Chapter XXII.) acidulated with acetic acid gives a salt which after

drying at 100° has the composition Co_{2}O_{3}10NH_{3}7MoO_{3}3H_{2}O. After ignition this salt leaves a residue having the composition 2CoO_{7}MoO_{3}. An analogous compound is also obtained for tungstic acid, having the composition Co_{2}O_{3}10NH_{3}10WO_{3}9H_{2}O. In this case after ignition there remains a salt of the composition CoO_{5}WO_{3} (Carnot, 1889). Professor Kournakoff, by treating a solution of potassium and sodium molybdates, containing a certain amount of suboxide of cobalt, with bromine obtained salts having the composition: 3K_{2}OCo_{2}O_{3}12MoO_{3}20H_{2}O (light green) and 3K_{2}OCo_{2}O_{3}10Mo_{3}10H_{2}O (dark green). Péchard (1893) obtained salts of the four complex phosphotungstic acids by evaporating equivalent mixtures of solutions of phosphoric acid and metatungstic acid (_see_ further on): phosphotrimetatungstic acid P_{2}O_{5}12WO_{3}48H_{2}O, phosphotetrametatungstic acid P_{2}O_{5}16WO_{3}69H_{2}O, phosphopentametatungstic acid P_{2}O_{5}20WO_{3}H_{2}O, and phosphohexametatungstic acid P_{2}O_{5}24WO_{3}59H_{2}O. Kehrmann and Frankel described still more complex salts, such as: 3Ag_{2}O_{4}BaOP_{2}O_{5}22WO_{3}H_{2}O,5BaO_{2} K_{2}OP_{2}O_{3}22WO_{3}48H_{2}O. Analogous double salts with 22WO_{3} were also obtained with KSr, KHg, BaHg, and NH_{4}Pb. Kehrmann (1892) considers the possibility of obtaining an unlimited number of such salts to be a general characteristic of such compounds. Mahom and Friedheim (1892) obtained compounds of similar complexity for molybdic and arsenic acids.

For tungstic acid there are known: (1) Normal salts--for example, K_{2}WO_{4}; (2) the so-called acid salts have a composition like 3K_{2}O,7WO_{3},6H_{2}O or K_{6}H_{8}(WO_{4})_{7},2H_{2}O; (3) the tritungstates like Na_{3}O,3WO_{3},3H_{2}O = Na_{2}H_{4}(WO_{4})_{3},H_{2}O. All these three classes of salts are soluble in water, but are precipitated by barium chloride, and with acids in solution give an insoluble hydrate of tungstic acid; whilst those salts which are enumerated below do not give a precipitate either with acids or with the salts of the heavy metals, because they form soluble salts even with barium and lead. They are generally called metatungstates. They all contain water and a larger proportion of acid elements than the preceding salts; (4) the tetratungstates, like Na_{2}O,4WO_{3},10H_{2}O and BaO,4WO_{3},9H_{2}O for example; (5) the octatungstates--for example, Na_{2}O,8WO_{3},24H_{2}O. Since the metatungstates lose so much water at 100° that they leave salts whose composition corresponds with an acid, 3H_{2}O,4WO_{3}--that is, H_{6}W_{4}O_{15}--whilst in the meta salts only 2 hydrogens are replaced by metals, it is assumed, although without much ground, that these salts contain a particular soluble metatungstic acid of the composition H_{6}W_{4}O_{15}.

As an example we will give a short description of the sodium salts. The normal salt, Na_{2}WO_{4}, is obtained by heating a strong solution of sodium carbonate with tungstic acid to a temperature of 80°; if the solution be filtered hot, it crystallises in rhombic tabular crystals, having the composition Na_{2}WO_{4},2H_{2}O, which remain unchanged in the air and are easily soluble in water. When this salt is fused with a fresh quantity of tungstic acid, it gives a ditungstate, which is soluble in water and separates from its solution in crystals containing water. The same salt is obtained by carefully adding hydrochloric acid to the solution of the normal salt so long as a precipitate does not appear, and the liquid still has an alkaline reaction. This salt was first supposed to have the composition Na_{2}W_{2}O_{7},4H_{2}O, but it has since been found to contain (at 100°) Na_{6}W_{7}O_{24},16H_{2}O--that is, it corresponds with the similar salt of molybdic acid.

(If this salt be heated to a red heat in a stream of hydrogen, it loses a portion of its oxygen, acquires a metallic lustre, and turns a golden yellow colour, and, after being treated with water, alkali, and acid, leaves golden yellow leaflets and cubes which are very like gold. This very remarkable substance, discovered by Wöhler, has, according to Malaguti's analysis, the composition Na_{2}W_{3}O_{9}; that is, it, as it were, contains a double tungstate of tungsten oxide, WO_{2}, and of sodium, Na_{2}WO_{4},WO_{2}WO_{3}. The decomposition of the fused sodium salt is best effected by finely-divided tin. This substance has a sp. gr. 6·6; it conducts electricity like metals, and like them has a metallic lustre. When brought into contact with zinc and sulphuric acid it disengages hydrogen, and it becomes covered with a coating of copper in a solution of copper sulphate in the presence of zinc--that is, notwithstanding its complex composition it presents to a certain extent the appearance and reactions of the metals. It is not acted on by aqua regia or alkaline solutions, but it is oxidised when ignited in air.)

The ditungstate mentioned above, deprived of water (having undergone a modification similar to that of metaphosphoric acid), after being treated with water, leaves an anhydrous, sparingly soluble tetratungstate, Na_{2}WO_{4},3WO_{3}, which, when heated at 120° in a closed tube with water, passes into an easily soluble metatungstate. It may therefore be said that the metatungstates are hydrated compounds. On boiling a solution of the above-mentioned salts of sodium with the yellow hydrate of tungstic acid they give a solution of metatungstate, which is the hydrated tetratungstate. Its crystals contain Na_{2}W_{4}O_{13},10H_{2}O. After the hydrate of tungstic acid (obtained from the ordinary tungstates by precipitation with an acid) has stood a long time in contact with a solution (hot or cold) of sodium tungstate, it gives a solution which is not precipitated by hydrochloric acid; this must be filtered and evaporated over sulphuric acid in a desiccator (it is decomposed by boiling). It first forms a very dense solution (aluminium floats in it) of sp. gr. 3·0, and octahedral crystals of _sodium metatungstate_, Na_{2}W_{4}O_{13},10H_{2}O, sp. gr. 3·85, then separate. It effloresces and loses water, and at 100° only two out of the ten equivalents of water remain, but the properties of the salt remain unaltered. If the salt be deprived of water by further heating, it becomes insoluble. At the ordinary temperature one part of water dissolves ten parts of the metatungstate. The other metatungstates are easily obtained from this salt. Thus a strong and hot solution, mixed with a like solution of barium chloride, gives on cooling crystals of barium metatungstate, BaW_{4}O_{13},9H_{2}O. These crystals are dissolved without change in water containing hydrochloric acid, and also in hot water, but they are partially decomposed by cold water, with the formation of a solution of metatungstic acid and of the normal barium salt BaWO_{4}.

In order to explain the difference in the properties of the salts of tungstic acid, we may add that a mixture of a solution of tungstic acid with a solution of silicic acid does not coagulate when heated, although the silicic acid alone would do so; this is due to the formation of a silicotungstic acid, discovered by Marignac, which presents a fresh example of a complex acid. A solution of a tungstate dissolves gelatinous silica, just as it does gelatinous tungstic acid, and when evaporated deposits a crystalline salt of silicotungstic acid. This solution is not precipitated either by acids (a clear analogy to the metatungstates) or by sulphuretted hydrogen, and corresponds with a series of salts. These salts contain one equivalent of silica and 8 equivalents of hydrogen or metals, in the same form as in salts, to 12 or 10 equivalents of tungstic anhydride; for example the crystalline potassium salt has the composition K_{8}W_{12}SiO_{42},14H_{2}O = 4K_{2}O,12WO_{3},SiO_{2},14H_{2}O. Acid salts are also known in which half of the metal is replaced by hydrogen. The complexity of the composition of such complex acids (for example, of the phosphomolybdic acid) involuntarily leads to the idea of polymerisation, which we were obliged to recognise for silica, lead oxide, and other compounds. This polymerisation, it seems to me, may be understood thus: a hydrate A (for example, tungstic acid) is capable of combining with a hydrate B (for example, silica or phosphoric acid, with or without the disengagement of water), and by reason of this faculty it is capable of polymerisation--that is, A combines with A--combines with itself--just as aldehyde, C_{2}H_{4}O, or the cyanogen compounds are able to combine with hydrogen, oxygen, &c., and are liable to polymerisation. On this view the molecule of tungstic acid is probably much more complex than we represent it; this agrees with the easy volatility of such compounds as the chloranhydrides, CrO_{2}Cl_{2}, MoO_{2}Cl_{2}, the analogues of the volatile sulphuryl chloride, SO_{2}Cl_{2}, and with the non-volatility, or difficult volatility, of chromic and molybdic anhydrides, the analogues of the volatile sulphuric anhydride. Such a view also finds a certain confirmation in the researches made by Graham on the _colloidal_ state of tungstic acid, because colloidal properties only appertain to compounds of a very complex composition. The observations made by Graham on the colloidal state of tungstic and molybdic acids introduced much new matter into the history of these substances. When sodium tungstate, mixed in a dilute solution with an equivalent quantity of dilute hydrochloric acid, is placed in a dialyser, hydrochloric acid and sodium chloride pass through the membrane, and a solution of tungstic acid remains in the dialyser. Out of 100 parts of tungstic acid about 80 parts remain in the dialyser. The solution has a bitter, astringent taste, and does not yield gelatinous tungstic acid (hydrogel) either when heated or on the addition of acids or salts. It may also be evaporated to dryness; it then forms a vitreous mass of the _hydrosol_ of _tungstic acid_, which adheres strongly to the walls of the vessel in which it has been evaporated, and is perfectly soluble in water. It does not even lose its solubility after having been heated to 200°, and only becomes insoluble when heated to a red heat, when it loses about 2-1/2 p.c. of water. The dry acid, dissolved in a small quantity of water, forms a gluey mass, just like gum arabic, which is one of the representatives of the hydrosols of colloidal substances. The solution, containing 5 p.c of the anhydride, has a sp. gr. of 1·047; with 20 p.c., of 1·217; with 50 p.c., of 1·80; and with 80 p.c., of 3·24. The presence of a polymerised trioxide in the form of hydrate, H_{2}OW_{3}O_{9} or H_{2}O_{4}WO_{3}, must then be recognised in the solution: this is confirmed by Sabaneeff's cryoscopic determinations (1889). A similar stable solution of molybdic acid is obtained by the dialysis of a mixture of a strong solution of sodium molybdate with hydrochloric acid (the precipitate which is formed is re-dissolved). If MoCl_{4} be precipitated by ammonia and washed with water, a point is reached at which perfect solution takes place, and the molybdic acid forms a colloid solution which is precipitated by the addition of ammonia (Muthmann). The addition of alkali to the solutions of the hydrosols of tungstic and molybdic acids immediately results in the re-formation of the ordinary tungstates and molybdates. There appears to be no doubt but that the same transformation is accomplished in the passage of the ordinary tungstates into the metatungstates as takes place in the passage of tungstic acid itself from an insoluble into a soluble state; but this may be even actually proved to be the case, because Scheibler obtained a solution of tungstic acid, before Graham, by decomposing barium metatungstate (BaO_{4}WO_{3},9H_{2}O) with sulphuric acid. By treating this salt with sulphuric acid in the amount required for the precipitation of the baryta, Scheibler obtained a solution of metatungstic acid which, when containing 43·75 p.c. of acid, had a sp. gr. of 1·634, and with 27·61 p.c. a sp. gr. of 1·327--that is, specific gravities corresponding with those found by Graham.

