193 the residue kept at a dull red heat and then lixiviated with water. Alkaliferous oxide of iridium, Ir 2 3 , remains as a blue-black powder, which needs only be heated in hydrogen to be reduced to metal, from which the alkali is now easily removed by washing with water. Such iridium is always contaminated with more or less osmium, ruthenium, rhodium, and platinum, to remove which the crude metal is fused up with ten parts of lead, and the alloy treated with dilute nitric acid to dissolve the bulk of the lead, when the polyxene metals remain in the shape of a black powder. From this the platinum is extracted by prolonged treatment with dilute aqua regia, and from the residue the rhodium by fusion with bisulphate of potash and subsequent treatment with water, which dissolves away the sulphate of rhodium formed. The residue now left is fused in a gold crucible with ten parts of caustic and three of nitrate of potash, when the ruthenium and osmium assume the form of soluble Me0 3 K 2 salts, which are extracted with water and thus removed. What remains is an alkaliferous (blue) sesquioxide of iridium, which as a rule still retains some iron, ruthenium, and traces of gold and silica (G. Matthey). For the iinal purification of the metal and the recovering of the ruthenium and rhodium see G. Matthey s memoir (Ghem. tioc. Journ., 1879, Abstr., p. 772) and chemical handbooks. The osmium, as already stated, is obtained at an early stage of the process in the shape of a solution of its volatile tetroxide in caustic potash. This solution is mixed with a little alcohol to bring the osmium into the state of osmite, K^O + rtOsO.,, which is insoluble in alcohol. This precipitate is digested in sal-ammoniac, to convert it into a yellow compound of the composition 2NlI 4 Cl + OsO.,(]SrH 3 ) 2 , which latter needs only be heated in hydro gen to be converted into finely divided metallic osmium. 2. The second residue consists of a solution of a variety of polyxene chlorides in sal-ammoniac. This liquor is kept in contact with metallic iron, when the dissolved polyxene metals, and any gold or copper present, come down as a black heavy precipitate. This precipitate includes all the palladium and part of the rhodium as principal components. Bunsen has worked out an exhaustive method for the extracting of all its polyxene metals in pure forms ; but it is too complicated to be reproduced here. 1 The customary method for extracting the palladium is to treat the metallic preci pitate with aqua regia, which dissolves the palladium and platinum along with some of the iridium and rhodium, to filter, evaporate the residue to a syrup (for bringing the palladium into the form of PdCl 2 ), redissolve and precipitate the palladium by addition of the exact (juantity of mercuric cyanide as cyanide Pd(NC). 2 . This cyanide needs only be ignited strongly to leave a residue of metal. But this metal includes at least part of the copper of the original material. To remove it and other impurities, the crude metal is dissolved in hydrochloric acid with the help of free chlorine, and the solution next evaporated to dryness to reduce the PdCl 6 H 2 to l dCl 2 . The chloride is redissolved, the solution mixed with enough of ammonia to redissolve the precipitate first produced, and Hydrochloric acid gas is now passed into the solution. Yellow palladiochloride of ammonium, PdCl 4 (XH 4 ) 2 , is precipitated, while copper and iron remain dissolved. After removal of the mother liquor the double salt is ignited and thus converted into palladium- sponge, which is easily fused up in the oxyhydrogen flame and thus brought into the form of regulus. Palladium, a silver-white metal of great ductility, is much used, notwithstanding its high price, in mechanical dentistry and occasionally also for the graduated limbs of theodolites and other instruments, because, unlike silver, it remains bright in sulphur etted hydrogen. Of all the properties of this metal the most remarkable is its extra ordinary power of "occluding" hydrogen. According to Graham (to whom we owe almost all our knowledge on the subject) the compact metal when immersed in cold hydrogen gas takes up none or at most very little of it ; but at higher temperatures very con siderable occlusions take place. A certain specimen of foil was found to occlude 526 volumes of the gas at 245 C., and 643 at 90 to 97 C. , measured at 17 5 to 18 and one atmosphere s pressure, per unit-volume of metal. The hydrogen, as in the case of platinum, is retained on cooling, and from the cold compound cannot be extracted by means of an absolute vacuum, which re- extracts the gas at a red heat. Far more striking results can be obtained by using palladium as a negative pole in the electrolysis of (acidulated) water. The coefficient of occlusion then assumes very high values ; in Graham s hands it attained its maximum when the palladiunTwas produced electrolytically from a 1 6 per cent, solution of its chloride, and thus hydrogenized while itself in the nascent state. The galvani- cally deposited sheet was found to contain 982 volumes of hydrogen (measured cold) per unit-volume of original metal, corresponding i Jaltresb. d. C/iemie, 1SG8, p. 280; Ann. d. Chemie, vol. cxlvi. 