ORGANIC AKALYSIS 255 INORGANIC ANALYSIS. Estimation of Carbonie Acid, Combined and Free, in Solution, partieu- larly in Natural Waters. J. Johnston. ( J . Amer. Chenz. Soc., 1916, 38, 947- 975.)-It is shown that the results of the ordinary titration method for free carbon dioxide cannot be interpreted unless the constitution of the solution is already known. Free carbon dioxide is readily determined by a gasornetric method, provided that the concentration of combined carbon dioxide is very small, and even when it is large by observing precautions discussed in the text.The most common method of estimating the relative proportions of bicarbonate and carbonate-viz., by titration first with phenolphthalein and then with methyl orange-only yields approximate results, though results of sufficient accuracy can in many cases be obtained by following the procedure recommended by Kiister (Zeitsch.anorg. Chenz., 1897, 13, 127), and discussed in the present paper. In general, however, it suffices to estimate (a) the total base equivalent to the combined carbonate, and (b) the total carbon dioxide, free and combined, both of which are easy, trustworthy, and unaffected by the presence of alkaline earths or iron.With a knowledge of these one can calculate (c) carbonate, ( d ) bicarbonate, and (e) hydrogen ion concentration, which is a measure of the acidity (or alkalinity) of the water ; for, since we are dealing with an equilibrium capable of fairly rapid readjustment, we are justified in applying the equilibrium constants to calculate the above quantities in the great majority of cases in which a knowledge of them is of real importance.G. C. J.256 ABSTRACTS OF CHEMICAL PAPERS Analysis of Aeid Calcium Bisulphite Solutions. E. Hagglund. ( Chem. Zed., 1916, 40, 433-434; through J. Xoc. Chenz. Im?., 1916, 35, 686.)-The usual method employed for the analysis of the fresh “acid” in the pulp-mill is that of Winkler-viz., titration of the total sulphur dioxide by iodine, and of the 6 ‘ free ” acid by sodium hydroxide in presence of phenolphthalein.I n the latter, titration errors are caused by the presence of carbonic acid derived from the limestone and of volatile organic acids in the recovered gases blown off from the digesters. The end-point with phenolphthalein also is not sharp.Any satisfactory method must be readily appli- sable to mill routine, andmust indicate the percentage of calcium oxide in the liquor with an error not exceeding - 10 per cent. or + 5 per cent. I t is more dangerous to employ too little than too much calcium oxide in the digestion. A method for which greater accuracy is claimed was proposed by Oman, consisting in precipitating the calcium as normal sulphite by adding to 100 C.C.of the ‘‘ acid” 30 C.C. of strong ammonia solution, filtering after fifteen minutes, and washing the precipitate three times with 50 C.C. of 15 per cent. ammonia each time. The precipitate is transferred to a 500 C.C. flask acidified with 10 C.C. of strong hydrochloric acid, the solution made up to the mark and titrated into 10 C.C.of iodine solution. Owing to the lesser solu- bility of calcium sulphite, practically the whole of the calcium sulphate is precipitated as sulphite, and a correction may be applied for this. The author has compared the Winkler and Oman methods with the results obtained by standard analytical methods, and found that both are liable to equally large errors, both frequently giving too low results for calcium oxide.The Oman method offers no advantage; apparently the calcium sulphite tends to oxidise on the filter, and possibly sulphur dioxide escapes during the acidification. A modification of the method is therefore proposed : 10 C.C. of clear sulphite “acid” are added to about.50 C.C. of water and 3 C.C. of strong ammonia. The liquid is made up to 100 c.c., the precipitate allowed to settle for -fifteen minutes, and 10 C.C.of the clear supernatant solution drawn off and titrated with iodine. A correction based on the average percentage of calcium sulphate in the liquors, = + 0.13 per cent. SO,, is applied. The maximum divergences recorded in a series of twelve analyses were equivalent to - 1 and + 4.1 per cent. of CaO. Rapid Method for Comparing the Decolorising Efficiency of Charcoals.L. Wickenden and J. W. Hassler. (J. Ind. and Eng. Chem., 1916, 8, 518419.)