tween the coasts of India and the Isle of France. He returned from that colony in 1772 with M. Poivre, that administrator whose wisdom and talents have left in his jurisdiction so high a reputation. "M. Rochon brought from that expedition the most beautiful crystals of quartz from Madagascar that had been at that time seen. He got some pieces of them cut, ascertained the double refraction which it possesses, and conceived the happy idea of applying it to the measurement of angles. Such is the origin of the ingenious micrometer, for the invention of which we are indebted to him. "Nobody knew better than our associate the wants of the province in which he had been born, and what was necessary to increase its prosperity; but the harbour of Brest fixed his constant predilection. Government approved of the plan which he proposed of cutting across Britanny a navigable canal between Brest and Nantes, which would in time of war serve to convey provisions without any risk to the first of our naval arsenals. The memoirs of M. Rochon on this important subject have the rare merit of pointing out at once the advantages, the difficulties to be overcome, and the means of surmounting them. "M. Rochon fully enjoyed during the whole of his life that reputation which his labours had justly acquired for him. He knew equally well how to make science useful in the society of men of the world with whom he was associated, and to render its application easy in the workshops of most of those arts with the processes of which he was familiar. It was by the utility of discoveries that he estimated their importance; and when a few days ago we heard him for the last time, at one of our meetings, it was to offer to the Academy the tribute of a useful investigation. "He was then in his 77th year. His strong constitution, though he had become a good deal weaker for some months past, left us the hope of preserving him, even when we heard that he was attacked by the disease under which he sunk. “After he had reached a mature age, M. Rochon had united himself with a widow lady, a relation of his own, and mother of two children. This union was during 25 years the source of mutual happiness, which was destroyed for ever by the fatal event which has collected us together-an event aggravated for his family by a deplorable circumstance. His respectable widow was obliged to divide her attention between her husband and her daughter, who were both seized at the same time with a fatal disease. Her care of both was useless, her vows were unavailing. The same instant deprived her of two objects, both most dear to her affections, and left her plunged in the deepest sorrow which virtue is capable of supporting." ARTICLE II. Appendix to the Essay on the Chemical Compounds of Axote and Oxygen. By John Dalton. (Concluded from p. 47.) CLASS III.-Experiments over Mercury with Caustic Alkalies. GAY-LUSSAC having recently stated that mixtures of nitrous and Oxygen gases over mercury to which caustic alkali was admitted exhibited always the same proportions of oxygen and nitrous gas; namely, one measure of oxygen uniting to four of nitrous gas; I was desirous to try if I could succeed in producing the union in like circumstances; for which purpose I made the following expe riments : 1. To 133 measures of nitrous gas of 97 per cent. = 129 real, 4 azote, put 32 measures caustic soda of 1·11 sp. gr. No diminution in two hours. Put 16 of 72 per cent. oxygen = 1,1·5 real, 4·5 azote. 149 102 in a few minutes, and remaining so for more than one hour. Put 16 more oxygen of same kind. 118 79 in a few minutes. 78 in 8 or 10 hours. 75 in 1 75 in 2 75 transferred, 63 nitrous by sulphate of iron. This gives 1 oxygen to 2.91 nitrous gas. 2. To 240 of 97 nitrous gas put 32 caustic potash of 1.45 sp. gr. It stood 12 hours without any change. Put 60 of 78 oxygen = 47 real, 13 azote. 300 155 in a few minutes. 152 soon after. 141 in 12 hours. 133 in 24 133 in 48 130 transferred over water, 109 nitrous. This gives 1 oxygen to 2.55 nitrous. The loss of 3 nitrous was occasioned chiefly by passing through the water as usual. It is obvious that these experiments are far from according with those of Gay-Lussac; and as he has not given us the detail of his, I cannot suggest the cause of the difference. My reason for subjecting the nitrous gas in the first place to the action of the alkalies was to show that these do not act on nitrous gas alone; and I had some ground for this; for in my first trials I had obtained a much greater reduction of the nitrous gas; but upon examination I found my potash contained a little sulphureted hydrogen, and this converted a part of the nitrous gas into nitrous oxide, and in this way reduced is volume. This was proved by admitting nitrous gas alone to the potash, when it was gradually reduced in volume, as if oxygen had been present; but when the residue was examined, it diminished rapidly, by passing a few times through water, and then left a residue of nitrous gas. CLASS IV.