third degree of oxidation of the same radical. It was left for Davy to demonstrate the elementary nature of chlorine and to discover the true relations of the hydrogen acids. Lavoisier regarded oxygen as the universal acidifying principle, and the facts known in his day admitted of this interpretation; and it is interesting to see how they were worked up in the table; but when a class of acids containing no oxygen came to be clearly recognized, they proved a serious embarrassment to the Lavoisierian system as it was developed by Berzelius and his associates.
With the exception of caloric and two of the radicals above referred to, Lavoisier's list of elements includes no substance not regarded as elementary at the present day; but the list is as remarkable for what it omits as for what it includes. There were then known, and had been known for a long time, two very well marked classes of bodies called alkalies and earths which readily combined with acids to form salts. In this respect these bodies closely resembled the known metallic oxides, as they did also in most cases in their general appearance, and they were classed by Lavoisier with the oxides under the general term of "bases salifiables." Still, they had never been decomposed, and, according to the spirit of Lavoisier's philosophy, ought to have been classed among elementary substances; but Lavoisier's classificatory instinct was altogether too acute to permit him to fall into any such error. He enumerates these bodies, and, although he speaks doubtfully in regard to them, he never for a moment questions their compound nature. In regard to the earths he says their composition is wholly unknown, implying, of course, that they were compounds, and under the head of "Des Substances métalliques" is this significant paragraph:
"It is probable that we only know a part of the metallic substances which exist in nature. All those, for example, which have more affinity for oxygen than for carbon can not be reduced or brought to a metallic state, and must appear to us as oxides which we mistake for earths. It is very probable that baryta, which we have classed as an earth, is a case in point. When experimented upon, it exhibits characters which closely approach those of metallic substances. It may be, indeed, that all the substances to which we give the name of earths are only metallic oxides that can not be reduced by the means which we use."
It will be noticed that the alkalies are not included under this remark, for their active qualities are very different from those of an insipid, earthy-looking, metallic oxide; and their resemblance to ammonia, the volatile alkali, a known compound of nitrogen, was constantly a confusing circumstance. Lavoisier discusses the question whether potash and soda pre-exist as such in the plants from whose ashes they are procured, and makes the suggestion