One point however remains on which I shall speak with some confidence, that, namely, of the inclusion among 'Comte's titles to immortal fame' of the creation of a Science of Sociology. 'What the law of gravitation is to astronomy, what the elementary properties of tissues are to physiology,' that, says Mr. Lewes, in the opinion of Comte's disciples, 'is the law of the three stages to sociology.' But if, as I have shown, there are not really three but only two stages, the so-called third stage being simply a return to either the second or the first, the law of the three stages cannot be much of a law, nor the science of which it is the essence much of a science. Mr. Lewes, nevertheless, maintains that M. Comte created Social Science. Mr. Mill considers that he did not create it, but only proved its creation to be possible. With all possible deference, I submit that what he really did was to prove its creation to be impossible. In a passage of Mr. Mill's 'Positivism,' quoted with approval in Mr. Lewes's 'History of Philosophy,' and presumably, therefore, expressing the sentiments of both writers, Comte is described as pronouncing inappropriate to the Science of Society, the method universally admitted to be proper to all other sciences--that, namely, of obtaining by induction the laws of the elementary phenomena, then, from these laws thinking out deductively those of the complex phenomena, and, finally, of verifying by specific observation the laws obtained by deduction. Among social phenomena, he is described as arguing, the elementary ones are human feelings and actions, the laws of which are the laws of universal human nature. But the human beings, on the laws of whose nature social facts depend, are not abstract or universal, but historical human beings, already shaped and made what they are, not by the simple tendencies of universal human nature, but by the accumulated influence of past generations of human society. This being the case, the laws of universal human nature evidently cannot serve as materials, whence it would be possible for any powers of deduction, starting from the bare conception of the Being Man, to predict beforehand how successive generations of men would feel and act. Wherefore, in order to get at social laws, we must reverse the ordinary method, seizing upon any generalizations which the facts of history, empirically considered, will supply, and then using the universal laws of human nature for the verification of these generalizations.