THE ANALYST. 381 INORGANIC ANALYSIS. H. A. Guess. (Trans. Am. Inst. Min. Eng., 1904; through Chem. Zeit. Rep., 1904, xxi., 287.)-The author suggests a chromic acid method as being superior to the usual titration with ammonium molybdate or ferrocyanide. Excese of standard potassium chromate is added to the lead solution, filtered, and to the filtrate 25 C.C. of 50 per cent. HC1 and 0.5 gramme KI are added. The liberated iodine is then titrated with thiosulphate, and this gives the excess of chromic acid. For a more rapid method 1 to 5 grains of the ore are heated with 3 to 5 C.C. of HNO, and 15 C.C. of HCl. To the mixture, evaporated to 8 C.C. and rendered alkaline with ammonia, excess of 80 per cent. acetic acid is added. Should any lead remain undissolved, concentrated ammonium acetate is added until solution is complete.After filtration, an excess of chromate is added, the whole again filtered, and the precipitate washed with 0.5 per cent. acetic acid. The filtrate is then titrated as above. Commercial Lead Analysis. Great accuracy is claimed for this method. H. A. T. The Assay of Alloys of Platinum, Gold, and Silver. A. Hollard and L. Bertiaux. (Bull. SOC. Chinz., 1904, xxxi., 1030-1034.)-The ordinary method of assaying makes no allowance for the solubility of the gold in acids or for the reten- tion of lead by the silver button obtained by cupellation, though the determination of the platinum and gold together by cupellation with lead and silver and treatment of the button with sulphuric acid gives accurate results. The loss in the determina- tion of gold on cupelling about 0.3 gramme of the alloy containing 0.2 gramme of gold and 0.1 gramme of platinum in the presence of 3 parts of silver, and treatment of the button with 20 C.C.of nitric acid of specific gravity 1-16 (20" Bif.), and then with 30 C.C. of nitric acid of specific gravity 1-28 (32" B6.), amounted on the average to 0.002 gramme. This amount ought, therefore, to be added as an empirical factor to the weight of gold obtained. Silver.-The following method is recommended : The finely-divided alloy is treated with aqua reyia consisting of 1 part by volume of nitric acid and 5 parts of hydro- chloric acid, and the solution evaporated to the consistency of a syrup on the water- bath. The residue is taken up with nitric acid and the liquid evaporated to dryness, this being twice repeated to eliminate the hydrochloric acid.Finally, the residue is taken up with 2 C.C. of nitric acid, 2 drops of hydrochloric acid, and several C.C. of water, and the liquid boiled, when the mixture ought practically to dissolve. The solution is now diluted to about 100 C.C. and again boiled, and the silver chloride collected, washed, and dissolved in 30 C.C. of a 20 per cent. solution of potassium cyanide. This solution is diluted to 150 c.c., and electrolysed with a feeble current382 THE ANALYST. (0.1 amp8re) in the authors’ small apparatus. dilute nitric acid and tha silver determined volumetrically. Lastly, the deposit is dissolved in C. A. M. The Separation of Palladium by Acetylene. H. Erdmann and 0.Makowka. (D. Chem. Ges. Ber., 1904, 2694; through Chem. Zeit. Rep., 1904, xxi., 253.)- Palladium is unique among metals in being precipitated from an acid solution by acetylene gas or by water saturated with acetylene. I n a strongly acid solution the separation is quantitative. The flocculent precipitate is pale brown in colour, very soluble in ammonia, is not explosive, and, on ignition, yields pure palladium. The authors state that the presence of copper does not interfere with the reaction. H. A. T. The Determination of Carbon and Sulphur in Iron and Steel. A. Mueller. (Chem. Zeit., 1904, lxviii., 795.)-The author has designed an improved and more compact form of the Corleis apparatus for carbon deter- mination. I n that apparatus any carbon given off as hydrocarbon vapour passes through a platinum combus- tion tube, and is absorbed by soda-lime as CO,. The platinum capillary combustion tube is given the shape shown in Fig.2, thus great heating surface is obtained in a small space. The short connecting portion takes up all strains, so that the life of the tube is much prolonged. A special compound gas-absorber is used for drying the gas. The lower, spiral part of this contains phosphor-sulphuric acid, and the short, straight portion (above the small bulb) is filled with glass-wool and phosphorus pentoxide. Long rubber connecting-tubes for the cooler are avoided by making the standard out of a piece of wide tube, into which a smaller is inserted. For the determination of sulphur, the gases pass through a combustion tube of quartz glass, and are finally absorbed in cadmium acetate solution.This apparatus, which has proved most satisfactory, is manufactured by Messrs. C. Gerhardt, Marquart’s Lager Chemischer Utensilien, Bonn, The construction can be seen from the figure. EL A. T.THE ANALYST. 383 A Method of Determining Chromium and Iron in the Presence of each Other. (Zeit. anal. Chem., 1904, xxxiv., 506-508.)