3. _The Uniformity of the Order of Nature_ may mean either "uniformity of co-existence" or "uniformity of succession." "Uniformity of co-existence" means that the same substances must always have the same essential properties[448] and the same permanent relations to other substances, as, for example, every molecule of hydrogen must have the same properties, the same definite mass, the same periodic vibrations, and the same chemical affinities. If these were to be altered in the least, it would no longer be a molecule of hydrogen.[449] This is uniformity in the ultimate _constitution_ of nature. "Uniformity of succession" means that the same or similar consequents will always be found to follow similar antecedents, or "the same causes will always be followed by the same effects,[450] as, for example, the combination of carbon and oxygen will always be followed by the evolution of heat, and heat will always melt ice." This is uniformity in the _course_ of nature or the procession of phenomena. Belief in the constancy of the course of nature or the uniformity of causation is the general expectation that "the future will resemble the past."[451] With a clearer apprehension of the terms, we may now discuss the first proposition with more precision, and hope to reach a logical conclusion. We approach the discussion by remarking-- 1. The constancy of the _course_ of nature or the uniformity of causation is not a self-evident and necessary truth. In so far as it is a scientific truth it is purely an induction from experience, an experience which is necessarily limited, and therefore does not warrant a universal conclusion. There is no rational _à priori_ ground for the assumption that the same or similar causes (even if we understand by physical causes all antecedent conditions) shall necessarily produce the same effects. In other words, there is no authority for the assertion that the course of nature or the procession of phenomena must be absolutely uniform. Science has succeeded in establishing a strong probability, but it is beyond her power to demonstrate an absolute certainty. This is generally conceded, alike by physicists and metaphysicians. J. S. Mill says, "The uniformity in the course of events ... must be received, not as a law of the universe, but of that portion of it which is within the range of our means of observation, with a reasonable degree of extension to adjacent cases."[452] "The uniformity of causation," says Murphy, "is not a truth of the reason, it is known by experience only; and the truth of a conclusion from experience can never be free from all possibility of limitation or exception."[453] And Professor Jevons asserts, "The conclusions of scientific inference appear to be always of a hypothetical and purely provisional nature. Given certain experience, the theory of probability yields us the true interpretation of that experience, and is the surest guide open to us.