predicates, but also that we should be able to analyse these even if they are abstractions, and so on until we reach knowledge of _perception_, and thus refer to concrete things through the distinct perception of which the final abstractions are verified and reality guaranteed to them, as well as to all the higher abstractions which rest upon them. Therefore the ordinary explanation that the conception is distinct as soon as we can give its predicates is not sufficient. For the separating of these predicates may lead perhaps to more conceptions; and so on again without there being that ultimate basis of perceptions which imparts reality to all those conceptions. Take, for example, the conception “spirit,” and analyse it into its predicates: “A thinking, willing, immaterial, simple, indestructible being that does not occupy space.” Nothing is yet distinctly thought about it, because the elements of these conceptions cannot be verified by means of perceptions, for a thinking being without a brain is like a digesting being without a stomach. Only perceptions are, properly speaking, _clear_, not conceptions; these at the most can only be distinct. Hence also, absurd as it was, “clear and confused” were coupled together and used as synonymous when knowledge of perception was explained as merely a confused abstract knowledge, because the latter kind of knowledge alone was distinct. This was first done by Duns Scotus, but Leibnitz has substantially the same view, upon which his “_Identitas Indiscernibilium_” depends. (See Kant’s refutation of this, p. 275 of the first edition of the Critique of Pure Reason.)