286 ABSTRACTS OF CHEMICAL PAPERS INORGANIC ANALYSIS. Estimation of Aluminium as Oxide. W. Blum, (J. Amer. Chem. Xoc., 1916, 38, 1282-1297.)-From observations made with a hydrogen electrode and with suitable indicators, it has been found that the precipitation of aluminium hydroxide by ammonium hydroxide is complete when [H+]= lO-G*’j to 10-75, points approxi- mately defined by the colour change of methyl red and of rosolic acid.The presence of ammonium chloride during precipitation is advantageous in limiting the alkalinity and in coagulating the precipitate. Solutions of ammonium nitrate and chloride are equally satisfactory for washing the precipitate. Solutions should be heated to boil- ing, a few drops of 0.2 per cent. solution of methyl red added, and then ammonia drop by drop until the colour of the solution changes to a distinct yellow. After boiling for one or two minutes, filtration is proceeded with without delay. If the original solution does not contain a considerabIe amount of hydrochloric acid, at least 2-5 per cent.of ammonium chloride is dissolved in it before precipitation. Crucibles containing ignited alumina should be kept covered in desiccators and on the balance.I n a large desiccator a small quantity of alumina contained in an uncovered crucible can readily absorb 1 per cent. of moisture from the air admitted to the desiccator when opening it for the reception of the crucible. Pre- cipitates of 0.1 to 0.2 grm. alumina require to be blowpiped for at least five minutes, but ten minutes is always sufficient.The presence of ammonium chloride during ignition causes no appreciable loss of alumina. G. C. J. Study of the Silver Arsenate Test for Arsenic. L. J. Curtman and P. Daschavsky. (J. dnzer. Chem. Suc., 1916, 38, 1280-1282.)-With pure solutions of arsenate the test with silver nitrate is sensitive to 1 part in 150,000. I n this con- ceutration a faint brown coloration is given which can only be recognised readily by comparison with a blank.I n a concentration of 1 part of arsenic (as arsenate) in 60,000 a decided and unmistakable precipitate is found. Ammonium nitrate, which is generally formed in carrying out the test, does not hinder the detection of arsenic in a concentration of 1 in 15,000. I n systematic analysis the following procedure is recommended.The solution is neutralised with ammonia, about 12 per cent. of concentrated hydrochloric acid is added, the solution heated to boiling and treated with hydrogen sulphide. The solution is then diluted, so that the acid concentrt- tion is reduced to about 2.5 per cent,, and treatment with hydrogen sulphide continued until precipitation is complete. The precipitate is collected on a filter, transferred to a beaker, heated for five minutes with 10 C.C.concentrated hydrochloric acid, and the solution is diluted and filtered. The residue, after washing free from chlorides with hot water, is transferred together with the filter to a dish, boiled with 2 C.C. nitric acid till nitrous fumes are no longer evolved, and the solution is diluted and filtered, The filtrate tmd washings are neutralised with ammonia, concentrated to 2 C.C.and mixed in a test-tube with 2 C.C. of 5 per cent. silver nitrate solution. The test will detect 0.5 mgrm. of arsenic with certainty. G. C. J.INORGANIC ANALYSIS 287 Estimation of Calcium. E. Cahen and W. H. Hurtley. (Biochtem. J., 1916, 10, 308-312.)-1n the estimation of the calcium in aortas of normal and diseased persons, a calcined residue is often obtained which is only soluble with great difficulty in strong sulphuric acid, whereas it is readily soluble in phosphoric acid (1 V O ~ .syrupy phosphoric acid and 3 vols. water). Such a solution can be used directly for the volumetric determination of the calcium as oxalate by decomposing with sulphuric acid, followed by permanganate titration ; but if the calcium is to be deter- mined gravimetrically the solution should be filtered after the solution of the calcium in phosphoric acid.I t was found best to precipitate with oxalic acid, followed by ammonia, rather than to use ammonium oxalate, which gives a precipitate which is less crystalline, and filters badly. On applying the method to test mixtures of calcium and magnesium, experiments show that when the amount of magnesium as oxide exceeds the amount of calcium as oxide, a little of the former is precipitated along with the calcium.In such a case the precipitate of calcium oxalate is ignited, redissolved in a little phosphoric acid, and precipitated again. An excellent separation of calcium and barium sulphate may be made by taking advantage of the complete solubility of the former arJd insolubility of the latter in phosphoric acid.Calcium fluoride is also readily soluble in pbosphoric acid. H. F. E. H. Modification of McCrudden’s Method for Calcium, for Estimation of Calcium and Strontium in Presence of Phosphoric Acid and a Small Amount of Iron. 0. B. Winter. (1. Ind. and Eizg. Chem., 1916, 8, 603-604.)