current of ammonia gas is passed into the cooled solution, the compound CdCl2,5NH, separates in small, anhydrous crystals. Another preparation gave crystals of the composition CdCl2,4NH, + H2O. Both compounds readily decompose. Divers prepared the compound ZnCl2,5NH, + H2O by a similar method, and some years ago the author obtained the anhydrous compound ZnBr2,5NH,. When cupric chloride is treated in the same way, it yields the compound (CuCl2,5NH3)3 + 3H2O.

As in Divers's process, the author dissolved the precipitated cadmium compound in ammonia at a gentle heat; when the liquid cooled it deposited large octahedra of the compound (CdCl2,3NH3) + H2O, which alter rapidly immediately on removal from the motherliquor. In presence of water, it is decomposed with formation of a bulky, white precipitate. Other experiments, in which heating was continued for a longer time and the conditions of cooling were somewhat different, gave the compound (CdCl2,4NH3)2 + H2O. When the copper compound is dissolved in ammonia, it recrystallises without alteration on cooling.

The compound (CdCl2,2NH3)2 + H2O, described by Crofts, is readily obtained by adding a saturated solution of cadmium chloride to well-cooled ammonia, and evaporating at a gentle heat. The author also confirms Hauer's statement that when cadmium oxide is boiled for a long time with ammonium chloride it yields the compound CdCl2,4NH,Cl. Cupric oxide, under the same conditions, yields no double salt; zinc oxide likewise yields no double salt, but forms the compound (ZnCl2,2NH3)2 + H2O. C. H. B.

Ammoniacal Compounds of Cadmium Sulphate and Nitrate. By G. ANDRÉ (Compt. rend., 104, 987-990).—When a current of ammonia gas is passed into a solution of cadmium sulphate in ammonia, the compound CdSO,,4NH3,2H2O separates as a crystalline precipitate composed of small needles. The same substance is obtained in somewhat larger crystals by dissolving the precipitate in the ammonia at a gentle heat, and allowing the solution to cool. The ratio of cadmium to ammonia is the same as that of copper to ammonia in the corresponding compound, but the cadmium is not completely precipitated. This compound is also obtained by pouring a layer of alcohol on a solution of cadmium sulphate in ammonia. When prepared in this way, it does not lose ammonia so readily as when prepared by the other methods.

Cadmium oxide dissolves somewhat readily in ammonium sulphate, but even after prolonged action the crystals which separate are variable mixtures of ammonium sulphate and cadmium ammonium sulphate. If, however, a solution of ammonium sulphate saturated in the cold is heated with cadmium oxide for several hours, the solution carefully concentrated, and the crystals of ammonium sulphate separated as they form, the compound CdSO,,3(NH,)2SO, + 10H20 is obtained in crystals. Copper oxide and ammonium sulphate under the same conditions yield crystals of ammonium sulphate coloured blue by a small quantity of copper. Zinc oxide treated in the same

way yields crystals of the compound ZnSO,(NH、)2SO、 + 7H2O, mixed with some ammonium sulphate. The salt obtained by mixing the two sulphates has the same composition.

If cadmium nitrate crystals are added to 20 per cent. ammonia, they at first dissolve rapidly, but the addition of a further quantity produces a bulky, crystalline precipitate which redissolves on gentle heating. When the solution cools, it deposits crystals of the composition Cd(NO3)2,6NH3 + 2H2O; these, when heated, melt with evolution of a little water, and then blacken with a slight explosion. The same substance is obtained in bulky, anhydrous crystals, when ammonia gas is passed into a solution of cadmium nitrate in ammonia. Both the anhydrous and the hydrated salts are decomposed by water with formation of an amorphous, white precipitate. Under similar conditions, copper nitrate yields an analogous compound. The author was unable to obtain the other ammoniacal copper nitrates described by Berzelius and by Kane.

The ammoniacal zinc chloride recently described by Thoms (this vol., p. 551) was prepared by the author three years ago (Ann. Chim. Phys. [6], 3, 84 and 98).

C. H. B.

Effect of Manganese and other Substances on the Properties of Steel. By F. OSMOND (Compt. rend., 104, 985—987).-Manganese retards the molecular change of the iron and recalescence during cooling from a high temperature, or in other words keeps the carbon in solution and the iron in the condition B, the effect being greater the greater the proportion of manganese. The same effect is produced by the rapid cooling of non-manganiferous steel, so that the presence of manganese exerts much the same influence as the process of tempering. Tungsten has the same property in a still more marked degree, but chromium seems to produce no similar effect. Silicon has no influence on the effect of the manganese, but sulphur seems to combine with part of the manganese, and thus diminishes its action. Phosphorus has no appreciable effect on the modification of the iron, nor on recalescence.

