(441) 210 200 (9.11" (11 45") (12′ 40′′) (12′ 39") (13' 15") (8' 8") 200 190 I' 32" 1/9 53" 50 491 41" 190 180 54" 47" 50 44 46+ 40+ 180 170 52 48 49 48 49 43- 170 Fixed point at 2049, Tin probably becomes solid much above 4009. Bismuth becomes solid at 264°, Tin at 228°, Fixed point at 143°. 140 (7' 2") (13′ 50) (18′ 5′′) (19′ 4′′ (12′ 17") (20′ 40") 140 1' 41" 130 1' 0" 120 zine; in Z T the two stationary points coincide at 204°; the relation of the variable point corresponds with that of the other alloys. For the combinations of lead and bismuth, I found the fixed point to be 129°, and the coincidence of the two stationary points to take place at L3 B; in L B the variable point is at 146°, and in L B2 at 143°, but this latter observation proved to be the result of a remarkable accident; viz., when the thermometer was examined, the ball was found to be compressed to such an extent as to raise the quicksilver in the tube by six degrees; this had, no doubt, been caused by the great expansion which bismuth undergoes when becoming solid, and which is such as generally to break the thermometer when immersed in fused bismuth, and left in it till it is completely solid. In the alloys of zinc and bismuth the fixed point was found to be at 251°; the proportion, at which the depression of the thermometer is regular, could not be ascertained, but I conceive it constitutes an alloy, in which the relative quantity of zinc is very small. It seems to follow from these combined observations, that whatever the proportion of the two metals may be which are fused together, an alloy is always produced, which is represented by a simple atomic ratio (and which might perhaps be properly called the chemical alloy); if the metals are combined in this proportion, the temperature of the mass regularly decreases till it arrives at the fixed point, which, under such circumstances, coincides with that at which the mass becomes solid, and which is generally lower than that at which either of the two simple metals solidify; if, on the contrary, one of the metals is in excess, the thermometer is rendered stationary at some point above that at which the chemical alloy becomes solid; for, as that portion of the metal which is in excess becomes solid before the chemical alloy, the latter derives from it the heat which becomes free by the congelation of the former. This must of course take place at a degree which will be the higher in the ratio of the quantity of the metal in excess. Within more or less time after the solidification of the metal in excess, the chemical alloy becomes also solid, and causes the thermometer to be stationary, in consequence of its latent heat becoming free: the latter is the fixed, the former the variable stationary point. The correctness of this view is also proved by the known fact, that if the mass in fusion is allowed to cool, solidification does not take place simultaneously, but it always, in more or less time, becomes of a mortar-like consistence; whilst, if the metals are fused in the proportion of the chemical alloy, the mixture will be found to become solid simultaneously, and almost in a moment. It appears that there are also ternary chemical alloys as well as binary ones; of the alloys of lead, tin, and bismuth, for instance, one of the points at which the thermometer is fixed is 98°; for the ternary alloys seem to have two fixed points, neither of which, nor the proportion of the metals in the chemical ternary alloys, I have yet been able to ascertain*. * Poggendorf's Ann. der Physik und Chemie, 1830, p. 240. 36. ON THE COMBINATION OF CHLORIDE OF GOLD WITH THE CHLORIDE OF POTASSIUM AND SODIUM.-(Berzelius.) The analysis of the oxide of gold and the other compounds of this metal has led to such various results, as to render it impossible to form any certain conclusion with regard to the proportion of their elements. The recent investigations of Berzelius on the subject are accordingly of much interest, especially as they, in some degree, tend to confirm his views on the atomic weight of gold. We give an extract from his paper in the Kongl. Vetensk. Handl. of 1829. It is known that Pelletier was led, by the results of his inquiry into the composition of the iodide of gold, to consider the atomic weight of gold different from that which had been adopted by Berzelius; but that the views of the Swedish philosopher were strengthened by the subsequent researches of Javal, and particularly by his analysis of the compound of the chlorides of gold and of potassium, which he found to consist of 24.26 of chloride of potassium, 68.64 of chloride of gold, and 7.10 of water, the gold being accordingly united to twice as much chlorine as the potassium. Figuier, who soon after Javal examined the same subject, discovered the compound of the chloride of gold and of sodium, the proportion of which he found to be 14.1 of chloride of sodium, 69.0 of chloride of gold, and 16.6 of water, and the chloride of gold to contain accordingly nearly three times as much chlorine as that of sodium. Dr. Thomson, who,' says Berzelius in the above paper, has lately undertaken to determine the atomic weights more accurately than others, was also led to the examination of that of gold *, the oxide of which he found to consist of one atom of gold and three atoms of oxygen, and the chloride of one atom of the oxide and two of muriatic acid. The analysis of the oxide corresponds with my own; that of the chloride is evidently erroneous, as may be seen from the decomposition of the salt by heat, where chlorine and oxygen gas are not formed in the proportion of 4 to 1, as would necessarily result from Dr. Thomson's analysis. The compound of the chloride of gold and of sodium is, according to the same chemist, composed of 14.85 of chloride of sodium, 49.51 of gold, 17.82 of