-McCrudden ( J .Bid. Chem., 1909, 7, 83) precipitates calcium very slowly as oxalate in boiling solution containing a small amount of free hydrogen chloride. No soluble salts appear to be carried down by occlusion, and the method is excellent for calcium in foods, urine, and faeces, where much phosphoric acid and a little iron are always present. The method is not applicable to the estimation of strontium without modi- fication, as strontium is not quantitatively precipitated under the above conditions.However, suitable addition of alcohol makes the precipitation of strontium complete without introducing any inconvenience. The solution to be analysed is diluted to 200 c.c., and made slightly alkaline with ammonia, using alizarin as indicator.It is then made just acid with hydrochloric acid, of which an excess of 10 C.C. is then added, together with 10 C.C. of 2.5 per cent. oxalic acid. The mixture is boiled until the precipitate is coarsely granular, and then ammonium oxalate solution, in amount about twice that necesssary to precipitate all the calcium and strontium, is added a few drops at a time with constant stirring.To the cooled solution 8 C.C. of 20 per cent. sodium acetate and 15 C.C. of alcohol are added, and the mixture allowed to stand overnight, or not less than four hours. The precipitate is filtered off and washed, first with 1 per cent. ammonium oxalate in 20 per cent. alcohol, and finally with a little 20 per cent. alcohol. The ignited oxides are dissolved in nitric acid, the solution evaporated to dryness, the calcium nitrate extracted with a mixture of absolute alcohol and ether, and calcium and strontium then each estimated in the usual manner.G. C. J.288 ABSTRACTS OF CHEMLCAL PAPERS Separation and Estimation of Polysulphides and Thiosulphate in Lime- Sulphur Solutions. s. D. Averitt. ( J . Incl. and Eitg. Chenz., 1916, 8, 623-627.) -The following method, due to J.E. Harris (Michigan Experiment Station, 1911, Tech. BUZZ. 6), which is much the simplest method yet described for the purpose, has been submitted to experimental comparison with other methods of unquestionable accuracy, and found to be thoroughly trustworthy. From 5 to 20 grms. of the sample, depending on ihs concentration, are diluted to 200 C.C.with freshly boiled and cooled distilled water, and transferred to small bottles, which are filled quite full and sealed. In routine work this sealing up is unnecessary, as the analysis may be completed immediately, but for the purpose of the author's experiments it was necessary to preserve the solutions, and sealed in this way they may be kept a long time unchanged. Ten C.C. of the dilution are run into a 200 C.C.conical flask with about 15 C.C. freshly boiled and cooled distilled water, and titrated immediately with & iodine, until the yellow colour due to polysulphide becomes very faint. Up to this point the titration should be conducted rapidly with constant shaking to prevent a local concentration of iodine oxidising a n appreciable amount of thiosulphate to tetrathionate.A very small crystal of sodium nitroprusside is next added, the solutioq shaken until the purple colour develops distinctly, and iodine solution added quickly but carefully until the colour is just discharged. The consumption of iodine up to this stage is the measure of the calcium present as sulphide. The titration is continued until 1 drop of undecom- posed iodine solution imparts a faint colour to the liquid.The consumption of iodine during this second stage is the measure of the thiosulphate present. The accuracy of the above method is shown by comparison with Thompson and Whittier's method (Delaware College Experiment Station, 1914, Bull. 105) and with two new methods, the theoretical accuracy of which are unquestionable. As these latter methods, however, like Thompson and Whittier's, are tedious, and were only developed to test the method now recommended, they will not be described in this abstract.Finally, to know the exact composition of lime-sulphur solution, which consists essentially of calcium polysulphide and thiosulphate, it is necessary to estimate the total sulphide sulphur. As this is all precipitated during the titration, and can be readily filtered after a few hours or at once after addition of 2 or 3 drops of hydro- chloric acid and warming, all that is necessary is to filter it off on a tared filter of close texture. G.C. J. Separation of Thorium from Iron with the Aid of the Ammonium Salt of Nitrosophenylhydroxylamine (" Cupferron "). W. M. Thornton jun.