THE ANALYST. 125 INORGANIC ANALYSIS. The Volumetric Determination of Zinc. W. Ueorge Waring. (Journ. Amer. Chem. SOC., xxvi., 4.)-The auther cites a number of errors likely to affect the ferro- cyanide method for zinc as ordinarily carried out. The most important of these are due to volatilization of zinc chloride at as low a temperature as 145" C. when heated with ammonium chloride and hydrochloric acid ; to the recombination of zinc with silica in ammoniacal solution when ammonia is added to the solution of a siliceous ore to precipitate iron and aluminium; to the insolubility in acids of zinc aluminates, which are perfectly decomposed only by fusion with potassium bisulphate ; to the use of too much hydrochloric acid in the final solution ; and to the employment of an incorrect standard for the ferrocyanide solution, The following method is said to give excellent results for even the most complex zinc ores; for simpler compounds some of the operations may of course be omitted or modified.The method depends on the separation of zinc from manganese, iron, and aluminium by means of hydrogen sulphide, under slight pressure, in a solution very slightly acidified with formic acid, the metals of the copper group having been previously separated by metallic iron or aluminium. For most ores, solution can be effected by treatment with hydrochloric acid or aqua regia, the nitric acid in the latter case being destroyed by evaporation with an excess of hydrochloric or sulphuric acid. If zinc spinels or aluminates are present, the insoluble residue is decomposed by fusion with sodium carbonate and borax, or, if no silica, is present, with potassium bisulphate, any lead sulphate having previously been dissolved out by ammonium acetate.The solution of the melt is added to the main solution. Some silicates, such as furnace cinders and slags, may be at once fused with sodium carbonate before treating them with acid. I n all cases the silica may be filtered off in the gelatinous state, thus saving the evaporation to dryness. I n the filtrate the metals of the copper group, except lead and cadmium, are next precipitated by boiling for fifteen minutes with a strip of clean iron or steel. Metallic aluminium may also be used for this reduction, and has the advantage of precipitating cadmium and lead as well as the other metals.When copper and aluminium are to be determined on the same portion, sodium sulphite or thiosulphate may also be used for the reduction. In any case, the liquid is filtered into a 300 C.C. flask, made slightly alkaline with caustic soda, using methyl orange as indicator, then just acid with 50 per cent. formic acid solution, after which another 0.5 C.C. formic acid is added, and the liquid diluted so as to contain not more than 0.15 to 0.20 gramme metallic zinc per 100 c.c., and heated to 80" C. Hydrogen sulphide is then led through the solution; as soon as zinc commences to come down, the flask is closed with- out interrupting the current of the gas, so as to saturate the solution with hydrogen sulphide under the pressure of the generating apparatus.The resulting precipitate settles easily, and can be washed with hot water. It is dissolved in 10 C.C. concen- trated hydrochloric acid, the filter-paper being washed with hot water until the total volume is about 125 C.C. If iron was used for the original precipitation, any cadmium present will be left undissolved at this stage; an equal volume of hydrogen sulphide water is added to the liquid, the cadmium sulphide filtered off, and determined with126 THE ANALYST acid ferric sulphate and perqnganate as usual.. The einc chloride solution is heated to 60" C., made up to 200 or 250 C.C. with hot water, a little ammonium chloride is added, and the zinc is titrated with ferrocyanide solution, made by dissolving 21.63 grammes potassium ferrocyanide m d 7 grammes sodium sulphite in 1 litre of water free from air.As a standard a solution of 6.23 grammes zinc oxide in 60 C.C. acetic acid made up to 1 litre with water is used. The amount of ferrocyanide solution required to give a colour with the uranium acetate solution when no zinc is present must always be deducted from the burette reading. If only small amounts of zinc are present, the author recommends igniting the sulphide precipitate obtained as above to oxide, and weighing instead of titrating. Or else, in the solution of the ore, the zinc is determined as phosphate after removing iron, aluminium, calcium, magnesium, and any heavy metals present as usual. A. G. L. Quantitative Reduction by Means of Aluminium. N. Tarugi. (Cfazz. chim. ital., xxxiii., vol.ii., 223 ; through Chem. Zeit. Rep., xxviii., 4.)