says that "euchlorine produces the phenomena which Chenevix in his paper on oxymuriatic acid ascribes to hyper-oxymuriatic acid :" and that "it is probably combined with the peroxide of potassium in the hyper-oxymuriate of potash." But I shall demonstrate that this is not the case.*

We must admit it as an incontestable principle, established by Berthollet, that an acid put into a saline solution acts on the base of the salt, and separates a portion of it from its acid. This principle holds especially with the strong acids when brought in competition with the weak acids. On the other side, we must recollect, that peroxide of potassium does not combine with sulphuric acid, and that as soon as these two bodies are brought in contact, oxygen is disengaged. Hence, if hyperoxymuriate of potash were produced by the combination of euchlorine with peroxide or potash, there ought to be disengaged oxygen gas, when diluted sulphuric acid is poured into a solution of this salt. Since at least the euchlo rine, a gaseous oxide, whose acid properties, if it has any, are very weak, will be partly separated by the sulphuric acid, and this acid is incapable of dissolving peroxide of potassium. But no oxygen is disengaged, and consequently the potassium is not in the state of peroxide in hyper-oxymuriate of potash.. Besides, even supposing potash super-oxygenated in the hyperoxymuriate, it ought to contain five times its usual quantity of oxygen, a conclusion which it would be very difficult to admit. The fact is, that potassium is in the same degree of oxydation in the hyper-oxymuriate as in the sulphate, as I shall now demonstrate, by giving an account of the real acid which forms the fulminating salts of chlorine.

In consequence of the above considerations, I was led to believe, that since sulphate of barytes is insoluble, and barytes is not superoxydated in this salt, if sulphuric acid be put into the hyper-oxymuriate of barytes, it would be easy to see if oxygen be disengaged, and perhaps even to obtain chloric acid. I accordingly prepared a certain quantity of this salt, employing the ingenious process of Mr. Chenevix, and I obtained it easily in fine rhomboidal prisms, quite exempt from muriate. Into a diluted solution of this salt Í poured weak sulphuric acid. Though I only added a few drops of acid, not nearly enough to saturate the barytes, the liquid became sensibly acid, and not a bubble of oxygen escaped. By continuing

* In a preceding memoir on oxymuriatic acid, Davy appears to doubt the existence of an acid in the hyper-oxymuriates. He expresses himself in this manner. "If we consider with attention the facts concerning the hyper-oxymuriate of potash, we can only consider it as a triple compound of oxymuriatic acid, potassium, and oxygen. We have no sufficient motive to conclude that any particular acid exists in that body, or that it contains a considerable quantity of water. It is perhaps more conformable to chemical analogy, to suppose the great quantity of oxygen to be combined with the potassium, the very great affinity of which for oxygen we know, rather than to consider this quantity of oxygen as in a state of combination with the oxymuriatic acid, which, as far as we know, has no affinity for that substance. And from some experiments, I am induced to believe that potassium may combine directly with more oxygen than exists in potash.”

to add sulphuric acid with caution, I succeeded in obtaining an acid liquid entirely free from sulphuric acid and barytes, and not precipitating nitrate of silver. It was chloric acid dissolved in water. Its characters are the following.

This acid has no sensible smell. Its solution in water is perfectly colourless. Its taste is very acid, and it reddens litmus without destroying the colour. It produces no alteration on solution of indigo in sulphuric acid. Light does not decompose it. It may be concentrated by a gentle heat without undergoing decomposition, or without evaporating. I kept it a long time exposed to the air, without perceiving that its quantity diminished sensibly. When concentrated it has somewhat of an oily consistency. When exposed to heat it is partly decomposed into oxygen and chlorine, and partly volatilized without alteration. Hydrochloric acid decomposes it in the same way at the common temperature. Sulphurous and hydro-sulphuric acids have the same property; but nitric acid produces no change upon it. I combined it with ammonia, and obtained a very fulminating salt, announced for the first time by Mr. Chenevix. With potash I produced hyper-oxymuriate with all its characters. It does not precipitate nitrate of silver nor any other metallic solution. It readily dissolves zinc, disengaging hydrogen; but it appeared to me to act slowly on mercury.* This acid without doubt cannot be obtained in the gaseous state. As it contains five times as much oxygen as the oxide of chlorine, which is so easily decomposed, we cannot doubt that it is the water which keeps its elements united, as is the case with nitric and sulphuric acids. In this point of view the water acts the same part as the salifiable bases. But as it does not neutralize the bodies which it holds in solution, on account of the perfect equilibrium which exists between the acidifying properties of the oxygen and the alkalifying properties of the hydrogen, and because its affinities are much weaker than those of the bases, it serves merely to unite the elements, and allows us to study the characters of the combinations which it forms, as if they were independent of its presence,

