tion at the ordinary temperature; at 50°, slight decomposition begins, and at 90° CaHPO, is precipitated, free phosphoric acid going into

solution.

Acid phosphate of iron, FeH,(PO4)3, is decomposed by water at the ordinary temperature into FePO, and free phosphoric acid; the greater the volume of water, the more complete the decomposition. Acid phosphate of aluminium on the other hand does not decompose, but dries to a white, colourless mass, again soluble in water. It decomposes on heating. Acid phosphate of manganese, which occurs in some superphosphates, forms a bright rose-coloured, crystalline product, which is easily soluble in water, and is not decomposed even by large quantities of water.

The whole of the soluble phosphates of aluminium and manganese existing in superphosphate are dissolved in estimating the soluble phosphates.

In consequence of the decomposition which acid phosphate of iron. undergoes, and the partial decomposition between the iron and monocalcic phosphates, the soluble phosphate obtained by the process recommended by the Magdeburg Conference (namely, extraction with a large volume of water at once, 1,000 c.c. to 20 grams superphosphate) is considerably less than that obtained by the usual process of extracting with successive small quantities.

A comparison of the two methods with a superphosphate containing 49.9 per cent. P2O, gave the following result: Magdeburg method, 45.8 per cent. P2Os; usual method, 47.2 per cent. P2Os. P205.

J. P. L. Anhydrous Magnesium Chloride. By W. HEMPEL (Ber., 21, 897). Crystallised magnesium chloride can be rendered anhydrous and the formation of basic salt prevented by heating it in a current of hydrogen chloride.

W. P. W.

4

Instability of the Double Sulphates, M"SO1, R'1⁄2SO + 6H2O, of the Magnesium Series. By W. DITTMAR (Proc. Roy. Soc. Edin., 124, 219-220).-When to a solution of a known weight of potassium sulphate, dissolved in a quantity of hot water less than sufficient to hold the intended double salt in solution, a known weight of the bivalent sulphate amounting to exactly 1 or 1.1, 1.2-1.5MgO, or FeO per 1K2O, is added, the resulting crystals are found to be coated with potassium sulphate; with 1.2 or 1-3MgO per 1K0, the crystals become more and more pure. The results with ferrous sulphate are similar. E. W. P.

Action of Potable Water on Lead Pipes. By E. REICHARDT (Arch. Pharm. [3], 25, 1049-1052; compare this vol., p. 344).— Present experience shows that waters which act on lead pipes, generally, if not exclusively, contain free carbonic anhydride. The view that lead pipes gradually become coated internally so as to resist this action is still wanting proof. It has been ascertained that the addition of finely-powdered calcium carbonate suffices perfectly to prevent the action; the free carbonic anhydride quickly attacks the carbonate, leaving any excess undissolved.

J. T.

Corrosion of Leaden Water Pipes. By T. CARNELLEY and W. FREW (J. Soc. Chem. Ind., 7, 15-20, and 78).-The results show that the corrosion is much greater with free exposure to the air than by exclusion of the latter; the action of water being diminished to about one-fourth, and that of calcium hydroxide to about one-fifth, when air is excluded. This confirms the observations of previous investigators. The most important result, however, is the great protective action exerted by sand, calcium carbonate, old mortar, calcium silicate, and a mixture of sand and calcium carbonate. The circumstance that the presence of sand has a most important effect in protecting lead from corrosion has already been shown by Crookes, Odling, and Tidy, in a report on the action of water on lead made to the Water Committee of the Corporation of Huddersfield in 1886. In this report, it is shown that the action of soft water on lead is regulated by the amount of silica contained in the water, and also that the most effectual way to silicate a water is to pass it over a mixture of flints and limestone. They do not, however, give any reason or experiment to show why a mixture of these two should be more effectual than either alone. The authors are of opinion that the true explanation is to be found in the formation of calcium silicate, and as a proof assign the fact that calcium silicate itself, either as such or in the form of old mortar, is more effective than either calcium carbonate or silica separately (compare Müller, this vol., p. 225, Reichardt, p. 344, and preceding Abstract).

D. B

Displacement of Copper by Zinc. By A. DESTREM (Compt. rend., 106, 489-492).-Pure zinc introduced into solutions of copper salts with strong acids produces a brown or blackish slightly adherent deposit of copper. In slightly alkaline solutions, the deposit is red and firmly adherent. In the case of copper salts of feeble acids the deposit has the colour of brass and is strongly adherent, especially if the solution is electrolysed for a short time with an anode of metallic

copper.