-Aluminium reduces silver completely in two to three hours from a solution of its nitrate acidulated with sulphuric acid ; under similar conditions copper is only incompletely reduced. A. G. L. Experiments on the Electrolytic Separation of the Metals of the Alkaline Earth Group. Alfred Coehn and Wilhelm Kettenbeil. (Zeits. .dnorg. Chem., xxxviii., 198.)-From the authors' experiments it appears that, using a mercury cathode, it is possible to separate, at least approximately, the metals of the alkaline earth group from each other by taking advantage of the fact that the current potentials necessary to separate them from solutions of their chlorides differ by several tenths of a volt. The method will, however, require further working out before it can lay claim to be quantitative, A.G. L. Determination of Alkalies in the Presence of Borstee. K. Jacobi. (Journ. dmer. Cheiu. Soc., xxvi., 88.)-Two grammes of the substance are dissolved in hydrochloric acid ; silica, iron, aluminium, and calcium are removed as usual, and magnesium is precipitated by ammonium phosphate. The excess of phosphoric acid is then removed by dding ferric chloride. The filtrate is evaporated to dryness twice with nitric acid to remove ammonium salts, the residue being evaporated twice with hydrochloric acid to remove nitric acid. The residue finally obtained is dissolved in water, the solution filterad into a weighed platinum dish, evaporated to dryness, and the residue ignited, at first very gently, and finally for ten minutes at a red heat, after which it is weighed.I t consists of anhydroue boric oxide, B,O,, sodium chloride, and sodium borate, Na,B,O,, due to the interaction of boric wid and sodium chloride at a high temperature. The residue in the dish is then dissolved in hot water, the sodium oxide present as borate is determined by titrating with sulphuric aoid, using methyl orange a8 indicator, and then the total boric acid by addingTHE ANALYST. 127 glycerin and titrating with potassium hydroxide, using phenolphthalein. The quantity of sodium chloride present is then obtained by subtracting from the total weight of residue the sum of sodium and boric oxides found. Should potassium or sulphuric acid also be present, they must be determined in separate portions, and their presence in the residue allowed for.The method appears to give good results. A. G. L. A New Method for the Determination of Free Lime, and on so-called " Dead-Burnt '' Lime. Edward H. Keiser and S. W. Forder. (,dmer. Chew&. Joum., xxxi., 153.)-The method described for the determination of free lime in Portland cement and similar substances depends on the fact that free lime combines almost instantly with water, whereas basic calcium silicates are acted upon much more slowly. It is carried out by first heating 0.2 to 0.5 grrtmme of the substance to be examined in a platinum crucible over the blast-lamp for a few minutes to remove water and carbon dioxide, weighing, and then adding a few drops of distilled water free from carbon dioxide.The crucible is next placed in a cylindrical brass box, D (see figure), fitted with a screw-top carrying a brass inlet and outlet tube, the screw- F h A-KOOR Bulb B--CaCl,TUbO C-Thermometer D-Brass Protector E-Porcelain dish for cmcfhle F-Platinum crucible G '-copperair- joint being made tight with white-lead and oil. The whole is put into an air-bath, G, and heated to 85" C. for thirty minutes. A slow current of purified air is then drawn through the apparatus for thirty minutes, the iiemperature being raised at the same time to 185" C . to expel the excess of water, after which the crucibls is placed in a desiccator containing caustic potash, and allowed to cool. From the increase in weight the quantity of free lime present is calculated. Results obtained in this way on pure lime vary from 99.26 to 101.00 per cent.of CaO. To ascertain the behaviour of so-called " dead-burnt " lime, pure lime was fused by means of the electric arc, and treated as above. At a temperature of 85' C. slaking was found to be complete at the end of two hours. The method was next tried on the compounds usually assumed to be present in Portland cement. For this128 THE ANALYST. Compound. (CaO)SiO, . . . ... 2(CaO) SiO, . . . purpose mixtures of pure lime and silica or alumina were fused by means of an OXY- coal-gas blowpipe and treated as above, the following results being obtained : Hydraulic Properties. None. Not quite as hard as cement. I 0.42 0.30 0.52 1.01 1.33 0.93 1-61 3-15 ... ~~(cGI)s~o, ... 