Péchard found that as much heat is evolved by neutralising metatungstic acid as with sulphuric acid.

Questions connected with the metamorphoses or modifications of tungstic and molybdic acids, and the polymerisation and colloidal state of substances, as well as the formation of complex acids, belong to that class of problems the solution of which will do much towards attaining a true comprehension of the mechanism of a number of chemical reactions. I think, moreover, that questions of this kind stand in intimate connection with the theory of the formation of solutions and alloys and other so-called indefinite compounds.

Hydrogen (which does not directly form compounds with Cr, Mo, and W) reduces molybdic and tungstic anhydride at a red heat; and this forms the means of obtaining metallic molybdenum and tungsten. _Both metals_ are infusible, and both under the action of heat form compounds with carbon and iron (the addition of tungsten to steel renders the latter ductile and hard).[9] Molybdenum forms a grey powder, which scarcely aggregates under a most powerful heat, and has a specific gravity of 8·5. It is not acted on by the air at the ordinary temperature, but when ignited it is first converted into a brown, and then into a blue oxide, and lastly into molybdic anhydride. Acids do not act on it--that is, it does not liberate hydrogen from them, not even from hydrochloric acid--but strong sulphuric acid disengages sulphurous anhydride, forming a brown mass, containing a lower oxide of molybdenum. Alkalis in solution do not act on molybdenum, but when fused with it hydrogen is given off, which shows, as does its whole character, the acid properties of the metal. The properties of tungsten are almost identical; it is infusible, has an iron-grey colour, is exceedingly hard, so that it even scratches glass. Its specific gravity is 19·1 (according to Roscoe), so that, like uranium, platinum, &c., it is one of the heaviest metals.[9 bis] Just as sulphur and chromium have their corresponding persulphuric and perchromic acids, H_{2}S_{2}O_{8} and H_{2}CrO_{8}, having the properties of peroxides, and corresponding to peroxide of hydrogen, so also molybdenum and tungsten are known to give _permolybdic_ and _pertungstic_ acids, H_{2}Mo_{2}O_{8} and H_{2}W_{2}O_{8}, which have the properties of true peroxides, _i.e._ easily disengage iodine from KI and chlorine from HCl, easily part with their oxygen, and are formed by the action of peroxide of hydrogen, into which they are readily reconverted (hence they may be regarded as compounds of H_{2}O_{2} with 2MoO_{3} and 2WO_{3}), &c. Their formation (Boerwald 1884, Kemmerer 1891) is at once seen in the coloration (not destroyed by boiling), which is obtained on mixing a solution of the salts with peroxide of hydrogen, and on treating, for instance, molybdic acid with a solution of peroxide of hydrogen (Péchard 1892). The acid then forms an orange-coloured solution, which after evaporation in vacuo leaves Mo_{2}H_{2}O_{8}4H_{2}O as a crystalline powder, and loses 4H_{2}O at 100°, beyond which it decomposes with the evolution of oxygen.[9 tri]

[9] Moissan (1893) studied the compounds of Mo and W formed with carbon in the electrical furnace (they are extremely hard) from a mixture of the anhydrides and carbon. Poleck and Grützner obtained definite compounds FeW_{2} and FeW_{2}C_{3} for tungsten. Metallic W and Mo displace Ag from its solutions but not Pb. There is reason for believing that the sp. gr. of pure molybdenum is higher than that (8·5) generally ascribed to it.

[9 bis] We may conclude our description of tungsten and molybdenum by stating that their sulphur compounds have an acid character, like carbon bisulphide or stannic sulphide. If sulphuretted hydrogen be passed through a solution of a molybdate it does not give a precipitate unless sulphuric acid be present, when a dark brown precipitate of _molybdenum trisulphide_, MoS_{3}, is formed. When this sulphide is ignited without access of air it gives the bisulphide MoS_{2}; the latter is not able to combine with potassium sulphide like the trisulphide MoS_{3}, which forms a salt, K_{2}MoS_{4}, corresponding with K_{2}MoO_{4}. This is soluble in water, and separates out from its solution in red crystals, which have a metallic lustre and reflect a green light. It is easily obtained by heating the native bisulphide, MoS_{2}, with potash, sulphur, and a small amount of charcoal, which serves for deoxidising the oxygen compounds. Tungsten gives similar compounds, R_{2}WS_{4}, where R = NH_{4}, K, Na. They are decomposed by acids, with the separation of tungsten trisulphide, WS_{3}, and molybdenum trisulphide, MoS_{3}. Rideal (1892) obtained W_{2}N_{3} by heating WO_{3} in NH_{3}. This compound exhibited the general properties of metallic nitrides.

[9 tri] When peroxide of hydrogen acts upon a solution of potassium molybdate well-formed yellow crystals belonging to the triclinic system separate out in the cold. When these crystals are heated in vacuo they first lose water and then decompose, leaving a residue composed of the salt originally taken. They are soluble in water but insoluble in alcohol. Their composition is represented by the formula K_{2}Mo_{2}O_{8}2H_{2}O. An ammonium salt is obtained by evaporating peroxide of hydrogen with ammonium molybdate. The following salts have also been obtained by the action of peroxide of hydrogen upon the corresponding molybdates: Na_{2}Mo_{2}O_{6}6H_{2}O--in yellow prismatic crystals; MgMo_{2}O_{8}10H_{2}O--stellar needles; BaMoO_{8}2H_{2}O--in microscopic yellow octahedra. A corresponding sodium pertungstate has been obtained by Péchard by boiling sodium tungstate with a solution of peroxide of hydrogen for several minutes. The solution rapidly turns yellow, and no longer gives a precipitate of tungstic anhydride when treated with nitric acid. When evaporated in vacuo the solution leaves a thick syrupy liquid from which ray-like crystals separate out; these crystals are more soluble in water than the salt originally taken. When heated they also lose water and oxygen. Their composition answers to the formula M_{2}W_{2}O_{8}2H_{2}O, where M = Na, NH_{4}, &c. The permolybdates and pertungstates have similar properties. When treated with oxygen acids they give peroxide of hydrogen, and disengage chlorine and iodine from hydrochloric acid and potassium iodide.

Piccini (1891) showed that peroxide of hydrogen not only combines with the oxygen compounds of Mo and W, but also with their fluo-compounds, among which ammonium fluo-molybdate MoO_{2}F_{2}2NH_{4} and others have long been known. (A few new salts of similar composition have been obtained by F. Moureu in 1893.) The action of peroxide of hydrogen upon these compounds gives salts containing a larger amount of oxygen; for instance, a solution of MoO_{2}F_{2}2KFH_{2}O with peroxide of hydrogen gives a yellow solution which after cooling separates out yellow crystalline flakes of MoO_{3}F_{2}2KFH_{2}O, resembling the salt originally taken in their external appearance. By employing a similar method Piccini also obtained: MoO_{3}F_{2}2RbFH_{2}O--yellow monoclinic crystals; MoO_{3}F_{2},2CsFH_{2}O,--yellow flakes, and the corresponding tungstic compounds. All these salts react like peroxide of hydrogen.

In speaking of these compounds I for my part think it may be well to call attention to the fact that, in the first place, the composition of Piccini's oxy-fluo compounds does not correspond to that of permolybdic and pertungstic acid. If the latter be expressed by formulæ with one equivalent of an element, they will be HMoO_{4} and HWO_{4}, and the oxy-fluo form corresponding to them should have the composition MoO_{3}F and WO_{3}F while it contains MO_{3}F_{2} and WO_{3}F_{2}, _i.e._ answers as it were to a higher degree of oxidation, MoH_{2}O_{3} and W_{3}HO_{3}. But if permolybdic acid be regarded as 2MoO_{3} + H_{2}O_{2}, _i.e._ as containing the elements of peroxide of hydrogen, then Piccini's compound will also be found to contain the original salts + H_{2}O; for example, from MoO_{2}F_{2}2KFH_{2}O there is obtained a compound MoO_{2}F_{2}2KFH_{2}O_{2}, _i.e._ instead of H_{2}O they contain H_{2}O_{2}. In the second place the capacity of the salts of molybdenum and tungsten to retain a further amount of oxygen or H_{2}O_{2} probably bears some relation to their property of giving complex acids and of polymerising which has been considered in Note 8 bis. There is, however, a great chemical interest in the accumulation of data respecting these high peroxide compounds corresponding to molybdic and tungstic acids. With regard to the peroxide form of uranium, _see_ Chapter XX., Note 66.

