considerations. I am not disposed to change them till it has been shown in a satisfactory manner that they are inconsistent with experiment. The proportion of oxygen which I have given differs only about th part from that given by Berzelius. Now I am very much afraid that the limits of unavoidable error in such experiments are greater thanth of the whole. Hence we have no means of coming at the truth except by theoretic views, which will guide us to new experiments; and when these are sufficiently multiplied, we shall obtain a mean approaching very near the truth. M. Arvedson, during his experiments, made a discovery of rather an interesting nature, and deserving the attention of the manufacturers of the bleaching salt and bleaching liquor. He found that there are two native black oxides of manganese. The first, the common peroxide; the second, the hydrated black oxide, which he found composed as follows : Oxidum manganoso-manganicum 89.92 . 10.08 100.00 This oxidum manganoso-manganicum is a compound of two atoms of peroxide and one atom of protoxide of manganese; or it contains th less oxygen than the peroxide. If the oxygen in the water be added to that of the oxidum manganoso-manganicum, the whole will be converted into peroxide of manganese.(Jour. de Phys. lxxxvii. 464.) There is reason to believe from the late experiments of Chevillot and Edwards, that manganese is capable of combining with an additional atom of oxygen, and of forming a new compound, which seems to possess acid properties, and to act with great energy on combustibles. They have not yet given us the proportions of manganese and oxygen which exist in this compound; but they have shown that red chameleon mineral is a compound of potash, black oxide of manganese, and oxygen, which are all present in definite proportions; that the quantity of oxygen depends upon that of the manganese present, and not upon that of the potash; that the combination is neutral, and possesses the characters of a salt; and that when an excess of potash is added, the chameleon assumes a green colour. When these crystals are heated in contact with hydrogen gas, they set it on fire. They detonate violently with phosphorus, set fire to sulphur, arsenic, and antimony, and indeed to all combustible bodies hitherto tried. Were we to suppose this manganesic acid (as Chevillot and Edwards have termed it) a compound of one atom manganese and three atoms oxygen, its constitution would be as follows: These facts claim the careful examination of chemists. If they be verified, they will exhibit the remarkable and hitherto unique example of the same base forming a perfect salifiable base and a perfect acid simply by uniting with different proportions of oxygen. This would be a fine confirmation of the theory advanced by Ersted respecting the cause of acidity and alkalinity, of which an account has been given in a late number of the Annals of Philosophy.-(See Ann. de Chim. et Phys. viii. 337.) 9. Cobalt and Nickel.-The most difficult problem, perhaps, in practical chemistry is the separation of these two metals from each other. A variety of methods have been proposed, all of which I have tried, with some additional ones of my own, without having yet hit upon one which is not either imperfect, or at least liable to some very serious objection. When into a concentrated solution of cobalt in sulphuric or muriatic acid, a solution of tartrate of potash is added, a triple salt is formed, consisting of tartaric acid, united at once with potash and with oxide of cobalt, which crystallizes in large flat rhomboidal prisms. These crystals, so far as I have examined them, contain no other metal except cobalt; but this method, though promising at first sight, I did not find to answer so well as I expected; for the tartrate of potash undergoes spontaneous decomposition when the solution is left to spontaneous evaporation; and if the evaporation is produced by the action of heat, the crystals formed are ill defined, and consequently liable to be impure. It was with great pleasure, therefore, that I perused a paper by M. Laugier, published in the Annales de Chimie et Physique for November, 1818, on the mode of analyzing the ores of cobalt and nickel, and on the best method of separating these two metals from each other. After trying every known method of separating these two metals from each other without succeeding, MM. Laugier and Silveira were on the point of abandoning the investigation, when it occurred to them to try the effect of a concentrated solution of ammonia on the impure oxalate of nickel. A solution took place of a fine azure colour. On expos ing this solution to the open air, the ammonia gradually made its escape, and at the same time the oxalate of nickel precipitated to the bottom of the vessel; while the whole of the oxalate of cobalt remained in solution. Thus it is easy to separate these two metals from each other by converting them into oxalates, treating the oxalates with.ammonia, and leaving the ammoniacal solution for some days in an open vessel. I applied this method as a test to ascertain the purity of the nickel and the cobalt which I had purified before M. Laugier's paper came into my possession. I had the satisfaction to find that it neither indicated the presence of nickel in my cobalt, nor of cobalt in my nickel; therefore, if M. Laugier's method be a good one, I had succeeded beforehand in accomplishing a complete separation of these two metals. The method of proceeding to analyze the ores of cobalt suggested to Laugier by the preceding facts, is the following: (1.) Let the ore be roasted to drive off as much of the arsenic as possible. (2.) Dissolve the roasted ore in nitric acid and evaporate nearly to dryness to get rid of the arsenious acid. (3.) Pass a current of sulphuretted hydrogen gas through the liquid till the whole of the arsenic and copper (if any be present) be thrown down. (4.) Heat the liquid to drive off the excess of sulphuretted hydrogen, and precipitate the metals by means of carbonate of soda. (5.) Treat the carbonates with oxalic acid to separate the iron. Then dissolve the oxalates of cobalt and nickel in ammonia to separate these two metals. Laugier informs us that he detected nickel in the cobalt ore of Tunaberg, though the presence of that metal had not hitherto been suspected in it. 10. Brass. I was much amused by a remark which Mr. Gill has thought proper to make upon an