102 THE ANALYST, INORGANIC ANALYSIS. The Iodometric Determination of Arsenious Acid. E. W. Washburn. (Journ. Amer. G'hem. SOL 1908 30 31-46.)-The equilibria in this well-known method are considered at some length and the calculation of the proper conditions to be observed at the end-point of the titration is given as well as the method for securing them conditions. It is considered preferable to weigh all the solutions instead of measuring them and the solutions are made up so that 1,000 grams of normal solution contain one equivalent weight of the substance. The quantity of solution required for an estimation is placed in a burette which can be hung on a balance the weight of solution used being ascertained by reweighing the burette at the end of the titration. The -:$ arsenious acid solution employed for standardising the iodine solution is prepared by dissolving 4-95 grams of pure arsenious oxide in a solution containing 10 to 12 grams of sodium hydroxide ; the solution is then diluted to 100 c.c.saturated with carbon dioxide and water is next added until the total weight of the solution is 206.73 times the weight of the arsenious oxide used. Sodium phosphate may be used in place of the sodium bicarbonate in preparing the solution THE ANALYST. 103 in this case the arsenious oxide is dissolved in a solution containing 12 grams of sodium hydroxide and when solution is complete a quantity of phosphoric acid equivalent to 0.15 molecule of H,PO is added the whole being then diluted to the weight before mentioned. The use of sodium phosphate is to be preferred as there is then no danger of loss of carbon dioxide during the titration a loss which may entail mechanical loss of the reagents.I n titrating unknown solutions of arsenious acid, the solution is neutralised with hydrochloric acid or sodium hydroxide as required, using phenolphthalein as indicator. Sodium hydrogen carbonate or sodium phosphate is then added and the titration carried out more of the neutralising agent being added from time to time until the quantity amounts to about 5 grams of the former or 11 grams of the latter (Na,H1’0,.12H20) for every 100 C.C. of ccr iodine solution used. Starch solution is used as indicator and the end-point of the titration is definite and permanent. w. P. s. Modified Volumetric Method for the Rapid Deterlnination of Carbon Dioxide in Mineral and especially in Acid Waters.J. Stransky. (Chem. h i t . , 1908,32 100-101.)-A layer of ether 2 or 3 rnm. in depth is poured into a measuring cylinder. Tbe sample of water is then introduced through a very fine tube terminat-ing below the level of the ether which prevents escape of carbon dioxide. The volume of water introduced having been ascertained the carbon dioxide is at once titrated with :& potassium hydroxide solution and phenolphthalein the tip of the burette used being also drawn out to a very fine tube terminating below the level of the ether. ,4. G. L. Action of Silver Nitrate and Mercuric Nitrate on Certain Inorganic Hydroxides. W. Biltz and F. Zimmermann. (Bey-. h 6 t . Chew. Ges. 1907 40, 4979-4984.)-When a colourless metallic hydroxide is treated with silver nitrate or mercuric nitrate the hydroxides of which are coloured the precipitate will become coloured or will remain colourless according to the relative basicities of the two hydroxides or according to the relative solubilities of the ions.The colourless hydroxide is precipitated hot or cold by means of ammonia ; the precipitate is collected and washed and then placed in a & solution of the silver or mercuric nitrate. With silver nitrate the hydroxides of magnesium and cadmium give a strong yellowish-brown colour; zinc gives only a trace of brown; lead gives a slight violet-brown colour ; and the hydroxides of beryllium aluminium indium zirconium antimony, bismuth and stannic hydroxide remain colourless.The violet tint observed in the case of lead is due to reduction of the silver hydroxide and the same phenomenon occurs still more strongly with manganous and stannous hydroxides which imme-diately turn black with silver nitrate. By using mercuric nitrate containing 5 C.C. of nitric acid per litre this complication due to reduction is not so much in evidence. The colour of the mercuric hydroxide ranges from pure yellow to yellowish red, according to the rapidity of its formation. Mercuric hydroxide possesses a lower solubility than silver hydroxide so that a greater number of metallic hydroxides show colours with mercuric nitrate than with silver. Thus with mercuric nitrate, beryllium and lead turn yellowish red; magnesium and cadmium show a stron 104 THE ANALYST.yellowish red ; zinc and aluminium give a faint yellow ; manganous hydroxide an egg-yellow ; whilst indium zirconium antimony and biBmuth remain colourless. These reactions may therefore be of use in certain cases in qualitative analysis and they are particularly useful for the identification of the rare earths the reactions of which are set forth in the following tables : SILVEH. NITRATE REACTIONS. Neo-dymium. Yttriiini. precipi -tated hot precipi-tated cold slowly becomes grey yellowish brown yellowish brown, slightly grey same as yttrium very faint violet -grey paler than y ttriuiii strongly grey to grey-brown