first turbid, but in a few minutes afterwards quite transparent. It should be kept at a temperature of 140° (at which point little nitrous acid is lost) for some hours, with addition, if necessary, of more nitrite of soda and acetic acid from time to time, until all ferridcyanide has disappeared. When this is the case, the whole mixture may be boiled down, until, on cooling, it solidifies to a thick paste. The right state of concentration has been attained, when on beating the paste up, and squeezing in linen, a pale syrup, chiefly of acetate of potash, is expressed. The pearly-looking mass freed from acetate of potash, as far as possible, must be redissolved in such an amount of boiling water, that on cooling a large proportion of cyanide of mercury separates in white pearly scales, quite free from nitroprusside crystals. On squeezing in linen, a deep-red solution of nitroprusside is expressed, and a white, pearly mass of cyanide of mercury remains on the linen. On concentrating the red filtrate, a large crop of crystals of nitroprusside of sodium is obtained in a mother-liquid containing more or less cyanide of mercury, in pearly scales, easily separated by throwing the whole on a moderately coarse hair sieve, which will retain the prisms of nitroprussides of sodium, and allow the cyanide of mercury to pass through. The prisms may be washed quite clean by allowing the cyanide of mercury to settle down in the filtrate, and using the clear supernatant fluid for washing. The operation can, of course, be continued as far as it may be deemed profitable. If the cyanide of mercury is not wanted as such, it can be made to furnish hydrocyanic acid and corrosive sublimate for use again, by boiling with hydrochloric acid.

It may be added, in conclusion, that nitroprussides react well only with monosulphides. The more of a persulphide the solution contains, and the deeper the yellow colour, the less distinct is the reaction. This difficulty can be overcome by warming the yellow persulphide with sufficient cyanide of potassium to decolorise it, when the beautiful carmine of the monosulphide will be obtained.

XXXIII.-The Platinum-bases: the best mode of obtaining and identifying them; some new Compounds.

By EDWARD ASH HADOW, King's College.

ON first attempting to prepare the platinum-bases as specimens for Dr. Miller's lecture table, I encountered various perplexing difficulties, arising from the endless number of compounds that could be formed, and I felt the want of ready methods for recognising the class to which any unknown salt belonged, whether platosamine or diplatosamine, platinamine, or diplatinamine, or chlorodiplatinamine, as Gros' or Raewsky's salts. The source, too, of all these salts is always the green salt of Magnus, the preparation of which, in any quantity, is attended with the greatest uncertainty, and if too much sulphurous acid has been used to reduce the bichloride of platinum, is often a total failure. Then again, when it is obtained and used for forming Gros' and Raewsky's nitrate, half the platinum is wasted as chloride or nitrate. Disheartened by such a mode of proceeding, I determined to proceed by an entirely different method, and to convert a certain weight of metallic platinum into its equivalent of pure dry protochloride by heat in the usual manner. When the pro

tochloride thus obtained is digested in warm, moderately strong ammonia, it dissolves gradually but completely, excepting a little iridium and oxide of iron, giving a brownish solution. On filtering and concentrating, a large crop of fine prisms of hydrochlorate of diplatosamine is obtained, which is the source of all the other bases. These crystals contain an atom of water, N2H PtCl,HO, which is not usually given in books, and which, to my cost, I did not discover for some time, and graduating my permanganate solutions by the salt as if anhydrous, I met with most perplexing results.

Gros' nitrate (N2H¿PtCIO.NO) is obtained from this salt by pouring a hot solution of it into hot, moderately concentrated nitric acid; abundance of red fumes are evolved, and a thick paste of small crystals of Gros' nitrate forms at the bottom of the acid, in which it is almost insoluble; it can be drained on asbestos, and dried at about 280° Fahr.; when re-crystallised from water it gives brilliant flat prisms.

Raewsky's nitrate is obtained easily and abundantly from

VOL. XIX.

2 B

Gros' salt by boiling the latter for some hours with nitrate of silver and some nitric acid; but no excess of nitrate of silver can remove the chlorine from Raewsky's nitrate, and reduce it to a nitrate of diplatinamine, from which it differs totally in all its properties; analysis completely confirms the correctness of Gerhardt's formula for Raewsky's nitrate, NH12Pt2CIO. 2NO,,HO= 442.-Raewsky's own formula, NH12Pt2CIO. 2NO, is obviously impossible, Pt, being brought to the platinic state by ClO3, leaving O, in excess. The above being the formula for Raewsky's nitrate, its derivation from Gros' nitrate by nitrate of silver is simple enough; 2 atoms of Gros' nitrate unite, and losing an atom of chlorine, gain one of oxygen

12

2(N2H¿PƖCIO.NO5) + AgO.NO + HO=N2H12Pt,CIO3.2NO5 + AgCl + HO.NO, so that 2 atoms of Gros' nitrate, or 462 parts, require 1 atom of nitrate of silver for the decomposition; but in practice it is better to use rather more, or about 190 parts of the nitrate of silver. After filtering from chloride of silver, it is well to boil the solution again, to make sure that the transformation is complete. After the decomposition has ceased, the filtered solution is somewhat concentrated by evaporation and allowed to cool, when Raewsky's nitrate crystallises abundantly; it can be treated as Gros' salt, washing away the solution of nitrate of silver with diluted nitric acid, and drying at 280° F.

