than half that of chlorine, and so little exceeding that of oxygen,
that those who would suppose it to contain oxygen combined with
an inflammable base, must suppose the base to be loss than one
twentieth part of the oxygen with which it combines.
The design of the Scale here proposed by the author is to save chemists the labour of many troublesome computations in estimating the ingredients of neutral salts, and the reagents and precipitates by which these ingredients might be ascertained.
For though certain laws to which chemical union is subjected have of late been discovered, and have enabled chemists to determine with greater precision than formerly the composition of bodies submitted to be examined, and to express numerically the relation of the several elementary chemical substances to each other; nevertheless the computations requisite for applying these results to many objects of inquiry are frequently attended with considerable trouble.
The author briefly sketches the history of proportional chemistry, beginning with Berg-man, who, perceiving that the same acid united to the same base, always in the same proportion, took pains to ascertain the composition of various salts. Kirwan followed the same line of endeavour to a greater extent, with a view to determine the proportions of various acids to different bases, as questions independent of each other. To these succeeded Richter, who gave connection to the subject by observing a new relation that had escaped the notice of Bergman, Kirwan, or any of his predecessors. They had observed only the constancy of the proportion of the same acid to the same base ; Richter observed, further, a fixed relation of acid to acid: namely, that when the proportional quantities of any two acids, that are each sufficient to saturate a given quantity of any one base is determined, the same proportional weights of these acids will also saturate equal quantities of any other base; and consequently that if any quantity of sulphuric acid be assumed as standard, then equivalent quantities of all other acids may be conveniently expressed by fixed numbers, adapted to each; and the several quantities of different alkalies and earths that would each saturate the standard quantity, might also be represented constantly by corresponding numbers.
The observation of other proportions, which are simple multiples of the preceding, by Mr. Dalton and others, are noticed as affording an important correction of the best analyses; but it is observed that
