depend upon an action’s total consequences and yet that this principle is untrue. It is also sometimes expressed by saying that if an action is once right, any precisely similar action, done in circumstances which are also precisely similar in all respects, must be right too. But this is both too narrow and too wide. It is too narrow, because our principle does not confine itself to an assertion about precisely similar actions. Our principle asserts that any action Y, whose effects are precisely similar to those of another X, will be right, if X is right, provided the effects of all the alternatives possible in the two cases are also precisely similar, even though Y itself is not precisely similar to X, but utterly different from it. And it is too wide, because it does not follow from the fact that two actions are both precisely similar in themselves and also done in precisely similar circumstances, that their effects must also be precisely similar. This does, of course, follow, so long as the laws of nature remain the same. But if we suppose the laws of nature to change, or if we conceive a Universe in which different laws of nature
