produce no consequence. But, though their succession is not really causal, it must none the less appear so, because it is regular. And it must be regular, since it depends on a series which is unalterably fixed by causation. And in this way (it may be urged) the alleged inconsistency is avoided, and all is made harmonious. We are not forced into the conclusion that the self-same cause can produce two different effects. A is not first followed by mere B, and then again by B—β, since α is, in fact, irremovable from A. Nor is it necessary to suppose that the sequence A—B must ever occur by itself. For a will, in fact, accompany A, and β will always occur with B. Still this inseparability will in no way affect our result, which is the outcome and expression of a general principle. A—B—C is the actual and sole thread of causation, while α, β, γ are the adjectives which idly adorn it. And hence these latter must seem to be that which really they are not. They are in fact decorative, but either always or usually so as to appear constructional.
This is the best statement that I can make in defence of my unwilling clients, and I have now to show that this statement will not bear criticism. But there is one point on which I, probably, have exceeded my instructions. To admit that the sequence A—B—C does not exist by itself, would seem contrary to that view which is more generally held. Yet, without this admission, the inconsistency can be exhibited more easily.
The Law of Causation is the principle of Identity, applied to the successive. Make a statement involving succession, and you have necessarily made a statement which, if true, is true always. Now, if it is true universally that B follows A, then that sequence is what we mean by a causal law. If, on the other hand, the sequence is not universally true,
