In Chemistry we regard the antecedents (elements) as still existing in and constituting the consequent (compound) because the latter is exactly similar to the former in weight, and because we can generally cause this compound to disappear and obtain the elements in its place. But we find nothing at all like this in the growth of mental phenomena: the psychical consequent is in no respect exactly similar to its antecedents, nor can it be resolved into them. I should explain that I am not here arguing the question whether the validity of moral judgments is affected by a discovery of their psychical antecedents. This question I reserve for subsequent discussion. See Book iii. chap. i. §4.