paler than lanthanum, yellowish brown saiiie as lanthanum same as lan t h mu tn MERCURIC NITEATE REACTIONS.Pi*ttseo-dymium. NCO-d y m iuni . Lan -thnnnm. Ceroiis Hydroxide. Erbium. Yttrium. Sairiarium. precipi -tated hot precipi-tated cold yellow yellow slightly yellow yellow yellow, slightly reddish yellowish red yellow, brown tinge yellow, brown tinge strongly yellowish red strongly yellowish red strongly yellowish red yellowish red strongly yellowish red slightly yellowish red The hydroxides of the cerium group are more soluble when precipitated cold than when precipitated hot ; they also show a stronger tendency to reduce the silver hydroxide than the last three hydroxides. I n fact a warm very slightly ammoniacal solution of silver nitrate is a delicate test for cerous salts which give a black pre-cipitate or if dilute a brown coloration.Thorium hydroxide appears to exist in two modifications a relatively soluble and an insoluble form. The hydroxide precipitated in the cold from the nitrate gives a strong coloration with both silver and mercuric nitrates but that precipitated from the sulphate only shows a colour after about twelve hours. Hot precipitation tends to retard the appearance of the colour and if the thorium hydroxide be precipitated hot and then boiled for fifteen minutes it is converted into a modification which gives no colour at all with silver or mercuric nitrate. J. F. B. Volumetric Estimation of Iron in Ferric Compounds. M. M. Pattison Muir. (Chem. News 1908 97 50.)-A known volume of the solution containing iron is placed in a flask fitted with a cork carrying a glass tube narrowed at its upper end THE ANALYST.105 20 grams of iron-free granulated zinc and about 200 C.C. of dilute sulphuric acid are added and the liquid warmed until a brisk current of hydrogen is evolved. When reduction is complete about 100 C.C. of a nearly saturated aqueous solution of mercuric chloride are added and the contents of the flask shaken for a few minutes when the liquid is cooled. The deposition of mercury on the surface of the zinc effectually stops the evolution of hydrogen. The liquid is finally titrated witb standard permanganate solution in the usual way. A. R. T. Examination of Red Lead. J. F. Sacher. (C'hem. Zeit. 1908 32 62-63.)-The insoluble residue should be determined by treating the sample of red lead with nitric acid and a sinall quantity of formaldehyde or of pure hydrogen peroxide, evaporating to dryness to remove excess of nitric acid and taking up in water.Partheil's method (ANALYST 1'307 32 395) of reduction with nitric acid and lactic acid is not to be recommended because lactic acid exerts a considerable solvent action on lead sulphate; moreover the excess of nitric acid which also dissolves lead sulphate cannot be removed by evaporation as large quantities of insoluble lead oxalate are formed. A. G. L. The Estimation of Manganese in Water. E. Ernyei. (Clrenz. Zeit. 1908, 32 41-42.)-The method is based upon Narshall's observation (ANALYST 1901 26, 195) that manganese compounds in acid solution are quantitatively oxidised to permanganates by persulphates in the presence of silver salts and upon the iodometric estimation of the permanganate thus formed.A few C.C. of the water are treated with a few drops of sulphuric acid and potassiuin iodide starch solution and if a blue colour is produced (in the absence of nitrites) too much iron is present. I n such a case the water should be acidified with sulphuric acid shaken with a slight excess of zinc oxide and filtered. 100 C.C. of the water freed from iron are mixed with about 5 C.C. of 30 per cent. sulphuric acid rather more silver sulphate solution than is required for precipitation of the chlorine and 1 to 2 grams of potassium persulphate and the whole boiled for 20 minutes. When cold the liquid is treated with a little potassium iodide followed by starch solution as soon as the rose coloration has disappeared aud the separated iodine titrated with & sodium thiosulphate solution.Each molecule of permanganate liberates from the potassium iodide 10 atoms of iodine and the silver iodide in suspension does not affect the end-point of the reaction. Natural waters contain so little manganese that only a few C.C. of the thiosulphste solution are required. Nitrates nitrites organic matter and ammonium salts have no appreciable influence upon the results even when present in greater quantities than are found in natural waters. For the estimation of manganese in ferruginous deposits the substance is dissolved in sulphuric acid the iron precipitated with zinc oxide and the filtrate treated as above described.C. A. N. Volumetric Estimation of Mercury in Minerals. J. A. Muller. (Bztll. SOC. Chim. 1907 4 [l] 1169-1173.)