In the case of the run of physical sciences this revision is less necessary; and for no very recondite reason. Every science by its very nature deals with a special, a limited topic. It is confined to a part or aspect of reality. Its propositions are not complete truths; they apply to an artificial world, to a part expressly cut off from the concrete reality. Its principles are generally cut according to their cloth,—according to the range in which they apply. The only danger that can well arise is if these categories are transplanted without due reservations, and made of universal application, i.e. if the scientist elects on his speciality to pronounce _de omnibus rebus_. But in the case of psychology and ethics the harmlessness of natural metaphysics will be less certain. Here a general human or universal interest is almost an inevitable coefficient: especially if they really rise to the full sweep of the subject. For as such they both seem to deal not with a part of reality, but with the very centre and purpose of all reality. In them we are not dealing with topics of secondary interest, but with the very heart of the human problem. Here the questions of reality and ideals, of unity and diversity, and of the evaluation of existence, come distinctly to the fore. If psychology is to answer the question, What am I? and ethics the question, What ought I to do? they can hardly work without some formulated creed of metaphysical character, without some preliminary criticisms of current first principles. (ii.) Herbart.