a gentle heat. This solution, when diluted to measure six fluid ounces, will contain to the fluid drachm two grains of the ferric succinate, Fe, O(C4 H4O4)2, or five grains of the double salt.

Two New Oxides of Bismuth. M. M. P. Muir. (From a paper read before the Chemical Society, June 16th, 1881.) The pure coloured compound formed during the action of potassium cyanide solution on a hot nitric acid solution of bismuth nitrate, is an oxide of the formula Big O, and not Bi, 05, 2 H, O, as asserted by Boedeker. If this heptoxide is treated with hot concentrated solution of potassium hydrate, it is partially dissolved, leaving a reddish brown residue, which, after washing with hot water, proves to be an oxide of the formula Bi 07.

The physical and chemical characters of both oxides are described in the paper.

Bismuth Subnitrate. A. Riche. (Journ. de Pharm. et de Chim. [5], 384.) The author finds that the majority of commercial samples of this preparation contain lead and arsenic, the former to the extent of 0.03-0.34 per cent. (sulphate), and the arsenious acid to 0.002-0.01 per cent.

Analysis of Bismuth Subnitrate. E. Baudrimont. (Journ. de Pharm. et de Chim. [5], 368.) The method recommended by the author for the estimation of the nitric acid in this preparation consists in the decomposition of a weighed quantity of the salt with a known volume of a standard soda solution, and the determination of the excess of alkali in a measured portion of the filtrate. The soda solution employed by him contains 7:407 grains of Na HO per litre, and the corresponding sulphuric acid 9.074 of H, SO4 per litre. 1 grain of the subnitrate to be examined is boiled for ten minutes in a flask with 20 c.c. of the soda solution and 30 c.c. of water, the boiled mixture, after cooling, made up to 100 c.c., and passed through a dry filter. 50 c.c. of the filtrate are now titrated with the sulphuric acid, the number of c.c. of acid used deducted from 50, and the difference calculated for nitric acid.

The bismuth is estimated by washing, drying, and weighing the precipitate, or by igniting a fresh portion of the subnitrate in a crucible, and weighing the residual oxide.

Separation of Copper from Cadmium. G. Vortmann. (Zeitschr. für analyt. Chem., 1881, 416.) The author's method is based on the different behaviour of sodium hyposulphite towards solutions of the two metals; the copper is precipitated as sulphide, while the cadmium remains in solution.

The diluted sulphuric or hydrochloric acid solution of the two

metals is mixed with sufficient hyposulphite to cause complete decolorization, and the mixture boiled for about five minutes. The precipitated cuprous sulphide is collected and washed, then mixed. with sulphur and ignited in a current of hydrogen. The cadmium may be precipitated from the filtrate either as carbonate or sulphide.

Decomposition of Verdigris by Water at High Temperatures. P. Cazeneuve. (Journ. de Pharm. et de Chim. [5], 409.) When ordinary verdigris is heated with water at 200° C. in sealed tubes, it is at first decomposed into cupric oxide and neutral acetate. As the heat is continued carbon dioxide is evolved, and the cupric oxide is ultimately converted into cuprous oxide. At the same time glycolic acid is formed, part of which combines with the lime which is nearly always present as an impurity in the verdigris.

Distribution of Copper in the Animal Kingdom. G. Bizio. (Gazz. Chim. Ital., x., 149.) This paper gives a complete historical survey of the labours of the various chemists who have investigated this subject.

A New Sulphate of Aluminium. P. Marguerite. (Comptes Rendus, xc., 1354-1357.) The new sulphate described by the author has a composition answering to the formula Al1⁄2 Og, 2 S O, 12 H, O. It is obtained by heating ammonium alum until it is reduced to anhydrous aluminium sulphate, and then carefully raising the temperature to expel one-third of the sulphuric acid. The salt crystallizes in rhombohedrons, and is soluble in water at an ordinary temperature to the extent of 45 per cent. It may be freed from any undecomposed portion of the ordinary sulphate by fractional crystallization.