From the singular proportion that exists between the acid and the oxygen in Raewsky's nitrate, I had the curiosity to ascertain what the composition of the hydrochlorate would be, and therefore added chloride of ammonium in excess to Raewsky's nitrate, collected the precipitate of minute prisms thus produced, and washed with alcohol till clean. The hydrochlorate was found to have the composition N4H12Pt2Cl2O.

With regard to discovering the class to which a base belongs, platosamine and diplatosamine salts correspond, in their reducing action on permanganate of potash, to the protochloride of platinum, from which they are derived. The diplatosamine-salts are readily recognised by their property of yielding Magnus' green salt when treated with a solution of protochloride of platinum. Magnus' green is only one of a large series of double salts that hydrochlorate of diplatosamine is capable of forming with other metallic chlorides, such as corrosive sublimate and the chlorides of cadmium, palladium, tin, and copper. The double salts formed

with cadmium and with mercury were analysed, and found to have the same composition-1 atom of each salt. The diplatosamine-salts are further characterised by the beautiful blue or green precipitate or solution, which they furnish when a current of nitrous acid is passed into their strongly acidified solution. None of these reactions are exhibited by the platosamine-salts, which possess little interest.

The diplatinamine, chlorodiplatinamine, and platinamine salts have, of course, no reducing action on solutions of the permanganates. The two series of most interest, viz., those of Gros and of Raewsky, are readily distinguished. Chloride of ammonium gives with the salts of Gros' series, an insoluble, and with those of Raewsky's, a soluble chloride; a single drop of solution of sulphate of soda with a drop of solution of Gros' nitrate, on stirring, gives hairy tufts, which are deposited on the lines of stirring, but no change occurs when Raewsky's nitrate is similarly treated. But the most characteristic reaction of Raewsky's nitrate is that of giving, with a very dilute solution of protochloride of platinum strongly acidulated with nitric acid, a beautiful, coppery, moss-like precipitate, exactly resembling in appearance the platinidcyanide of potassium; Gros' nitrate has no such reaction.

The diplatinamine-salts are quite uninteresting; they give with chlorides, after some time, large crystals of the hydrochlorate; they may readily be recognised by being reduced by sulphurous acid to salts of diplatosamine, whereas platinamine-salts, by the same agent, are reduced to salts of platosamine.

The reaction between nitrous acid and acidulated salts of diplatosamine seems most singularly to have been overlooked. I have examined two of the compounds, one of which is blue and the other green. The blue compound is obtained with nitrate of diplatosamine (obtained from the hydrochlorate by double decomposition with nitrate of silver), by acidulating its solution strongly with nitric acid, and passing in nitrous acid vapours. A beautiful smalt-blue precipitate then falls in abundance, which, under the microscope, is seen to consist of dodecahedrons. The green precipitate is less beautiful, and is obtained by passing nitrous acid into hydrochlorate of diplatosamine strongly acidulated with hydrochloric acid. These precipitates can be drained on asbestos and afterwards washed with dilute hydrochloric and nitric acids respectively; they may be dried without visible change at 212°. Some very simple experiments served to show qualitatively the

composition of these precipitates. Placed in pure water, both precipitates soon dissolve, furnishing blue and green liquids respectively, smelling strongly of nitrous acid; on gently warming, the nitrous acid goes off, leaving unchanged diplatosamine-salts strongly acid with hydrochloric and nitric acids. If, instead of expelling the nitrous acid by heat, the corresponding acids be added to each solution, the original precipitate will be obtained unaltered. These precipitates appear, therefore, to consist, the blue of nitrate of diplatosamine with nitric and nitrous acids, and the green of the hydrochlorate of diplatosamine with hydrochloric and nitrous acids. The nitrous acid manifestly acts in these compounds the part of a base, nor is this the only instance in which it appears to do so. In the crystalline body of the leaden chambers, NO3.2SO3, as well as in Frémy's sulphammonates, it appears to act a basic part; further, the analogy between SbO3 and NO, would lead us to expect in the latter signs of a basic character. Peroxide of nitrogen has all the characters of a nitrate of the oxide NO,. It may be well to mention here that the instability of nitrous acid is in most chemical works greatly over-estimated, it being stated to be instantly decomposed by water. My own experience indicates quite the reverse, and if the water be acidulated, the nitrous acid appears to be remarkably stable.

These nitrous compounds were analysed by determining the amount of platinum in each, and then, by means of standard permanganate, determining the amount of oxygen capable of being absorbed. One atom of platinum (99), as diplatosamine, can only take up 1 atom (8) of oxygen. All absorbed beyond this is due to nitrous acid.

The blue nitrate dried at 212° gave 40.8 per cent. of platinum, and in one experiment it was found capable of absorbing 15:06 by weight of oxygen, for every one atom of platinum (99), evidently pointing to 16 parts, or 2 atoms, of oxygen; but as I atom of nitrous acid absorbs 2 atoms of oxygen, there can be only 1 atom of nitrous acid present to 2 atoms of platinum as diplatosamine: this points to the formula 2(N,H,PO.NO)

{2(
2(N2H

NO3.NO

= 482,

which is sufficiently confirmed by the following percentages :

Found. Calculated from Formula.

Oxygen absorbable =
Platinum

6.2 = 40·8

6.6

41.0