubt it is possible that the savage has arrived at his theory of animal immortality by some such process of reasoning, but the supposition seems at least more far-fetched and improbable than in the case of human immortality. And if we admit the insufficiency of the explanation in the one case, we seem bound to admit it, though perhaps in a less degree, in the other case also. In short, we conclude that the theory of dreams appears to be hardly enough by itself to account for the widespread belief in the immortality of men and animals; dreams have probably done much to confirm that belief, but would they suffice to originate it? We may reasonably doubt it.