1 1: 3( CaO) SiO, ... Hard as cement.4( CaO) SiO, ... None. Y, 9, $ 9 1, 2(Ca0)A1,03 . . . 3(CaO)A1,0, . . . Like a cement.. None. I of 1 Percentage of Lime Water taken UB. calculated from Water. Percentage Composition of Cement. 5A12O3, 25Sio2, 70Ca0 ... 9A1,O3, 21Si02, 70CaO ... 15Al2o3, 15SiO,, 70Ca0 ... Percentage of Set slowly; did not become quite as 1.43 Set slowly, and became quite as hard 2.16 Set slowly, and became very hard. ~ 4.65 Hydraulic Properties. Water taken up. -- hard as Portland cement. as Portland cement. 3 molecules of lime to 1 of sil& take up water only with great diflculty ; aluminates, on the other hand, are readily decomposed, and behave like free lime. Three synthetic cements were next prepared from pure lime, silica, and alumina, and tested by the same method, with the following results : I I ~ It appears that, as the percentage of alumina present increases, the percentage of water taken up increases ; but if the quantity of alumina does not exceed 10 per cent.-as is the rule in commercial cements-then the amount of water taken up does not exceed 3 per cent These conclusions were confirmed on testing a number of '' sound " commercial cements, the quantity of water taken up varying from 1-81 to 3.04 per cent.One sample of commercial cement examined took up 10.17 per cent. of water, indicating the presence of 7 per cent. of free lime ; a pat made from the neat cement proved to be altogether unsound. Finally, the method was tested by adding 15.40 per oent. of lime to a cement which took up 1.16 per cent. of water, corresponding to the presence of 3.60 per cent.of free lime. After adding the lime, the cement took up 6.25 per cent. of water, corresponding to 19-44 per cent. of free lime instead of 19.00 per cent. A. G. L.THE ANALYST. 129 On the Determination of Small Quantities of Sodium Hydroxide and Car- bonate in Soap. p. Heermann. (Chem. Zeit., xxviii., 53 and 60.)-The author condemns the older methods as being inexact, and quotes results obtained in favour of the following method for determining the cawtic soda : From 5 to 10 grammes of the soap are dissolved in 250 C.C. recently boiled distilled water; 10 to 15 C.C. of neutral concentrated barium chloride solution are added to precipitate the carbonate and fatty acids, and the solution is warmed until the barium soap collects into one mass.The liquid is then filtered or, preferably, decanted from the soap, which is washed, the solution being then titrated with & acid, using phenolphthalein as indicator. To determine the carbonate, the soap is either salted out with sodium chloride from its aqueous solution, which is then titrated for total free alkali with methyl orange and acid; or else the finely-divided soap is dried, dissolved in absolute alcohol, and carbon dioxide passed through the solution to preuipitate the total free alkali as carbonate, which is washed with hot alcohol, dissolved in water, and titrated as before, In either case, the sodium hydrate found accordiug to the first method must of course be subtracted from the NaOH plus Na,CO, obtained. The author prefers this last method. A.G. L. The Determination of Thiocyanates in the Presence of Salts precipitated by Silver, (Ann. de Chim. anal., 1904, ix., 45-46.)-The electrolysed solution of a pure chloride furnishes a neutral oxidizing agent that rapidly converts thiocyanates into sulphates without affecting cyanides. Suitable solutions for the purpose, containing 18 to 20 grammes of chlorine per litre, are readily obtained by cooling the anode during the electrolysis of sodium chloride or magnesium chloride. In the determination the solution of the thiocyanate, the strength of which should be sufficient to yield 0.2 to 0.4 gramme of barium sulphate, is mixed with a solution of 2 gramrues of barium chloride in 50 C.C. of water, and 50 C.C. of the oxidizing agent added little by little with constant stirring.After ten minutes the liquid is acidified with 10 C.C. of pure hydrochloric acid and heated to the boiling-point, and the beaker left for an hour on the water-bath for the barium sulphate to settle. The supernatant liquid is then decanted, and the precipitate washed two or three times with boiling water, collected, and weighed. The method, however, can also be used for the determination of sulphides. The author states that he has obtained closely concordant results in the determination of thiocyanic acid in the residual liquors from the manufacture of ferrocyanides. A. Dubosc. C. A. M. On the Determination of Silica in Silicate Mixtures. Ed. Donath. (Oester. Chem. Zeit., vi., 561.)