(Chem. News, 1916,114,13-14.)--The ammonium salt of nitrosophenylhydroxyl- amine (" cupferron '7 was first introduced into analytical chemistry by Baudisch (ANALYST, 1910, 35, 78), and affords a clean separation in the case of certain difficult mixtures. It affords a quantitative precipitation of titanium and an easy separation of that metal from aluminium. It precipitates titanium in presence of tartaric acid, so that after throwing down iron as ferrous sulphide, and acidifying, the titanium may be separated from aluminium and phosphoric acid.In a similarINORGAKlC ANALYSIS 289 manner ‘( cupferron ” may be used to precipitate zirconium, and for the separation of zirconium from iron and aluminium. In the present investigation the conditions for the quantitative precipitation of thorium have been studied.Contrary to the case of the cupferron compounds of titanium and zirconium, which are quantitatively pre- cipitated in presence of free sulphuric acid, it is found that with the thorium com- pound the presence even of a small amount of free mineral acid prevents complete precipitation. In this case the mineral acid must be neutralised by the addition of ammonium acetate, so that precipitation takes place in presence of acetic acid. Twenty- five C.C.of a standard acidified solution of thorium sulphate, equivalent to 0.092 grm. of thorium oxide, were treated with 15 grrns. of ammonium acetate and the volume made up to 500 C.C. A 5 per cent. solution of ( 6 cupferron ” was then added gradually with constant stirring, 15 C.C.being used. The precipitate, being thoroughly coagu- lated by stirring, was collected on a filter, washed with a 1 per cent, solution of ammonium acetate, and dried on the paper at 100°-llOo C. in a platinum crucible, then ignited to constant weight, using the blast lamp. The results for thorium oxide were within the ordinary limits of error for gravimetric work.The thorium salt of nitrosophenylhydroxylamine, [C,H,(NO).N.O],Th, differs somewhat from the titanium and zirconium salts. I t resembles the latter in appearance, but, besides being sensibly soluble in presence of mineral acid, it differs in texture? so that it cannot be collected on a, suction filter without appreciable loss. The separation of thorium from iron is effected by adding suficient tartaric acid to hold up the bases in ammo- niacal solution; the liquid is made slightly alkaline with ammonia, the iron pre- cipitated by colourless ammonium sulphide, the filtrate boiled with a little sulphuric acid, and the thorium precipitated after the addition of 25 grms.of ammonium acetate in the manner described. I t is not contended, however, that this method offers any general advantage over the well-known and satisfactory oxalate method for the estimation of thorium.J. F. B. Qualitative and Quantitative Analysis of Tungsten. M. L. Hartman. (Chew. News, 1916, 114, 27-28.)-The qualitative reduction test for tungsten is thus described: Boil at least 0-2 grm. of finely powdered material in a test-tube with concentrated hydrochloric acid until about half of the acid is evaporated.Dilute with an equal volume of water, add a piece of metallic tin, and heat if necessary. A fine blue colour in the solution, or a blue residue, indicates the presence of tungsten. If this test gives negative results, 0-5 grm. of the substance should be fused with 4 grms. of sodium carbonate, the melt dissolved in water, acidified with an equal volume of strong hydrochloric acid, treated with tin, and warmed if necessary.The volume of the solution should not exceed 10 to 20 C.C. In either case, if reduction is continued long enough, the blue colour is replaced by brown. These tests will show the presence of tungsten in materials containing 2 per cent., or, with special precautions, even less. Columbium (niobium) gives a blue colour under similar con- ditions, which disappears on dilution.Vanadium gives a blue colour in reduced solutions, but tartaric acid will cause this reduction, whereas it will not reduce tungstic oxide. Molybdenum goes through a series of colour changes from violet to blue or black. Titanium gives a violet colour. Three alternative methods for the290 ABSTRACTS O F CHEMICAL PAPERS quantitative estimation of tungsten are given.The hydrofluoric acid method and the alkali fusion method are quoted from A. H. Low’s “Technical Methods of Ore Analysis.” It may be necessary to purify the