_Uranium_, U = 240, has the highest atomic weight of all the analogues of chromium, and indeed of all the elements yet known. Its highest salt-forming oxide, UO_{3}, shows very feeble acid properties. Although it gives sparingly-soluble yellow compounds with alkalis, which fully correspond with the dichromates--for example, Na_{2}U_{2}O_{7} = Na_{2}O,2UO_{3},[10]--yet it more frequently and easily reacts with acids, HX, forming fluorescent yellowish-green salts of the composition UO_{2}X_{2}, and in this respect uranic trioxide, UO_{3}, differs from chromic anhydride, CrO_{3}, although the latter is able to give the oxychloride, CrO_{2}Cl_{2}. In molybdenum and tungsten, however, we see a clear transition from chromium to uranium. Thus, for example, chromyl chloride, CrO_{2}Cl_{2}, is a brown liquid which volatilises without change, and is completely decomposed by water; molybdenum oxychloride, MoO_{2}Cl_{2}, is a crystalline substance of a yellow colour, which is volatile and soluble in water (Blomstrand), like many salts. Tungsten oxychloride, WO_{2}Cl_{2}, stands still nearer to uranyl chloride in its properties; it forms yellow scales on which water and alkalis act, as they do on many salts (zinc chloride, ferric chloride, aluminium chloride, stannic chloride, &c.), and perfectly corresponds with the difficultly volatile salt, UO_{2}Cl_{2} (obtained by Peligot by the action of chlorine on ignited uranium dioxide, UO_{2}), which is also yellow and gives a yellow solution with water, like all the salts UO_{2}X_{2}. The property of uranic oxide, UO_{3}, of forming salts UO_{2}X_{2} is shown in the fact that the hydrated oxide of uranium, UO_{2}(HO)_{2}, which is obtained from the nitrate, carbonate, and other salts by the loss of the elements of the acid, is easily soluble in acids, as well as in the fact that the lower grades of oxidation of uranium are able, when treated with nitric acid, to form an easily crystallisable uranyl nitrate, UO_{2}(NO_{3})_{2},6H_{2}O; this is the most commonly occurring uranium salt.[11]

[10] Uranium trioxide, or uranic oxide, shows its feeble basic and acid properties in a great number of its reactions. (1) Solutions of uranic salts give yellow precipitates with alkalis, but these precipitates do not contain the hydrate of the oxide, but compounds of it with bases; for example, 2UO_{2}(NO_{3})_{2} + 6KHO = 4KNO_{3} + 3H_{2}O + K_{2}U_{2}O_{7}. There are other _urano-alkali compounds_ of the same constitution; for example, (NH_{4})_{2}U_{2}O_{7} (known commercially as uranic oxide), MgU_{2}O_{7}, BaU_{2}O_{7}. They are the analogues of the dichromates. Sodium uranate is the most generally used under the name of uranium yellow, Na_{2}U_{2}O_{7}. It is used for imparting the characteristic yellow-green tint to glass and porcelain. Neither heat nor water nor acids are able to extract the alkali from sodium uranate, Na_{2}U_{2}O_{7}, and therefore it is a true insoluble salt, of a yellow colour, and clearly indicates the acid character (although feeble) of uranic oxide. (2) The carbonates of the alkaline earths (for instance, barium carbonate) precipitate uranic oxide from its salts, as they do all the salts of feeble bases; for example, R_{2}O_{3}. (3) The _alkaline carbonates_, when added to solutions of uranic salts, give a _precipitate, which is soluble in_ _an excess of the reagent_, and particularly so if the acid carbonates be taken. This is due to the fact that (4) the uranyl salts _easily form double salts_ with the salts of the alkali metals, including the salts of ammonium. Uranium, in the form of these double salts, often gives salts of well-defined crystalline form, although the simple salts are little prone to appear in crystals. Such, for example, are the salts obtained by dissolving potassium uranate, K_{2}U_{2}O_{7}, in acids, with the addition of potassium salts of the same acids. Thus, with hydrochloric acid and potassium chloride a well-formed crystalline salt, K_{2}(UO_{2})Cl_{4},2H_{2}O, belonging to the monoclinic system, is produced. This salt decomposes in dissolving in pure water. Among these double salts we may mention the double carbonate with the alkalis, R_{4}(UO_{2})(CO_{3})_{3} (equal to 2R_{2}CO_{3} + UO_{2}CO_{3}); the acetates, R(UO_{2})(C_{2}H_{3}O_{2})_{3}--for instance, the sodium salt, Na(UO_{2})(C_{2}H_{3}O_{2})_{3}, and the potassium salt, K(UO_{2})(C_{2}H_{3}O_{2})_{3},H_{2}O; the sulphates, R_{2}(UO_{2})(SO_{4})_{3},2H_{2}O, &c. In the preceding formula R = K, Na, NH_{4}, or R_{2} = Mg, Ba, &c. _This property of giving comparatively stable double salts indicates feebly developed basic properties_, because double salts are mainly formed by salts of distinctly basic metals (these form, as it were, the basic element of a double salt) and salts of feebly energetic bases (these form the acid element of a double salt), just as the former also give acid salts; the acid of the acid salts is replaced in the double salts by the salt of the feebly energetic base, which, like water, belongs to the class of intermediate bases. For this reason barium does not give double salts with alkalis as magnesium does, and this is why double salts are more easily formed by potassium than by lithium in the series of the alkali metals. (5) The most remarkable property, proving the feeble energy of uranic oxide as a base, is seen in the fact that when their composition is compared with that of other salts those of uranic oxide _always appear as basic salts_. It is well known that a normal salt, R_{2}X_{6}, corresponds with the oxide R_{2}O_{3}, where X = Cl, NO_{3}, &c., or X_{2} = SO_{4}, CO_{3}, &c.; but there also exist basic salts of the same type where X = HO or X_{2} = O. We saw salts of all kinds among the salts of aluminium, chromium, and others. With uranic oxide no salts are known of the types UX_{6} (UCl_{6}, U(SO_{4})_{3}, alums, &c., are not known), nor even salts, U(HO)_{2}X_{4} or UOX_{4}, but it always forms salts of the type U(HO)_{4}X_{2}, or UO_{2}X_{2}. Judging from the fact that nearly all the salts of uranic oxide retain water in crystallising from their solutions, and that this water is difficult to separate from them, it may be thought to be water of hydration. This is seen in part from the fact that the composition of many of the salts of uranic oxide may then be expressed without the presence of water of crystallisation; for instance, U(HO)_{4}K_{2}Cl_{4} (and the salt of NH_{4}, U(HO)_{4}K_{2}(SO_{4})_{2}, U(HO)_{4}(C_{2}H_{3}O_{2})_{2}. Sodium uranyl acetate however does not contain water.

[11] _Uranyl nitrate_, or uranium nitrate, UO_{2}(NO_{3})_{2},6H_{2}O, crystallises from its solutions in transparent yellowish-green prisms (from an acid solution), or in tabular crystals (from a neutral solution), which effloresce in the air and are easily soluble in water, alcohol, and ether, have a sp. gr. of 2·8, and fuse when heated, losing nitric acid and water in the process. If the salt itself (Berzelius) or its alcoholic solution (Malaguti) be heated up to the temperature at which oxides of nitrogen are evolved, there then remains a mass which, after being evaporated with water, leaves uranyl hydroxide, UO_{2}(HO)_{2} (sp. gr. 5·93), whilst if the salt be ignited there remains the dioxide, UO_{2}, as a brick-red powder, which on further heating loses oxygen and forms the dark olive uranoso-uranic oxide, U_{3}O_{8}. The solution of the nitrate obtained from the ore is purified in the following manner: sulphurous anhydride is first passed through it in order to reduce the arsenic acid present into arsenious acid; the solution is then heated to 60°, and sulphuretted hydrogen passed through it; this precipitates the lead, arsenic, and tin, and certain other metals, as sulphides, insoluble in water and dilute nitric acid. This liquid is then filtered and evaporated with nitric acid to crystallisation, and the crystals are dissolved in ether. Or else the solution is first treated with chlorine in order to convert the ferrous chloride (produced by the action of the hydrogen sulphide) into ferric chloride, the oxides are then precipitated by ammonia, and the resultant precipitate, containing the oxides Fe_{2}O_{3}, UO_{3}, and compounds of the latter with potash, lime, ammonia, and other bases present in the solution (the latter being due to the property of uranic oxide of combining with bases), is washed and dissolved in a strong, slightly-heated solution of ammonium carbonate, which dissolves the uranic oxide but not the ferric oxide. The solution is filtered, and on cooling deposits a well-crystallising _uranyl ammonium carbonate_, UO_{2}(NH_{4})_{4}(CO_{3})_{3}, in brilliant monoclinic crystals which on exposure to air slowly give off water, carbonic anhydride, and ammonia; the same decomposition is readily effected at 300°, the residue then consisting of uranic oxide. This salt is not very soluble in water, but is readily so in ammonium carbonate; it is obvious that it may readily be converted into all the other salts of oxides of uranium. Uranium salts are also purified in the form of _acetate_, which is very sparingly soluble, and is therefore directly precipitated from a strong solution of the nitrate by mixing it with acetic acid.

We may also mention the _uranyl phosphate_, HUPO_{6}, which must be regarded as an orthophosphate in which two hydrogens are replaced by the radicle uranyl, UO_{2}, _i.e._ as H(UO_{2})PO_{4}. This salt is formed as a hydrated gelatinous yellow precipitate, on mixing a solution of uranyl nitrate with disodium phosphate. The precipitation occurs in the presence of acetic acid, but not in the presence of hydrochloric acid. If moreover an excess of an ammonium salt be present, the ammonia enters into the composition of the bright yellow gelatinous precipitate formed, in the proportion UO_{2}NH_{4}PO_{4}. This precipitate is not soluble in water and acetic acid, and its solution in inorganic acids when boiled entirely expels all the phosphoric acid. This fact is taken advantage of for removing phosphoric acids from solutions--for instance, from those containing salts of calcium and magnesium.

_Uranium_, which gives an oxide, UO_{3}, and the corresponding salt UO_{2}X_{2} and dioxide UO_{2}, to which the salts UX_{4} correspond, is rarely met with in nature. Uranite or the double orthophosphate of uranic oxide, R(UO_{2})H_{2}P_{2}O_{8},7H_{2}O, where R = Cu or Ca, uranium-vitriol U(SO_{4})_{2},H_{2}O, samarakite, and æschynite, are very rarely found, and then only in small quantities. Of more frequent and abundant occurrence is the non-crystalline, earthy brown uranium ore known as _pitchblende_ (sp. gr. 7·2), which is mainly composed of the intermediate oxide, U_{3}O_{8} = UO_{2},2UO_{3}. This ore is found at Joachimsthal in Bohemia and in Cornwall. It usually contains a number of different impurities, chiefly sulphides and arsenides of lead and iron, as well as lime and silica compounds. In order to expel the arsenic and sulphur it is roasted, ground, washed with dilute hydrochloric acid, which does not dissolve the uranoso-uranic oxide, U_{3}O_{8}, and the residue is dissolved in nitric acid, which transforms the uranium oxide into the nitrate, UO_{2}(NO_{3})_{2}.