Every foreign substance in the steel produces its own peculiar effects, but it is to carbon alone that steel owes its characteristic differences from the other forms of iron. C. H. B.

Alkaline Vanadates. By A. DITTE (Compt. rend., 104, 902— 905, and 1061-1064).-When a solution of vanadic anhydride (1 mol.) in a solution of potassium oxide (1 mol.) is evaporated in a vacuum, it deposits slender needles of the salt 2K2V2O。 + 5H2O, and the mother-liquor yields white, silky, nacreous needles of the hexahydrate K2V2O + 6H2O. If potassium oxide is in excess, the vanadate separates in brilliant, silky, elongated prisms, which only contain 4 mols. H2O. A solution of vanadic anhydride in an equivalent quantity of potassium carbonate deposits the hydrate K2V2O + 3H2O in stellate groups of needles.

All these hydrates lose their water when heated, and melt to a pale-yellow liquid which solidifies to a white, nacreous mass of the anhydrous vanadate, which has a lamellar fracture. The anhydrous

vanadate is only slightly soluble in water, whilst the hydrates are readily soluble.

A solution of potassium carbonate saturated with an excess of vanadic anhydride at 80° forms a garnet-red liquid, which on cooling deposits the compound K,0,2V20s + 10H2O in orange-red plates. At a higher temperature, the crystals contain 8H2O, and have a deeper colour. If the solution is acidified with acetic acid and concentrated at 80°, the salt containing 10H2O is deposited in brilliant, transparent, hexagonal plates, whilst if crystallisation takes place at a higher temperature, the crystals are orange, and contain 3H2O. If these crystals are boiled in the mother-liquor they alter as the liquid becomes more concentrated, and form small, deep-red, very brilliant crystals of the anhydrous divanadate K2O,2V20s. When the motherliquor from a previous crystallisation, containing a certain proportion of potassium acetate, is concentrated by boiling and allowed to cool, it deposits orange-red plates of the composition 2K,O,3V2O, + 6H2O.

When an excess of vanadic anhydride is dissolved in potassium carbonate and mixed with a large quantity of acetic acid, a garnet-red solution is obtained, and if this is heated at 70° it deposits orange crystals of the compound K,0,3V20, + 5H2O. If the mother-liquor is filtered off and allowed to cool, transparent, garnet-red crystals of the hydrate K20,3V2O, + 5H2O separate. When heated, all these hydrated acid salts lose their water and melt to a brown liquid, which solidifies to an almost black, crystalline mass, only slightly soluble in water.

If a solution of vanadic anhydride (1 mol.) in water containing rather more than 2 mols. of potassium oxide is filtered and evaporated in a vacuum, it yields colourless, transparent crystals of the salt 2K2O,V2O + 4H2O, which lose water when heated, then melt, and solidify on cooling to a crystalline mass of the anhydrous salt, 2K2O,V2OS.

A solution of 3 mols. potassium oxide and 1 mol. of vanadic anhydride, yields very deliquescent, colourless, transparent, channelled crystals of the vanadate 3K,O,V2Os with 9 or 12 mols. H2O, according to the temperature at which crystallisation takes place. It readily forms supersaturated solutions.

A solution of vanadic anhydride containing a large excess of potash deposits no crystals, but if it is agitated with alcohol an oily layer separates, and when this is washed with alcohol it solidifies to a very hygroscopic crystalline mass with a radiating structure. When this is dried on porous plates in a vacuum, it forms brilliant white needles of the composition 4K,O,V2O, + 20H2O; these lose their water when heated, and yield the white anhydrous salt which fuses with difficulty even at a red heat.

The compound Na2O,V2Os, formed when vanadic anhydride (1 mol.) is dissolved in a solution of sodium oxide (1 mol.), crystallises with difficulty, but if a large quantity of the solution is slowly concentrated it deposits translucent nodules consisting of slender, radiating needles of the composition Na2O,V2O5 + 4H2O. If the syrupy solu tion which will not crystallise is treated with a mixture of alcohol and water, it deposits an oily layer which quickly solidifies and can be

recrystallised from dilute alcohol, when it forms brilliant white, silky needles of the hydrate Na2O,V2O5,5H2O.

When vanadic anhydride is dissolved in a solution of an equivalent quantity of sodium carbonate, the liquid crystallises with difficulty, but when treated with alcohol it yields white needles of the composition Na2O,V2O, + 6H2O or Na2O,V2O, + 8H2O, according to the temperature.

If somewhat more than an equivalent quantity of vanadic anhydride is dissolved in a boiling solution of sodium carbonate, the liquid on cooling deposits a crystalline crust consisting of distinct red transparent crystals of the composition Na2O,2V2Os + 5H2O. If a solution of the normal vanadate is acidified with acetic acid and concentrated, it yields garnet-red, channelled needles of the composition Na20,2V20 + 10н2O.

