Gay Lussac likewise examined the precipitation of metals, in their metallic state, by other metals. The re sults of his examination confirm what Bergman and Richter had already advanced on this subject.
But the most detailed and extensive course of experi ments concerning chemical proportions has been performed by M. Berzelius of Stockholm. The works of Richter had engaged this chemist to repeat the experiments detailed in them, and correct their results. He endeavoured to obtain data sufficiently exact for enabling him to calculate, according to the rules established by Richter, the compo sition of most saline bodies. During this enterprise, the late analyses of the alkalis induced him to give greater compass to his researches, in order to determine the quantity of oxygen contained in those bodies, and chiefly in ammonia, by means of what Bergman, and still more what Richter had discovered concerning the precipitation of metals by each other. These experiments presented difficulties at the very outset. Assuming as a groundwork those analyses, which he had every reason to consider as the best, he found them rather to contradict than to con firm the laws already looked upon as proved. But having studied to observe and avoid every circumstance which might affect the result of an analysis, he obtained at last a certain number of analytical results, exact enough to cor respond with the laws under consideration. Of all the analyses performed before him, none but those of Wenzel were found to coincide with his own. The experiments of Dr. Wollaston, concerning the multiple proportions of Dalton, having been published in the Philosophical Trans actions,(1808,) M. Berzelius here entered upon a new field, deserving more minute examination. The first series of M. Berzelius' experiments was printed in a Su edish periodical work, conducted by himself, along with M. Hisinger, and entitled, ?lfhandlingar i Fysik, Kenai och ]t]ineralogi. The various memoirs which he afterwards published on this subject, are to be met with in the different journals of physics and chemistry ; such as Dr. Thomson's ?Innals of PhilosoPhy ; Mr. Nicholson's Journal ; Mr. Tilloch's Philosophical Magazine ; ?nizales de Chytnie ; the Journal de Physique ; nl. Gilbert's Annulen -der Physik ; Schweigger's Journal ; no one of which, however, contains the complete collection. It would be tedious in this place to exhibit the entire series of his experiments on chemical proportions; it is enough to mention, that in order to arrive at the results, of which NV C are immediately to present an exposition, he examined nearly all the oxyds having simple radicals ; many saline combinations, among the greater part of acids and bases, having an excess of base, or existing in a neutral state ; a multitude of salts, with two bases, as well as salts with two acids ; the chemi cal combinations of water with acids, bases and salts ; some combinations of metals with each other, and of metallic Dxyds with each other; he extended his researches to minerals, of which lie analyzed several himself, and ex amined the analyses made of others by the most celebrated chemists of our age ; and, finally, he likewise analyzed some vegetables. The experiments of M. 13 erzclius differ from
those of the chemists already mentioned, in not having been made to establish any preconceived hypothesis. On the contrary, they form a course of study regarding deter minate proportions, from which the laws have emerged by degrees, as the facts including them augmented the number. From his experiments, M. Berzelius concludes, that the laws, according to which elementary substances combine, may, in so far as concerns inorganic nature, be reduced to two principal rules. First, when a body A combines with a body B in several proportions, the numbers expressing those proportions are integer multiples of the smallest quantity of B that A can absorb; so that if this quantity of B were = 6, the other proportions must be some of the following : 24, 36, 48, &c. It is from this law that Mr. Dalton's hypothesis acquires so much probability. Second ly, when two oxydized bodies combine, the oxygen of the one is an integer multiple of that contained by the other; or if the number of oxydized bodies is greater, the oxygen of the body containing least is an integer submultiple of the oxygen found in any of the rest. In the sulphate of potass, for example, the oxygen of the sulphuric acid is three times that of the potass; and in crystallized alum the oxygen contained by the potass, which is the smallest quantity, is a submultiple by three of that contained by the alumina, by 12 of that contained by the sulphuric acid, and by 24 of that contained by the water of crystallization. This last rule, however, may be expressed in a more gene ral manner : Two compound bodies, the electronegative element of which is common, combine in such a ratio, that the quantity of that common element contained by the one is an integer multiple of the quantity contained by the other ; or when the number of compound bodies is greater, Ste. &c. It is conformable to this law, that in fossil combination of the different metallic sulphurets, the sulphur of the one is always a multiple of that of the other. It happens sometimes that compound bodies, of which the eleetto-positive element is common, may enter into com bination. Those combinations have been little cnamined; but it appears that the elect•o-negative elements divide the electropositive one among them, in some multiple pro portion. Such, for example, are mispickel, compounded of sulphuret and arseniate of iron, in the iron is equally divided between the sulphur and the arsenic ; the double salt having for its base the oxyd of lead, combined with nitric acid and phosphoric acid ; in which the latter occupies twice as much of the base as the former.