the Greeks called θεοἱ or gods, but it meant something common to all these θεοἱ, and thus could naturally rise to express what the Greeks wanted to express by that word. There was as yet no necessity for defining deva or θεὁς, when applied to what was meant by gods, but of course the most opposite meanings had clustered round it. While a philosophical Greek would maintain that θεὁς meant what was one and never many, a poetical Greek or an ordinary Greek would hold that it meant what was by nature many. But while in such a case philosophical analysis and historical genealogy would support each other, there are ever so many cases where etymological analysis is as hopeless as logical analysis. Who is to define _romantic_, in such expressions as romantic literature. Etymologically we know that romantic goes back finally to Rome, but the mass of incongruous meanings that have been thrown at random into the caldron of that word, is so great that no definition could be contrived to comprehend them all. And how should we define _Gothic_ or _Romanic_ architecture, remembering that as no Goths had anything to do with pointed arches, neither were any Romans responsible for the flat roofs of the German churches of the Saxon emperors. Enough to show what I meant when I said that Professor Drobisch, in his Lectures on Herbart, gave one great encouragement in the special work in which I was already engaged as a mere student, the Science of Language and Etymology. If Herbart declared philosophy to consist in a thorough examination (_Bearbeitung_) of concepts, or conceptual knowledge, my answer was, Only let it be historical, nay, in the beginning, etymological; I was not so foolish as to imagine that a word as used at present, meant what it meant etymologically. _Deus_ no longer meant brilliant, but it should be the object of the true historian of language to prove how _Deus_, having meant originally brilliant, came to mean what it means now.