-The mineral is decomposed with aqua regia, the solution evaporated to dryness a t 50' C. and the residue dissolved'in hot water 106 THE ANALYST, Potassium iodide sodium carbonate and finally a little sodium hydroxide are added to the liquid which is then filtered the insoluble residue being washed first with a little potassium iodide solution and then with hot water. The filtrate is made up to 100 C.C. with water and an aliquot part containing about 0.1 gram of mercury is treated in the cold with 20 C.C. of 20 per cent. sodium hydroxide solution and an excess of formaldehyde. After standing for twenty hours the solution is poured off from the precipitated mercury which is then washed finally with a little alcohol air being blown in to remove traces of formaldehyde.The mercury is next treated with an excess of a standard solution of iodine in potassium iodide in an atmosphere of carbon dioxide. The last traces of mercury are dissolved if necessary, by adding a little fresh iodine solution and gently warming. The excess of iodine is then titrated with sodium thiosulphate. The only metals which interfere with the method are gold and platinum. If either of these is present the mineral is decom-posed by heating in a current of chlorine instead of treating with aqua regia and mercury is estimated in the distillate as above. The test results quoted are slightly low. A. G.L. Colour Test for Molybdenum in Ores etc. W. Bettel. (Chem. Nws. 1908, 97 40.)-The author observed in 1903 that if to an acid solution of molybdic acid in which hydrogen peroxide had given the well-known yellow colour (permolybdic acid) sufhjent dilute ammonia solution was carefully added to faint alkalinity the colour changed to an intense brownish-red presumably due to the formation of a permolybdate. This red colour is partially discharged on dihtion and disappears on standing oxygen being evolved. A large excess of ammonia destroys the colora-tion as also do more than traces of fixed alkalies and thus the reaction is a good indicator of neutrality in forming neutral molybdates. For the detection and approximate estimation of molybdenum the liquid is evaporated nearly to dryness, and if alkaline is neutralised with sulphuric acid and treated with hydrogen peroxide.If a yellow colour be produced a very little dilute ammonia solution is next added when the brown-red coloration will appear in the presence of 0.001 mgm. of molybdic acid. -1. R. T. The Detection of Nickel. E. Pozzi-Escot. (A12n. d e Clziiit. mzcil. 1908 13, 16.)-The double molybdate of nickel and ammonium the formation of which may be used as a test for nickel in the presence of cobalt (ANALYST 1907 32 432) is not only insoluble in ammonium rnolybdate solution but also in solutions of other ammonium salts. Thus in testing a neutral or slightly acid solution for nickel ;t small quantity of a saturated solution of ammonium molybdate solution is introduced, followed by a large excess of a saturated solution of ammonium chloride and the liquid gently heated and shaken.In the presence of nickel a turbidity and eventually a precipitate appears in a few minutes. Or a reagent for the detection of nickel may be prepared by adding 10 per cent. of ammonium molybdate solution to a saturated solution of ammonium chloride. C. A. M THE ANALYST. 107 Volumetrie Estimation of Nickel. H. Cantoni and M. Rosenstein. (BUZZ. SOC. Chim. 1907 4 [l] 1163-1169.)-The authors examined the volumetric method of estimating nickel by titrating with potassium ferrocyanide solution using ferric chloride or uranium acetate as indicators and with potassium ferricyanide, using ferrous sulphate as indicator in the presence of acetic acid sodium and ammonium acetates and potassium sodium and ammonium sulphates.They find it best to titrate with potassium ferrocyanide solution in the presence of a small but constant amount of acetic acid and to standardise the solution in the same way against pure nickel. The indicator should be applied to a drop of the solution filtered through a single or double filter-paper according to the fineness of the precipitate. A. G. L. On the Detection of Ozone Nitrogen Peroxide and Hydrogen Peroxide in Gas Mixtures. E. H. Keiser and L. McMaster. (Arne? Chenz. Jown. 1908, 39 96-104.)-Ozone may be detected in presence of hydrogen peroxide and nitrogen peroxide by passing the mixed gases through a dilute solution of potassium perman-ganate and through a solution containing potassium iodide and starch.Both hydrogen peroxide and nitrogen peroxide are instantly reduced by the permanganate, which has no action on ozone. To detect nitrogen peroxide the mixed gases are passed through a tube containing powdered manganese dioxide which decomposes both ozone and hydrogen peroxide but leaves the nitrogen peroxide unchanged ; this gas may then be identified by its action on permanganate or by the formation of nitrite on leading it into pure sodium hydroxide. Hydrogen peroxide may be detected by the formation of Prussian blue in a solution containing potassium ferri-cyanide and ferric chloride ; neither ozone nor hydrogen peroxide causes this reaction, which may however be obtained if the solution is exposed to air for some hours. By means of the above tests the authors show that hydrogen peroxide and nitrogen peroxide are formed when hydrogen is burnt in air.A silent electric discharge at 9,160 volts produced only ozone in air. The action of concentrated sulphuric acid on barium dioxide resulted only in the production of ozone unless barium nitrate was present when nitrogen peroside also was obtained. Moist phosphorus osidising ~ 1 0 ~ 1 ~ in air gave both ozone and nitrogen peroxide but not hydrogen peroxide. Air passed through an electric arc contained all three gases. A. G. L. Modification of Petermann’s Method for the Estimation of Citrate-soluble Phosphoric Acid in Phosphate Fodder Meals. G. Fingerling and A. Grombach. (Zeds. a n d Chem. 1907 46 756-760.)