absorption is constant for a given temperature. The trough was heated by means of an air-bath. Curves having the temperatures in degrees as abscissæ, and the percentages of dehydrated salt as ordinates, were then traced for each sodium chloride solution.

These curves show that in each case the dehydration is continuous; no intermediate hydrate is formed. The actual composition of the final salt, CoCl2 or CoCl2 + Aq, is unknown. The rate at which dehydration takes place is more rapid the higher the temperature. At low temperatures, the curves for different salt solutions are congruent; and as a rule, one curve can be converted into another by moving each point of it along the abscissa through a definite number of units. The course of the conversion is thus always the same, and is independent of the amount of cobalt in solution; the presence of sodium chloride simply transfers it to a lower range of temperature. In a pure cobalt chloride solution, the conversion is very slight below 100°.

Experiments were also made with sodium bromide and iodide. The greatest effect is produced by the chloride, the least by the iodide. With the bromide and iodide, the action is not very regular; probably because the degree of saturation is lowered as the temperature rises.

Сн. В.

Condensation of Water Vapour by Solid Substances. By T. IHMORI (Ann. Phys. Chem. [2], 31, 1006—1014).-A continuation of previous experiments (Warburg and Ihmori, ibid., 27, 481). The author has improved the construction of the balance there described, avoiding the use of sealing-wax, which is hygroscopic, and of brass, which becomes so by oxidation. Platinum is used instead.

Sealing-wax, shellac, and metallic surfaces varnished with shellac, are very hygroscopic. Bright metal surfaces take up very little water; but if oxidation occurs, the surface may become rather hygroscopic. Water taken up by such a surface is only partly given off in a dry vacuum.

Agate is also very hygroscopic: the absorption appears to vary with the colour. Rock crystal is very slightly hygroscopic, especially after cleaning with leather, or, still better, washing with hot water. After the latter treatment, the absorption becomes zero or negative.

Platinum condenses water very slightly; the condensation disappears entirely after rubbing with leather. Old platinum may require heating to redness, probably in order to destroy a film of grease.

In the author's opinion, a sensitive balance should have all metallic parts platinised; neither shellac varnish nor agate should be used in its construction; rock crystal might, perhaps, be substituted for the latter. Platinum or platinised brass is preferable to rock crystal for weights.

Experiments with glass are also described.

CH. B.

Rate of Oxidation of Carbon Compounds by Potassium Per. manganate. By DREYFUS (Compt. rend., 105, 523-525). When the oxidation of carbon compounds by potassium permanganate reaches

its limit, the quantities of oxygen absorbed by equal weights of different compounds are not identical, but are of the same order of magnitude. When, however, the action is restrained, and the rate of oxidation is measured, the results vary widely with different compounds.

The reagents employed consisted of a potassium permanganate solution equivalent to a solution of 0.1 gram of crystallised oxalic acid per litre, and a solution of indigocarmine, 10 c.c. of which was equal to 5 c.c. of the permanganate. A 1 per cent. solution of ethyl alcohol was used as a standard liquid, and all the other solutions were compared with it, the rate of oxidation of ethyl alcohol being taken as unity. 50 c.c. of the alcohol solution was placed in a cylinder, and an equal volume of the solution to be examined in another cylinder, 25 drops of sulphuric acid was added to each, and after two minutes 10 c.c. of potassium permanganate. The cylinders were allowed to remain in diffused light at about 15°, and 10 minutes after the permanganate had been added to the alcohol, the latter was run into 10 c.c. of the indigocarmine solution until the colour of the latter changed to yellow. Two minutes later, the other solution was treated in the same way. A simple calculation gives the quantity of oxygen absorbed by each substance under conditions which are strictly comparable, except that the weights of the substances are not identical. The second solution is diluted to a suitable extent, and after two or three comparative experiments, it is easy to calculate the amount of water to be added to the second substance, in order that the quantity of oxygen absorbed may be the same in both cylinders, and from this we get the weight p of the substance which will absorb as much oxygen in a given time as 1 gram of ethyl alcohol. If the 1 two numbers 1 and. are not actually proportional to the rates of

p

oxidation of equal weights of the two substances, they are of the same order, and may be regarded as coefficients of the relative rates of oxidation of the various compounds.

A large number of substances were examined, and the rates of oxidation vary from 0.2 in the case of saccharose to 10,000 in the case of pyrogallol; next in order to pyrogallol comes catechol, quinol, and resorcinol, with 5000, 3333, and 2000 respectively, phenol 786-0, a-naphthol 769-00, and B-naphthol 666·0. Hydrocarbons, sugars, alcohols of the ethyl series, and acids of the acetic and benzoic series, have much lower rates, varying from 10 to 60. Ether and alcohol show identical velocities.

