It has lately been shown, by detailed and exact experi ments, that Lavoisicr was not wrong in considering these phenomena as different. It is known, for example, that pure lime and magnesia combine with water, for which they have a great affinity; but the one is very sparingly soluble in water, whilst the other is not soluble at all. Several bodies which contain great quantities of water in chemical combination, the metallic hydrates for example, the carbonates of various earths, Ste. are not at all soluble in water ; whilst other bodies which do not combine chemi cally with water are abundantly soluble in it ; such as salt petre, and culinary salt. When a substance combines chemically with water, it disengages caloric ; when merely dissolved in water, its caloric is absorbed, and the tem perature falls. A body possessing the property of com bining with water, and at the same time soluble in it, first evolves heat, and afterwards cold. These entirely opposite phenomena must arise from a different cause ; and conse quently there must he some circumstance in solution which does not find place in chemical combination, or inversely.
M. Berthollet's opinions were first examined by M. Proust. He proved, from a series of ingenious experi ments, that the new theory, when applied to any thing but solutions, and mixtures in a liquid state, led to consider able errors. It admitted, for example, that metals, in com bining with oxygen, exhibit an infinite number of de grees of oxydation, between their minimum and maximum. Proust chose antimony and iron to show that this is not the case; and that, on the contrary, those metals do not com bine with oxygen in more than two proportions; but that, if their minimum oxyd is exposed to the action of oxygen, a portion being converted into the maximum oxyd, re mains mechanically mixed with the former. From which it appears, said Proust, that Berthollet has deceived him self by considering mechanical mixtures of the two oxyds, as particular degrees of oxydation. Proust extended his experiments to the metallic sulphurets, and found that a similar principle may be applied to them likewise ; so that each new addition of oxygen or sulphur takes place per saltum without any intermediate stage. Berthollet de fended himself with so much sagacity, that the impartial reader, though in the secret and perhaps obscure senti ment derived from a general survey of chemistry, he might decide for Proust, yet felt himself compelled to suspend his final judgment. Proust however enjoyed at last the triumph of observing that, in proportion as this depart ment of science began to receive a more sedulous cultiva tion, as doubtful points were cleared up, the results ob tained by impartial chemists confirmed his statement.
Some times before the works of Messrs. Richter and Berthollet appeared, Mr. Higgins, professor of chemistry at Dublin, in a publication named, .4 Comparative View of the Phlogistic and .Inti/lhlogtstic Theories (17890 started the idea of explaining the different degrees in which cer tain bodies combine—azotc and sulphur, for example, with oxygen—by regarding them as united for each new degree of oxydation to an additional particle of oxygen ; so that nitrous gas should contain two particles of oxygen, nitrous acid three, and nitric acid four particles of azote But as Mr. Higgins attached little importance to this happy
idea, and applied it very sparingly, it excited no attention whatever among chemists.
About fifteen years later, Mr. Dalton, another English chemist, struck out the same idea ; tried it by a compa rison with the best analyses then known ; and finding suf ficient reasons to consider it as just, made it the basis of a chemical system, the details of which he published in a work, entitled, A New System of Chemical Philosophy, (1808.) According to this system, elementary substances combine together, in such a manner, that an atom or in destructible particle of the one always unites itself to 1, 2, 3, 4, Lc. whole atoms of the other ; each new addition taking place in a multiple ratio. Numerous experiments afterwards confirmed this hypothesis; none has contradicted it ; and without exaggeration, it may be marked as one of the greatest improvements which chemistry has ever re ceived. In his new system, Mr. Dalton supposes that elementary molecules unite most readily one to one ; and hence, if but a single proportion is known in which two bodies unite, it must be considered as that of one atom to one atom. If several are known, the minimum of one element is considered as the proportion of one to one ; the second as that of one to two, and so on. But when it hap pens that in the second combination the element added is multiplied only by 14 ; the compound is looked upon as formed by two atoms of the one and three of the other. In the second volume of the work just quoted, (printed in 1810,) Mr. Dalton has examined the simple combustibles with their oxyds, and given the number of atoms, which, in his opinion, those oxyds must contain. To the method of Mr. Dalton, it may be objected that his principle is hypothetical ; that he has shown little scruple in the ap plication of it ; • that his analytical experiments are not always very exact; and that a desire, on the part of the operator, to obtain a preconceived result, seems often to have influenced the actual result ; a circumstance which, in these researches, it is impossible too carefully to guard against, and which often misleads, above all, when the system is framed first, and the proofs sought afterwards. Mr. Dalton's new system is, moreover, mingled, like the work of M. Richter, with results less solidly founded ; to which, no less than his predecessor, he has endeavoured to add probability by giving them a mathematical colour. Mr. Dalton was even of opinion that he had found the mathematical laws, according to which gasiform substances are absorbed and retained by liquids. He speaks of having determined by experiments, which are very exact, and, to judge from his own account of them, very conclusive, that gases, excepting such as disengage a great quantity of caloric in combining with water, are absorbed by that sub stance, either in a volume equal to that of the water, or else in a volume equal to 4, or 31/4 of that of the water. These numbers being the cubes of i, 1, z, he infers that the distance between the molecules of gasiform sub stances, absorbed by water, is always some multiple of their distance when out of water. Every liquid absorbs the same quantity of gas as water ; with this single dif ference, that viscous liquids require more time for satura tion.