Preparation of Potassium Iodide from Kelp. E. Allary and J. Pellieux. (Bull. de la Soc. Chim. [2], xxxiv., 627-630. From Journ. Chem. Soc.) The mother-liquors, from which the chlorides and sulphates have been separated, are evaporated to dryness in a furnace of special construction, and the residue is carefully roasted, until all the sulphur is either expelled or oxidized. No iodine is lost in this operation. It is necessary to stop the roasting as soon as the sulphur compounds are completely oxidized; this may be ascertained by withdrawing samples from time to time. The calcined mass is broken up and subjected to methodical lixiviation with cold water in a small Shanck's apparatus. The solution, when evaporatd to dryness, gives a white salt which contains about 50 per cent. of iodides. This salt is powdered, and introduced into a Dorvault's digester, which may be used either as an extractor

or as a still. In this it is treated with warm alcohol, which dissolves out the iodides. When the extraction is complete, the alcohol is distilled off in the same apparatus, and used again. The salt thus obtained is a mixture of potassium and sodium iodides, containing on an average 34 per cent. of the former and 66 per cent. of the latter. To convert the sodium iodide into the potassium salt, its amount is determined, and to the saturated aqueous solution of the mixed salts is added a saturated solution of potassium carbonate, containing an amount equivalent to the sodium iodide present, and a stream of cooled carbonic anhydride from the furnaces is passed into the mixture. The following reactions take place: (1) 2 Na I + K2 CO2 = 2 KI+ Na, C Og; (2) KI+Na2 CO3 + H2 O + CO2 = KI+ 2 Na HC Og. When the conversion of the sodium carbonate into bicarbonate is complete, the crystals of the latter are removed, and the small quantity of bicarbonate remaining in solution is exactly decomposed by means of hydrochloric acid. The solution then contains potassium iodide, mixed with a small quantity of sodium chloride, which may be separated by repeated crystallization. All the residues are worked up in the treatment of fresh quantities of kelp, and thus loss of iodide is avoided.

2

2

To prepare pure potassium iodide, the salt obtained by the above method is treated with alcohol, and the dissolved iodide recrystallized.

Estimation of Chlorine in Potassium Iodide and Potassium Bromide. O. Schlickum. (Pharmaceut. Zeitung, 1881, 45.) The principle of the author's method of determining the chlorine present as an impurity in potassium iodide, consists in the precipitation with an excess of silver nitrate, treatment of the washed precipitate with solution of ammonia to extract the silver chloride, and the addition to the ammoniacal filtrate of a standard solution of potassium iodide until the silver is completely precipitated. The standard solution he employs contains one per cent. of potassium iodide. As it would be impracticable to insist on the absolute freedom from chloride, he considers an allowance of one per cent. of chlorine in the iodide as a fair limit, which ought not to be exceeded. The presence of more than one per cent. may be readily detected in the following manner :

One gram of the iodide to be tested is dissolved in distilled water, and the solution made up to 100 c.c. Of this solution 10 c.c.= 0.1 gram of KI, are precipitated with an excess of silver nitrate; the precipitate is collected on a small filter, well washed with distilled water, then repeatedly exhausted with small quantities of ammonia

solution (amounting in all to about 5 or 6 c.c.), mixing the entire ammoniacal filtrate with one-half c.c. of the same potassium iodide solution (the quantity corresponding to one per cent. of chlorine), filtering, and testing one-half of the clear filtrate with another drop of the KI solution, and the other half with silver nitrate. The latter ought to produce a precipitate, but not the former.

Potassium bromide is tested in a similar manner. Of the solution containing one per cent. of the salt, 30 c.c. are submitted to the test, and the ammoniacal filtrate obtained as above is mixed in this case with 1 c.c. of the K Br solution. The filtered mixture ought to form a precipitate with silver nitrate, but none with a further drop of the bromide solution.