It must be observed that the oxide of uranium, first distinguished by Klaproth (1789), was for a long time regarded as able to give metallic uranium under the action of charcoal and other reducing agents (with the aid of heat). But the substance thus obtained was only the _uranium dioxide_, UO_{2}. The compound nature of this dioxide,[12] or the presence of oxygen in it, was demonstrated by Peligot (1841), by igniting it with charcoal in a stream of chlorine. He thus obtained a volatile _uranium tetrachloride_, UCl_{4},[13] which, when heated with sodium, gave _metallic uranium_ as a grey metal, having a specific gravity of 18·7, and liberating hydrogen from acids, with the formation of green uranous salts, UX_{4}, which act as powerful reducing agents.[14]

[12] Uranium dioxide, or _uranyl_, UO_{2}, which is contained in the salts UO_{2}X_{2}, has the appearance and many of the properties of a metal. Uranic oxide may be regarded as uranyl oxide, (UO_{2})O, its salts as salts of this uranyl; its hydroxide, (UO_{2})H_{2}O_{2}, is constituted like CaH_{2}O_{2}. The green oxide of uranium, uranoso-uranic oxide (easily formed from uranic salts by the loss of oxygen), U_{3}O_{8} = UO_{2},2UO_{3}, when ignited with charcoal or hydrogen (dry) gives a brilliant crystalline substance of sp. gr. about 11·0 (Urlaub), whose appearance resembles that of metals, and decomposes steam at a red heat with the evolution of hydrogen; it does not, however, decompose hydrochloric or sulphuric acid, but is oxidised by nitric acid. The same substance (i.e. uranium dioxide UO_{2}) is also obtained by igniting the compound (UO_{2})K_{2}Cl_{4} in a stream of hydrogen, according to the equation UO_{2}K_{4}Cl_{4} + H_{2} = UO_{2} + 2HCl + 2KCl. It was at first regarded as the metal. In 1841 Peligot found that it contained oxygen, because carbonic oxide and anhydride were evolved when it was ignited with charcoal in a stream of chlorine, and from 272 parts of the substance which was considered to be metal he obtained 382 parts of a volatile product containing 142 parts of chlorine. From this it was concluded that the substance taken contained an equivalent amount of oxygen. As 142 parts of chlorine correspond with 32 parts of oxygen, it followed that 272 - 32 = 240 parts of metal were combined in the substance with 32 parts of oxygen, and also in the chlorine compound obtained with 142 parts of chlorine. These calculations have been made for the now accepted atomic weight of uranium (U = 240, _see_ Note 14). Peligot took another atomic weight, but this does not alter the principle of the argument.

[13] _Uranium tetrachloride_, uranous chloride, UCl_{4}, corresponds with uranous oxide as a base. It was obtained by Peligot by igniting uranic oxide mixed with charcoal in a stream of _dry_ chlorine: UO_{3} + 3C + 2Cl_{2} = UCl_{4} + 3CO. This green volatile compound (Note 12) crystallises in regular octahedra, is very hygroscopic, easily soluble in water, with the development of a considerable amount of heat, and no longer separates out from its solution in an anhydrous state, but disengages hydrochloric acid when evaporated. The solution of uranous chloride in water is green. It is also formed by the action of zinc and copper (forming cuprous chloride) on a solution of uranyl chloride, UO_{2}Cl_{2}, especially in the presence of hydrochloric acid and sal-ammoniac. Solutions of uranyl salts are converted into uranous salts by the action of various reducing agents, and among others by organic substances or by the action of light, whilst the salts UX_{4} are converted into uranyl salts, UO_{2}X_{2}, by exposure to air or by oxidising agents. Solutions of the green uranyl salts act as powerful reducing agents, and give a brown precipitate of the uranous hydroxide, UH_{4}O_{4}, with potash and other alkalis. This hydroxide is easily soluble in acids but not in alkalis. On ignition it does not form the oxide UO_{2}, because it decomposes water, but when the higher oxides of uranium are ignited in a stream of hydrogen or with charcoal they yield uranous oxide. Both it and the chloride UCl_{4}, dissolve in strong sulphuric acid, forming a green salt, U(SO_{4})_{2},2H_{2}O. The same salt, together with uranyl sulphate, UO_{2}(SO_{4}), is formed when the green oxide, U_{3}O_{8}, is dissolved in hot sulphuric acid. The salts obtained in the latter instance may be separated by adding alcohol to the solution, which is left exposed to the light; the alcohol reduces the uranyl salt to uranous salt, an excess of acid being required. An excess of water decomposes this salt, forming a basic salt, which is also easily produced under other circumstances, and contains UO(SO_{4}),2H_{2}O (which corresponds to the uranic salt).

[14] The atomic weight of uranium was formerly taken as half the present one, U = 120, and the oxides U_{2}O_{3}, suboxide UO, and green oxide U_{3}O_{4}, were of the same types as the oxides of iron. With a certain resemblance to the elements of the iron group, uranium presents many points of distinction which do not permit its being grouped with them. Thus uranium forms a very stable oxide, U_{2}O_{3}(U = 120), but does not give the corresponding chloride U_{2}Cl_{6} (Roscoe, however, in 1874 obtained UCl_{5}, like MoCl_{5} and WCl_{5}), and under those circumstances (the ignition of oxide of uranium mixed with charcoal, in a stream of chlorine), when the formation of this compound might be expected, it gives (U = 120) the chloride UCl_{2}, which is characterised by its volatility; this is not a property, to such an extent, of any of the bichlorides, RCl_{2}, of the iron group.

The alteration or doubling of the atomic weight of uranium--_i.e._ the recognition of U = 240--was made for the first time in the first (Russian) edition of this work (1871), and in my memoir of the same year in Liebig's _Annalen_, on the ground that with an atomic weight 120, uranium could not be placed in the periodic system. I think it will not be superfluous to add the following remarks on this subject: (1) In the other groups (K--Rb--Cs, Ca--Sr--Ba, Cl--Br--I) the acid character of the oxides decreases and their basic character increases with the rise of atomic weight, and therefore we should expect to find the same in the group Cr--Mo--W--U, and if CrO_{3}, MoO_{3}, WO_{3} be the anhydrides of acids then we indeed find a decrease in their acid character, and therefore uranium trioxide, UO_{3}, should be a very feeble anhydride, but its basic properties should also be very feeble. Uranic oxide does indeed show these properties, as was pointed out above (Note 10). (2) Chromium and its analogues, besides the oxides RO_{3}, also form lower grades of oxidation RO_{2}, R_{2}O_{3}, and the same is seen in uranium; it forms UO_{3}, UO_{2}, U_{2}O_{3} and their compounds. (3) Molybdenum and tungsten, in being reduced from RO_{3}, easily and frequently give an intermediate oxide of a blue colour, and uranium shows the same property; giving the so-called green oxide which, according to present views, must be regarded as U_{3}O_{8} = UO_{2}2UO_{3}, analogous to Mo_{3}O_{8}. (4) The higher chlorides, RCl_{6}, possible for the elements of this group, are either unstable (WCl_{6}) or do not exist at all (Cr); but there is one single lower volatile compound, which is decomposed by water, and liable to further reduction into a non-volatile chlorine product and the metal. The same is observed in uranium, which forms an easily volatile chloride, UCl_{4}, decomposed by water. (5) The high sp. gr. of uranium (18·6) is explained by its analogy to tungsten (sp. gr. 19·1). (6) For uranium, as for chromium and tungsten, yellow tints predominate in the form RO_{3}, whilst the lower forms are green and blue. (7) Zimmermann (1881) determined the vapour densities of uranous bromide, UBr_{4}, and chloride, UCl_{4} (19·4 and 13·2), and they were found to correspond to the formulæ given above--that is, they confirmed the higher atomic weight U = 240. Roscoe, a great authority on the metals of this group, was the first to accept the proposed atomic weight of uranium, U = 240, which since Zimmermann's work has been generally recognised.

As the salts of uranic oxide are reduced in the absence of organic matter by the action of light, and as they impart a characteristic coloration to glass,[15] they find a certain application in photography and glass work.

[15] Uranium glass, obtained by the addition of the yellow salt K_{2}U_{2}O_{7} to glass, has a green yellow fluorescence, and is sometimes employed for ornaments; it absorbs the violet rays, like the other salts of uranic oxide--that is, it possesses an absorption spectrum in which the violet rays are absent. The index of refraction of the absorbed rays is altered, and they are given out again as greenish-yellow rays; hence, compounds of uranic acid, when placed in the violet portion of the spectrum, emit a greenish-yellow light, and this forms one of the best examples (another is found in a solution of quinine sulphate) of the phenomenon of fluorescence. The rays of light which pass through uranic compounds do not contain the rays which excite the phenomena of fluorescence and of chemical transformation, as the researches of Stokes prove.

