which I estimated by the formation of muriate of silver, founding my calculations on the analysis of that salt given by Proust.* 7.5 cubic inches, or 5 4375 grs. of the gas were agitated with distilled water, until it had absorbed the whole. Nitrate of silver was then added, and, after being several times agitated, the phial was set aside in a dark place for 24 hours. The mixture of muriate of silver and acid liquor was then thrown on a filtre, previously dried with care, and weighed, and the residuum was washed with distilled water, until it passed through tasteless. The filtre with the muriate of silver was next wrapped up, and laid in a dark place, having a temperature of about 70°, until it appeared perfectly dry. It was then found to have gained between 22.5 and 23 grains.

According to the old theory, 5:4375 grs. of oxymuriatic gas should be resolved into 1.323467 gr. of oxygen, and 4·114033 grs. of muriatic acid, which can combine with 19-28073 grs. of oxide of silver, forming 23.394763 grs. of muriate.

In consonance with Sir H. Davy's views, the chlorine and the silver combine, while the oxygen previously in union with the silver is set at liberty. Now 5.4375 grs. of chlorine are capable of forming, by union with hydrogen, 5'64645 grs. of muriatic acid, which can in their turn produce 31.37917 grs. of horn silver; therefore 5.4375 grs. of chlorine are capable of forming 31-37917 grs. of horn silver.

To this experiment it may be objected, that, if Sir H. Davy's theory be correct, the analysis of muriate of silver assumed must be incorrect in regard to proportions. This objection it will be necessary to examine somewhat minutely. The proportions of acid assigned to the salt by the analyses of Gay-Lussac and Thenard, Berzelius, Rose, Proust, Zaboada, Chenevix, and Kirwan, § do not differ from each other more than between 19-44 and 16.54. Of these I have in this paper preferred that of Proust, which is very nearly the mean of them. If, then, 100 grs. of muriate of silver yield, on analysis, 18 grs. of muriatic acid and 82 grs, of oxide of silver, it seems probable that it contains 17.323 grs. of chlorine, united to 756 grs. of silver. But this supposition leaves a deficiency of 7.077 grs.; and to fill up this we must suppose the salt to contain a proportion of water, which, being decomposed during its analysis, yields its oxygen to the silver, and its hydrogen to the chlorine. Now 75.6 grs. of silver take 6-4 grs. of oxygen, which can combine with 1.07 gr. of hydrogen, while 17.323 of chlorine cannot

Journal de Phys. xlix. p. 221.

+ Rech. Phys. Chem, ii. 123. Ann. de Chim. lxxix. 133; lxxviii. 114; 1xxvii. 84. See Dr. Thomson's System of Chemistry, iii. 153,

unite with more than 677 of a grain to form muriatic acid; so that an evolution of hydrogen must take place to the extent of 393 of a gr. or 15.04 cubic inches. Should it be preferred to suppose horn silver composed solely of chlorine and silver, it must yield, on analysis, 20-02 of muriatic acid, and 86-648 of oxide of silver

Mr. J. Davy has, in his paper on the combinations of chlorine, stated another analysis of horn silver, which, he observes, agrees very nearly with that given by Klaproth, viz. 24.5 of chlorine and 75.5 of silver. On this basis, the result of the analysis should be 25.4224 of muriatic acid, and 82·4 of oxide of silver. It must, however, be allowed, that if this difficulty, and some others that will be stated immediately as attaching to Mr. J. Davy's analysis, can be got over, it will answer very nearly to the result of the above detailed experiment; for 5.4375 grs. of chlorine ought, on this estimate, to form 22.194 grs. of horn silver.

• If oxide of silver, at the precise degree of oxidation in which it has been stated to be found in muriate of silver (viz. containing about 7.8049 per cent. of oxygen) be added to common muriatic acid, there ought to ensue, on Sir H. Davy's theory (if Proust's analysis be correct), an evolution of oxygen to the amount of 7.188 cubic inches for each 100 grs. of the salt formed, because the hydrogen previously united to the chlorine, with which 75.6 grs. of silver can combine, is incapable of entering into combination with more than 3.95598 grs. of the 6'4 grs. of oxygen which that quantity of silver holds in union with it. Or, adopting Mr. J. Davy's analysis of horn silver, the evolution of oxygen should amount to 2.97 cubic inches; because 75.5 grs, of silver hold in combination 6.5 grs. of oxygen, of which only 5.51 grs. can enter into union with the hydrogen (9224 of a gr.) set free from its combination with the chlorine.