-1t is recognised that dicalcium phosphate is assimilated by animals far more readily than tricalcium phosphate and that phosphatic fodder meals should consist of precipitated calcium phosphate and not merely of steamed or extracted bonemeal.Petermann’s process originally used for the estimation of citrate-soluble phosphoric acid in superphosphates is therefore applicable for the valuation of these fodder meals but it has the dis-advantage that the substance is to be digested with the citrate solution for fiftee 108 THE ANALYST. hours at the ordinary temperature with occasional shaking.” This is generally impossible without night work and the authors’ experiments have been directed to finding a more convenient procedure which will give results practically identical with those afforded by the original method.The modified process is as follows One gram of the substance is placed in a 200 C.C. flask with 5 C.C. of alcohol and 100 C.C. of Petermann’s citrate solution. The mixture is agitated by means of a rotary shaking apparatus for half an hour After shaking it is digested for one hour at a temperature of 40” C. with agitation then cooled diluted to 200 c.c. and filtered; 100 C.C. of the filtrate are treated with 20 C.C. of concentrated nitric acid evaporated to half the original volume cooled neutralised with ammonia and treated with 50 C.C. of Hallens,’ solution. After cooling 30 C.C. of magnesia mixture a.re added drop by drop the liquid is shaken for half an hour and the phosphoric acid is estimated in the usual manner. J. F. B. The Detection of Ruthenium in Platinum Alloys.N. A. Orlow. (Chem Zcit. 1908 32 77.)-A sample of the alloy is fused with lead the mass extracted with nitric acid and the residue ignited in the air to expel osmium. The ignited mixture which may contain iridium rhodium and ruthenium (in addition to platinum) is fused with potassium nitrate and potassium hydroxide the mass, when cold extracted with cold water and the liquid treated with excess of nitric acid to bring about the reaction-~KARuO + 4HN0 = RuO + Ru(OH) + 4IiN0,. The brown liquid is left for twelve to twenty-four hours in a vessel covered with filter-paper and in the presence of ruthenium the under side of the paper will have become black through absorption of the vapours of the ruthenium tetroxide (RuO,). In this way it is possible to detect 0.01 gram of ruthenium in an alloy.The blackened paper may be ignited the ash fused with potassium nitrate and potassium hydroxide and the orange-coloured ruthenate extracted with water from the mass. The darkening of the paper occurs more rapidly on heating but there is then a risk of the vapours escaping. The preliminary removal of all osmium (which gives a similar reaction) is essential. C. A. M. The Analysis of Commercial Silicon. F. Limmer. (ClLenz. Zeit. 1908 32, 42.)-Silicon may be quantitatively separated from silica by volatilisation as silicon chloride in a current of chlorine. From 0.25 to 0-50 gram of the finely divided sample is heated at a moderately red heat in a porcelain boat in a glass tube through which is passed a slow current of perfectly dry chlorine free from all trace of oxygen.The heating is not started until all air has been expelled from the tube by the chlorine and after the end of the reaction (one to three hours) the boat is allowed to cool in a current of the gas. The silicon and iron and aluminium chlorides con-dense on the colder parts of the tube whilst the residue in the boat may contain in addition to silica traces of carborundurn (especially in the case of silicon prepared by electrolysis) magnesium chloride and calcium chloride. The separation of silicoc from silica cannot be regarded as complete until the residue becomes constant i THE ANALYST. 109 weight. From thirty minutes’ to one hour’s heating is sufficient in the case of crystalline and fused silicon but with amorphous silicon as long as three and a half hours’ heating may be required before the residue becomes white and constant in weight.This is possibly due to the large proportion of silica in the amorphous pre-paration. Thus commercial s ~ ~ ~ p l e s examined by the author yielded from 23.24 to 65.56 per cent. of residue whilst two samples of fused silicon gave 11.90 and 2.35 per cent. respectively and two samples of crystalline silicon 1.47 and 1.56 per cent. respectively. The silica in the residue may be estimated in the usual way whilst the amount of silicon is found by fusing the sample in a platinum crucible for thirty minutes with sodium and potassium hydroxides so as to obtain the total silicon and deducting from the result the silica found in the residue from the chlorine volatilisa-tion. A complete analysis of a sample of commercial silicon gave the following results Residue from chlorine treatment 2.35 per cent. (consisting of 0.90 per cent. of silica 1-40 per cent. of magnesium chloride and traces of calcium and carbon silicides) ; iron 2.35 per cent. (including a little aluminium) ; silicon 95.27 per cent. ; and traces of a phosphorus compound. C. A. 1%