If we compare together the highly acid elements, sulphur, selenium, and tellurium, of the uneven series, with chromium, molybdenum, tungsten, and uranium of the even series, we find that the resemblance of the properties of the higher form RO_{3} does not extend to the lower forms, and even entirely disappears in the elements, for there is not the smallest resemblance between sulphur and chromium and their analogues in a free state. In other words, this means that the small periods, like Na, Mg, Al, Si, P, S, Cl, containing seven elements, do not contain any near analogues of chromium, molybdenum, &c., and therefore their true position among the other elements must be looked for only in those large periods which contain two small periods, and whose type is seen in the period containing: K, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Ge, As, Se, Br. These large periods contain Ca and Zn, giving RO, Sc, and Ga of the third group, Ti and Ge giving RO_{2}, V and As forming R_{2}O_{5}, Cr and Se of the sixth group, Mn and Br of the seventh group, and the remaining elements, Fe, Co, Ni, form connective members of the intermediate eighth group, to the description of the representatives of which we shall turn in the following chapters. We will now proceed to describe _manganese_, Mn = 55, as an element of the seventh group of the even series, directly following after Cr = 52, which corresponds with Br = 80 to the same degree that Cr does with Se = 79. For chromium, selenium, and bromine very close analogues are known, but for manganese as yet none have been obtained--that is, it is the only representative of the even series in the seventh group. In placing manganese with the halogens in one group, the periodic system of the elements only requires that it should bear an analogy to the halogens in the higher type of oxidation--_i.e._ in the salts and acids--whilst it requires that as great a difference should be expected in the lower types and elements as there exists between chromium or molybdenum and sulphur or selenium. And this is actually the case. The elements of the seventh group form a higher salt-forming oxide, R_{2}O_{7}, and its corresponding hydrate, HRO_{4}, and salts--for example, KClO_{4}. Manganese in the form of potassium permanganate, KMnO_{4}, actually presents a great analogy in many respects to potassium perchlorate, KClO_{4}. The analogy of the crystalline form of both salts was shown by Mitscherlich. The salts of permanganic acid are also nearly all soluble in water, like those of perchloric acid, and if the silver salt of the latter, AgClO_{4}, be sparingly soluble in water, so also is silver permanganate, AgMnO_{4}. The specific volume of potassium perchlorate is equal to 55, because its specific gravity = 2·54; the specific volume of potassium permanganate is equal to 58, because its specific gravity = 2·71. So that the volumes of equivalent quantities are in this instance approximately the same whilst the atomic volumes of chlorine (35·5/1·3 = 27) and manganese (55/7·5) are in the ratio 4 : 1. In a free state the higher acids HClO_{4} and HMnO_{4} are both soluble in water and volatile, both are powerful oxidisers--in a word, their analogy is still closer than that of chromic and sulphuric acids, and those points of distinction which they present also appear among the nearest analogues--for example, in sulphuric and telluric acids, in hydrochloric and hydriodic acids, &c. Besides Mn_{2}O_{7} manganese gives a lower grade of oxidation, MnO_{3}, analogous to sulphuric and chromic trioxides, and with it corresponds potassium manganate, K_{2}MnO_{4}, isomorphous with potassium sulphate.[16] In the still lower grades of oxidation, Mn_{2}O_{3} and MnO, there is hardly any similarity to chlorine, whilst every point of resemblance disappears when we come to the elements themselves--_i.e._ to manganese and chlorine--for manganese is a metal, like iron, which combines directly with chlorine to form a saline compound, MnCl_{2}, analogous to magnesium chloride.[17]

[16] The comparison of potassium permanganate with potassium perchlorate, or of potassium manganate with potassium sulphate, shows directly that many of the physical and chemical properties of substances do not depend on the nature of the elements, but on the atomic types in which they appear, on the kind of movements, or on the positions in which the atoms forming the molecule occur.

[17] If, however, we compare the spectra (Vol. I. p. 565) of chlorine, bromine, and iodine with that of manganese, a certain resemblance or analogy is to be found connecting manganese both to iron and to chlorine, bromine, and iodine.

Manganese belongs to the number of metals widely distributed in nature, especially in those localities where iron occurs, whose ores frequently contain compounds of manganous oxide, MnO, which presents a resemblance to ferrous oxide, FeO, and to magnesia. In many minerals magnesia and the oxides allied to it are replaced by manganous oxide; calcspars and magnesites--_i.e._ R´´CO_{3} in general--are frequently met with containing manganous carbonate, which also occurs in a separate state, although but rarely. The soil also and the ash of plants generally contain a small quantity of manganese. In the analysis of minerals it is generally found that manganese occurs together with magnesia, because, like it, manganous oxide remains in solution in the presence of ammoniacal salts, not being precipitated by reagents. The property of this manganous oxide, MnO, of passing into the higher grades of oxidation under the influence of heat, alkalis, and air, gives an easy means not only of discovering the presence of manganese in admixture with magnesia, but also of separating these two analogous bases. Magnesia is not able to give higher grades of oxidation, whilst manganese gives them with great facility. Thus, for instance, an _alkaline_ solution of sodium hypochlorite produces a precipitate of manganese dioxide in a solution of a manganous salt: MnCl_{2} + NaClO + 2NaHO = MnO_{2} + H_{2}O + 3NaCl; whilst magnesia is not changed under these circumstances, and remains in the form of MgCl_{2}. If the magnesia be precipitated owing to the presence of alkali, it may be dissolved in acetic acid, in which manganese dioxide is insoluble. The presence of small quantities of manganese may also be recognised by the green coloration which alkalis acquire when heated with manganese compounds in the air. This green coloration depends on the property of manganese of giving a green alkaline manganate: MnCl_{2} + 4KHO + O_{2} = K_{2}MnO_{4} + 2KCl + 2H_{2}O. Thus _the faculty of oxidising in the presence of alkalis_ forms an essential character of manganese. The higher grades of oxidation containing Mn_{2}O_{7} and MnO_{3} are quite unknown in nature, and even MnO_{2} is not so widely spread in nature as the ores composed of manganous compounds which are met with nearly everywhere. The most important ore of manganese is its dioxide, or so-called _peroxide_, MnO_{2}, which is known in mineralogy as _pyrolusite_. Manganese also occurs as an oxide corresponding with magnetic iron ore, MnO,Mn_{2}O_{3} = Mn_{3}O_{4}, forming the mineral known as _hausmannite_. The oxide Mn_{2}O_{3} also occurs in nature as the anhydrous mineral _braunite_, and in a hydrated form, Mn_{2}O_{3},H_{2}O, called _manganite_. Both of these often occur as an admixture in pyrolusite. Besides which, manganese is met with in nature as a rose-coloured mineral, _rhodonite_, or silicate, MnSiO_{3}. Very fine and rich deposits of manganese ores have been found in the Caucasus, the Urals, and along the Dnieper. Those at the Sharapansky district of the Government of Kutais and at Nicopol on the Dnieper are particularly rich. A large quantity of the ore (as much as 100,000 tons yearly) is exported from these localities.

Thus manganese gives oxides of the following forms: MnO, manganous oxide, and manganous salts, MnX_{2}, corresponding with the base, which resembles magnesia and ferrous oxide in many respects; Mn_{2}O_{3}, a very feeble base, giving salts, MnX_{3}, analogous to the aluminium and ferric salts, easily reduced to MnX_{2}; MnO_{2}, dioxide, generally called peroxide, an almost indifferent oxide, or feebly acid;[18] MnO_{3}, manganic anhydride, which forms salts resembling potassium sulphate;[18 bis] Mn_{2}O_{7}, permanganic anhydride, giving salts analogous to the perchlorates.

[18] The name 'peroxide' should only be retained for those _highest_ oxides (and MnO_{2} stands between MnO and MnO_{3}) which either by a direct method of double decomposition are able to give hydrogen peroxide or contain a larger proportion of oxygen than the base or the acid, just as hydrogen peroxide contains more oxygen than water. Their type will be H_{2}O_{2}, and they are exemplified by barium peroxide, BaO_{2}, and sulphur peroxide, S_{2}O_{7}, &c. Such a dioxide as MnO_{2} is, in all probability, a salt--that is, a manganous manganate, MnO_{3}MnO, and also, as a basic salt of a feeble base, capable of combining with alkalis and acids. Hence the name of manganese peroxide should be abandoned, and replaced by manganese dioxide. PbO_{2} is better termed lead dioxide than peroxide. Bisulphide of manganese, MnS_{2}, corresponding to iron pyrites, FeS_{2}, sometimes occurs in nature in fine octahedra (and cube combinations), for instance, in Sicily; it is called Hauerite.

[18 bis] On comparing the manganates with the permanganates--for example, K_{2}MnO_{4} with KMnO_{4}--we find that they differ in composition by the abstraction of one equivalent of the metal. Such a relation in composition produced by oxidation is of frequent occurrence--for instance, K_{4}Fe(CN)_{6} in oxidising gives K_{3}Fe(CN)_{6}; H_{2}SO_{4} in oxidising gives persulphuric acid, HSO_{4}, or H_{2}S_{7}O_{8}; H_{2}O forms HO or H_{2}O_{2}, &c.

_All the oxides of manganese when heated with acids give salts_, MnX_{2}, corresponding with the lower grade of oxidation, _manganous oxide_, MnO. Manganic oxide, Mn_{2}O_{3}, is a feebly energetic base; it is true that it dissolves in hydrochloric acid and gives a dark solution containing the salt MnCl_{3}, but the latter when heated evolves chlorine and gives a salt corresponding with manganous oxide MnCl_{2}--_i.e._ at first: Mn_{2}O_{3} + 6HCl = 3H_{2}O + Mn_{2}Cl_{6}, and then the Mn_{2}Cl_{6} decomposes into 2MnCl_{2} + Cl_{2}. None of the remaining higher grades of oxidation have a basic character, but _act as oxidising agents in the presence of acids_, disengaging oxygen and passing into salts of the lower grade of oxidation of manganese, MnO. Owing to this circumstance, _the manganous salts_ are often obtained; they are, for instance, left in the residue when the dioxide is used for the preparation of oxygen and chlorine.[19]

