the crystals in the cold, but dissolution is complete on heating, with evolution of hydrogen. The crystals have the composition Al 71-06; Ti 2665; Si 2·19; loss (? C) 0·10, which agrees with the formula (Ti Si) Al, and hence it is probable that they are mixtures of the isomorphous compounds, TiAl, and SiAl.

If zinc or magnesium is substituted for aluminium, no crystals are obtained. C. H. B.

Titanium Trioxide. By A. CLASSEN (Ber., 21, 370-372).-The action of hydrogen peroxide on titanium dioxide (compare Piccini, Abstr., 1882, 808; 1883, 1054; Weller, Abstr., 1883, 295) has been studied by the author, who recommends the following method for obtaining the product :-Pure titanium chloride is added drop by drop to dilute alcohol, and the clear and very dilute solution is treated with a large excess of hydrogen peroxide. Ammonia, ammonium carbonate, or aqueous potash is added to the solution with the production of a yellow or, in the case of ammonia, of a reddish-yellow liquid, which after some time yields a yellow precipitate. This is allowed to subside, the clear solution siphoned off, and the precipitate repeatedly washed by decantation; the compound, however, tends to retain water and salts in considerable quantities. When dried on a tile, it approximates to the composition TiO, + 3H2O.

W. P. W.

Antimony Pentachloride. By R. ANSCHÜTZ and P. N. EVANS (Proc. Roy. Soc., 42, 379-387; compare Trans., 1886, 379).—The authors find that, contrary to the statement of Daubrawa (this Journal, 1877, ii, 406), SbOCl, is not formed when water is added to antimony pentachloride, nor is any hydrogen chloride evolved. The antimony pentachloride is best dissolved in chloroform, and the calculated amount of water added. Under these circumstances a crystalline substance, SbCl, H2O, soluble in chloroform, is obtained, melting at 87-92°. It is very hygroscopic, and diliquesces to a clear liquid, which crystallises over sulphuric acid in broad crystals, described by Daubrawa as the oxychloride. When distilled, it gives SbCl,,Sbci, and a waxy residue.

A chloroform solution of antimony pentachloride, when heated with water in a sealed tube, gives antimony trichloride and phosgene gas. Phosgene gas is also produced by heating a chloroform solution of the monohydrate at 100°.

Antimony pentachloride tetrahydrate can be produced in the same way as the monohydrate. It is a crystalline mass, insoluble in chloroform. By adding anhydrous oxalic acid to a chloroform solution of antimony pentachloride, the authors obtain a substance, Sb2Cl,C2O1, probably CO2(O.SbCl)2. It crystallises from chloroform in tables melting at 148·5-149°, and is decomposed by water with liberation of oxalic acid.

The difference in behaviour with carbon compounds between phosphorus pentachloride and antimony pentachloride is attributed to the property of the latter of combining with water instead of decomposing it. H. K. T.

4

Method.

Redetermination of the Atomic Weight of Platinum. By W. DITTMAR and J. MCARTHUR (J. Soc. Chem. Ind., 6, 799–803).— The value Pt 1948 which Seubert (Abstr., 1881, 514) deduced from his analyses of platinochlorides, is too low; his own analyses, if properly interpreted, show that the true value lies, by a considerable fraction of a unit, higher. According to the authors' analysis of potassium platinochloride, the true "Pt," although probably a shade below, lies close to 1955. Taking "Pt" as meaning the number which must be substituted for Pt in the calculation of the ratios 2KC1: PtCl,K,; 2KC1: Pt, &c., in order to obtain the correct factors for reducing analytically obtained platinochloride to potassium chloride, &c., even the number 195.5 is too low, 196 affording in general a better approximation. But Pt if taken in this sense is no constant at all. Those factors must be determined directly by standard experiments. The results of the authors' own standard experiments are given and contrasted with the theoretically calculated ratios in the subjoined table. The entries "Ta" refer to Tatlock's methods; those "F" to the authors' form of Finkener's method described in detail in the original paper, and those marked "N" to the usual platinum process for the determination of ammonia :

Notes. (1) Refers to the potassium chloride in the substance, (2) to that in the platinochloride precipitate, and (3) and (4) to the ammonium chloride to be determined, not to that contained in the platinochloride precipitate.