The compound 2Na2O,3V2O, + 18H2O is obtained in friable hexagonal tables, together with the preceding compound, by concentrating an acidified solution of the normal vanadate. It is also obtained in red, prismatic crystals with only 16H2O, by dissolving an excess of vanadic anhydride in sodium hydroxide solution, acidifying the cold liquid with acetic acid, and concentrating at about 50°.

A hot solution of soda saturated with excess of vanadic anhydride and concentrated by boiling, deposits brilliant, orange-red plates of the composition Na2O,3V205 + 3H 0.

All the hydrated acid salts, when heated, lose water and become deep brown, melt at a higher temperature, and solidify on cooling to a crystalline mass of the very slightly soluble anhydrous salt.

When vanadic anhydride (1 mol.) is dissolved in a solution containing 2 mols. of sodium oxide, filtered, a small quantity of alkali added, and the solution concentrated in a vacuum, it yields colourless, transparent plates of the composition 2Na,O,V,O,+ 18H,O. They melt to a colourless liquid which loses water, forming a white substance; this afterwards becomes pale yellow, and is finally converted into the white, very deliquescent, anhydrous salt. If the solution is evaporated to dryness, and the residue extracted with water and alcohol at about 60°, the liquid deposits brilliant needles of the composition 2Na2O,V2Os + 8H2O on cooling.

If vanadic anhydride is dissolved in a solution of three equivalents of sodium oxide and the liquid slowly concentrated, it deposits bulky colourless, transparent prisms of the disodium salt, and afterwards brilliant white, silky needles of the composition 3Na2O,V2O, + 26H2O. This salt is also often deposited in nodules which contain only 24H2O.

When vanadic anhydride is dissolved in a large excess of soda, the solution yields brilliant white needles of the salt 4Na2O,V2O¿ + 30H2O. If crystallisation takes place in a warm liquid, the crystals contain only 26H2O. When the crystals are heated they effloresce, melt in their water of crystallisation, and are finally converted into the very soluble anhydrous salt, which does not melt even at a bright red heat. In this respect, it resembles the corresponding potassium salt. C. H. B.

VOL. LII.

2 น

Crystallised Niobic Anhydride. By A. KNOP (Zeit. Kryst. Min., 12, 610-615).—-The author has obtained niobic anhydride in crystals. Borax glass was fused in a platinum crucible, and saturated with niobic anhydride, of which material 30 grams were at his disposal. On cooling slowly, the excess of anhydride crystallised out. It was found that 9.42 per cent. of niobic anhydride was contained in the clear borax glass, and separated out in an amorphous condition when the mass was dissolved in hot water. The larger crystals are in the form of cubes, and appear to belong to the regular system. Their optical properties, however, do not point to the regular system, but show that the crystals must be regarded as rhombic combinations of the three pinacoids, OP, P∞, and coco. All the plane angles are exactly 90°. It is therefore doubtful whether niobic anhydride crystallises in the regular system with optical anomalies, or in the rhombic with a molecular system approaching that of the regular. Ebelmen (this Journal, 1848, 181), who prepared crystallised niobic anhydride from boric acid, was unable to measure the crystals he obtained on account of the small quantity of material at his disposal. He merely stated that they were prismatic. B. H. B.

Ammoniacal Platinum Compounds. By A. Cossa (Gazzetta, 17, 1-11). After reference to the experiments of Reisel, Cleve, and others on ammoniacal platinum compounds, the results of experiments are given on the admixture of solutions of platinic chloride or sodium platinochloride with platosodiamine chloride. At ordinary temperatures, if the solutions are neutral, a yellow, amorphous compound, of the composition and properties of a platosodiamine platinochloride, is produced. This changes slowly at ordinary temperatures, more rapidly on boiling, into platinodiamine platinochloride (Cleve's salt), when the original substances are mixed in equimolecular proportions, or into the green salt of Magnus and platinodiamine chloride if in the original mixture the proportion of platosodiamine chloride predominates. The results of experiments with various mixtures are given in full. It is evident that the platosodiamine platinochloride has the same percentage composition as platinodiamine platinochloride; thus the yellow, amorphous substance may only be a so-called physical isomeride of the red, crystalline salt obtained directly at temperatures above 60°. In order to determine this point, the reaction between the yellow substance and potassium platinochloride is examined; for if the former is platosodiamine platinochloride it should give by double decomposition platosodiamine platinochloride (Magnus' salt) and potassium platinochloride, thus: Pt(NH3),Cl2,PtCl + 2KCl,PtCl2 = Pt(NH3),Cl2, PtCl2 + 2KCl,PtCl. This was confirmed by experi

ment.

V. H. V.