chlorine, and 17.82 of water, which is also erroneous; for if it were correct, the third part of the gold would be precipitated as an oxide, during the preparation. of the compound from chloride of sodium and (what Dr. Thomson considers as a muriate) of the oxide of gold; but, according to his own experiments, such a precipitate is not formed. He concludes, however, that I am wrong in supposing that hydracids decompose oxidized bases, and that the muriate of gold gives a striking proof of the incorrectness of my ideas, as the oxide in this salt contains a third more oxygen than could be united to the hydrogen of the acid. After this rapid sketch of the previous labours on the subject, I come to my own analyses, which were made in the presence of Mr. Johnson, a pupil of Dr. Thomson's. * Transact. of the Royal Soc. of Edinb., vol. xi., p. 23. Chloride of Gold and Potassium crystallizes in square prisms and rhombo-hedrons; is of an orange colour, and very efflorescent in a dry atmosphere; at 212° F. they part with their water of crystallization, but without losing any chlorine; two grammes of the salt were found to have lost 0.2125 gr. of water. On bringing the remainder in contact with hydrogen gas at a gentle heat, the gold was reduced, and 0.501 gr. of chlorine was found to have united with the hydrogen: of the remainder, 0.3505 gr. of chloride of potassium was dissolved, and .0936 gr. of gold was the residuum. If the compound be considered as composed of one equivalent of chloride of potassium (equal to one atom of chlorine and one of potassium), and one of chloride of gold (equal to one atom of gold and three of chlorine), the result of the analysis will be found pretty nearly to correspond with that of calculation: 'Chloride of Gold and Sodium crystallizes in orange-red prisms, and cannot be freed from its water of crystallization without losing its chlorine. On reducing it by hydrogen gas, the residue of 100 parts consisted of 14.466 of chloride of sodium, and 49.51 of gold. In order to determine the proportion of chlorine in the compound, 3.026 grammes of the crystallized salt were mixed with 6 gr. of efflorescent carbonate of soda, and heated in a platina vessel, until the compound was decomposed and the gold reduced, which, when separated from the soluble salt, was found to weigh 1.4978 gr.: it formed, therefore, 49.497 per cent. of the compound. The solution was saturated with nitric acid, and nitrate of silver added to it; the precipitate was 4.3347 gr. of chloride of silver, which corresponds to 35.54 of chlorine in 100 parts of the salt. Of this chlorine 8.835 must accordingly belong to 14.466 of the chloride of sodium, and the remainder 26.501 to the gold; and if the compound be considered as consisting of one equivalent of chloride of gold (equal to one atom of gold and three atoms of chlorine), one equivalent of chloride of sodium (equal to one atom of chlorine and one of sodium), and four equivalents of water, the results of analysis and calculation will correspond perfectly. Poggendorf's Ann, der Physik und Chemie, 1830, p. 597. 37. ON MAGNESIUM.-(Justus Liebig.) The Annales de Chimie, of March, 1830, contain a paper by M. Bussy, on magnesium, which he obtained by the action of chloride of magnesium on potassium; the properties of this metal appeared to M. Liebig to be so very extraordinary, that he was induced to make some experiments on it. M. Bussy's method of obtaining the chloride consists in passing chlorine gas over a mixture of magnesia and charcoal, whilst in a state of ignition; it may, however, also be obtained by evaporating equal parts of the muriates of ammonia and magnesia, and heating the dry residuum in a platina vessel, until the muriate of ammonia is completely expelled, and the mass becomes fused. The remainder is chloride of magnesium, and if left to cool, forms white transparent leaf-like crystals. In order to reduce the chloride of magnesium, from about 10 to 20 small globules of potassium are put into a glass tube, three or four lines in diameter, the chloride is placed over them, and heated over charcoal, until it begins to flow; the tube is then slightly inclined, so that the potassium runs through the chloride, which is thus reduced to magnesium with the evolution of light. If the mass, when cold, be treated with water, a large quantity of small metallic globules will be collected at the bottom of the vessel; they are of a silver-white colour, have much metallic lustre, and, though malleable, are very hard; neither cold nor hot water acts on them. If mixed with chloride of potassium, and heated in a crucible, they may be fused into one mass, and their point of fusion does not apparently exceed that of silver. The metal is dissolved by diluted acetic acid, as well as by sulphuric and nitric acids, with the evolution of hydrogen gas, and sulphurous and nitrous vapours: the solutions are found to contain no other oxide besides magnesia. When heated in atmospheric air or oxygen gas, the metal burns with the most vivid light; the vessel is covered with magnesia; and at the place where the metal was, a black spot remains, which seems to be silicium, as it was not destroyed by boiling hot acids. Sulphur did not seem to unite with the metal, when both were fused together. The solution in sulphuric acid yielded, on evaporation, crystals of sulphate of magnesia*. 38. ON THE EXPANSION OF BISMUTH AND ITS ALLOYS DURING CONGELATION. (Professor Marx, of Brunswick.) Bismuth is known to be a very remarkable instance of apparent exception from the general rule, that fluids contract when becoming solid; and it corresponds with water in this respect also, that it communicates this property to other bodies, particularly metals, if it forms a certain proportion of the alloy. Where the maximum of Poggendorf's Aunalen, |