nine inches long. From it passed to the pneumatic trough a bent tube, of which the extremity opened, above the level of the water, under a narrow graduated jar, so adjusted as that it should rise as gas passed into it. The pure metallic copper having been first fully ignited, heat was applied to the extremity of the tube, and the nitric acid of the salt was so completely deoxidized by the red hot copper, that the gas which passed over contained no sensible trace of nitric oxide. When the operation was concluded the apparatus was allowed to cool, and the proper correction being made for the change of temperature which had taken place in the air of the room during the experiment, the volume of the gas collected in the graduated jar was considered to represent the quantity of nitrogen which the substance contained. After the proper corrections for temperature, pressure, and moisture, it measured 2.319 cubic inches, weighing 0.7452 grains, or 4.31 per cent.

By this method the composition of the basic nitrate of palladium is found to be

These results lead to the simple formula NO, + 4 Pd O + 4

A. When a solution of nitrate of palladium is heated with a considerable excess of water of ammonia, or when ammoniacal gas is passed into the solution, a colourless liquor is ultimately obtained, the precipitate which first forms being perfectly redissolved. By careful evaporation, this solution deposits a pure white salt in rhombic crystals (prisms or plates), of a brilliant pearly lustre. When heated, this salt fuses, and deflagrates violently with a brilliant white flame, depositing metallic palladium, and evolving water and nitrogen gas. The existence of this salt and its property of so exploding has been long known, and its composition only remained to be determined by me.

I. 20.696 grains of this salt in good chrystals were mixed with powdered glass, and placed in a tube: in front of the mixture was put oxide of copper, and in front of that again clean metallic copper, in

thin turnings. The latter part of the tube having been heated to redness, heat was applied to that part of the tube containing the mixture, and the gas and watery vapour brought into contact with the oxide and metallic copper. The water was collected in a chloride of calcium tube, the last portions being obtained by breaking open the far extremity of the analysis tube which had been formed to a point for the purpose, and drawing by the mouth a current of air through the apparatus: the quantity of water was 8.180, or 39.53 per cent., containing 4.39 of hydrogen.

11. 16.036 grains were mixed with powdered glass, and heated in a platinum crucible, at first gently, until all volatile products had heen expelled, but then to dull redness. The weight of the glass used being known, it was found that there remained of metallic palladium 5.725, or 35.71 per cent.

III. 16-237 grains being mixed with powdered glass and placed in a tube with oxide of copper and metallic copper as in the last experiment, but from which a narrow tube passed to a graduated jar in the water-pneumatic trough, the decomposition was effected, and the volume of the nitrogen gas evolved determined as already noticed in the analysis of the basic nitrate. It measured 14.7075, cubic inches, weighing 47149 grains, equivalent to 29:04 per cent. The composition of this salt is, therefore,

Palladium.........

35.71

The rational formula of this salt is, therefore, NO, . N H, O + Pd. N H2, and it is completely analogous in constitution to the corresponding ammonaical nitrate of copper.

B. If the quantity of ammonia added to the solution of nitrate of palladium be not sufficient to redissolve all of the brown basic nitrate which is first formed, a deep yellow liquor is produced, which deposits on standing, or by moderate concentration, small yellow crystals, whose form appears to be rhombic-octohedral. When heated, these crystals deflagrate, producing water and nitrogen gas with copious fumes of nitrous acid, and some clouds of nitrate of ammonia. The residue after ignition is metallic palladium.

The analysis of this body was conducted as follows:

I. 6:065 grains were ignited in a tube of Bohemia glass, and the volatile products having been conducted over red-hot metallic copper to decompose any red fumes, were passed through a tube containing

chloride of calcium, which collected water amounting to 1.219 grains, or 20.10 per cent., corresponding to 2.23 of hydrogen in 100'00.

The tube being cut the residual palladium was found to weigh 2.623 grains, or 43.6 per cent.

II. 5.016 grains ignited in a platinum crucible gave 2.285, or 45.54 per cent. of metal.

III. 5.571 grains gave when treated exactly in the manner described in the preceding salt, nitrogen gas, which after correction measured 4.211 cubic inches, weighing 1.355 grains, or 24:10 per cent. The decomposition did not appear to proceed so regularly, nor to give so pure gas as in other cases.