Again, in the process of separating chlorine from silver, there ought to be an evolution of a little more than 15 cubic inches of hydrogen for every 100 grs. because the silver requires 1.44402 gr. of oxygen more than can be furnished by the water sufficient to afford 677 of a gr. of hydrogen to the chlorine. But, by Mr. J. Davy's analysis, this evolution of hydrogen ought not to exceed 6.99 cubic inches, for 24.5 grs. of chlorine take 9224 of a gr. of hydrogen, and this quantity is capable of combining with 5.51 grs. of the 6.5 grs. of oxygen required by the silver.

The former part of the reasoning employed to obviate the probable objection against the experiment which I have detailed, would be in some measure incomplete, were I to overlook Sir H. Davy's theory of the nature of the muriates. This I shall notice as briefly as possible.

Taking, then, the instance of muriate of potash, it appears

by calculation, that while the oxygen of 100 of potash unites with the hydrogen of 52.4748 of muriatic acid, the residual potassium (861) combining with the chlorine (50-16) forms 136.26 of muriate of potash. But the materials employed were weight 152-4748; so that a deficiency in the weight of the product, of not less than 16 2148, remains to be accounted for. It may, indeed, be said, that on account of the water contained in potash, a deduction is requisite on either theory; but as this operates equally on both, while the other is entirely unnecessary on the old theory, we are thus furnished with an apparently easy mode of discrimination.

Nor does the analysis of the salt seem less decisive: according to Sir H. Davy's theory, 100 parts of muriate of potash should yield 35.976 of muriatic acid, and 71-93 of anhydrous potash; in all, 107 906.

Moreover, in decomposing muriate of potash by nitrate of silver, the potassium must be oxidated by decomposing either the water or the oxide of silver. If by the former means, there must be an evolution of 88.6 cubic inches of hydrogen from the decomposition of 136.26 grs. of the salt. If in the latter way, then, as 50 16 grs. of chlorine can decompose 239-2741 grs. of oxide of silver, the oxygen of this quantity, which amounts to nearly 22.085 grs., must unite with the potassium; but 81-6 grs. of potassium can take 13.9 grs. of oxygen only: therefore there must be an extrication of 8.185 grs. of this gas, or about 24 0735 cubic inches. Beside, the residual muriate of silver will not weigh 291-7489 grs., but 267-3491 grs. only.

Before concluding, I have only to remark, that the calculations above detailed are rendered in a great measure uncertain, from the want of accurate and consistent statements of the composition of potash, and of water, as well as of the quantity of oxygen which may be obtained from a given portion of oxymuriatic gas.

With regard to potash, the various statements given by different chemists, and on different occasions, are full of perplexity.

Sir H. Davy details two experiments, the mean result of which was 13.9 per cent. of oxygen; * and on the basis afforded by this estimate, the calculations in the foregoing paper were made the method practised in these experiments was the combustion of potassium in oxygen gas, the result of which process is afterwards† said by Sir H. Davy to be peroxide of potassium.

* Phil. Trans. 1808, p. 28.

+ Phil. Trans. 1811, p. 6.

Yet the experiments detailed in page 4 of the Phil. Trans. for 1811 seem to indicate that the peroxide of potassium contains about 30 per cent. of oxygen; and MM. Gay-Lussac and Thenard (Recherches, &c. i. p. 132) state that it contains three times as much oxygen as exists in potash.