-This is based upon the fact recorded by Zimmermann that chromic salts are quantitatively reduced by zinc and sulphuric acid to chromous salts, and also on the fact that ferric salts are reduced by sulphur dioxide to ferrous salts, whilst chromic salts remain unaltered. The solution of the mixed salts, which must not contain more than 0.05 gramme of chromic oxide, is treated with sulphur dioxide, the excess of which is expelled by boiling the liquid, while a current of carbon dioxide is passed through it.After cooling, the amount of iron is determined by titration with standard permanganate solution. The solution is now treated with zinc and sulphuric acid, and heated on the sand-bath until it assumes a clear blue colour, after which it is again oxidized by permanganate, as in Zimmermann’s method (Liebig’s Am., 1882, ccxiii., 322), the amount of chromium being obtained from the difference between the two titrations. The reactions underlying the method are shown in the following equations : B. Glasmann. II. REDUCTION. (a) Cr,O, + Ee,O, + SO, = SO, + 2Fe0 + Cr20,. ( b ) Cr,O, + 2Fe0 + H, = 2Cr0 + 2Fe0 + H,O. 11. OXIDATION.( a ) Cr,O, + 2Fe0 + 0 = Ee,O, + Cr,03. ( b ) 2Cr0 + 2Fe0 + 0, = Cr,O, + Fe,O,. The test results given show that the method is extremely accurate, and the author considers that it will be found valuable in the analysis of chrome iron stone and chrome iron ores. C. A. M. A Rapid Method of Determining Metallic Aluminium in Aluminium ( A m . de Chim. mal., 1904, ix., 381, 382.)-This is an Powder. application of Webl’s method of analysing zinc powder, and is based on reaction : E. Kohn-Abrest. 3Fe,(SO,), + 2A1= AlJSO,), + 6FeS0,. I n the determination 0.5 gramme is heated with an excess (20 grammes) of ferric sulphate and about 50 C.C. of water in a flask through which a current of carbon dioxide is made to pass throughout the reduction. The flask is kept on the water- bath for about an hour, after which time the aluminium powder will have completely disappeared and the ferric sulphate have been almost dissolved.After rapid cooling in a current of carbon dioxide, solution will be complete, and the liquid can now be acidified with 20 C.C. of concentrated sulphuric acid, then diluted to 250 c.c., and an aliquot portion titrated with standard permanganate solution containing, c.g., 3.163 grammes per litre, 1 C.C. of which corresponds to 0.00062968 gramme of aluminium. A commercial sample of the powder yielded by this method 91.4 per cent. of aluminium. The author has found that the bulk of the silicon in aluminium powder is present in the state of silica, and of the other common impurities only iron in the metallic state would affect the results.C. A. M.384 THE ANALYST. A Colour Reaction of Tungsten. C. Frabot. ( A m . de C h i w anal., 1904, ix., 371, 372.)-The blue coloration produced on adding a few drops of a solution of uric acid, followed by sodium hydroxide solution to tungstic hydrate (WO:,.H,O), is due to the formation of the intermediate blue tungsten oxide by reduction. The reaction is capable of detecting 1 part of uric acid in 100,000, and conversely can be used for the detection of tungsten in metallurgical compounds, such as ferro-tungsten. For this purpose the tungsten should be converted into tungstic acid by slow ignition in a current of air and the residue tested after elimination of other metals. Traces of tungsten can be detected in this way in the presence of a large excess of silica.Molybdenum, which gives an analogous reaction, can be separated from tungsten before applying the test for the latter. C. ,4. M. Borax of Abnormal Composition. L. Spiegel. (Chem. Zed,, 1904, xxviii., 750, 751.)--A sample of borax examined by the author gave consistently abnormal results on titration with + acid. As the borax had been previously fused, it was at first supposed that this had altered its composition, but experimental proof was obtained that such was not the case. A complete analysis of the sample showed that it consisted of sodium triborntc. E’urther experiments were carried out to ascertain under what conditions the triborate is formed. I t was found that aqueous solutions of borax containing an excess of boric acid deposit crystals of sodium triborate.The above sample was the only abnormal one in a large number of samples of commercial borax examined, showing that the substitution of the triborate for the ordinary biborate occurs but seldom. w. P. s. Standardization of Permanganate Solution. F. Dupr6. (Zeits. f. nngew. Chem., 1904, xxv., 815.)-The author states that iron wire, ferrous ammonium sulphate, and oxalic acid are all unsatisfactoryfor this purpose when a high degree of accuracy is required. Treadwell’s electrolytic iron method is very reliable, but is tedious, and requires special apparatus, which in many technical laboratories is unavailable. The author finds that Volhard’s iodometric method is as accurate as Treadwell’s, and far more rapid. riceording to Volhard, the potassium perman- ganate should be recrystallized repeatedly to free it from perchlorate, but Dupre found that samples as purchased were already free from this impurity. To a solution containing 0.5 granime of potassium iodide in 1LO C.C. of water, 5 C.C. of pure hydrochloric acid are added, and 20 C.C. of the permsnganate solution are run in with constant stirring. Iodine is set free in accordance with the equation KMnO, + 5KI + 8HC1= 51 + 6KC1+ MnCl, + 4H,O, and is titrated directly with decinormal thiosulphate solution, using starch as indicator. The thiosulphate solution should be kept for at least a, week before using, in order to give time for all the carbonic acid present to react with the thiosulphate. When the whole of this carbonic acid has disappeared the solution will retain its strength indefinitely. A. If.