From these experiments this compound appears to contain in 100 parts,

The formula to which these numbers lead is is NO, + Pd. N H2, which gives

If this formula were perfectly established it would constitute the first example of a direct combination of a metallic amidide with an acid. The only other supposition at all possible is, that it may contain the elements of an atom of water; the formula should then become NO, + Pd O. NH3, and correspond to the yellow crystalline form of the ammonia-chloride Pd. Cl + NH3, or the yellow ammonia-sulphate, S O,. Pd O+ NH3. Adopting this formula, the compositior of the salt should have been

And with this the hydrogen and nitrogen determinations might be considered as perfectly agreeing. However, from the irregular way in which this salt is decomposed I attach little value to the nitrogen result, and, on the other hand, the quantity of palladium found exceeds that given by the formula as last written, too much to be

accounted for by any error of experiment, as the crystals analysed were very pure. I hence consider that although the formula NO, + Pd O. N H, agrees best with our received ideas, and with the analogies of other compounds, yet the former, NO, + Pd. NH2, is more strictly deducible from the analytical results, and deserves at least provisional adoption until an opportunity presents itself of obtaining more decisive results, for I have found that this substance is not formed in all cases when its elements are brought together. I have procured it crystallized and sufficiently pure for analysis but twice, and then only in such small quantity as limited my power of experiment, as has been already seen.

Double Oxalate of Palladium and Ammonia.

This salt, which may be formed by adding oxalic acid to a solution of any colourless ammonia-salt of palladium, or by dissolving the freshly-percipitated hydrated oxide, or basic carbonate in a solution of binoxalate of ammonium, crystallizes in very beautiful bronzeyellow rhombic prisms, which are of two kinds, differing in the quantity of water of crystallization which they contain. I shall describe them as being, the one form in short prisms, the other in needles.

The short prismatic salt gave on analysis the following results :31.225 grains gave, when very gently heated until the evolution of watery vapour ceased, 27 624 of dry salt, being 88.43 per cent. This dry substance being then ignited there remained 9.429 grains of metallic palladium, or 30 20 per cent.

A quantity of the salt having been mixed with oxide of copper and introduced into a tube of German glass, with clean metallic copper in front, it was burned as for an organic analysis, and the relative volumes of nitrogen and carbonic acid in the gaseous mixture, which came over, were determined in the usual way. This relation varied very much in the different tubes full of gas, showing that the constituents of the salt were not uniformly acted on by the heat, but the relation of the total volume of nitrogen to the total volume of carbonic acid was very exactly 1 : 4.

These results show that the formula of this salt is Pd. O. C2 03 + NH, O. C2 O3 + 2 Aq, giving

The salt crystallizing in long needles, gives, when dried, about 70 per cent. of dried salt, having lost 30 per cent. of water. By ignition it yields from 23 to 23.26 of metallic palladium. These

numbers point out that it contains eight atoms of water of crystallization, from which the numbers should be 31.13 of water, and 23.04 of metal.

SECTION II.-PLATINUM COMPOUNDS.

Protoxychloride of Platinum.

During experiments conducted for the purpose of preparing sulphate of the deutoxide of platinum, and the results of which I shall on another occasion more specially describe, I had occasion to remark the formation of the substance now in question, and to determine its properties and composition. The bichloride of platinum being boiled with strong sulphuric acid in a retort, nearly to dryness, much muriatic acid gas is driven off, and on washing out the residue with water, a black powder remains undissolved, which may be collected on a filter. It is anhydrous. At a red heat it evolves oxygen and chlorine, leaving metallic platinum. In solution of potash it appears to dissolve. By ammonia it is converted into a fulminating substance, of which the analysis is not yet completed. By muriatic acid it is converted into the ordinary protochloride of platinum. Its analysis was effected as follows::

29.402 grains were perfectly fused with an excess of carbonate of soda. When the saline matter was dissolved out with water, metallic platinum remained which weighed 25 620 grains, or 87.14 per cent. The solution having been precipitated by nitrate of silver, after being acidulated by nitric acid, gave 8:995 of chloride of silver, or 30.55 per cent., containing 7.54 of chlorine.

These results lead to the formula Pt Cl + 3 Pt O, which gives

The material which is dissolved off from this black powder is by no means a pure sulphate of platinum. It contains also bichloride, the decomposition of which appears to have a definite limit regarding sulphuric acid.

Action of Ammonia on Biniodide of Platinum.

When the biniodide of platinum is digested in water of ammonia, either cold or hot, it becomes gradually changed to a bright cinnabar red powder, which is insoluble in water. This powder may also be formed by adding water of ammonia to a solution of the double iodide of platinum and potassium. The liquor becomes nearly colourless, and the red substance precipitates.