D. B.

Hydroxylamine Platinum Bases. By H. ALEXANDER (Chem. Centr., 1887, 1254-1255).-As the platinum bases mentioned below are all explosive, the platinum and chlorine cannot be determined by heating the compound. The platinum was determined by moistening a quantity of the substance with concentrated sulphuric acid in a platinum capsule, evaporating off the acid and finally dispelling any remaining acid by the addition of ammonium carbonate. The chlorine was determined by distilling the compound with sulphuric acid, and passing the resulting hydrogen chloride into a solution of silver nitrate.

Platoso-dihydroxylamine hydrochloride, Pt(NH,O-NH2OCl)2, which Lössen has already described, is obtained by the action of potassioplatinous chloride on hydroxylamine hydrochloride. Strong bases precipitate platoso-dihydroxylamine hydroxide, Pt(OH)2,4NH,Ŏ, from solutions of the chloride. The oxalate, PtC2O,,4NH,O, is prepared from neutral potassium oxalate and the chloride; an acid salt does not seem to exist. Platino-dihydroxylamine sulphate, PtSO,,4NH3O+ H2O, can be prepared from this oxalate by the action of weak sulphuric acid; the sulphate is more easily prepared from the free base and sulphuric acid. A phosphate was obtained, but no nitrate.

Platoso-dihydroxylamine hydrochloride platinous chloride is prepared by the addition of platinous hydrochloride to the chloride.

Platoso-hydroxylamine hydrochloride, Pt(NH3OCI)2, is formed by the action of hydrochloric acid, on either platino-dihydroxylamine hydrate when warm, or on the corresponding chloride.

The compound formed by the action of ammonia on platinous hydrochloride and on platoso-hydroxylamine hydrochloride, which has been described by Jörgensen, the author regards as platosohydroxylamine ammonium chloride, Pt(NH,O.NH Cİ)2.

The double salt, Pt(NH,O-NH,Cl)2, PtCl2, is obtained from platinous hydrochloride and platoso-hydroxylamine ammonium chloride, as well as from the above-named mixed chlorides and potassio-platinous chloride.

Free hydroxylamine acting on platinous chloride gives rise to the compound OH PtC1,4NH30 + 2H2O, which explodes at 140— 150°. Another compound which may be regarded as platinum nitrogen chloride, PtNCI, together with the double salt,

2(OH PtC1,4NH,O) + Pt(OH)2,2NH,O, or 2[Pt(OH)2,4NH30] + PtCl1⁄2 + 2ÑH ̧0.

were also obtained.

Free hydroxylamine acting on platinous hydrochloride gives rise to a compound corresponding approximately with the formula PtCl2,4NH3O, and by its action on potassio-platinous chloride to a base Pt(OH)2,2NH,O, which has not been obtained quite pure as yet.

J. P. L. Ruthenium Oxides. By H. DEBRAY and A. JOLY (Compt. rend., 106, 100-106).-The authors prepared crystallised ruthenium dioxide by heating the amorphous oxide in a vacuum or by heating the metal in oxygen. It forms quadratic prisms with the faces m, h', a', and b'. The ratios being b: h:: 1000: 692:43, and hence is isomorphous with cassiterite and rutile. The faces in the zone m h' are often striated as in those minerals, and very frequently the crystals are macled after the cassiterite and rutile types.

When heated to bright redness in a muffle, ruthenium absorbs oxygen, but after some time absorption takes place very slowly, and the metal is never fully oxidised to the dioxide. If, however, the product is powdered and again heated, the dioxide is obtained crystalline, with an indigo-blue colour. The authors were unable to obtain the sesquioxide in the manner described by Claus.

When ruthenium is heated in oxygen at a temperature above the

melting point of silver, it is entirely converted into crystalline products, and a portion volatilises and condenses in crystals. If the current of gas is rapid, the odour of ozone or ruthenium peroxide is perceived, and a certain quantity of the peroxide can be condensed in a flask cooled by ice. The inside of the tube is lined with ruthenium dioxide and a small quantity of a black substance which seems to contain more oxygen than the dioxide. The products are similar to those obtained when ruthenium peroxide and nitrogen are passed through a red-hot porcelain tube, and they are distributed in the same manner. The dioxide is found in the cooler parts of the tube, which indicates that the peroxide formed at about 1000° decomposes at a lower temperature. At temperatures above 1000°, ruthenium dioxide has a considerable tension of dissociation, and in a vacuum is partially reduced to the metal, a small quantity of the peroxide being formed. At a bright red heat, the phenomena are similar.

