The true nature of a scientific law has never been better explained than by the writer already quoted as Mr. Buckle's dexterous apologist. A scientific law is not an ordinance, but a record. It simply professes to describe the order in which certain phenomena have been observed uniformly to recur. It differs from a legislative enactment, in that the one would be a law although it were never obeyed, whereas the other would cease to be a law if one single exception to its statement could be pointed out. Thus the Act of Parliament enjoining the registration of births, would be equally a law although no births were ever registered; whereas the law, that in a body moving in consequence of pressure the momentum generated is in proportion to the pressure, would entirely forfeit its legal character if, on any one occasion or in any circumstances, momentum were generated in any other proportion. It is essential, then, to the existence of a scientific law that there should be uniformity of phenomena. But in human affairs uniformity is impossible. No doubt, in exactly the same circumstances exactly the same events must happen; but exactly the same aggregation of circumstances cannot possibly be repeated. Such repetition is inconsistent with the very theory, which is based on the assumption that the repetition is continually happening. 'In the moral as well as the physical world' there are, say the exponents of the new theory, not only 'invariable rule' and 'inevitable sequence,' but 'irresistible growth' and 'continual advance.' In other words, things can never be twice in precisely the same condition--never, at least, within the same cycle. It has, indeed, been suggested that there may be in human affairs the same sort of regularity as is observed by the hands of a clock; and that, as the latter, at the end of every twenty-four hours, recommence the movement which they have just concluded, so at the end of, say 'every ten thousand years,' all the same events which have been happening throughout the period may begin to happen over again in the same order as before. Such a succession, however, would involve quite as much of retrogression as of progression, and the continual advance so boastfully spoken of would be nothing else than a tendency of society to return to the condition from which it had originally emerged. But, even on this uncomfortable hypothesis, there could be no regularity of occurrences within the same cycle; no clue as to the future could be obtained from investigation of the past. On the contrary, the only certainty would then, as now, be that no combination of events which had happened once could happen again, as long as the existing order of things continued. The inference here follows necessarily from the premises. If there be continual advance--if things are constantly moving forward--they cannot remain in the same state; and if not in the same state, they cannot produce the same effects. For, if