the former part of the apparatus. The larger iron cylinder exposes only about three square inches of surface; yet two pairs of this description, is sufficient to heat to whiteness two inches of fine platinum wire, and to decompose water very rapidly ! ! !

There is a trifling liberation of hydrogen gas at the plate immersed in the sulphuric acid; but it may be avoided by having the iron tinned. The coat of tin on the iron answers to mercury on the surface of zinc, and produces a similar effect: it seems even preferable !!! Probably the apparatus would be better if made of cast iron. The porous cylinders which are employed in this battery are those made at Berlin, and are of excellent quality, and apparently of calcined porcelain. When these cannot be had, they may be substituted by Hessian crucibles.

The nitric acid employed in this battery should be sufficiently concentrated to prevent its attacking the iron. A mixture of one part of fuming nitric acid, and one and a half of common nitric acid, will answer very well.

Note. This is not the first time of our hearing of voltaic batteries of "constant action," with iron as the only metal; and, we believe, all of them have been planned from the idea of nitric acid of a certain strength not acting on iron without ever considering that the acid by voltaic action soon becomes altered. But this is the first time of our hearing of the substitution of tinned iron for amalgamated zine!!!-EDIT.

:

On Metallic Acids. By M. E. FREMY.

(Abstract from a Memoir read before the Academy of Sciences)." FROM a general examination of the metallic acids, I have been led to the discovery of some new compounds of metals and oxygen, and consequently of new salts, some of which are remarkable for their beautiful crystalline forms. The metallic acids may be classified under two distinct heads: the first of which comprehends all those which are formed by the immediate combination of metals with oxygen, and are soluble in the alkalies at common temperatures. The second class consists of those acids which are formed by the combined action of an alkali, an oxydizing body, and a metallic oxide, when simultaneously in contact with one another.

These two classes of metallic acids, display properties essentially different from each other. Those of the first class are generally stable; and in combination with certain bases are productive of well-defined crystallized salts. The second class, on the contrary, suffer partial decomposition by a loss of oxygen, when exposed to the feeblest influences.

To give an example of the first class of these acids, I have selected the last combination of tin with oxygen, which has obtained

• Comptes Rendus.

the appellation of stannic acid. The second class of metallic acids I shall illustrate by a new compound of iron with oxygen, which I have named ferric acid. In making these selections of acids, into which the known important metals enter as essential elements, I have been desirous of showing that similar compounds are obtainable with metals less frequent to observation.

I have commenced with the examination of ferric acid, and have given a minute description in my memoir of the different processes by which I have obtained the ferrutes. The compounds of ferric acid with bases may be formed either by the dry or by the humid way. The dry method is simply that of calcining the peroxide of potassium with a sesquioxide of iron, in any vessel not operated on by the produced ferrate. The ferrate of potassa may be easily obtained by the dry process, by throwing ten parts of dry powdered nitre on five parts of red hot iron filings. The red mass thus obtained contains a large share of ferrate of potassa. I have obtained the ferrate of potassa, by the humid process by availing myself of the results of M. Bertheir's experiments, which show the action of chlorine on the metallic oxides. I transmit chlorine through a concentrated solution of potassa, which holds the hydrate of peroxide of iron in suspension. By these means the ferrate of potassa is produced.

In this part of my memoir I enter into some details on the action of chlorine, and strong solutions of potassa; and show, in this case, contrary to the general opinion, that there is no formation of chlorate of potassa and chloride of potassium; but that there is formed a peculiar compound, which I have named potasse chlorée, which at a slight elevation of temperature becomes resolved into chloride of potassium, oxygen, and potassa. It is this compound which, by reacting on the hydrate of the peroxide of iron, gives rise to the formation of ferrate of potassa. In my memoir I dwell upon the influential action of chlorated potassa, in the production of new compounds, constituted of metallic acids and suitable bases; and, in illustration, I prove that under the influence of chlorated potassa, the oxide of copper becomes transformed into a new metallic acid, which I have called cupric acid, and that this acid forms a new compound with the potassa. In other respects it has not been my object to pay particular attention to the action of chlorine on the alkalies; this study belonging more properly to those chemists who have recently brought to light so much interesting matter in this branch of science.