The theory of the chlorates will not now present any difficulty. They are salts formed by the combination of chloric acid with bases, and are entirely analogous to the iodates. Some obscurity, however, may remain about the circumstances of their formation, when an alkaline solution is saturated with chlorine.. I shall therefore endeavour to throw some light on the subject. I shall commence by determining theoretically the ratio of the quantities of chloruret of potassium and chlorate of potash which form at the same time, and then I shall inquire if it agrees with that which experience gives.

It is composed of 1 volume of chlorine and 2·5 of oxygen, or by weight of 100 chlorine and 113-95 oxygen, supposing the specific gravity of chlorine to be 2-421.

I have already remarked, that from 100 parts of chlorate of potash we may obtain 38.88 of oxygen and 61 12 of chloruret of potassium, and that this chloruret is composed of 28.924 chlorine, and 32.196 metal. Further, as I have demonstrated that potassium is in the state of potash in the chlorate, we must give it 6.576 of the 38.88 of oxygen. There will remain 32.304 to convert the 28.924 of chlorine into chloric acid. But what hypothesis soever we adopt with respect to the existence of the hydrochlorates, the oxygen can only have been furnished to the chlorine either by the potash or the water. On the first supposition there will be formed evidently a quantity of chloruret of potassium, proportional to the quantity of oxygen which the potash has furnished to the chlorine. And as that obtained from the decomposition of the chlorate is neutral, and that it is equally proportional to the oxygen which its potassium would take to be converted into potash, we see that the quantity of chloruret of potassium which forms at the same time with the chlorate, will be to that of the chloruret obtained from the decomposition of the same chlorate, as 32.304 to 6.576, or nearly as 5 to 1; and the quantity of chlorate will be to that of the chloruret formed at the same time as 100 to 300 2.

On the second supposition such a quantity of water will be decomposed, that there will result 32.304 of oxygen for the chlorine, that is to say, 36.59, and the corresponding hydrogen will form with the chlorine hydrochloric acid, which will saturate the potash. We will then have for the proportion of chlorate to hydrochlorate, 100 to 300-2 +36.59; or 100 to 336-79. We must suppose that the hydrochlorate remains in solution in water; for I have demonstrated, that as soon as the water is removed, even by a very gentle evaporation, it is converted into chloruret of potassium.

The proportion of 100 chlorate to 300-2 chloruret, which I have just determined is very different from that found by experience. Mr. Chenevix, in his paper on oxymuriatic acid (Phil. Trans. xcii. 132) finds that there are formed 16 parts of chlorate for 84 of chloruret. Correcting this ratio from his data, and the results which I have just established, I find 14.4 of chlorate to 85 6 of chloruret, or 100 to 595-4. M. Berthollet (Stat. Chim. ii. 198) says, that he obtained a proportion still weaker. If these proportions were nearly exact, it would follow, that we have not attended to all the circumstances which accompany the formation of chlorate and chloruret; for otherwise what hypothesis soever we adopt, the proportions of chlorate to chloruret and to hydrochlorate, cannot differ from those which I have just established, supposing the data correct from which I set out. To determine this point I made the following experiments.

