In treating of the ultimate foundations of Inductive Certainty (Logic of Deduction, p. 273), I laid it down as essential that we should postulate or beg the Uniformity of Nature; maintaining that we could give no reason for the future resembling the past, but must simply risk it. Observation can prove that what has been, has been; but it cannot prove that what has been will be. When we run the risk and find, after the thing has happened, that our anticipation is correct, we feel re-assured, and think less and less of the danger of being found wrong; but this hardening operation does not make a logical proof.(11 § 2 ¶ 1)

Against this view of the postulate of Uniformity, Mr. Lewes brings the view, that to say Nature is uniform, is an identical proposition; there is no hazard in it at all (Problems of Life and Mind, ii. p. 99). Now, to oppose an identical proposition is to bring about a contradiction in terms. Yet, at first sight, there seems no such contradiction, in saying that Nature follows one course to-day and another to-morrow; does one thing in London and another in Pekin. I should call Nature inconsistent with herself, in that loose sense of consistency that we apply to human actions; but I do not see any self-contradiction in saying that, a million of years hence, the boiling point of water at the ordinary pressure of the air is to be raised to 250° Fahrenheit.(11 § 2 ¶ 2)

According to Mr. Lewes, the true expression of Nature's uniformity is: the assertion of identity under identical conditions; whatever is, is and will be, so long as the conditions are unchanged; and this is not an assumption, but an identical proposition. But now as to the conditions, in what light does Mr. Lewes view Time and Place? Are these among the conditions, or are they not? If these are conditions, I fully grant the identity; because the assertion the nis that what is happening here and now, is happening; and nothing else is happening. But is he prepared to set aside time and place as not being conditions, as not needing to be taken account of at all? If he does, he gets the advantage of being able to affirm the Uniformity of Nature in the full extent required as a basis of Induction; but I deny that he affirms an identical proposition. It seems to me that to pass the bounds of time and place, is a hazard; and this is the real point at issue. I can only repeat that, as it seems to me, there is no self-contradiction in supposing that, though the physical conditions of an effect remain as they are, the effect may not be constant through all the eternity of years and all the infinitude of space. For this reason I call the Uniformity of Nature a postulate or an assumption, and refuse to call it an identical truth.(11 § 2 ¶ 3)

A. Bain