Colorimetric Determination of Chlorine in Potassium Bromide. C. Roth. (Chem. News, from Correspondenz Blatt des Ver. deutsch. chem., 1880, No. 15.) One gram of potassium bromide is ground to a powder with an approximately equal quantity of potassium bichromate, placed in a flask holding 100 c.c., and covered with 5 c.c. concentrated sulphuric acid. The flask is then connected air-tight, by means of an adaptor ground to fit its mouth, with a receiver containing 100 c.c. very dilute ammonia (five or six drops of caustic ammonia to 100 c.c. of water). A gentle heat is applied and raised to about 128°. There should be two large bulbs blown on the connection tube, to prevent the reflux of the liquid. When all the chlorine has thus been expelled, the distillate is compared with solutions of ammonium chromate of known strength prepared for the purpose.

Impure Potassium Bromide. O. Maschke. (Pharmaceut. Zeitung, 1880, 728.) The author calls attention to the occasional presence of not inconsiderable quantities of lead in this salt. The crystals thus contaminated do not form a perfectly clear solution with water. Sulphuric acid fails to indicate the lead, as lead sulphate is soluble in the solution of potassium bromide; but potassium chromate or ammonium sulphide at once reveals its presence.

Potassium Tetrachromate. G. Wyrouboff. (Bull. de la Soc. Chim. [2], xxxv., 162.) The author shows that the salts described by Darmstaedter as potassium nitro-dichromate and nitro-trichromate consist of potassium tetrachromate containing some nitrate as an impurity, which it is difficult entirely to remove.

Preparation of Potassium Ferricyanide. K. Seuberlich. (Dingl. polyt. Journ., ccxxxviii., 482, and Amer. Journ. of Pharm., 1881, 233.) The author has tested the conditions under which ferrocyanide of potassium is changed by the action of lead peroxide

in alkaline solution into the ferricyanide of potassium. His results agree with those previously obtained by Lunge, who found that a complete change was only reached when the liberated alkali was neutralized by an acid. The change succeeds perfectly in the cold when the solution of ferrocyanide is treated with lead peroxide, and then a slight excess of dilute hydrochloric acid added with constant stirring. When red lead is used higher temperatures and larger excess of acid are necessary.

The change to ferricyanide can also be effected by the aid of manganese dioxide, even in the cold, if to 1 molecule of ferrocyanide of potassium 1 of the manganese dioxide is used. In both cases there is obtained from the filtrate, after neutralizing with soda, a very pure salt, although in the second case the liquid is difficult to filter. The author believes that by the addition of carbonates and by the blowing in of a current of air, the manganese sesquioxide precipitate can be oxidized so that it will give no trouble in washing. Upon adding a base, this manganese precipitate is readily converted into dioxide again. Either of the above-mentioned methods seems adapted to replace the old chlorine method.

A New Test for Potassium. L. L. de Koninck. (Zeitschr. für analyt. Chem., 1881, 390.) A ten per cent. solution of sodium nitrite, mixed with a little cobalt chloride and acetic acid, forms a yellow precipitate with potassium solutions, even if the latter be so weak that platinum perchloride fails to precipitate them. This test is a reversed application of the well-known reaction between cobalt and potassium nitrite. With ammonium salts the test is much less delicate than with salts of potassium. With magnesium, calcium, barium, strontium, aluminium, and iron, this reagent produces no precipitates.

Preparation of Chemically Pure Soda. MM. Endemann and Prochazka. (Chem. Industrie, iii., 273.) The method recommended by the authors is based upon the fact, observed by Gerresheim, that any chlorine or sulphuric acid present in soda may be completely removed from it by shaking the solution with Millon's base, the product obtained by the action of ammonia on mercuric oxide. For the purpose of this purification, 2 litres of soda solution are agitated with 30 grams of the said base, twice daily for a week, after which the separation of chlorine and sulphuric acid is complete. As Millon's base always contains some free ammonia, however well it may be washed, the authors suggest the addition of some mercuric oxide to the soda solution, for the removal of this ammonia.