observation of mine in my Historical Sketch of Chemical Science for 1817. I stated the well-known fact that old Dutch brass was much more valued by watchmakers than British brass, and gave my reasons for the difference between them. The Dutch brass is a compound of two atoms copper and one atom zinc; while English brass is a compound of one atom copper and one atom zinc. I think I generally write so perspicuously that my meaning can hardly be mistaken; yet Mr. Gill insinuates, in pretty broad terms, that I considered the partiality of watchmakers for the Dutch brass as a prejudice (Annals of Philosophy, xii. 125); though I had stated, in as clear a manner as I could, the reason of the superiority of the Dutch over the British. The prejudice of my friend, the watchmaker, did not consist in considering the Dutch brass as better for his purpose than the English, which is really the case; but in supposing that the art of making that good kind of brass is lost. I pointed out how it might be easily manufactured at the pleasure of the brass maker; and Mr. Gill informs us in the article already quoted that his father in-law intended to set up a manufactory of this old superior kind of brass. I am glad to hear it. He will prove the truth of what I ventured to assert on general grounds, that modern brass makers may, if they think proper, make as good brass as that which the watchmakers value so highly. Whatever Mr. Gill may think upon the subject, I must be allowed to consider my observations as of some importance. They were founded on experiment, and they explained a fact generally known, but not previously accounted for, that old Dutch brass is superior in ductility, &c. to English brass. 11. Bismuth. It was observed many years ago by Dufoy that bismuth may be substituted for lead in the process of purifying gold and silver by cupellation; but no accurate experiments had been made to determine whether that metal could be employed to ascertain exactly the quantity of alloy contained in gold and silver. We are indebted to M. Chaudet for solving this problem. He has ascertained that the bismuth to be employed in such cases must be free from silver; that if it contains arsenic, which is commonly the case, a portion of the silver is driven off the cupel along with the arsenic and lost that a smaller proportion of bismuth must be employed than is required of lead; and that the cupels employed must be less porous than those used when lead is used to separate the alloys from gold or silver; because bismuth has the property of inducing so great a degree of fluidity into those metals that they are apt to penetrate into the pores of an ordinary cupel and to be lost. It follows from the experiments of Chaudet, that if these precautions be attended to, bismuth may be employed as well as lead to determine the purity of gold and silver. The following table exhibits the quantity of bismuth requisite for purifying one part of silver of the degrees of purity marked in the table: (Ann. de Chim. et Phys. viii. 113.) 12. Tin. This metal has so great a tendency to unite with a maximum of oxygen that the preparation of its protoxide is attended with some difficulty. I have generally succeeded by keeping the permuriate of tin in a close vessel in contact with a quantity of metallic tin, and then precipitating the protomuriate by ammonia; but this method is not always attended with the desired success. M. Cassola has given us a process which he assures us never fails. Upon filings of tin, he pours nitric acid diluted with ten times its volume of water, and leaves the two substances in contact for 48 hours. The tin acquires a brownish black colour, and is completely converted into protoxide. The nitric acid contains in solution a portion of protoxide. When kept for some time, it lets fall an insoluble subnitrate, which is gradually changed into peroxide of tin. Besides the protoxide, there is a yellowish light matter which floats about in the liquid, and which may be separated by the filter. It is a protohydrate of tin. Acetic acid, when left in contact with tin filings, dissolves a portion, and converts it into protacetate of tin, but the residual tin filings are not oxidized.*-(See Giornale di Fisica, Chimica, &c. 1818, p. 378.) The observations which M. Cassola makes on the peroxide of tin contain nothing which has not been long known to chemists. This peroxide is not white, as is stated in some recent systems of chemistry, but yellow; and it is insoluble in all the acids which I have tried. Its hydrates (for there are several) are of a fine white colour, and dissolve readily in muriatic, but not in nitric acid. 13. Mercury.-M. de Blainville has made an observation which seems entitled to attention, and which, therefore, I notice here. It is well known to chemists that mercury amalgamates very easily with gold, silver, lead, tin, zinc, bismuth, and arsenic; but it does not amalgamate with iron, cobalt, and nickel; or at least the amalgams of these metals cannot be formed without considerable difficulty. Now the observation of M. de Blainville is, that when these metals are united to arsenic, the alloy amalgamates very readily; so that by the intervention of this metal, we can easily procure amalgams of those metals which do not, in other circumstances, unite to mercury.-(Jour. de Phys. lxxxiv. 267.) I was rather surprised to find (Annals of Philosophy, xii. 67) that Mr. Donovan had concluded from his experiments that the composition of the oxides of mercury is as follows: Protoxide..... 100 mercury + 4.12 oxygen Peroxide. 100 mercury + 7.82 oxygen .... These numbers are inconsistent with the doctrine of definite proportions, which has been perfectly well established. The experiments of Fourcroy, Sefstrom, &c. have shown that the composition of these oxides is as follows: .... Protoxide. 100 mercury + 4 oxygen 14. Copper.—It seems hardly worth while to recall the attention * I may observe here that I had the curiosity to try this process, but did not find it to answer. The mode which I have usually followed to obtain protoxide of tin is to dissolve that metal by means of heat in muriatic acid. I put the solution into a well-stopped phial, placing in it a number of slips of tin. These slips gradually reduce the whole of the dissolved metal to the state of protoxide. I have sometimes seen the dissolved tin precipitated upon the tin in crystalline plates, having the metallic lustre. |