-Confirming the results obtained by Lunge and Milberg (Zeit.ungew. Chem., 1897, 393), the author finds that when complex silicates are treated with hot 10 per cent. sodium carbonate solution, not only silica, but alumina and, occasionally, ferric oxide pass into the solution. The amount of alumina, dissolved may reach 1 per cent. and upwards. A. G. L.130 THE ANALYST. On the Quantitative Determination of Fluorine in Fluorides. A Critical Study of the Wohler-Freeenius Method. Karl Daniel. (Zeits. Auorg. Chem., xxxviii., 257.)-As the result of a long and elaborate investigation, the author comes to the conclusion that the Wihler-Fresenius method for fluorine gives accurate results, provided that a number of precautions are used, chief amongst which appears to be the destruction of all organic matter accidentally present, by heating the sulphuric acid to be used for the determination to its boiling-point in the apparatus itself before introducing the fluoride.For other details the original article rnust be consulted. A. G. L. On the Reduction of Alkali Iodates and Chlorates with Hydrazine Sulphate. Max Schlotter. ( Z e d s . Anory. Chem., xxxviii., 184.)-The author shows that potassium iodate is completely and instantaneously reduced to iodide if its hydrochloric acid solution is allowed to flow gradually into an excess of a solution of hydrazine sulphate. If the hydrazine sulphate solution is added to the iodate solution, on the other hand, iodine is set free, the reduction not being so complete. Chlorates are also converted quantitatively into chlorides by hydrazine sulphate, but the reaction is much slower, requiring several hours’ heating for its completion.A. G. L. A New Chlorometric Method. J. Pontius. ((‘hem. Zeit., xxviii., 59.)-This method for the determination of available chlorine in bleaching-powder depends on the oxidation of iodides to iodates by means of hypochlorous acid, and the liberation of iodine from iodides by means of iodates. For a determination, 7.1 grammes of the bleaching-powder are treated with water, the whole is made up to 1 litre, and 50 C.C. are measured out into a beaker containing 3 gramrnes solid sodium bicarbonate, which reacts with the calcium salts with formation of calcium carbonate, sodium chloride, and hypochlorous acid. After stirring well, 1 or 2 C.C. of starch solution are added, and the liquid is immediately titrated with ->G potassium iodide solution (containing 2.7667 grammes::: potassium iodide per litre), until the blue colour produced is permanent.3CaOC1, + GNaHCO, + KI = KIO, + 3CaC0, + 6NaC1+ 3C0, + 3H,O. If other bleaching agents, such as e m d e Javelle, are examined by this method, any free alkali they may contain must first be neutralized ; this is best done by adding an excess of boric acid, which is without influence on the above reaction. The potassium iodide solution used should be standardized against arsenious acid and a sample of bleaching-powder. From the test analyses given the method appears to give very good results, and it is also much more rapid than the original one of Penot. A. G. L. The reactions which take place may be expressed by the equation : - __ On Eschka’s Method for Sulphur Determinations.Otto Pfeiffer. (Chesn. Zeit., xxviii., 38.)--A Bunsen burner is used to heah the crucible containing the One-sixth of a niolecnle of l i I being relliiired for one atom of Cl.THE ANALYST. 131 mixture of coal and Eschka's reagent. To prevent contamination by the sulphur in the gas, the crucible is placed in one hole cut in the 'top of an inverted tray of sheet- iron, which is placed on the top of a tripod. To a second hole a chimney is fitted, which draws the burnt gases away from the crucible. The latter is made to fit tightly on the eheet-iron by means of an asbestos ring cemented to the iron by means of sodium silicate. The author uses a porcelain crucible in place of the more usual platinum one, and also uses this method for determining sulphur in indiarubber.A. G. L. The Iodometric Determination of Phoaphorus. E. Rupp. (Arch. Phccrm., 1903, ccxli., 321-326 ; through Zeit. fiir Untersiuh. cler Ndtr. u?ul Gewssmittel, 1904, vii., 158.)