appears to me capable of forming regular salts with acids. I am of opinion that in this compound manganese is in its lowest degree of oxidation. The existence of the sub-oxides described by Berzelius and Dr. John of Berlin, has never been satisfactorily demonstrated; and I have reason to suspect that one or other of them would in some of my experiments have been generated, did there exist any tendency to their formation.

The protoxide may be formed, as was shown by M. Berthier in the 20th volume of the Annales de Chimie et de Physique, by exposing the peroxide, deutoxide, or red oxide of manganese to the combined agency of charcoal and a white heat; and Dr. Forchhammer has in the Annals of Philosophy described an elegant method of preparing it by means of hydrogen gas at a red heat. Arfwedson has likewise had recourse to this method, and I have employed it very extensively during the course of the present investigation. The mode of performing the experiment is as follows. The material for yielding the protoxide was either the red oxide, deutoxide, or peroxide of manganese; and, occasionally, the carbonate was used. When it was wished to employ a red heat only, the material was placed in a small tray of platinum foil, which was introduced into a tube of green glass, through which the hydrogen gas was transmitted. The heat was applied by means of a pan of burning charcoal. To prevent the tube from bending while softened by the heat, two or three pieces of tobacco-pipe were tied to it longitudinally by means of ironwire. But when it was wished to prepare the oxide at a very high temperature, the material was put into a small tube of porcelain, and then introduced into a gun-barrel which was exposed to a full white heat in a common wind-furnace. · A supply of hydrogen gas was procured in the usual manner from zinc and dilute sulphuric acid; but before coming in contact with the oxide of manganese, it was purified by being transmitted through a strong solution of potash, and then dried by the chloride of calcium. At the close of the process, the protoxide was of course preserved in an atmosphere of hydrogen gas until it was quite cold.

2

1

The abstraction of oxygen commences at a temperature below that of redness; and when the peroxide is employed, it becomes red hot by the caloric evolved during the formation of water, considerably before the tray which supports it is rendered luminous by the heat of the fire. It appears nevertheless from all my experiments that a strong heat is requisite in order to convert all the red oxide into the protoxide. When the process is conducted at a low red heat, I uniformly

found

found that on putting the product into dilute sulphuric acid, which instantly dissolved all the protoxide, a portion of the red oxide came into view. This affords a sure criterion of the operation being complete; for the pure protoxide dissolves without residue in dilute sulphuric acid, and yields with it a perfectly colourless solution. There seems to be no risk of decomposing the protoxide by the employment of a high-temperature. I have exposed the recently prepared protoxide a second time to the action of hydrogen gas and a long continued bright red heat without the weight being changed in the slightest degree; and after exposure to the same gas and a full white heat for an hour, it dissolves in dilute sulphuric acid without the slightest effervescence.

The protoxide of manganese is described by Forchhammer as being of a beautiful light-green, and by Arfwedson as of a pistachio-green colour. I have seen specimens with a tint very near the pistachio-green, but these always contained an admixture of red oxide. The colour of the pure protoxide is very near the mountain-green.

With respect to the action of air, my observations differ from those of Forchhammer, who found that recently prepared protoxide attracted oxygen from the atmosphere before he could weigh it. The protoxide procured in my experiments is far more permanent. I exposed fifteen grains of recently prepared protoxide to the free action of the air. during the space of nineteen days, when it was found to have undergone no change either in appearance or weight. If, therefore, it does attract oxygen at all from the air, the operation must proceed very tardily. It absorbs oxygen very slowly, even at a temperature of 400° F.; for 7.269 grains of the protoxide, after an hour's exposure to that degree of heat, did not gain in weight more than 0.021 of a grain. At a temperature of 600° F. it absorbs oxygen much more rapidly; and at a low red heat it loses its green tint, and becomes almost black in an instant. I have repeated this process frequently, but in no case did the protoxide take fire, as occurred in the experiments of Forchhammer and Arfwedson. I entirely agree with M. Arfwedson, however, in the statement, that the protoxide is converted, by simultaneous exposure to heat and air, into the red oxide. This is the uniform result at whatever temperature the oxidation is effected.

I have already mentioned my opinion, that, of the oxides of manganese, the protoxide is the only one which forms definite compounds with acids. It unites readily with this class of bodies, without effervescence, producing with them the same salt which is formed when the same acids act on the carbonate

of

of manganese. When it comes in contact with concentrated sulphuric acid, an intense heat is instantly evolved; and the same phænomenon is produced, though in a less degree, by strong muriatic acid. This oxide is likewise the base of the salts which are formed when sulphuric or muriatic acid is heated with the peroxide, deutoxide, or red oxide of manganese. As the accuracy of this statement, as respects sulphuric acid, has been denied by an acute chemist and good observer, I have been induced to examine the question with considerable care. I mentioned in my Elements of Chemistry, in explaining the process for procuring oxygen gas by means of sulphuric acid and the black oxide of manganese, that the peroxide loses a whole proportion of oxygen, and is converted into the protoxide, which unites with the acid, forming a sulphate of the protoxide of manganese. The gentleman who has done me the honour to review that work in the Annals of Philosophy, I apprehend Mr. Richard Phillips, has made the following remark on the preceding passage. "This statement is at variance with both Dr.Thomson's and also with the results of our experiments; for we find that 44 or one atom of peroxide of manganese yield 4-2 of oxygen, which is so much nearer 4 than 8, that there is no question but that the deutoxide, and not the protoxide is obtained by the action of sulphuric acid; that this is the case is further proved by the deep red colour of the solution of the sulphate, and by its losing that colour, as stated by Dr. Thomson, when mixed with sulphurous or nitrous acid."