205. approximately to the formula Pd 4 H 3 for the compound. When palladium unites with (nascent or free) hydrogen it suffers a very appreciable expansion which on the removal of the hydrogen is followed by a contraction beyond the original volume of the plain metal. This can be most beautifully illustrated by electrolysing water in an apparatus in which the negative electrode consists of a long strip of palladium-foil of which one side is covered over with varnish or electrolytically deposited platinum. The hydrogen goes in at the bare side of the electrode ; this side consequently expands more strongly than the other and the originally straight strip of metal becomes curved. When the current is reversed, hydrogen bubbles at once rise from what is now the negative pole, but the oxygen due at the palladium plate is for a time taken up by the hydrogen occluded there ; this hydrogen is gradually consumed, and as it diminishes the plate unbends more and more completely and at last gets bent over in the opposite sense. Palladium by being hydrogenized does not lose any of its metallic properties, but (in the case of complete saturation) its density sinks from 12 38 to 1179, its tenacity to 82 per cent, of its original value, its electric conductivity in the ratio of 8 1 to 5 "9. Graham views hydrogenized palladium as a true alloy, containing its hydrogen in the form of a metal "hydrogenium. He found that certain palladium alloys take up hydrogen as readily (though less abundantly) as the pure metal does with corresponding expan sion, but when dehydrogenized shrink back into exactly their ori ginal volume. He calculated that the density of hydrogenium lies somewhere about the value 0733 (water=l), which of course means only that the weight of the occluded hydrogen, measured by the weight of a volume of water equal to the expansion observed, is = 733. Dewar arrived at 620 as being probably nearer the truth, and for the specific heat of hvdrogenium found values from 379 to 5 88. Osmium. According to Deville and Debray, powdery osmium is most readily obtained by mixing the vapour of the tetroxide with that gas (CO + C0 2 ) which is prepared by the decomposition of oxalic acid with oil of vitriol, and passing the mixture through a red-hot porcelain tube. The powdery metal readily fuses up with 3 or 4 parts of tin into a homogeneous alloy. When this alloy is treated with hydrochloric acid most of the tin dissolves, and the rest of it can be driven off by heating the residue in HC1 gas. There remains ultimately pure osmium in the form of blue crystals endowed with a grey to violet reflex, which are hard enough to scratch glass. Their specific gravity is 22 "48, so that osmium, besides being the most infusible of metals, is the heaviest of all known bodies. Osmiridium. Native osmiridium forms crystalline plate-shaped grains, distinguished by an extraordinary degree of hardness, which certainly exceeds that of hard-tempered steel. Most of the grains are very minute ; the larger ones are utilized for making the so-called "diamond points" of gold pens. Osmiridium would lend itself for endless other applications if it were possible to unite the native dust into large compact masses. From a series of articles in the Chemical Neius (Jan. 2, 9, and 16, 1885), by Kelson W. Perry,, it would appear that this problem has been solved, in a sense. John Holland, an American pen-maker, starting from the long- known fact that platinum metals readily unite with phosphorus into relatively easily fusible alloys, succeeded in producing a phos- phorized osmiridium which can be cast (and pressed while liquid) into thin continuous slabs even harder than the native substance, and susceptible of being wrought into drills, knife-edges, &c. Statistics. The production of platinum-ore in Eussia was 2327 kilogrammes in 1862, 492 in 1863, 397 in 1864, 2273 in 1865, 1768 in 1867, and 2050 in 1871, a total in those six years of 9307. The average production of platinum metal, from 1828 to 1845, amounted to 2623 - 8 kilogrammes per annum. In 1870 it was only 2005 8 kilos, of which about 80 per cent, came from the Ural Mountains. 2 The manufacture of platinum utensils is in the hands of a very few firms, of which that of Messrs Johnson, Matthey, & Co. of London is generally understood to be the most important. Even the total amount of metal which passes through these works in the aggregate is difficult of ascertainment, the more so as some of them at least are discounting large reserves of old metal, including more or less of the obsolete coins. According to an approximate estimate which a very competent authority has kindly furnished, the consumption during the last five years fell little short of 100,000 It troy, 3 of which from 75 to 80 per cent, are believed to have passed through the hands of London manufacturers. The price of the metal during the last ten or twelve years has ranged from four to eight times that of silver. It is very high at present (1885) in consequence of the constantly increasing demand for platinum utensils. (W. D. ) 2 From Kamarsch and Heeren s Tcchnisches WorterbucJi. 3 Equal to 7464 kilogrammes per annum, which is 37 times the amount given above for 1*7.
XIX. zs