-Yellow phosphorus is quantitatively oxidized by :G iodine solution within twenty-four hours, if about 5 C.C. of carbon disulphide be added to the mixture. About 5 grammes of sodium-potassium tartrate are also added to neutralize the hydriodic acid formed. In titrating the excess of iodine solution, the thiosulphate solution should be added slowly with constant agitation. I t is unnecessary to add starch solution, as the characteristic colour of the iodine diasolved in the carbon disulphide serves as an indicator.On0 C.C. of & iodine solution is equivalent to 0*00062 gramme of phosphorus. ,4ny araenic in the latter is counted as phosphorus, but if necessary the areenic may be separately determined by Gooch and Morris's method, in which the arsenic acid is reduced by hydriodic acid, and the arsenious acid produced titrated with iodine solution in the presence of sodium hydrogen carbon ate. w. P. s. The reaotion is shown by the equation : P, + 51, + 8H,O = 2H3P0, + 10HT. A New Method for the Detection and Determination of Phosphorus in Oils. W. Straub. (.4rcli. Pharm., 1903, ccxli., 335-340 ; through Zeit. j'iir liittersuch. dele Nnhr. mid Geiz~~ss~)~ittel, 1904, vii., 158.)-By shaking 10 C.C. of a 0.1 per cent. solution of phosphorus in oliveoil with 25 C.C.of a 1 per cent. copper sulphate solution a brown- black emulsion is obtained. On continuing the agitation for from four to five hours, the brown coloration disappears, and when allowed to stand, the mixture separates into two layers, the whole of the phosphorus being found in the aqueous layer as phos- phoric acid, together with the copper sulphate. After removing the aqueous portion, the phosphoric acid in it is determined as usual. The disappearance of the brown colour indicates that the oxidation of the phosphorus is complete. w. P. s. On the Determination of Citric- Acid-Soluble Phosphoric Acid in Thomas- Slag. (Ckent. Zeit., xxvii., 1125.)-The author proposes the following rapid method for the examination of Thomas-slag comparatively poor in silica (cf.ANALYST, xxviii., 50, 159): The citric acid extract is prepared as usual; 100 C.C. are placed in a 200 C.C. flask, 75 C.C. of ammoniacal citrate solution are added, the whole is boiled up rapidly and allowed to stand five or ten minutes; if no 0. Bottcher.132 THE ANALYST. precipitate soluble in hydrochloric acid appears, the solution does not contain a injurious amount of silica, and the phosphoric acid may be determined in 50 C.C. as usual. On the other hand, if a precipitate of silica is obtained, the liquid is made faintly acid with hydrochloric acid, made up to the mark, and filtered through ti large filter-paper, 100 C.C. of the filtrate being then precipitated with 25 C.C. of magnesia mixture. A. G. L. Contributions to Technical Chemical Analysis.G. Lunge. (Zeits. f: aiigew. C h m . , 1904, vii., viii., ix., 195-203, 225-236, 265-270.)-Indicators.-Methyl orange and phenolphthalein are again recommended (see ANALYST, 1903, 307). Objections are raised to the use of para- and ortho-nitro-phenols on the ground that they are distinctly affected by carbonic acid and that the end-points are not sharp. Jferric salicylate, recommended by J. Wolff for the titration of boric acid, Lunge finda to be in no way preferable to methyl orange, the results given being identical. Stmulard Substances for Alknlimtric Titmtwns.--Potassium biniodate does not require the theoretical quantity of acid for neutralization when methyl orange is used as indicator, because iodic is but a weak acid. With phenolphthalein in a boiling solution the results agree with the theoretical, but are not very sharp.Potassium tetroxalate cannot be readily obtained with the correct proportion of water of crystallization; all the samples prepared according to the directions of 0. Kiihling or J. Wagner contained too much water. Sodium oxalate, prepared after the method of Sorensen, gives accurate and reliable results (see ANALYST, 1903, 306) ; so also does sodium carbonate, if made by gently igniting the bicarbonate. Iot7ometry.-Potassium biniodate and tetroxalate are rejebted, because, as is stated, it is difficult to obtain them in a state of sufficient purity. The method suggested by Kalman of standardizing iodine solution by adding an excess of sodium sulphite and titrating the acid set free is rendered unsatisfactory from the fact that it is difficult to obtain sulphite free from carbonate; the titration is consequently troublesome.The author, therefore, prefers to standardize thiosulphate solution against pure iodine, freshly weighed. Permnngaiiate Solutio)i.-This solution may be titrated with hydrogen peroxide, the strength of which has been determined by measuring the volume of oxygen evolved on shaking with permanganate solution and sulphuric acid. A. &I. Technical Analysis of Water. H. R. Procter. (Joitrii. SOC. Chem Ind., 1904, i., 8-lO.)