1

To decide this point between the reviewer and myself, it is only necessary to heat the peroxide of manganese with concentrated sulphuric acid, so as to form a solution highly charged with the oxide of manganese, and decant off the solution while hot from the undecomposed peroxide. The liquid on cooling deposits a perfectly white salt, which possesses every property of the protosulphate of manganese. If the acid, which retains an amethyst-tint even when cold, be again heated, the red colour speedily disappears; because the red oxide, which is dissolved in small quantity by the sulphuric acid, is then also converted into the protoxide with the evolution of oxygen gas. The red colour disappears gradually even without the aid of heat; for the solution will be found after a few days to be almost and sometimes quite colourless, while a minute quantity of red oxide has subsided to the bottom. On applying a very gentle heat, the red oxide is redissolved, and the acid acquires a lively amethyst-red colour. It is easy, by operating in this way, to obtain satisfactory proof, that a minute portion of red oxide suffices to communicate a rich colour to a considerable

quantity

quantity of sulphuric acid. The acid may be made to retain its red colour, either by diluting it with water, or by keeping it in contact with undissolved oxide. I

On the Red Oxide.I have followed the usage of most chemists in applying the term red oride to that compound which Arfwedson has described under the name of Oxidum manganoso-man→ ganicum (Annals of Philosophy, N. S. vol. vii. p. 267), and which is uniformly produced when the nitrate, peroxide, or deutoxide of manganese is exposed to a white heat. In my early experiments on this oxide, I entertained considerable doubt as to the uniformity of its composition. This opinion origi nated in the remark, that, on exposing the peroxide of manganese to a white heat, the quantity of oxygen lost by different portions of it, though agreeing perfectly in some experiments; differed widely in others; and that, on one occasion, I pro cured the green oxide almost in a state of purity. I subse quently discovered, however, that the disagreement in the results was occasioned by the want of a free current of air within the furnace. In some of the experiments the draft was unguardedly cut off, and consequently an atmosphere of carbonic oxide gas, collecting around the heated manganese, reduced it more or less nearly to the state of protoxide. On avoiding this source of fallacy, the results were no longer discordant; and I am now quite satisfied that the red oxide formed at a white heat and with free exposure to atmospheric air, is uniform in its composition. The accuracy of this inference is established by the occurrence of the red oxide in nature, as will appear in the sequel of the present communi

The red oxide, when formed at a white heat and rubbed in a mortar to the same degree of fineness, is always of a brownishred colour when cold, and nearly black while warm. The powder of the native red oxide has a reddish-brown tint, and the colour of the red oxide prepared by exposing the precipitated protoxide or the carbonate to a moderate red heat, has most commonly an admixture of yellow, something like rhubarb, though of a deeper hue; but both of these acquire the red colour when heated to whiteness.

The red oxide manifests little tendency to pass into a higher degree of oxidation by abstracting oxygen from the atmo sphere, even by the aid of heat. Thus a portion of the red oxide, preserved for an hour at a low red heat, and freely exposed to the air at the same time, did not acquire any appreciable addition to its weight. The protoxide of manganese precipitated from the sulphate by an excess of pure potash, collected on a filter and washed, fully exposed to the air in its

moist state for twenty-four hours, and then heated in an open vessel to a moderate red heat, which was insufficient to decompose the deutoxide, lost only 0.218 per cent by subsequent exposure to a white heat. The quantity of deutoxide present, therefore, must have been very minute. The anhydrous protoxide, as already mentioned, always yields the pure red oxide when heated to redness in the open air. The carbonate, also, in similar circumstances, is converted into a red oxide containing but a very small proportion of the deutoxide. It will appear from these experiments that it is unsafe in analyses to heat the precipitated protoxide or carbonate to redness, and consider the product as the deutoxide; a practice which is calculated to lead analytical chemists into considerable errors, and indeed has actually done so. If it is wished to procure the deutoxide, the precipitate should be moistened with nitric acid, and then exposed to heat.

I have endeavoured to ascertain the composition of the red oxide by several methods. The first is by the combined agency of heat and hydrogen gas. In the first experiments 100 parts of pure red oxide, in being thus converted into the protoxide, lost. 6.802 and 6.817 parts of oxygen; but as the resulting green oxide, when put into dilute sulphuric acid, was found to contain a little red oxide, the loss in oxygen must be rather below the truth. To avoid this error I exposed 44 256 grains of red oxide to hydrogen gas and a white heat for the space of one hour, when the loss amounted to 3·153 grains on 7.125 per cent.

Judging by the increase in weight which the protoxide acquires when heated in the open air, 100 parts of the red oxide consist of 93.05 parts of protoxide and 6'95 of oxygen. According to a similar experiment made by Arfwedson, the red oxide is composed of 93.153 protoxide and 6.847 parts of

oxygen.

In an analysis already described, the carbonate of manganese was found to contain 56.853 per cent of the protoxide of manganese. When 100 parts of the same carbonate are exposed to air and a white heat, 61.18 parts of red oxide are obtained. From these data it may easily be calculated that the red oxide consists of 92.927 parts of protoxide, and 7·073 of oxygen.

As a mean of the numbers afforded by these three methods, it follows that the red oxide is composed of 92.951 parts of the green oxide and 7.049 of oxygen, or of 72-291 parts of metallic manganese and 27.709 of oxygen. According to M. Berthier*,

* Ann. de Chimie et de Physique, tom. xx.

who