Pure carefully cleaned zinc was immersed for brief equal intervals of time in solutions of cuprammonium sulphate, cupric acetate, cupric formate, and cupric picrate. At first the weight of the zincincreases, but afterwards it gradually diminishes with each successive immersion and tends to become constant. If there were simply an exchange of copper for zinc, the weight of the zinc should always decrease. The yellow colour of the deposit, its firm adherence to the zinc, and the increase in weight indicate that on the first immersion of the zinc into neutral or feebly alkaline solutions, alloys of zinc and copper resembling brass are formed. If the yellow deposit is touched with a rod moistened with hydrochloric acid, the red colour of copper appears, owing to the more rapid dissolution of the zinc. When the solution of the copper salt is acid, there is no increase in the weight of the zinc.

Similar phenomena are observed in the case of cadmium.

C. H. B. Action of Sulphur Vapour on Copper. By H. N. WARREN (Chem. News, 57, 95).—If sulphur is thrown on to molten copper and

the mass allowed to cool when the reaction has ceased, a button is produced consisting of a core of pure copper enclosed in a uniform shell of copper sulphide; the thickness varying with the quantity of sulphur employed. In a similar manner, a rod of copper 1 inch in width was exposed at a dull red heat for half an hour to the action of sulphur vapour, in this case also, a layer of copper sulphide of uniform thickness was formed, leaving a core of pure copper. Good commercial copper must be used. D. A. L.

Absorption of Carbonic Oxide by Cuprous Chloride. By W. HEMPEL (Ber., 21, 898—900).—Cuprous chloride solution absorbs noteworthy quantities of all gases, and it is therefore necessary to saturate a fresh solution with those gases in a mixture with which it does not enter into combination, otherwise the diminution in volume is greater than that due to the absorption of carbonic oxide; the results of the first analysis in which this solution is employed should consequently be rejected. When once saturated, the cuprous chloride solution gives very concordant results with gaseous mixtures whose composition is tolerably constant as, for example, with coal-gas. If, however, a solution, which has been used with one kind of gas, is employed for the analysis of a mixture of different composition, errors come in owing to the evolution of a portion of those dissolved gases which are absent from or are present in smaller proportions in the new mixture, and to the absorption of others; it is on this account that Drehschmidt found that the volume of the hydrogen increased in his experiments (this vol., p. 88; Ber., 20, 2753). W. P. W.

Basic Aluminium Sulphate. By C. BÖTTINGER (Annalen, 244, 224-227). A basic aluminium sulphate of the composition Al,O,,SO3,6H2O, is deposited when a solution containing 5 grams of aluminium sulphate and 1 grams of common salt is heated at 140° in sealed tubes. It is insoluble in water, and loses 2 mols. H2O at a dull red heat. W. C. W.

Preparation of Potassium Manganate. By A. JOLLER (Arch. Pharm. [3], 25, 970; from Rep. anal. Chem., 1887, No. 33).-Potassium hydroxide (2 mols.) is placed in a crucible, some water is added, and finely divided potassium permanganate (2 mols.) is gradually added with constant stirring and heating. After two hours at a faint red heat, the crucible is cooled, and the manganate placed in a well-stoppered flask to prevent access of air and contact with organic matter. The value of manganate as a reagent depends on its solubility in alkalis without change. The alkaline solution has the greatest tendency to produce with certain oxides precipitates which settle rapidly, and are constant in composition. Again, the solution is so powerful an oxidising agent that certain compounds are readily and perfectly oxidised at ordinary temperatures, whilst permanganate frequently requires heat and a large excess of reagent. Lastly, the end of the reaction is sharply defined, the green colour disappears, and the solution becomes colourless.

J. T.

Combination of Carbon with Iron under Pressure. By W. HEMPEL (Ber., 21, 903).—When tested for its tensile strength, steel is found to give a fracture which is grey in the centre and brighter in colour towards the edge, whilst a fracture of the same specimen obtained in the ordinary way without extension has a uniform appearance. Inasmuch as contraction occurs in the former case previous to fracture, the author considers that owing to the resulting pressure a portion of the carbon present passes into the combined form (compare Spring and van't Hoff, this vol., p. 341), and ascribes the increase in the hardness of iron which occurs in wire-drawing or when it is hammered cold to a similar cause. W. P. W.

Cobaltic Alums. By H. MARSHALL (Proc. Roy. Soc. Edin., 123, 203-204). When mixed solutions of cobalt sulphate and ammonium or potassium sulphate are electrolysed in a divided cell so arranged that the two electrodes are practically in separate vessels, the solution changes to greenish-blue, due to the oxidation of the cobaltous to a cobaltic compound. After crystallisation, a blue ammonium alum is obtained, which when dry is stable, but in solution rapidly becomes reduced, the oxygen which is liberated being in part ozone. potassium alum was obtained but with less ease, but the salt was always mixed with potassium sulphate from which it could not be freed. Nickel salts of a corresponding composition have not hitherto been obtained. E. W. P.

