These remarks are founded upon distinctions which have no real existence. We cannot separate _things_ from their _properties_, nor can we consider their properties and analogies in any other way than by having _opinions_ about them. By contrasting _analogies_ with _opinions_, it might appear as if the author maintained that there were certain analogies about which there was no room for erroneous opinions. Yet the analogies of chemical compounds, are, in fact, those points which have been most the subject of difference of opinion, and on which the revolutions of theories have most changed men's views. As an example of analogies which are still recognized under alterations of theory, the writer gives the relation of a metal to its oxide or calciform state. But this analogy of metallic oxides, as Red Copper or Iron Ore, to Calx, or burnt lime, is very far from being self-evident;--so far indeed, that the recognition of the analogy was a great step in chemical _theory_. The terms which he quotes, _oxygenated muriatic acid_ (and the same may be said of _dephlogisticated marine acid_,) if improper, are so not because they involve theory, but because they involve false theory;--not because those who framed them did not endeavour to express analogies, but because they expressed analogies about which they were mistaken. Unconnected names, as _metals_, _earths_, _alkalis_, are good as the _basis_ of a systematic nomenclature, but they are not substitutes for such a nomenclature. A systematic nomenclature is an instrument of great utility and power, as the modern history of chemistry has shown. It would be highly unphilosophical to reject {298} the use of such an instrument, because, in the course of the revolutions of science, we may have to modify, or even to remodel it altogether. Its utility is not by that means destroyed. It has retained, transmitted, and enabled us to reason upon, the doctrines of the earlier theory, so far as they are true; and when this theory is absorbed into a more comprehensive one, (for this, and not its refutation, is the end of a theory _so far as_ it is true,) the nomenclature is easily translated into that which the new theory introduces. We have seen, in the history of astronomy, how valuable the theory of _epicycles_ was, in its time: the nomenclature of the relations of a planet's orbit, which that theory introduced, was one of Kepler's resources in discovering the _elliptical_ theory; and, though now superseded, is still readily intelligible to astronomers.