Chapter XVII. (Note 43) we already pointed out this higher oxide of

cerium. As the above-mentioned elements are rather rare in nature, have but little practical application, and do not present any new forms of combination, it is unadvisable to dwell on them in this treatise.

_Titanium_ is found in nature in the form of its anhydride or oxide, TiO_{2}, mixed with silicon in many minerals, but the oxide is also found separately in the form of semi-metallic _rutile_ (sp. gr. 4·2). Another titanic mineral is found as a mixture in other ores, known as _titanic iron ore_ (in the Thuensky mountains of the southern Ural; it is known as _thuenite_), FeTiO_{3}. This is a salt of ferrous oxide and titanic anhydride. It crystallises in the rhombohedric system, has a metallic lustre, grey colour, sp. gr. 4·5. The third mineral in which titanium is found in considerable quantities in nature is _sphene_ or _titanite_, CaTiSiO_{5} = CaO,SiO_{2},TiO_{2}, sp. gr. 3·5, colour yellow, green, or the like, crystallises in tablets. The fourth, but rare, titanic mineral is _peroffskite_, calcium titanate, CaTiO_{3}; it forms blackish-grey or brown cubic crystals, sp. gr. 4·02, and occurs in the Ural and other localities. It may be prepared artificially by fusing sphene in an atmosphere of water vapour and carbonic anhydride. At the end of the last century Klaproth showed the distinction between titanic compounds and all others then known.[57]

[57] The compounds of titanium are generally obtained from rutile; the finely-ground ore is fused with a considerable amount of acid potassium sulphate, until the titanic anhydride, as a feeble base, passes into solution. After cooling, the resultant mass is ground up, dissolved in cold water, and treated with ammonium hydrosulphide; a black precipitate then separates out from the solution. This precipitate contains TiO_{2} (as hydrate) and various metallic sulphides--for example, iron sulphide. It is first washed with water and then with a solution of sulphurous anhydride until it becomes colourless. This is due to the iron sulphide contained in the precipitate, and rendering it black, being converted into dithionate by the action of the sulphurous acid. The titanic acid left behind is nearly pure. The considerable volatility of titanium chloride may also be taken advantage of in preparing the compounds of titanium from rutile. It is formed by heating a mixture of rutile and charcoal in dry chlorine; the distillate then contains _titanium chloride_, TiCl_{4}. It may be easily purified, owing to its having a constant boiling point of 136°. Its specific gravity is 1·76; it is a colourless liquid, which fumes in the air, and is perfectly soluble in water if it be not heated. When hot water acts on titanic chloride, a large proportion of titanic acid separates out from the solution and passes into metatitanic acid. A similar decomposition of acid solutions of titanic acid is accomplished whenever they are heated, and especially in the presence of sulphuric acid, just as with metastannic acid, which titanic acid resembles in many respects. On igniting the titanic acid a colourless powder of the anhydride, TiO_{2}, is obtained. In this form it is no longer soluble in acids or alkalis, and only fuses in the oxyhydrogen flame; but, like silica, it dissolves when fused with alkalis and their carbonates; as already mentioned, it dissolves when fused with a considerable excess of acid potassium sulphate--that is, it then reacts as a feeble base. This shows the basic character of titanic anhydride; it has at once, although feebly developed, both basic and acid properties. The fused mass, obtained from titanic anhydride and alkali when treated with water, parts with its alkali, and a residue is obtained of a sparingly-soluble poly-titanate, K_{2}TiO_{3}_n_TiO_{2}. The hydrate, which is precipitated by ammonia from the solutions obtained by the fusion of TiO_{2} with acid potassium sulphate, when dried forms an amorphous mass of the composition Ti(OH)_{4}. But it loses water over sulphuric acid, gradually passing into a hydrate of the composition TiO(OH)_{2}, and when heated it parts with a still larger proportion of water; at 100° the hydrate Ti_{2}O_{3}(OH)_{2} is obtained, and at 300° the anhydride itself. The higher hydrate, Ti(OH)_{4}, is soluble in dilute acid, and the solution may be diluted with water; but on boiling the sulphuric acid solution (though not the solution in hydrochloric acid), all the titanic acid separates in a modified form, which is, however, not only insoluble in dilute acids, but even in strong sulphuric acid. This hydrate has the composition Ti_{2}O_{3}(OH)_{2}, but shows different properties from those of the hydrate of the same composition described above, and therefore this modified hydrate is called _metatitanic acid_. It is most important to note the property of the ordinary gelatinous hydrate (that precipitated from acid solutions by ammonia) of dissolving in acids, the more so since silica does not show this property. In this property a transition apparently appears between the cases of common solution (based on a capacity for unstable combination) and the case of the formation of a hydrosol (the solubility of germanium oxide, GeO_{2}, perhaps presents another such instance). If titanium chloride be added drop by drop to a dilute solution of alcohol and hydrogen peroxide, and then ammonia be added to the resultant solution, a yellow precipitate of _titanium trioxide_, TiO_{3}H_{2}O, separates out, as Piccini, Weller, and Classen showed. This substance apparently belongs to the category of true peroxides.