I next proceed to an examination of the properties of the ferrates, and prove that heat, organic substances, or bodies in a state of minute division, are capable of decomposing the ferrates. I also compare these reactions with those which take place in peroxide of hydrogen under similar circumstances. The composition of ferric acid is represented by the formula Fe 03. This acid may therefore be classed with the chromic, manganic, sulphuric, &c. By analysis I prove that the ferrates, by whatever way produced, have precisely

the same composition, but that those obtained by the dry process are generally mixed with nitrite of potassa, which absorbs oxygen at the moment the ferrate commences decomposition, and thus becomes a nitrate of potash.

Lastly, I describe all the experiments that I have made whilst in pursuit of an acid with a larger proportion of oxygen than is contained in the ferric acid, or of an oxide corresponding with the peroxide of manganese. I allude to the action of the binoxide of barium on the sesquioxide of iron, and advance reasons which lead to the opinion that, in that case, there is formed a compound of iron and oxygen, intervening sesquioxide of iron and ferric acid. Such are the various topics of which I have treated in the first part of my memoir. The second part is devoted to an examination of stannic acid. I precede this examination by reviewing all that has been made public concerning that acid, and especially the remarkable researches of Berzelius, and the interesting observations of Gay Lussac thereon. I also allude to a note in Liebig's Journal, by Mr. Graham, in which he explains the modification of stannic acid, previously made known by Berzelius. The first experiments that I made on stannic acid were intended to ascertain what part it plays in its combinations: for chemists have hitherto held different opinions respecting its influence in that capacity. Ought stannic to be considered as an acid, or as a base? May it not alternately operate as an acid and as a base? I have directed my attention to these inquiries.

The result of all the experiments to which I have submitted stannic acid are decidedly unfavourable to its being considered as a base. If, for instance, it be formed by decomposing chloride of tin with an insoluble carbonate, a well developed acid, capable of reddening turnsole, is precipitated. But if the chloride of tin be treated with carbonate of potassa, no stannic acid is produced, the result being a stannate of potassa, which is insoluble. In examining the compounds consisting of stannic acid and the ordinary acids, I find that they ought not to be regarded as salts of peroxide of tin, but rather as combinations of staunic acid with those acids. It is known that chemistry affords many examples of combinations of acids by which double acids are formed. M. Chevreul has proved that stannic acid placed in contact with the colouring matter of logwood, operates in the capacity of an acid; whilst the metallic oxides, properly speaking, even the peroxide of tin itself, operate in the capacity of bases. Hence the compound in which oxygen unites with tin in the highest degree should always be regarded as an acid.

After the examination of this first topic in the history of stannic acid, I proceed to the consideration of its properties. The experiments which I first describe are intended to elucidate the cause of the modifications presented by stannic acid. The inquiry also applies to other metallic acids, and gave rise to the investigations of Berzelius; and, from its generality, involves a problem whose resolu

tion is of high importance. From the results of my experiments I learn that the two modifications of stannic acid, are, in fact, distinct acids, to which I have given different names. To that produced by nitric acid, I have continued the name stannic acid, and that obtained from the chloride of tin I name metastannic acid.