tungstic oxide from co-precipitated stannic oxide by volatilisation of the stannic oxide in the form of stannic chloride on heating with ammonium chloride.To do this effectively, the crucible containing the mixed oxides and ammonium chloride should be placed inside a larger one with a cover, and the whole ignited with the cover on until the chloride is completely volatilised. Ignition in absence of air darkens the colour of the tungstic oxide, and the inner crucible should then be ignited with full access of air until the pure yellow colour is restored.The operations are repeated until the weight becomes constant. Another method for the estimation of tungsten in ores, preferred by some chemists, is the aqua regia method. One grm. of the finely powdered ore is digested below the boiling-point with 50 C.C. of aqua regia. When the solution has evaporated to 10 to 15 c.c., it is diluted with 50 C.C. of hot water and allowed to stand for half an hour.It is decanted through a filter, and the residue washed several times by decantation with 50 C.C. of hot water each time. The wash water should be slightly acidified with hydrochloric acid. The residue is treated with 20 C.C. of dilute ammonia, prepared by mixing 200 C.C. of strong ammonia solution with 1,000 C.C. of water and adding 10 c.c, of hydrochloric acid to form a little ammonium chloride.This dissolves the yellow tungstic oxide, and the solution is also decanted through the filter. When the residue is finally extracted and washed, the silica should be white. The tungstic oxide is recovered by evaporating the filtrate, igniting to expel the ammonium salts, and heating with hydrofluoric acid to free the residue from silica.J. F. B. Estimation of Vanadic Acid after Reduction by Metallic Silver. G. Edgar. (J. Amer. Chenz. SOC., 1916,38,1297-1302.)-Vanadium is quantitatively reduced from the pentavalent to the tetravalent condition by metallic silver in acid solution under proper conditions. The reaction may be used to estimate vanadium ( a ) gravimetrically, from the loss in weight of the silver ; (b) oxidimetrically, by titra- tion with permanganate; or (c) by titration of the dissolved silyer with thiocyanate.All these operations may be carried out on the same solution and serve to check each other. Solutions of sodium vanadate are acidified with 2 C.C. of sulphuric acid, dilated to about 75 c.c., and treated with a known weight (1-2 grms.) of silver reduced by zinc, silver prepared by heating the oxide, or preferably, perhaps, by electrolytic silver, as this is more apt to be pure. The solutions are boiled for twenty to thirty minutes, or for ten minutes after the appearance of a pure blue colour.They are then filtered through porcelain Gooch crucibles, and the silver washed with hot water. The crucibles are placed in larger nickel crucibles and heated with the full flame of a large Bunsen burner for thirty minutes.From the loss in weight of the silver the vanadic acid can be calculated. The filtrate from the silver can, alter- natively or in addition to the gravimetric estimation, be titrated hot with & per- manganate, and at the completion of this titration the silver in the cooled solution can be titrated with 2% ammonium thiocyanate, using ferric alum as indicator. TheAPPARATUS, ETC.291 volumetric modifications are as exact as the gravimetric, and serve to estimate 0.13 grm. of vanadic acid with an error not exceeding 0.4 mgrm. G. C. J. APPARATUS, ETC. (J. Amer. Chem. SOC., 1916, 38, 1270-1272.)-Large quantities of inorganic hydrosols, containing only minute amounts of electrolytes, may be prepared in a comparatively short time if the difference in concentration of diffusible substances in the internal and external liquids is large, and if the temperature of the two liquids is raised. To obtain this acceleration in the process, the author suspends a parchment paper mem- brane container of about 1 litre capacity in a 2-litre beaker containing about 1 litre of the solution to be dialysed. Distilled water is run a t a fairly constant rate into the membrane, which is maintained a little more than half full by means of an automatic siphon. The colloidal solution in the beaker is heated to 70"-90" C. by a burner placed under the beaker. This apparatus will give in twenty days a hydrated ferric oxide hydrosol as pure as if not purer than the same hydrosol prepared by dialysis in the cold for six months ; a colloidal solution, satisfactory for most purposes, may be prepared in eight to ten days, whilst the usual method requires about one month. w. P. s. Temperature Effect in Dialysis, and a Simple Rapid Dialyser. M. Neidle.