A

New Hydrate of Molybdic Acid. By A. VIVIER (Compt. rend., 106, 601-602).-White crystals of the composition H2MoO., separated from a solution of ammonium molybdate in nitric acid They may also be obtained by heating the ordinary molybdate solution. at 50-60° with its own volume of water for several days.

C. H. B. Titanium Chloride and Titanic Acid. By R. WAGNER (Ber., 21, 960-962).—Commercial titanium chloride has always a yellowish or greenish colour; the former is usually due to ferric chloride, the latter to chlorine and sometimes to certain decomposition products of titanium chloride. A little stannic chloride was found in one sample, but with this exception, no foreign element other than iron could be detected. Chlorine, hydrochloric acid, and ferric chloride can be removed by slow distillation over sodium, other impurities are eliminated by fractional distillation.

A clear solution of orthotitanic acid can be obtained by adding water, in small quantities at a time, to titanium chloride, the mixture being well shaken during the process.

The hydrate of orthotitanic acid, precipitated in the cold from a clear solution in hydrochloric or sulphuric acid, yields, when dried in the air or at a moderate temperature, not only a white but also a green, brown, or black hydrate according to the treatment adopted; the latter is the final product of the orthohydrate.

When moist orthotitanic acid is heated, a remarkable play of colours is observed, and the oxide on further careful heating becomes almost white, but appears citron-yellow while hot; by continued

VOL. LIV.

2 P

heating, the colour becomes more and more brownish. Metatitanic acid becomes brownish-coloured when ignited for a quarter of an hour.

If orthotitanic acid is allowed to remain under water for a long time it is partially converted into metatitanic hydrate.

The colour of titanium dioxide depends on the treatment to which it has been subjected.

F. S. K.

Action of Vanadic Anhydride on Alkaline Fluorides. By A. DITTE (Compt. rend., 106, 270-272).-Sodium fluoride when heated in a closed crucible with an excess of vanadic anhydride yields a crystalline, red mass, and when this is treated with tepid water, it yields an orange solution, which deposits orange-red crystals of the composition 3V2O5,4NaF + 18H2O. The mother-liquor when concentrated yields lemon-yellow crystals of the compound V,O,,4NaF. With an excess of the alkaline fluoride, the product is pale-yellow, and is only slowly attacked by water. If extracted with warm water and the solution evaporated in a vacuum, the compound 3V2O5,4NaF + 18H2O is obtained in transparent, orange-red crystals, and the motherliquor yields small, yellowish-white crystals of the compound V2O5,8NaF + 3H20.

If a boiling saturated solution of sodium fluoride is mixed with excess of soluble vanadic anhydride and filtered, the filtrate deposits brilliant red crystals of the compound V2O,,NaF + 5H2O, and the mother-liquor yields brilliant, lemon-yellow crystals of the compound V2O5,6ÑaF + 5H2O. All these sodium compounds are analogous to those previously obtained with potassium fluoride.

A warm concentrated solution of ammonium fluoride dissolves a large quantity of soluble vanadic anhydride, and the warm liquid deposits the compound V2O5,4NH,F + 4H2O in pale-yellow crystals, very soluble in water. When heated, they lose water and hydrogen fluoride, and yield the compound V2O,2NH,F. If the mother-liquor from the first crystals is mixed with excess of vanadic anhydride, filtered and cooled, it deposits the compound V2O5,4NH,F+ 4H2O in almost white, brilliant, nacreous needles, readily soluble in cold water. With a cold solution of ammonium fluoride and excess of vanadic anhydride, the first product consists of indistinct yellow crystals of the composition V2O,8NH,F+ 4H,O, followed by crystals of the compound V2O5,4NH,F+ 4H2O.

Vanadic anhydride reacts with many other metallic fluorides, and the products are under investigation.

C. H. B.

Action of Hydrogen Chloride on Cupric Chloride. By ENGEL (Compt. rend., 106, 273-275).-Hydrochloric acid precipitates cupric chloride from its solution in accordance with the author's law-that is to say, the sum of the equivalents of acid and salt in solution remains practically constant. At first the sum of the equivalents slightly diminishes, but when 10 c.c. of the solution contains 70 equivalents of hydrogen chloride the law is exactly obeyed. Beyond this point, the sum of the equivalents increases slowly as the proportion of free acid increases.