-The determination of the hardness in water is of importance in many industries, but has hitherto been considered too exclusively from the point of view of the soap user. The soap test may with advantage be replaced by others of a more scientific and reliable character.The ‘‘ temporary hardness,” which is due to bicarbonates, is best estimated by Hehner’s method. The water is titrated with decinormal hydrochloric acid in the cold, using a minimum quantity of methyl orange as indicator. A blank determination should be made with distilled water titrated until the same tint is observed. The end reaction is, however, sharper if alizarinTHE ANALYST. 133 he used as indicator, the titration being conducted at 100” C. This indicator is prepared by mixing 1 gramme of pure alizarin paste with 200 C.C. of water. The colour-change is from violet in alkaline solution to clear pale lemon-yellow in neutral or acid. As this indicator is affected by carbonic acid, it is necessary to conduct, the whole of the titration at the boiling-point, and to use a vessel of either silver, platinum, or porcelain. The ‘‘ permanent hardness,” which is due to various salts, such as magnesium sulphate, should be determined by Pfeifer and Wartha’s modification of Hehner’s method (Zeits.fi angezi!. c‘hem., 1902, 198), in which a fairly large excess of a mixture of decinormal solutions of sodium carbonate and hydroxide are added to 200 C.C. of the water. The liquid is then reduced in volume to 200 c.c., or even less by boiling in a platinum dish or a flask of Jena glass ; it is then cooled, made up to 200 c.c., allowed to settle, 100 C.C. syyhoned off, and titrated with decinormal acid, The difference between this titre and that of the alkali originally added gives the amount of magnesium sulphate or other similar salt present in 100 C.C.of the water. The same water may be used for this operation which was neutralized with acid in the estimation of the temporary hardness; in this case, however, the titration will give the ‘‘ total hardness.” In some cases the permanent hardness as determined by these methods appears to be negative; this is due to the presence of sodium carbonate in the waters in question. Since, in the ordinary methods of water softening, lime is precipitated as carbonate, and magnesia as hydroxide, double the quantity of caustic alkali or lime being consumed, it is desirable to know the amount of magnesia present. This may be ascertained by rendering the water neutral and precipitating with lime-water, which throws down the magnesium, but does not affect the calcium salts.Pfeifer neutralizes 100 C.C. of the boiling water, using alizarin as indicator, as in the determination of the temporary hardness, which determination may be combined with that of the magnesia. A known quantity of clear lime-water (25 or 50 c.c.), which should be at least 50 per cent. in excess of that required for precipitating the magnesia, is added, the whole mixed and allowed to settle, and an aliquot portion of t h e clear liquid taken and titrated with decinormal acid either in the cold, using phenolphthalein, or a t boiling-point, using alizarin as indicator. The diminution from the original titre of the lime-water gives the quantity of magnesia in the water. Free carbonic acid (i.e., CO, not combined as bicarbonate) may be estimated by Archbutt’s method” of titrating with sodium carbonate solution and using phenolphthalein as indicator. A. 11 The Defects of Uncarburetted Water-Gas as Fuel for Laboratory Use. Masume Chikashige and Histoshi Matsumoto. (Joum. SOC. C‘hem. I d . , 1905, ii., 50.)-The flame of this gas is so small that it is necessary in many cases to use a far larger quantity of gas than is actually required to heat a vessel, otherwise the requisite heating surface is not obtained. The flame rapidly destroys air- and water-baths made of copper, the metal being oxidized at the high temperature of the flame. Nickel and platinum vessels are speedily damaged, probably through the alternate * Scr Jour. Sol. (,‘hoiL. Ittd., ssiii., 245.134 THE ANALYST. formation and decomposition of carbide. Porcelain vessels are also unavailable ; they become coated with a coherent deposit consisting largely of iron carbide. When burnt under a sand-bath a quantity of carbon monoxide escapes unconsumed, varying, according to the relative position of the fiame and the bath, from 0.3 to 0.9 volume per 100 volumes of gas burnt. As the volume of water-gas that must be burnt to secure the same heating effect is several times greater than that of coal-gas, the vitiation of the atmosphere is greater. In the discussion that followed Mr. David Howard suggested that many of the defects of the gas might be avoided by using Argand burners instead of Bunsen burners. A. M.