By the evolution of hydrogen from water by potassium, the proportion is fixed at 16.* At p. 34 of the Phil. Trans. for 1810, is a reference to the Bakerian lecture for 1807, as warranting an estimate of 15.254. On one occasion† calculations are made from a proportion of 14.895, and on another ‡ 15.6 is assumed. MM. Gay-Lussac and Thenard give 16-629 as the result of their experiments; § aud M. Berzelius states at first || 17.007, but afterwards** either 16.978 or 17.153. It may perhaps be objected to the accuracy of my calculations, that I have founded them on a proportion authorized by the mean result of only two experiments, and those the first that had been performed on the subject. To this I would reply, that the detail given of these experiments appeared to me sufficiently minute and satisfactory; and that the other results have all of them been obtained by calculations founded on a view of the constitution of water, with which I must confess myself not perfectly satisfied.

The statement of the composition of water given by Dr. Thomson, tt on the authority of Fourcroy, Vauquelin, and Sequin, has been adopted in the preceding calculations: this proportion is 85-662 of oxygen to 14.338 of hydrogen. GayLussac and Humboldt give 87.4 and 126;11 and in the Recherches Physico-Chimiques, the calculation of the constituents of turpentine §§ is made from the proportion of 87-998152 and 12:001847, while in the analysis of oxalic acid |||| the proportions asssumed are 88.367260 and 11·632739. The proportions assigned by Berzelius are 88.246 and 11.754,*** or 87-587 to 12-413. ttt

Mr. Murray 11 gives 16 of oxygen and 84 of muriatic acid as the constituents of oxymuriatic gas: Dr. Thomson §§§ details an experiment by Chenevix, apparently very accurate, the results of which were 22.5 of oxygen, and 77.5 of muriatic acid; and to this estimate I have adhered in the foregoing paper. Mr. Dalton states 24 of oxygen, and 76 of muriatic acid; and Berzelius **** assigns 23.37 and 76.63, but afterwards †††† states 22.768 and 77-232.

Phil. Trans. 1808, p. 30. Phil. Trans. 1810, p. 245. Ann. de Chim, lxxvii. p. 84. ++ System, &c. i. p. 123. §§ Ann. de Chim, ii, p. 313. *** Ann. de Chim. Ixxvii. p. 84; +++ Ann. de Chim. lxxix. p. 246. Syst. of Chem. ii. p. 257. **** Ann, de Chim. Ixxvii, p. 84.

ARTICLE XI.

Memoir on the Determination of the Specific Heat of the different Gases. By MM. F. Delaroche, M.D. and J. E. Berard.*

THE subject proposed by the Institute as a prize at the meeting of the 7th January, 1811; namely, the determination of the specific heat of the gases; had previously attracted the attention of different philosophers, some of whom have treated of it in detail: yet so little progress has been made in the investigation, though the question proposed be sufficiently simple, that we are almost as far from being able to answer it with precision as we were before it became the subject of investigation.

Crawford, as far as we know, is the first person who began the investigation. He published the result of his researches in 1788. At that time a great number of experiments had been made upon the specific heat of bodies in general. MM. Lavoisier and de Laplace had already published the result of their experiments on that subject, and had given a description of their calorimeter of ice yet Dr. Crawford preferred the method of mixtures of water, or other bodies whose specific heat was considered as known. After many unsuccessful attempts, which it would be too tedious to describe here, he thought that he succeeded by the following method. He procured two copper vessels, very thin, and of the same shape, size, and weight. He filled one of them with the gas that he wished to examine, and made a vacuum in the other. He then heated both in boiling water, and plunged both suddenly into cylinders containing a small quantity of cold water, but sufficient to cover them. He subtracted the heat communicated to this water by the exhausted vessel from that communicated by the vessel full of gas, and considered the remainder as the effect produced by the gas, or as its specific heat. He had taken great precautions to ensure the accuracy of his results; but it is quite evident, from the smallness of the change, that no confidence could be placed in them. The rise in the temperature of the water never exceeded 0.4° Fahrenheit. The following is the table of his results :—

* Translated from the Annales de Chimie for January, 1813, vol. Ixxxv. p. 72. This memoir gained the prize proposed by the Institute, and deserves particular attention. It overturns the theory of animal heat advanced by Crawford, and Lavoisier's theory of combustion.