The authors were unable to obtain an oxide lower than the dioxide. Ruthenium peroxide is formed at about 1000°, and decomposes with explosion when cooled to 108°, but can be isolated by rapid cooling. It affords another instance of a compound which is decomposed by heat, and yet is formed at a high temperature. The crystallisation of the dioxide is a phenomenon of apparent volatilisation, and the formation of the peroxide is analogous to the formation of ozone, silicon hexachloride, silver oxide, hydrogen selenide or telluride, &c. The formation of ruthenium peroxide, like the decomposition of water, is endothermic. C. H. B.

Mineralogical Chemistry.

A New Zealand Sulphur Island. By R. W. E. MACIVOR (Chem. News, 56, 251-253).-White Island, in the Bay of Plenty, New Zealand, is part of the crater of a huge, submerged, conical volcano. Part of the island disappeared during the volcanic disturbances in New Zealand a year or two ago. The mineralogical and other characters of the place in 1883 were such as are general in volcanic regions; special features were heaps of gypsum with from 30 to 85 per cent. of sulphur, and curious hollow spheres, some as large as old-fashioned bombshells, consisting of an exterior crust of gypsum and a little sulphur, with transparent crystals of prismatic sulphur inside. There were active geysers and fumeroles sending out hydrogen chloride and steam, and ejecting with great violence transparent lumps of sulphur varying in colour from orange to a reddish-brown, although the sulphur deposited around the orifices was pale yellow and opaque; the depth of colour is due to selenium; some specimens contain as much as 1.75 to 21 per cent., which render it unfit for the manufacture of sulphuric acid. In the neighbourhood of the fumeroles there were many large holes full of thin, black, boiling mud consisting of clay, siliceous matter, gypsum, and iron pyrites suspended in strongly acid

water containing more mineral matter than the water from the lake. Lake Hope, in the midst of the central amphitheatre, had a fairly uniform depth, but was unfathomable near the centre; its temperature was 68°; it was muddy in the centre, but transparent, and of bluegreen colour elsewhere; the atmosphere over it was very irritating owing to the constant evolution of hydrogen chloride; the bottom of the lake wherever visible was covered with oblique crystals of selenite. Analysis of the water yielded the following results per 1000 parts:

FeSO4. Al2(SO4)3. CaSO4. MgSO4. K&SO4. NagSO4. Al Cl6. 15.254 1.350 0.931 4-715 10.033

25.557

3.605

It is suggested that the sulphur and hydrogen chloride result from the action of sea-water on hot beds of pyrites inside the mountain. D. A. L.

Graphite from the Bagoutal Mountains, Siberia. By W. H. COLLINS (Chem. News, 57, 36).-Four samples of graphite from this region were analysed and gave the following mean results per cent. :C, 38 91; SiO2, 38.83; Al2O3, 13-86; CaO and MgO, 2·12; Fe2O3, 4·52; loss and volatile matter, 177. This graphite is employed for crucible making with excellent results. It is somewhat similar in character to the graphite from the Stephanovsky mine. D. A. L.

Gold Quartz from the Transvaal. By P. HOLLAND (Chem. News, 56, 271-272).-The specimen weighed 18 grams; it was a greenish-grey, dark-coloured, massive quartz with iron-stained fissure. In sections, under the microscope, F. Rutley detected "allotriomorphous" quartz crystals, cubes of pyrites, gold, numerous liquid lacunæ, particles of mica, and patches, probably, of felspar. On heating in an open tube, sulphurous anhydride was given off. For analysis the mineral was decomposed with hydrofluoric acid, and gave the following numbers :

SiO2. Al2O3. Fe2O3. FeS2. Au. MgO. KO. NagO. HO. 92.86 2.86 1.22 0.27 0.03 0.18 1:40 041 0.58

C.

a little. D. A. L.

Braunite from Jakobsberg. By M. SCHUSTER (Zeit. Kryst. Min., 13, 621-623, from Tschermak's Min. Mitth., 7, 443-451).—This mineral, discovered by Igelström (Abstr., 1887, 643) at the Jakobsberg manganese mines, Wermland, Sweden, has been subjected by the author to a careful crystallographical examination. Braunite, he finds, is not tetragonal, as stated by Igelström, but rhombohedral, tetartohedral, and isomorphous with iron glance and titaniferous iron B. H. B.

ore.

Chrome Iron Ore in Australasia. By R. W. E. MACIVOR (Chem. News, 57, 1—2).—Chrome iron ore is found in New Zealand in two or three places, but chiefly near Nelson, in the middle island,