[19] In the preparation of oxygen from the dioxide by means of H_{2}SO_{4}, MnSO_{4} is formed; in the preparation of chlorine from HCl and MnO_{2}, MnCl_{2} is obtained. These two manganous salts may be taken as examples of compounds MnX_{2}. Manganous sulphate generally contains various impurities, and also a large amount of iron salt (from the native MnO_{2}), from which it cannot be freed by crystallisation. Their removal may, however, be effected by mixing a portion of the liquid with a solution of sodium carbonate; a precipitate of manganous carbonate is then formed. This precipitate is collected and washed, and then added to the remaining mass of the impure solution of manganous sulphate; on heating the solution with this precipitate, the whole of the iron is precipitated as oxide. This is due to the fact that in the solution of the manganese dioxide in sulphuric acid the whole of the iron is converted into the ferric state (because the dioxide acts as an oxidising agent), which, as an exceedingly feeble base precipitated by calcium carbonate and other kindred salts, is also precipitated by manganous carbonate. After being treated in this manner, the solution of manganous sulphate is further purified by crystallisation. If it be a bright red colour, it is due to the presence of higher grades of oxidation of manganese; they may be destroyed by boiling the solution, when the oxygen from the oxides of manganese is evolved and a very faintly coloured solution of manganous sulphate is obtained. This salt is remarkable for the facility with which it gives various combinations with water. By evaporating the almost colourless solution of _manganous sulphate_ at very low temperatures, and by cooling the saturated solution at about 0°, crystals are obtained containing 7 atoms of water of crystallisation, MnSO_{4},7H_{2}O, which are isomorphous with cobaltous and ferrous sulphates. These crystals, even at 10°, lose 5 p.c. of water, and completely effloresce at 15°, losing about 20 p.c. of water. By evaporating a solution of the salt at the ordinary temperature, but not above 20°, crystals are obtained containing 5 mol. H_{2}O, which are isomorphous with copper sulphate; whilst if the crystallisation be carried on between 20° and 30°, large transparent prismatic crystals are formed containing 4 mol. H_{2}O (see Nickel). A boiling solution also deposits these crystals together with crystals containing 3 mol. H_{2}O, whilst the first salt, when fused and boiled with alcohol, gives crystals containing 2 mol. H_{2}O. Graham obtained a monohydrated salt by drying the salt at about 200°. The last atom of water is eliminated with difficulty, as is the case with all salts like MnSO_{4}nH_{2}O. The crystals containing a considerable amount of water are rose-coloured, and the anhydrous crystals are colourless. The solubility of MnSO_{4},4H_{2}O (Chapter I., Note 24) per 100 parts of water is: at 10°, 127 parts; at 37°·5, 149 parts; at 75°, 145 parts; and at 101°, 92 parts. Whence it is seen that at the boiling-point this salt is less soluble than at lower temperatures, and therefore a solution saturated at the ordinary temperature becomes turbid when boiled. Manganous sulphate, being analogous to magnesium sulphate, is decomposed, like the latter, when ignited, but it does not then leave manganous oxide, but the intermediate oxide, Mn_{3}O_{4}. It gives double salts with the alkali sulphates. With aluminium sulphate it forms fine radiated crystals, whose composition resembles that of the alums--namely, MnAl_{2}(SO_{4})_{4},24H_{2}O. This salt is easily soluble in water, and occurs in nature.

_Manganous chloride_, MCl_{2}, crystallises with 4 mol. H_{2}O, like the ferrous salt, and not with 6 mol. H_{2}O like many kindred salts--for example, those of cobalt, calcium, and magnesium; 100 parts of water dissolve 38 parts of the anhydrous salt at 10° and 55 parts at 62°. Alcohol also dissolves manganous chloride, and the alcoholic solution burns with a red flame. This salt, like magnesium chloride, readily forms double salts. A solution of borax gives a dirty rose-coloured precipitate having the composition MnH_{4}(BO_{3})_{2}H_{2}O, which is used as a drier in paint-making. Potassium cyanide produces a yellowish-grey precipitate, MnC_{2}N_{2}, with manganous salts, soluble in an excess of the reagent, a double salt, K_{4}MnC_{6}N_{6}, corresponding with potassium ferrocyanide, being formed. On evaporation of this solution, a portion of the manganese is oxidised and precipitated, whilst a salt corresponding to Gmelin's red salt, K_{3},MnC_{6}N_{6} (_see_ Chapter XXII.), remains in solution. Sulphuretted hydrogen does not precipitate salts of manganese, not even the acetate, but ammonium sulphide gives a flesh-coloured precipitate, MnS; at 320° this sulphide of manganese passes into a green variety (Antony). Oxalic acid in strong solutions of manganous salts gives a white precipitate of the oxalate, MnC_{2}O_{4}. This precipitate is insoluble in water, and is used for the preparation of manganous oxide itself because it decomposes like oxalic acid when ignited (in a tube without access of air), with the formation of carbonic anhydride, carbonic oxide, and manganous oxide. _Manganous oxide_ thus obtained is a green powder, which however oxidises with such facility that it burns in air when brought into contact with an incandescent substance, and passes into the red intermediate oxide Mn_{3}O_{4}. In solutions of manganous salts, alkalis produce a precipitate of the hydroxide MnH_{2}O_{2}, which rapidly absorbs oxygen in the presence of air and gives the brown intermediate oxide, or, more correctly speaking, its hydrate.

Manganous oxide, besides being obtained by the above-described method from manganous oxalate, may also be obtained by igniting the higher oxides in a stream of hydrogen, and also from manganese carbonate. The manganous oxide ignited in the presence of hydrogen acquires a great density, and is no longer so easily oxidised. It may also be obtained in a crystalline form, if during the ignition of the carbonate or higher oxide a trace of dry hydrochloric acid gas be passed into the current of hydrogen. It is thus obtained in the form of transparent emerald green crystals of the regular system, and in this state is easily soluble in acids.

Manganous oxide in oxidising gives the _red oxide of manganese_, Mn_{5}O_{4}. This is the most stable of all the oxides of manganese; it is not only stable at the ordinary but also at a high temperature--that is, it does not absorb or disengage oxygen spontaneously. When ignited, all the higher oxides of manganese pass into it by losing oxygen, and manganous oxide by absorbing oxygen. This oxide does not give any distinct salts, but it dissolves in sulphuric acid, forming a dark red solution, which contains both manganous and manganic (of the _oxide_, Mn_{2}O_{3}) sulphates. The latter with potassium sulphate gives a manganese alum, in which the alumina is replaced by its isomorphous oxide of manganese. But this alum, like the solution of the intermediate oxide in sulphuric acid, evolves oxygen and leaves a manganous salt when slightly heated.

_Manganese dioxide_ is still less basic than the oxide, and disengages oxygen or a halogen in the presence of acids, forming manganous salts, like the oxide. However, if it be suspended in ether, and hydrochloric acid gas passed into the mixture, which is kept cool, the ether acquires a green colour, owing to the formation of tetrachloride of manganese, MnCl_{4}, corresponding with the dioxide which passes into solution. It is however very unstable, being exceedingly easily decomposed with the evolution of chlorine. The corresponding fluoride, MnF_{4}, obtained by Nicklés is much more stable. At all events, manganese dioxide does not exhibit any well-defined basic character, but has rather an acid character, which is particularly shown in the compounds MnF_{4} and MnCl_{4} just mentioned, and in the property of manganese dioxide of combining with alkalis. If the higher grades of oxidation of manganese be deoxidised in the presence of alkalis, they frequently give the dioxide combined with the alkali--for example, in the presence of potash a compound is formed which contains K_{2}O,5MnO_{2}, which shows the weak acid character of this oxide. When ignited in the presence of sodium compounds manganese dioxide frequently forms Na_{2}O,8MnO_{2} and Na_{2}O,12MnO_{2}, and lime when heated with MnO_{2} gives from CaO,3MnO_{2} to (CaO)_{2},MnO_{2} (Rousseau) according to the temperature. Besides which, perhaps, MnO_{2} is a saline compound, containing MnOMnO_{3} or (MnO)_{3}Mn_{2}O_{7}, and there are reactions which support such a view (Spring, Richards, Traube, and others); for instance it is known that manganous chloride and potassium permanganate give the dioxide in the presence of alkalis.

Manganese dioxide may be obtained from manganous salts by the action of oxidising agents. If manganous hydroxide or carbonate be shaken up in water through which chlorine is passed, the hypochlorite of the metal is not formed, as is the case with certain other oxides, but manganese dioxide is precipitated: 2MnO_{2}H_{2} + Cl_{2} = MnCl_{2} + MnO_{2},H_{2}O + H_{2}O. Owing to this fact, hypochlorites in the presence of alkalis and acetic acid when added to a solution of manganous salts give hydrated manganese dioxide, as was mentioned above. Manganous nitrate also leaves manganese dioxide when heated to 200°. It is also obtained from manganous and manganic salts of the alkalis, when they are decomposed in the presence of a small amount of acid; the practical method of converting the salts MnX_{2} into the higher grades of oxidation is given in Chapter II., Note 6.

As the salts of manganous oxide MnX_{2} closely resemble (and are isomorphous with) the salts of magnesia MgX_{2} in many respects (with the exception of the fact that MnX_{2} are rose coloured and are easily oxidised in the presence of alkalis), we will not dwell upon them, but limit ourselves to illustrating the chemical character of manganese by describing the metal and its corresponding acids. The fact alone that the oxides of manganese are not reduced to the metal when ignited in hydrogen (whilst the oxides of iron give metallic iron under these circumstances), but only to manganous oxide, MnO, shows that manganese has a considerable affinity for oxygen--that is, it is difficult to reduce. This may be effected, however, by means of charcoal or sodium at a very high temperature. A mixture of one of the oxides of manganese with charcoal or organic matter gives fused _metallic manganese_ under the powerful heat developed by coke with an artificial draught. The metal was obtained for the first time in this manner by Gahn, after Pott, and more especially Scheele, had in the last century shown the difference between the compounds of iron and manganese (they were previously regarded as being the same). Manganese is prepared by mixing one of its oxides in a finely-divided state with oil and soot; the resultant mass is then first ignited in order to decompose the organic matter, and afterwards strongly heated in a charcoal crucible. The manganese thus obtained, however, contains, as a rule, a considerable amount of silicon and other impurities. Its specific gravity varies between 7·2 and 8·0. It has a light grey colour, a feebly metallic lustre, and although it is very hard it can be scratched by a file. It rapidly oxidises in air, being converted into a black oxide; water acts on it with the evolution of hydrogen--this decomposition proceeds very rapidly with boiling water, and if the metal contain carbon.[20]

[20] Other chemists have obtained manganese by different methods, and attributed different properties to it. This difference probably depends on the presence of carbon in different proportions. Deville obtained manganese by subjecting the pure dioxide, mixed with pure charcoal (from burnt sugar), to a strong heat in a lime crucible until the resultant metal fused. The metal obtained had a rose tint, like bismuth, and like it was very brittle, although exceedingly hard. It decomposed water at the ordinary temperature. Brunner obtained manganese having a specific gravity of about 7·2, which decomposed water very feebly at the ordinary temperature, did not oxidise in air, and was capable of taking a bright polish, like steel; it had the grey colour of cast iron, was very brittle, and hard enough to scratch steel and glass, like a diamond. Brunner's method was as follows: He decomposed the manganese fluoride (obtained as a soluble compound by the action of hydrofluoric acid on manganese carbonate) with sodium, by mixing these substances together in a crucible and covering the mixture with a layer of salt and fluor spar; after which the crucible was first gradually heated until the reaction began, and then strongly heated in order to fuse the metal separated. Glatzel (1889) obtained 25 grms. of manganese, having a grey colour and sp. gr. 7·39, by heating a mixture of 100 grms. of MnCl_{2} with 200 grms. KCl and 15 grms. Mg to a bright white heat. Moissan and others, by heating the oxides of manganese with carbon in the electric furnace, obtained carbides of manganese--for example, Mn_{3}C--and remarked that the metal volatilised in the heat of the voltaic arc. Metallic manganese is, however, not prepared on a large scale, but only its alloys with carbon (they readily and rapidly oxidise) and _ferro-manganese_ or a coarsely crystalline alloy of iron, manganese and carbon, which is smelted in blast-furnaces like pig-iron (_see_ Chapter XXII.) This ferro-manganese is employed in the manufacture of steel by Bessemer's and other processes (see Chapter XXII.) and for the manufacture of manganese bronze. However, in America, Green and Wahl (1895) obtained almost pure metallic manganese on a large scale. They first treat the ore of MnO_{2} with 30 p.c. sulphuric acid (which extracts all the oxides of iron present in the ore), and then heat it in a reducing flame to convert it into MnO, which they mix with a powder of Al, lime and CaF_{2} (as a flux), and heat the mixture in a crucible lined with magnesia; a reaction immediately takes place at a certain temperature, and a metal of specific gravity 7·3 is obtained, which only contains a small trace of iron.