In determining the proportions of water contained in these two ininsulated acids, I have ascertained that the metastannic acid is more hydrated than the stannic acid; and as the only difference in their constitutions is in the proportions of water which they contain, it is obvious that by gentle dessication, the metastannic acid may be transformed into stannic acid. Applying to these acids the ingenious idea of Graham, with reference to phosphoric acid, I was led to infer that the stannates and the metastannates would be found to differ from each other only in the proportions of base, which has proved to be the fact; for by representing the neutral stannates by the general formula, Sn3 06 MO, the composition of the metastannates would be represented by Sn3 06 3 MO. Hence, accordingly with this hypothesis, which I discuss at some length in my memoir, the stannic acid should be regarded as a monobasic acid, and the metastannic as a tribasic acid. The relation which exists between the stannates and the metastannates, explains a curious fact which I have observed, viz.: the stannates when heated with an excess of alkali become transformed into metastannates. The stannates may be obtained by dissolving the alkalies in cold stannic acid which has been prepared by the action of nitric acid on tin. The metastannates may be produced by two distinct processes. By dissolving, in alkaline solutions, the metastannic acid which is obtained from chloride of tin, by the employment of an insoluble carbonate; and, secondly, by calcining in a silver crucible, the stannic acid with an excess of base. The metastannates of potassa and of soda crystallize with facility. These compounds are not surpassed, in any respect, even by the most perfectly defined salts, and probably are the best specimens of crystalline structure exhibited by the combinations of tin.

Whilst studying stannic acid, I have discovered a compound of tin and oxygen intervening the protoxide of tin and stannic acid, which must not be identified with the sesquioxide of tin recently discovered by M. Fuchs. This compound is obtained by treating stannic acid, at common temperatures, with protochloride of tin. The acid immediately assumes a fine orange yellow colour, and pure hydrochloric acid remains in solution. This compound, whose properties I have described in my memoir, should be considered as a stannate of protoxide of tin, corresponding to the molybdate of oxide of molybdenum, to the tungstate of oxide of tungsten, to the chromate of oxide of chrome, &c.

Finally from an examination of the results obtained by the decomposition of the stannates by heat, and extending the process to other metallic salts, I have arrived at the following general conclusion, viz.: certain combinations of metals with oxygen are

not acid, unless hydrated; but when thus acidified, the water enters as a constituent element; and, contrary to that which occurs with other acids when combining with bases, it is not displaced by them, but continues as an essential part of the resulting salt. If by an elevation of temperature, the metallic acid loose its water whilst in combination, it will no longer remain united with bases, but is precipitated in the anhydrous state.

On the Formation of Ferric Acid, by Voltaic Electricity. By Professor J. C. POGGENDORFF.*

An examination of the chemical action of the voltaic pile has led me to discover a novel fact, equally interesting to physical and chemical science. When wrought iron of the best quality is placed in a solution of potassa, and united to platinum immersed in nitric acid, it liberates oxygen gas from the solution, without suffering oxidation itself. Oxygen is also liberated in the gaseous form, when plumbago, platinum, palladium, gold, nickel, cobalt, or tin, is substituted for the iron. Silver, copper, antimony, bismuth, lead, cadmium, and, what is still more remarkable, even zinc, liberates gaseous oxygen: though, not without suffering a partial oxidation, and consequent alteration on the surface of the metals. Silver and lead, which soon become covered with a film of black oxide, are striking examples of this action; and it is not until after this covering is formed, that these metals liberate oxygen gas.

With cast iron, which, by accident, I was led to use in my experiments, the case is very different. This metal is immediately enveloped in a wine-red coloured cloud, which gradually expands till the whole of the liquid assumes a red tint; growing darker as the voltaic action proceeds, and eventually becomes of a deep brown colour, approaching perfect opacity, excepting at the edges, where the light still shows the beautiful medoct colour. The concentration of the alkaline solution appears to be of little consequence in the production of this phenomenon. I have produced it in all its splendour and beauty in a solution of one part of hydrate of potassa in four parts of distilled water.

By a close examination of the alkaline solution, a feeble cracking noise amongst the air bubbles will be heard; and, at the same time, a change of colour in the liquid will be observed to be taking place. The colour assumes a reddish brown, which gradually becomes deeper, but, in about half an hour, the colouring matter having all fallen to the bottom of the vessel, the liquid becomes quite colourless. A continuance of the electric current does not prevent, nor even retard this change; for it commences and goes on whilst

• Annalen der Physik und Chemie.

The French name for a beautiful red coloured flint pebble, found in Madoc.-EDIT.