I passed chlorine into a somewhat concentrated solution of potash, till it refused any longer to dissolve in it. The liquid was green

* It has been believed that the property which the chlorates have of being

ish, and had a strong odour of chlorine, which it lost when heated. I observed that during this process a little oxygen was disengaged, and the liquid became alkaline. Having evaporated it to dryness, I put a certain quantity of the residual saline mass into a small glass retort, to the beak of which was fitted a syphon-shaped tube, rising to the upper part of the vessel in which the oxygen gas was collected. I heated the retort gradually nearly to redness. When no more oxygen was disengaged, and when the apparatus had sunk to its original temperature, I brought the water in the jar to a level with that in the cistern, and withdrew the tube which had conducted the gas into the jar. By this method, the oxygen which remained in the tube and retort was replaced by an equal quantity of common air. Knowing the quantity of oxygen disengaged, and of chloruret remaining in the retort, it was easy, on the supposition that 100 parts of chlorate contain 38.88 of oxygen, to determine the quan tity of chlorate of potash mixed at first with the chloruret of potassium, and to calculate the ratio of the one to the other. By this method I found that 100 of chlorate corresponded in this mixture to 356.5 of chloruret. On suturating with chlorine a solution of potash more concentrated than the preceding, the proportion of chlorate to chloruret was still found sensibly the same. But when the potash was dissolved in about 30 times its weight of water, ratio of the chlorate to the chloruret was then 100 to 512. It results then from these experiments, that the more concentrated the potash is, the more chlorate do we obtain relatively to the chloruret; but that the ratio always differs from that of 1 to 3, which calculation gives us. As I remarked that the solution of potash, though super-saturated with chlorine, is alkaline, when the excess of chlorine is disengaged by heat, I determined the quantity of alkali in excess, by saturating it with hydrochloric acid of a given strength. By this means I reduced the ratio of 100 chlorate to 356-5 chloruret, to that of 100 to 349. I observe further, that oxygen is disengaged when we heat a solution of potash saturated with chlorine, and even during the saturation of the potash, according to the observation of M. Berthollet. But as I have not determined the quantity, I cannot say what modification it will introduce into the ratio. However, as it is evident that on decomposing by heat the saline mass produced by the saturation of potash with chlorine, we must obtain a quantity of oxygen equal to that contained in the alkali, whether chloric acid be formed, or any other combination of chlorine and oxygen, we cannot ascribe to any other causes than those of which I have just spoken, the

the

easily decomposed by heat, and of burning most combustible bodies, depends on the chlorine preserving all its caloric when it combines with potash. As a proof, it was stated, that during the combination of these two bodies, the temperature of the solution did not sensibly vary. This cause cannot be true, for in the experiment of which I have just spoken, the temperature at the commencement of the saturation rose from 64° to 174°.

difference between the quantity of oxygen which I ought to have obtained, and what I actually obtained by experiment.

The action of chlorine on the oxides is entirely analogous to that of iodine; and chloric acid is produced nearly in the same circumstances as iodic acid. Thus we obtain with peroxide of mercury and chlorine, chloruret and chlorate, in the same manner as with iodine and this peroxide we form ioduret and iodate of mercury. These different objects require new researches, and it is to be desired that they may fix the attention of chemists.

The chloruret of azote, from its analogy with the ioduret, ought to be composed of three parts of chlorine and one part of azote; but Davy instead of this found four to one. When we see azote forming with chlorine and iodine very fulminating compounds, we may ask whether fulminating gold and silver, and even mercury, are not binary combinations of azote and the metal. This is the more probable, as gold, silver, and mercury, having very little affinity for oxygen, seem by this property to approach chlorine and iodine.

From the analogies which I have established in this memoir, the reader must be convinced that oxygen, chlorine, and iodine do not form an insulated group to which belong exclusively the property of acidifying. We have seen that this property belongs likewise to sulphur and azote, and to a great number of other bodies. However, oxygen may be always considered as the principal acidifying substance, both from the energy with which it possesses it, and from the numerous acids which it forms; and because we are only able to employ as solvents liquids containing oxygen or hydrogen, capable of changing the nature of the compounds which they dissolve. Though chlorine does not disengage oxygen from all its combinations, I think it should be placed before it, on account of the energy of its properties. But fluorine, which has not hitherto been obtained in a separate state, will, without doubt, stand before chlorine, because it disengages oxygen from all its combinations. It is to M. Ampere that we owe the first idea that fluoric acid is analogous to hydrochloric acid; that is to say, that it is composed of hydrogen and a body analogous to chlorine, which he proposed to call fluorine. Davy, to whom he communicated that theory did not adopt it nor endeavour to verify it till long after, when M, Ampere had answered his objections.

(To be continued.)