Titanium chloride absorbs ammonia and forms a compound, TiCl_{4},4NH_{3}, as a red-brown powder which attracts moisture from the air and when ignited forms _titanium nitride_, Ti_{3}N_{4}. Phosphuretted hydrogen, hydrocyanic acid, and many similar compounds are also absorbed by titanium chloride, with the evolution of a considerable amount of heat. Thus, for example, a yellow crystalline powder of the composition TiCl_{4},2HCN is obtained by passing dry hydrocyanic acid vapour into cold titanium chloride. Titanium chloride combines in a similar manner with cyanogen chloride, phosphorus pentachloride, and phosphorus oxychloride, forming molecular compounds, for example TiCl_{4},POCl_{3}. This faculty for further combination probably stands in connection, on the one hand, with the capacity of titanium oxide to give polytitanates, TiO(MO)_{2},_n_TiO_{2}; on the other hand, it corresponds with the kindred faculty of stannic chloride for the formation of poly-compounds (Note 41), and lastly it is probably related to the remarkable behaviour of titanium towards nitrogen. Metallic titanium, obtained as a grey powder by reducing potassium titanofluoride, K_{2}TiF_{6}, (sp. gr. 3·55 K. Hofman 1893), with iron in a charcoal crucible, combines directly with nitrogen at a red heat. If titanic anhydride be ignited in a stream of ammonia, all the oxygen of the titanic oxide is disengaged, and the compound TiN_{2} is formed as a dark violet substance having a copper-red lustre. A compound Ti_{5}N_{6} is also known; it is obtained by igniting the compound Ti_{3}N_{4} in a stream of hydrogen, and is of a golden-yellow colour with a metallic lustre. To this order of compounds also belongs the well-known and chemically historical compound known as _titanium nitrocyanide_; its composition is Ti_{5}CN_{4}. This substance appears as infusible, sometimes well-formed, cubical crystals of sp. gr, 4·3, and having a red copper colour and metallic lustre; it is found in blast furnace slag. It is insoluble in acids but is acted on by chlorine at a red heat, forming titanium chloride. It was at first regarded as metallic titanium; it is formed in the blast furnace at the expense of those cyanogen compounds (potassium cyanide and others) which are always present, and at the expense of the titanium compounds which accompany the ores of iron. Wöhler, who investigated this compound, obtained it artificially by heating a mixture of titanic oxide with a small quantity of charcoal, in a stream of nitrogen, and thus proved the direct power for combination between nitrogen and titanium. When fused with caustic potash, all the nitrogen compounds of titanium evolve ammonia and form potassium titanate. Like metals they are able to reduce many oxides--for example, oxides of copper--at a red heat. Among the alloys of titanium, the crystalline compound Al_{4}Ti is remarkable. It is obtained by directly dissolving titanium in fused aluminium; its specific gravity is 3·11. The crystals are very stable, and are only soluble in aqua regia and alkalis.

The comparatively rare element _zirconium_, Zr = 90, is very similar to titanium, but has a more basic character. It is rarer in nature than titanium, and is found principally in a mineral called _zircon_, ZrSiO_{4} = ZrO_{2}.SiO_{2}, crystallising in square prisms, sp. gr. 4·5. It has considerable hardness and a characteristic brownish-yellow colour, and is occasionally found in the form of transparent crystals, as a precious stone called hyacinth.[58] Metallic zirconium was obtained, by Berzelius and Troost, by the action of aluminium on potassium zirconofluoride in the same way that silicon is prepared; it forms a crystalline powder, similar in appearance to graphite and antimony, but having a very considerable hardness, not much lustre, sp. gr. 4·15. In many respects it resembles silicon; it does not fuse when heated, and even oxidises with difficulty, but liberates hydrogen when fused with potash. When fused with silica it liberates silicon. With carbon in the electrical furnace it forms ZrC_{2}, with hydrogen it gives ZrH_{2} (like CaH_{2}, Winkler, Vol. I., p. 621); hydrochloric and nitric acids act feebly on it, but aqua regia easily dissolves it. It is distinguished from silicon by the fact that hydrofluoric acid acts on it with great facility, even in the cold and when diluted, whilst this acid does not act on silicon at all.