Manganese gives two compounds with _nitrogen_, Mn_{5}N_{2} and Mn_{3}N_{2}. They were obtained by Prelinger (1894) from the amalgam of manganese Mn_{2}Hg_{5} (obtained on a mercury anode by the action of an electric current upon a solution of MnCl_{2}); the mercury may be removed from this amalgam by heating it in an atmosphere of hydrogen, and then metallic manganese is obtained as a grey porous mass of specific gravity 7·42. If this amalgam be heated in dry nitrogen it gives Mn_{5}N_{2} (grey powder, sp. gr. 6·58), but if heated in an atmosphere of NH_{3} it gives (as also does Mn_{5}N_{2}) Mn_{3}N_{2}, (a dark mass with a metallic lustre, sp. gr. 6·21), which, when heated in nitrogen is converted into Mn_{5}N_{2}, and if heated in hydrogen evolves NH_{3} and disengages hydrogen from a solution of NH_{4}Cl. At all events, manganese is a metal which decomposes water more easily than iron, nickel, and cobalt.

It has been shown above that if manganese dioxide, or any lower oxide of manganese, be heated with an alkali in the presence of air, the mixture absorbs oxygen,[21] and forms an alkaline manganate of a green colour: 2KHO + MnO_{2} + O = K_{2}MnO_{4} + H_{2}O. Steam is disengaged during the ignition of the mixture, and if this does not take place there is no absorption of oxygen. The oxidation proceeds much more rapidly if, before igniting in air, potassium chlorate or nitre be added to the mixture, and this is the method of preparing _potassium manganate_, K_{2}MnO_{4}. The resultant mass dissolved in a small quantity of water gives a dark green solution, which, when evaporated under the receiver of an air-pump over sulphuric acid, deposits green crystals of exactly the same form as potassium sulphate--namely, six-sided prisms and pyramids. The composition of the product is not changed by being redissolved, if perfectly pure water free from air and carbonic acid be taken. But in the presence of even very feeble acids the solution of this salt changes its colour and becomes red, and deposits manganese dioxide. The same decomposition takes place when the salt is heated with water, but when diluted with a large quantity of unboiled water manganese dioxide does not separate, although the solution turns red. This change of colour depends on the fact that potassium manganate, K_{2}MnO_{4}, whose solution is green, is transformed into potassium permanganate, KMnO_{4}, whose solution is of a red colour. The reaction proceeding under the influence of acids and a large quantity of water is expressed in the following manner: 3K_{2}MnO_{4} + 2H_{2}O = 2KMnO_{4} + MnO_{2} + 4KHO. If there is a large proportion of acid and the decomposition is aided by heat, the manganese dioxide and potassium permanganate are also decomposed, with formation of manganous salt. Exactly the same decomposition as takes place under the action of acids is also accomplished by magnesium sulphate, which reacts in many cases like an acid. When water holding atmospheric oxygen in solution acts on a solution of potassium manganate, the oxygen combines directly with the manganate and forms potassium permanganate, without precipitating manganese dioxide, 2K_{2}MnO_{4} + O + H_{2}O = 2KMnO_{4} + 2KHO. Thus a solution of potassium manganate undergoes a very characteristic change in colour and passes from green to red; hence this salt received the name of _chameleon mineral_.[22]

[21] Volume I. p. 157, Note 7.

[22] It was known to the alchemists by this name, but the true explanation of the change in colour is due to the researches of Chevillot, Edwards, Mitscherlich, and Forchhammer. The change in colour of potassium manganate is due to its instability and to its splitting up into two other manganese compounds, a higher and a lower: 3MnO_{3} = Mn_{2}O_{7} + MnO_{2}. Manganese trioxide is really decomposed in this manner by the action of water (see later): 3MnO_{3} + H_{2}O = 2MnHO_{4} + MnO_{2} (Franke, Thorpe, and Humbly). The instability of the salt is proved by the fact of its being deoxidised by organic matter, with the formation of manganese dioxide and alkali, so that, for instance, a solution of this salt cannot be filtered through paper. The presence of an excess of alkali increases the stability of the salt; when heated it breaks up in the presence of water, with the evolution of oxygen.

The method of preparing _potassium permanganate_ will be understood from the above. There are many recipes for preparing this substance, as it is now used in considerable quantities both for technical and laboratory purposes. But in all cases the essence of the methods is one and the same: a mixture of alkali with any oxide of manganese (even manganous hydroxide, which may be obtained from manganous chloride) is first heated in the presence of air or of an oxidising substance (for the sake of rapidity, with potassium chlorate); the resultant mass is then treated with water and heated, when manganese dioxide is precipitated and potassium permanganate remains in solution. This solution may be boiled, as the liquid will contain free alkali; but the solution cannot be evaporated to dryness, because a strong solution, as well as the solid salt, is decomposed by heat.

By adding a dilute solution of manganous sulphate to a boiling mixture of lead dioxide and dilute nitric acid, the whole of the manganese may be converted into permanganic acid (Crum).

_Potassium permanganate_, KMnO_{4}, crystallises in well-formed, long red prisms with a bright green metallic lustre. In the arts the potash is frequently replaced by soda, and by other alkaline bases, but no salt of permanganic acid crystallises so well as the potassium salt, and therefore this salt is exclusively used in chemical laboratories. One part of the crystalline salt dissolves in 15 parts of water at the ordinary temperature. The solution is of a very deep _red colour_, which is so intense that it is still clearly observable after being highly diluted with water. In a solid state it is decomposed by heat, with evolution of oxygen, a residue consisting of the lower oxides of manganese and potassium oxide being left.[22 bis] A mixture of permanganate of potassium, phosphorous and sulphur takes fire when struck or rubbed, a mixture of the permanganate with carbon only takes fire when heated, not when struck. The instability of the salt is also seen in the fact that its solution is decomposed by peroxide of hydrogen, which at the same time it decomposes. A number of substances reduce potassium permanganate to manganese dioxide (in which case the red solution becomes colourless).[23] Many organic substances (although far from all, even when boiled in a solution of permanganate) act in this manner, being oxidised at the expense of a portion of its oxygen. Thus, a solution of sugar decomposes a cold solution of potassium permanganate. In the presence of an excess of alkali, with a small quantity of sugar, the reduction leads to the formation of potassium manganate, because 2KMnO_{4} + 2KHO = O + 2K_{2}MnO_{4} + H_{2}O. With a considerable amount of sugar and a more prolonged action, the solution turns brown and precipitates manganese dioxide or even oxide. In the oxidation of many organic bodies by an alkaline solution of KMnO_{4} generally three-eighths of the oxygen in the salt are utilised for oxidation: 2KMnO_{4} = K_{2}O + 2MnO_{2} + O_{3}. A portion of the alkali liberated is retained by the manganese dioxide, and the other portion generally combines with the substance oxidised, because the latter most frequently gives an acid with an excess of alkali. A solution of potassium iodide acts in a similar manner, being converted into potassium iodate at the expense of the three atoms of oxygen disengaged by two molecules of potassium permanganate.

[22 bis] The solution of this salt with an excess of impure commercial alkali generally acquires a green tint.

[23] A solution of potassium permanganate gives a beautiful absorption spectrum (Chapter XIII.) If the light in passing through this solution loses a portion of its rays in it (if one may so account for it), this is partially explained by the increased oxidising power which the solution then acquires. We may here also remark that a dilute solution of permanganate of potassium forms a colourless solution with nickel salts, because the green colour of the solution of nickel salts is complementary to the red. Such a decolorised solution, containing a large proportion of nickel and a small proportion of manganese, decomposes after a time, throws down a precipitate, and re-acquires the green colour proper to the nickel salts. The addition of a solution of a cobalt salt (rose-red) to the nickel salt also destroys the colour of both salts.

_In the presence of acids, potassium permanganate acts as an oxidising agent_ with still greater energy than in the presence of alkalis. At any rate, a greater proportion of oxygen is then available for oxidation, namely, not 3/8, as in the presence of alkalis, but 5/8, because in the first instance manganese dioxide is formed, and in the second case manganous oxide, or rather the salt, MnX_{2}, corresponding with it. Thus, for instance, in the presence of an excess of sulphuric acid, the decomposition is accomplished in the following manner: 2KMnO_{4} + 3H_{2}SO_{4} = K_{2}SO_{4} + 2MnSO_{4} + 3H_{2}O + 5O. This decomposition, however, does not proceed directly on mixing a solution of the salt with sulphuric acid, and crystals of the salt even dissolve in oil of vitriol without the evolution of oxygen, and this solution only decomposes by degrees after a certain time. This is due to the fact that sulphuric acid liberates free permanganic acid from the permanganate,[24] which acid is stable in solution. But if, in the presence of acids and a permanganate, there is a substance capable of absorbing oxygen--for instance, capable of passing into a higher grade of oxidation--then the reduction of the permanganic acid into manganous oxides sometimes proceeds directly at the ordinary temperature. This reduction is very clearly seen, because the solutions of potassium permanganate are red whilst the manganous salts are almost colourless. Thus, for instance, nitrous acid and its salts are converted into nitric acid and decolorise the acid solution of the permanganate. Sulphurous anhydride and its salts immediately decolorise potassium permanganate, forming sulphuric acid. Ferrous salts, and in general salts of lower grades of oxidation capable of being oxidised in solution, act in exactly the same manner. Sulphuretted hydrogen is also oxidised to sulphuric acid; even mercury is oxidised at the expense of permanganic acid, and decolorises its solution, being converted into mercuric oxide. Moreover, the end point of these reactions may easily be seen, and therefore, having first determined the amount of active oxygen in one volume of a solution of potassium permanganate, and knowing how many volumes are required to effect a given oxidation, it is easy to determine the amount of an oxidisable substance in a solution from the amount of permanganate expended (Marguerite's method).