Determinations of the rate of oxidation may be used to determine the class to which a substance belongs, and also to detect impurities in such compounds as acetone. The constitution of compounds affects the rate of oxidation more than their composition. Other conditions being the same, a saturated compound is less active than a nonsaturated compound. Substances of the same chemical function show comparable rates of oxidation. The aldehyde function is more active than the alcoholic, and the phenolic function shows a very much greater activity. The rates of oxidation of isomerides are not the In the case of the dihydroxybenzenes and the toluidines, the

same.

ortho-derivative absorbs the most oxygen, and the meta-derivative the least, the para-derivative occupying an intermediate position. C. H. B. Laboratory Fittings. By J. GIBSON (J. Soc. Chem. Ind., 6, 205 -211). The following is a list of fittings and apparatus in use at the new chemical laboratories of the University of Edinburgh:(1) System of ventilation and fume extraction, the draught being produced by a Blackman air propeller; the great power of the draught has rendered it possible to dispense with draught cupboards almost entirely; (2) draught arrangements on the working tables; (3) utilisation of the draught in connection with steam-baths; (4) arrangement for supplying the laboratories with water at a high and constant pressure; (5) gas, water, and steam supply to a stone table intended for combustion and similar operations; (5) convenient arrangement of draught, gas, and steam supply to a small draught cupboard; (7) new substitute for a sand-bath; (8) improved filtering apparatus; (9) apparatus for preserving gaseous hydrogen sulphide. D. B.

Inorganic Chemistry.

Compound of Iodine with Ammonia. By F. RASCHIG (Annalen, 241, 253-255).-Iodine absorbs dry ammonia gas, forming a darkblue liquid; Bineau (Ann. Chim. Phys. [3], 15, 80) assigns to this substance the formula 3NH3,21, but according to Millon (Annalen, 62, 54) iodine absorbs less than half the volume represented by this formula. The author finds that the volume of ammonia absorbed by the iodine varies with the temperature. At 20°, the amount of ammonia absorbed corresponds with Bineau's formula, but at 80° it corresponds with NH,1, at 0° with I(NH3)2, and at -10° with I2(NH3). The liquid is decomposed by water, yielding ammonium iodide and nitrogen iodide, but it dissolves completely in alcohol without undergoing any change. W. C. W.

Method for Decomposing Arsenical Sulphides. By H. WARREN (Chem. News, 56, 193-194).-The cobalt speiss or arsenical alloy is digested in hydrochloric acid containing copper nitrate, and after a day or so, the insoluble portions are calcined at a low red heat with plentiful access of air; the calcined mass is then easily dissolved by hydrochloric acid and mixed with the other solution. The copper is separated by means of metallic iron, which also removes some bismuth and arsenic; and the iron and remaining arsenic are precipitated by adding milk of lime. The calcium salts are removed by treatment with sulphuric acid, and the solution containing nickel and cobalt is precipitated by means of sodium carbonate. The precipitate is then suspended in water and chlorine passed to saturation, when the nickel goes into solution, whilst the cobalt remains undissolved. The solu

tion is boiled, and on adding caustic soda the nickel is obtained as hydroxide, which is ignited and reduced in the usual way.

D. A. L.

Zinc Titanates. By L. Lévy (Compt. rend., 105, 378-380).— When a mixture of 6 grams of titanic oxide, 2.5 grams of zinc oxide, and 5 to 10 grams of anhydrous zinc chloride is heated in a glass tube, the reaction is incomplete, and a violet product is obtained which contains an excess of titanic oxide. With excess of zinc chloride, the product is yellowish. If the mixture is heated in a Perrot's furnace in a crucible brasqued with charcoal and rutile, the zinc chloride volatilises, but if the heating takes place in a long porcelain tube closed at one end, a violet or green, crystalline mass is formed, which contains titanium, zinc, silicon, and potassium.