-Experiments on the Analyses of Nitrous Gas, Nitrous Oxide, and Ammonia, by exploding their Mixtures over Mercury. Proust, I believe, was the first person who pointed out the analysis of ammonia, by exploding it with oxygen in Volta's eudiometer. (Jour. de Phys. 1799, vol. xlix.) A. B. Berthollet used the process in 1808, and Dr. Henry in 1809. On these modes of analysis I have already animadverted (Chemistry, p. 434), and have seen no reason since for changing my opinions. Dr. Henry at the same time discovered that the analysis of ammonia was capable of being effected by nitrous oxide and nitrous gas severally as well as by oxygen. This was scarcely to have been expected, especially by nitrous gas, which is not decomposed by hydrogen alone; but it should seem that the azote of the ammonia, repelling that of the nitrous gas, contributes to the separation of the elements as much perhaps as the attraction of the oxygen for the hydrogen. Whatever may be the true explanation, the fact is a curious and important one; namely, that two compounds, in each of which azote is an element, mutually decompose each other, the oxygen of the one uniting to the hydrogen of the other, and the azote of both being liberated. It may enable us to investigate the proportions of the constituents in both compounds. If chemists were agreed respecting the composition of one of the two compounds in such mixtures (namely, nitrous gas and ammonia), it would be an easier task to ascertain that of the other; but unfortunately the proportions in both are yet subject to dispute. GayLussac, and I apprehend some others, hold that 100 measures of ammoniacal gas are constituted of 50 azote and 150 hydrogen; whereas, according to my experiments, as well as those of Davy and Henry (see my Chemisty, p. 429, 430, 432), 100 ammonia produce only 186, or from that to 190, of mixed gases by electricity; of which I find 28 or 29 per cent. azote, and the rest hydrogen. For the sake of those who may not be much conversant in this subject, it may be proper further to state that, supposing (for in(for`instance) nitrous gas and ammoniacal gas to be mixed in such po portions as, when fired, they may be mutually saturated, by which we mean the oxygen of the one may saturate the hydrogen of the other, and the azote of both be liberated, then, according to Gay-Lussac, 20 ammonia = 10 azote +30 hydrogen, and they require 30 nit. gas = 15 azote + 15 oxygen for their saturation; but, according to my view, 20 ammoniacal gas 10 azote+26 hydrogen, and require nearly 23 nitrous gas = = 10 azote + 13 oxygen for their saturation; all the numbers being understood to denote measures. That is, if a given volume of azote be united to oxygen so as to form nitrous gas, and the same volume of azote be united to hydrogen to form ammonia; then the oxygen of the one will just equal the hydrogen of the other to form water. But it is otherwise in Gay-Lussac's view; for he insists that one-third of the hydrogen will remain in excess. In mixtures of nitrous oxide and ammonia the disproportion between us is still greater; for, according to Gay-Lussac, 20 ammoniacal gas = 10 azote + 30 hydrogen, and they require 30 nitrous oxide = 30 azote + 15 oxygen for saturation; but my view is, that 20 ammoniacal gas 10 azote + 26 hydrogen, and that they require 21 nitrous oxide = 21 azote+ 13 oxygen for saturation; that is, the azote of the nitrous oxide is double that of the ammonia, whereas Gay-Lussac says it is triple. No detail of experiments on the mutual decomposition of the nitrous compounds and ammonia has been published that I know of, besides that of Dr. Henry soon after the discovery; and it is professedly too limited for a complete exposition of the facts. Dr. Henry has only selected one experiment on the decomposition of nitrous oxide and ammonia, the result of which he has explained according to the facts previously discovered by Davy and himself, as the above two theoretic views were not then published. We shall now see whether it supports either of them. 41 measures ammoniacal gas. 38 pure nitrous oxide. together 79 73 when fired; and then found to consist of 16 hydrogen +57 azote. According to the theory of volumes, hydrogen azote+ 19 oxygen. 41 ammoniacal gas = 20 azote+ 61 and 38 nitrous oxide = 38 There should then be found 58 57 azote and 16 hydrogen. But, according to my view, 41 ammoniacal gas = and 38 nitrous oxide azote + 23+ hydrogen, instead of 21 azote + 53 hydrogen There should then be found 59 azote+ 6 hydrogen, instead of 57 azote and 16 hydrogen. Here both theories appear at variance with experiment, and mine rather more of the two; but the accuracy of the experiment cannot fairly be doubted. We must, therefore, see which theory can be most easily bent to accommodate it. According to Dr. Henry, when the nitrous oxide is in excess, there is evidence of nitrous or nitric acid being formed; and I have reason to suppose, further, that in every explosion, whatever may be the proportions, less or more of the acid is formed; because I always find a small diminution of the residuary gas on passing it through water. This being admitted, we are to see how the supposition will influence the preceding calculations. It will evidently decrease the residuary azote, and increase the hydrogen. Let us suppose that two measures azote unite to five oxygen, forming the coloured or mixed acid; then will the residuary gas on my hypothesis be found to be 57 azote + 16 hydrogen, exactly agreeing with the experiment; but the residuary gas, according to GayLussac, would be 564 azote + 334 hydrogen, differing from the experiment by 17 measures of hydrogen. The error increases as the quantity of acid increases; and I cannot conceive any probable cause to be assigned for the differences observed upon this hypothesis. Two experiments on the mixture of nitrous gas and ammonia are related by Dr. Henry in the paper above referred to : 96 fired, left 484 azote + S01 hydrogen and 10 ammonia. Before adverting to these experiments, I may observe that the quantity of ammonia present in any experiment, even if measured in the detonating tube, is very uncertain. Such is the effect of the smallest unperceived moisture in the tube or mercury, that I have known 30 measures of ammonia decomposed when there were only |