[24] If sulphuric acid is allowed to act on potassium permanganate without any special precautions, a large amount of oxygen is evolved (it may even explode and inflame), and a violet spray of the decomposing permanganic acid is given off. But if the pure salt (_i.e._ free from chlorine) be dissolved in pure well-cooled sulphuric acid, without any rise in temperature, a green-coloured liquid settles at the bottom of the vessel. This liquid does not contain any sulphuric acid, and consists of permanganic anhydride, Mn_{2}O_{7} (Aschoff, Terreil). It is impossible to prepare any considerable quantity of the anhydride by this method, as it decomposes with an explosion as it collects, evolving oxygen and leaving red oxide of manganese. _Permanganic anhydride_, Mn_{2}O_{7}, in dissolving in sulphuric acid, gives a green solution, which (according to Franke, 1887) contains a compound Mn_{2}SO_{10} = (MnO_{3})_{2}SO_{4}--that is, sulphuric acid in which both hydrogens are replaced by the group MnO_{3}, which is combined with OK in permanganate of potassium. This mixture with a small quantity of water gives Mn_{2}O_{7}, according to the equation: (MnO_{3})_{2}SO_{4} + H_{2}O = H_{2}SO_{4} + Mn_{2}O_{7}, and when heated to 30° it gives _manganese trioxide_, (MnO_{3})_{2}SO_{4} + H_{2}O = 2MnO_{2} + H_{2}SO_{4} + O. Pure manganese trioxide is obtained if the solution of (MnO_{3})_{2}SO_{4} be poured in drops on to sodium carbonate. Then, together with carbonic anhydride, a spray of manganese trioxide passes over, which may be collected in a well-cooled receiver, and this shows that the reaction proceeds according to the equation: (MnO_{3})_{2}SO_{4} + Na_{2}CO_{3} = Na_{2}SO_{4} + 2MnO_{3} + CO_{2} + O (Thorpe). The trioxide is decomposed by water, forming manganese dioxide and a solution of _permanganic acid_: 3MnO_{3} + H_{2}O = MnO_{2} + 2HMnO_{4}. The same acid is obtained by dissolving permanganic anhydride in water.

Barium permanganate when treated with sulphuric acid gives the same acid. This barium salt may be prepared by the action of barium chloride on the difficultly soluble silver permanganate, AgMnO_{4}, which is precipitated on mixing a strong solution of the potassium salt with silver nitrate. The solution of permanganic acid forms a bright red liquid which reflects a dark violet tint. A dilute solution has exactly the same colour as that of the potassium salt. It deposits manganese dioxide when exposed to the action of light, and also when heated above 60°, and this proceeds the more rapidly the more dilute the solution. It shows its oxidising properties in many cases, as already mentioned. Even hydrogen gas is absorbed by a solution of permanganic acid; and charcoal and sulphur are also oxidised by it, as they are by potassium permanganate. This may be taken advantage of in analysing gunpowder, because when it is treated with a solution of potassium permanganate, all the sulphur is converted into sulphuric acid and all the charcoal into carbonic anhydride. Finely-divided platinum immediately decomposes permanganic acid. With potassium iodide it liberates iodine (which may afterwards be oxidised into iodic acid) (Mitscherlich, Fromherz, Aschoff, and others). Ammonia does not form a corresponding salt with free permanganic acid, because it is oxidised with evolution of nitrogen. The oxidising action of permanganic acid in a strong solution may be accompanied by flame and the formation of violet fumes of permanganic acid; thus a strong solution of it takes fire when brought into contact with paper, alcohol, alkaline sulphides, fats, &c.

We may add that, according to Franke, 1 part of potassium permanganate with 13 parts of sulphuric acid at 100° gives brown crystals of the salt Mn_{2}(SO_{4})_{3},H_{2}SO_{4},4H_{2}O, which gives a precipitate of hydrated manganese dioxide, H_{2}MnO_{3} = MnO_{2}H_{2}O, when treated with water.

Spring, by precipitating potassium permanganate with sodium sulphite and washing the precipitate by decantation, obtained a soluble colloidal manganese oxide, whose composition was the mean between Mn_{2}O_{3} and MnO_{2}--namely, Mn_{2}O_{3},4(MnO_{2}H_{2}O).

The oxidising action of KMnO_{4}, like all other chemical reactions, is not accomplished instantaneously, but only gradually. And, as the course of the reaction is here easily followed by determining the amount of salt unchanged in a sample taken at a given moment,[25] the oxidising reaction of potassium permanganate, in an acid liquid, was employed by Harcourt and Esson (1865) as one of the first cases for the investigation of the laws of the _rate of chemical change_[26] as a subject of great importance in chemical mechanics. In their experiments they took oxalic acid, C_{2}H_{2}O_{4}, which in oxidising gives carbonic anhydride, whilst, with an excess of sulphuric acid, the potassium permanganate is converted into manganous sulphate, MnSO_{4}, so that the ultimate oxidation will be expressed by the equation: 5C_{2}H_{2}O_{4} + 2MnKO_{4} + 3H_{2}SO_{4} = 10CO_{2} + K_{2}SO_{4} + 2MnSO_{4} + 8H_{2}O. The influence of the relative amount of sulphuric acid is seen from the annexed table, which gives the measure of reaction _p_ per 100 parts of potassium permanganate, taken four minutes after mixing, using n molecules of sulphuric acid, H_{2}SO_{4}, per 2KMnO_{4} + 5C_{2}H_{2}O_{4}:

_n_ = 2 4 6 8 12 16 22 _p_ = 22 36 51 63 77 86 92

showing that in a given time (4 minutes) the oxidation is the more perfect the greater the amount of sulphuric acid taken for given amounts of KMnO_{4} and C_{2}H_{2}O_{4}. It is obvious also that the temperature and relative amount of every one of the acting and resulting substances should show its influence on the relative velocity of reaction; thus, for instance, direct experiment showed the influence of the admixture of manganous sulphate. When a large proportion of oxalic acid (108 molecules) was taken to a large mass of water and to 2 molecules of permanganate 14 molecules of manganous sulphate were added, the quantity x of the potassium permanganate acted on (in percentages of the potassium permanganate taken) in t minutes (at 16°) was as follows:

_t_ = 2 5 8 11 14 44 47 53 61 68 _x_ = 5·2 12·1 18·7 25·1 31·3 68·4 71·7 75·8 79·8 83·0

These figures show that the rate of reaction--that is, the quantity of permanganate changed in one minute--decreases proportionally to the decrease in the amount of unchanged potassium permanganate. At the commencement, about 2·6 per cent. of the salt taken was decomposed in the course of one minute, whilst after an hour the rate was about 0·5 per cent. The same phenomena are observed in every case which has been investigated, and this branch of theoretical or physical chemistry, now studied by many,[27] promises to explain the course of chemical transformations from a fresh point of view, which is closely allied to the doctrine of affinity, because the rate of reaction, without doubt, is connected with the magnitude of the affinities acting between the reacting substances.

[25] For rapid and accurate determinations of this kind, advantage is taken of those methods of chemical analysis which are known as 'titrations' (volumetric analysis), and consist in measuring the volume of solutions of known strength required for the complete conversion of a given substance. Details respecting the theory and practice of titration, in which potassium permanganate is very frequently employed, must be looked for in works on analytical chemistry.

[26] The measurements of velocity and acceleration serve for determining the measure of forces in mechanics, but in that case the velocities are magnitudes of length or paths passed over in a unit of time. The velocity of chemical change embodies a conception of quite another kind. In the first place, the velocities of reactions are magnitudes of the masses which have entered into chemical transformations; in the second place, these velocities can only be relative quantities. Hence the conception of 'velocity' has quite a different meaning in chemistry from what it has in mechanics. Their only common factor is time. If _dt_ be the increment of time and _dx_ the quantity of a substance changed in this space of time, then the fraction (or quotient) _dx/dt_ will express the rate of the reaction. The natural conclusion, come to both by Harcourt and Esson, and previously to them (1850) by Wilhelmj (who investigated the rate of conversion, or inversion, of sugar in its passage into glucose), consists in establishing that this velocity is proportional to the quantity of substances still unchanged--_i.e._ that _dx/dt_ = C(A - _x_), where C is a constant coefficient of proportionality, and where A is the quantity of a substance taken for reaction at the moment when _t_ = 0 and _x_ = 0--that is, at the beginning of the experiment, from which the time _t_ and quantity _x_ of substance changed is counted. On integrating the preceding equation we obtain log(A/(A - _x_)) = _kt_, where _k_ is a new constant, if we take ordinary (and not natural) logarithms. Hence, knowing A, _x_, and _t_, for each reaction, we find _k_, and it proves to be a constant quantity. Thus from the figures cited in the text for the reaction 2KMnO_{4} + 108C_{2}H_{2}O_{4} + 14MnSO_{4}, it may be calculated that _k_ = 0·0114; for example, _t_ = 44, _x_ = 68·4 (A = 100), whence _kt_ = 0·5004 and _k_ = 0·0114, (_see also_ Chapter XIV., Note 3, and Chapter XVII., Note 25 bis).

[27] The researches made by Hood, Van't Hoff, Ostwald, Warder, Menschutkin, Konovaloff, and others have a particular significance in this direction. Owing to the comparative novelty of this subject, and the absence of applicable as well as indubitable deductions, I consider it impossible to enter into this province of theoretical chemistry, although I am quite confident that its development should lead to very important results, especially in respect to chemical equilibria, for Van't Hoff has already shown that the limit of reaction in reversible reactions is determined by the attainment of equal velocities for the opposite reactions.