[58] The formula ZrO was first given to the oxide of zirconium as a base, in this case Zr = 45 whilst the present atomic weight is Zr = 90--that is, the formula of the oxide is now recognised as being ZrO_{2}. The reasons for ascribing this formula to the compounds of zirconium are as follows. In the first place, the investigation of the crystalline forms of the zirconofluorides--for example, K_{2}ZrF_{6}, MgZrF_{6},5H_{2}O--which proved to be analogous in composition and crystalline form with the corresponding compounds of titanium, tin, and silicon. In the second place, the specific heat of Zr is 0·067, which corresponds with the combining weight 90. The third and most important reason for doubling the combining weight of zirconium was given by Deville's determination of the vapour density of _zirconium chloride_, ZrCl_{4}. This substance is obtained by igniting zirconium oxide mixed with charcoal in a stream of dry chlorine, and is a colourless, saline substance which is easily volatile at 440°. Its density referred to air was found to be 8·15, that is 117 in relation to hydrogen, as it should be according to the molecular formula of this substance above-cited. It exhibits, however, in many respects, a saline character and that of an acid chloranhydride, for zirconium oxide itself presents very feebly developed acid properties but clearly marked basic properties. Thus zirconium chloride dissolves in water, and on evaporation the solution only partially disengages hydrochloric acid--resembling magnesium chloride, for example. Zirconium was discovered and characterised as an individual element by Klaproth.

Pure compounds of zirconium are generally prepared from zircon, which is finely ground, but as it is very hard it is first heated and thrown into cold water, by which means it is disintegrated. Zircon is decomposed or dissolved when fused with acid potassium sulphate, or still more easily when fused with acid potassium fluoride (a double soluble salt, K_{2}ZrF_{6}, is then formed); however, zirconium compounds are generally prepared from powdered zircon by fusing it with sodium carbonate and then boiling in water. An insoluble white residue is obtained consisting of a compound of the oxides of sodium and zirconium, which is then treated with hydrochloric acid and the solution evaporated to dryness. The silica is thus converted into an insoluble form, and zirconium chloride obtained in solution. Ammonia precipitates _zirconium hydroxide_ from this solution, as a white gelatinous precipitate, ZrO(OH)_{2}. When ignited this hydroxide loses water and in so doing undergoes a spontaneous recalescence and leaves a white infusible and exceedingly hard mass of _zirconium oxide_, ZrO_{2}, having a specific gravity of 5·4 (in the electrical furnace ZrO_{2} fuses and volatilises like SiO_{2}, Moissan). Owing to its infusibility, zirconium oxide is used as a substitute for lime and magnesia in the Drummond light. This oxide, in contradistinction to titanium oxide, is soluble, even after prolonged ignition, in hot strong sulphuric acid. The hydroxide is easily soluble in acids. The composition of the salts is ZrX_{4}, or ZrOX_{2}, or ZrOX_{2},ZrO_{2}, just as with those of its analogues. But although zirconium oxide forms salts in the same way with acids, it also gives salts with bases. Thus it liberates carbonic anhydride when fused with sodium carbonate, forming the salts Zr(NaO)_{4}, ZrO(NaO)_{2}, &c. Water, however, destroys these salts and extracts the soda.

The very similar element _thorium_ (Th = 232) was distinguished by Berzelius from zirconium. It is very rarely met with, in _thorite_ and _orangeite_, ThSiO_{4},2H_{2}O. The latter is isomorphous with zircon (sp. gr. 4·8).[59]

[59] Thorium has also been found in the form of oxide in certain pyrochlores, euxenites, monazites, and other rare minerals containing salts of niobium and phosphates. The compounds of thorium are prepared by decomposing thorite or orangeite with strong sulphuric acid at its boiling point; this renders the silica insoluble, and the thorium oxide passes into solution when the residue is treated with cold water, after having been previously boiled with water (boiling water does not dissolve the oxide of thorium). Lead and other impurities are separated by passing sulphuretted hydrogen through the solution, and the thorium hydroxide is then precipitated by ammonia. If this hydroxide be dissolved in the smallest possible amount of hydrochloric acid, and oxalic acid be then added, thorium oxalate is obtained as a white precipitate, which is insoluble in an excess of oxalic acid; this reaction is taken advantage of for separating this metal from many others. It, however, resembles the cerite metals (Chapter XVII., Note 43) in this and many other respects. The thorium hydroxide is gelatinous; on ignition it leaves an infusible oxide, ThO_{2}, which, when fused with borax, gives crystals of the same form as stannic oxide or titanic anhydride; sp. gr. 9·86. But the basic properties are much more developed in thorium oxide than in the preceding oxides, and it does not even disengage carbonic acid when fused with sodium carbonate--that is, it is a much more energetic base than zirconium oxide. The hydrate, ThO_{2}, however, is soluble in a solution of Na_{2}CO_{3} (Chapter XVII., Note 43). Thorium chloride, ThCl_{4} is obtained as a distinctly crystalline sublimate when thorium oxide, mixed with charcoal, is ignited in a stream of dry chlorine. When heated with potassium, thorium chloride gives a metallic powder of thorium having a sp. gr. 11·1. It burns in air, and is but slightly soluble in dilute acids. The atomic weight of thorium was established by Chydenius and Delafontaine on the basis of the ismorphism of the double fluorides.