If a mixture of 7 grams of titanium oxide, 5 grams of zinc oxide, and a small quantity of zinc fluoride, or 7 grams of titanium oxide and 30 grams of zinc fluoride is heated under a thin layer of potassium fluoride in a graphite crucible in a Perrot's furnace for an hour and a half, washed with water, and then treated with concentrated sulphuric acid to remove zinc oxide and titanium fluoride, beautiful violet needles are obtained. With potassium chloride in place of the fluoride, the product is a greenish mass. With a mixture of potassium and sodium chlorides the violet needles are obtained mixed with yellowish needles of potassium titanate. The violet crystals are zinc trititanate, ZnO,3TiO2, a small quantity of the zinc being displaced by iron. They are insoluble in water, alcohol, and ether, are not affected by hot dilute sulphuric, nitric, and hydrochloric acids, nor by boiling concentrated solutions of alkaline hydroxides, but are attacked with difficulty by boiling concentrated sulphuric acid, and are decomposed by fusion with potash. They are infusible before the blowpipe, but change to a greenish mass without loss of weight; sp. gr. at 15°: 4-92. The crystals are not attacked by hydrogen at a red heat, but partially volatilise in a mixture of chlorine and hydrogen chloride. When treated with acidified hydrogen peroxide, the latter acquires a characteristic yellow colour, but decomposition is never complete.

C. H. B.

=

Electrolytic Method of preparing Metallic Alloys, &c. By H. WARREN (Chem. News, 56, 153–154).—The following method is recommended for the preparation of alloys such as phosphor-bronzes, silicides, &c. The metal and substance containing alloying material are placed in a deep, conical crucible, through the bottom of which passes a rod of graphite, extending about one inch within the crucible and protected on the outside by an iron tube. The metal is melted and the graphite put in connection with the negative pole, whilst the molten substance on the surface is connected with the positive pole of a battery of two large ferric chloride cells. In this manner silicon copper and silicon-eisen are easily prepared from potassium silicofluoride and the respective metal; the salt being taken in sufficient quantity to form a molten layer 2 inches deep. By some slight variation in the details, phosphor-bronzes can be produced; moreover, native cryolite can be decomposed in contact with metallic zinc, and on

subsequently volatilising the zinc, pure aluminium is obtained. Magnesium, barium, strontium, and calcium have not yielded satisfactory alloys as yet. D. A. L.

Process for obtaining the Rare Earths from the Ceriferous Hainstadt Clays. By J. R. STROHECKER (Chem. News, 56, 175–176). -The author attributes the failure of other chemists to obtain cerium from these clays (compare Abstr., 1886, 678) to the fact that in the presence of more than 0.5 per cent. of iron, the cerium precipitated by oxalic acid and potassium sulphate is much contaminated with iron, the oxide prepared therefrom being so coloured by iron oxide as to be mistaken for that substance. He describes the processes by which he affirms that he has separated the various rare earths from these clays. A. J. G.

Reduction of Aluminium Oxide. By G. A. FAURIE (Compt. rend., 105, 494).-Two parts of pure finely-powdered aluminium oxide is made into a paste with one part of petroleum or some other hydrocarbon, and then mixed with one part of sulphuric acid. When the mass is homogeneous with a pale yellow tint, and begins to give off sulphurous anhydride, it is wrapped in paper and thrown into a crucible heated to above 800° in order to decompose the hydrocarbon. The compact product thus obtained is powdered and mixed with its own weight of a finely-divided metal, the mixture being then heated to a white heat in a plumbago crucible. The regulus after being allowed to cool is found to contain grains of an aluminium alloy in the midst of a metallic powder.

This method of reduction is applicable to silica, calcium oxide, magnesium oxide, &c.

C. H. B.

Halogen Compounds of Gold. By G. KRÜSS and F. W. SCHMIDT (Ber., 20, 2634—2643).-Experiments made with a view to prepare aurous chloride and bromide by the action of chlorine and bromine respectively on gold, gave negative results. The gold is converted into auric compounds in both cases; it is difficult to complete the reaction. It is suggested that the numbers obtained by Thomsen, pointing to the formula Au Cl, were obtained from a product from which the adbering chlorine had not been removed. When gold is warmed in bromine vapour, a black compound is formed which decomposes into its constituents when heated at 100°, even when kept in bromine vapour. The product contains a large amount of unattacked gold together with auric bromide. The compound Au,Br, is not

formed.

The authors conclude that Thomsen's auro-auric chloride and bromide (this Journal, 1877, ii, 485) do not exist. N. H. M.

Gold Sulphides. By L. HOFFMANN and G. KRÜSS (Ber., 20, 2704-2710).-Oberkampf believed that he obtained an auro-auric sulphide, Au2S2, by heating a solution of auric chloride with hydrogen sulphide, but the product, according to Levol and Fellenberg, had a composition varying between that of Au S, and Au2S,, whilst Schrötter and Priwoznik state that an auro-auric sulphide of constant composi