- One hundred C.C. of a solution of Aniline Red, Soudan III., or of Oil Red RN, in kerosene oil (0.25 grm. per litre), are mixed with 1 grm. of a charcoal of average quality, and the beaker heated in a steam-bath. The contents are constantly stirred and filtered after ten minutes, and the colour of the filtrate given the arbitrary value 10.In like manner filtrates with 2, 3, 4 grms., etc., up to 10 are prepared, and numbered 20, 30, etc., up to 100. I n testing a sample of charcoal, 5 grms. are ltreated with the aniline solution as described, and the colour of the filtrate compared with those of the standard filtrates.If, for example, the filtrate matches No. 70, 5 grms. of the charcoal will possess 40 per cent. more efficiency than an average char- coal. Charcoals which show high decolorising efficiency in this test show equally high results as decolorisers of cottonseed, coconut, and palm kernel oils. The solutionINORGANIC ANALYSIS 257 of the dyestuff can be kept for several weeks without perceptible fading, but it is advisable to test it at intervals with charcoal selected afi the average standard.C . A. N, Separation of Erbium from Yttrium. P. S. Willand and C. James. ( J . Amer. Chem. Xoc., 1916, 38, 1198-1202.)-Experiments are described which had for their object the discovery of the best reagent for the fractional separation of erbium from a larger quantity of yttrium.The best reagents, among those tried, were sodium nitrite and potassium cobalticyanide. When using sodium nitrite, the oxides are dissolved in hydrochloric acid, the solution is diluted and stirred with steam. Sodium nitrite solution in moderate quantity is then added drop by drop and the precipitate filtered off. More nitrite is added to the filtrate and the process continued, the final filtrate being treated with concentrated oxalic acid to recover the earths not precipitated by nitrite.This final precipitate is freed from sodium after ignition by boiling with water. When 10 grms. of mixed oxides were treated in this way and separated into six fractions, the first corresponded to an oxide of a metal with an atomic weight of 116, and the last to an oxide of a metal with an atomic weight of 90.The separation effected by cobalticyanide was about as good as with nitrite. Sodium phosphate was somewhat less satisfactory as a fractional precipitant, whilst the other methods tried, which included fractional crystallisation of the ammonium double sulphates, diphenylmonosulphonates, picrates, and precipitation by means of ferrocyanides, were much less satisfactory. G.C. J. Estimation of Lead as Lead Sulphite. H. Pellet. (Ann. Chim. anal., 1916, 21, 114-116.)-The lead solution is rendered slightly alkaline with sodium hydroxide, then acidified with acetic acid, and treated with an excess of sulphur dioxide. The latter is most conveniently obtained from a siphon of the compressed gas. The precipitated lead sulphite is collected on a filter, washed, dried, and weighed.The weight obtained is multiplied by 0.721 to obtain the equivalent quantity of metallic lead. This method of estimating lead is particularly useful in the case of sugar solutione which have been clarified with lead acetate. It also aftords a means of separating lead from other metals which Elre precipitated by hydrogen sulphide.w. P. s. Behaviour of Metals towards Certain Aeids eontaining Hydrogen Peroxide. E. Salkowski. (Chem. Zeit., 1916, 40, 448-449.)--Copper, bismuth, nickel, gold, platinum, and antimony, are dissolved by a mixture of hydrochloric acid (sp. gr. 1.125) and 30 per cent. hydrogen peroxide solution. Silver and mercury do not dissolve in the mixture, and lead is attacked only slightly owing to the forma- tion of a protective coating of lead chloride.Dilute sulphuric acid containing hydrogen peroxide dissolves copper, silver, nickel, and bismuth, but not tin, lead, gold, platinum, or antimony. A mixture of acetic acid and hydrogen peroxide dissolves copper, silver, mercury, lead, and bismuth, but not tin, nickel, gold, or platinum, Aluminium is soluble in the three acids mentioned, and the presence of258 ABSTRACTS O F CHEMICAL PAPERS hydrogen peroxide is without effect.The insolubility of mercury in the mixture of hydrochloric acid and hydrogen peroxide is difficult to explain, seeing that the solvent action of the mixture is that of chlorine itself. I t cannot be due to the catalytic action of the metal on the peroxide, or gold and platinum would not be dis- solved, nor to the reducing action of the hydrogen peroxide on mercury chloride.Although hydrogen peroxide reduces mercuric acetate to mercurous acetate, it has no action on mercuric chloride. w. P. s. Titrations with Permanganate in Strongly Alkaline Solutions. B. Brauner. (Zeitsch. anal. Chein., 1916, 55, 225-267.)-Arsenious acid, in strongly alkaline solution, reduces permanganate to the manganic condition state, provided that an electrolyte such as potassium sulphate is present to cause the colloidal manganese hydroxide, Mn(OH),, to separate as a precipitate. If the manganese hydroxide remains in colloidal suspension, the reduction proceeds as far as the manganous state.When manganous salts are oxidised in alkaline solution, hydrated manganese dioxide is produced, both from the oxidised salt and the reduced permanganate.Thallous salts are oxidised quantitatively to thallic (T1,0,) salts, and cerous to ceric salts. Selenious and tellurous acids are converted into selecic apd telluric acids respectively, and ferrous into ferric salts. Partial oxidation takes place in the case of lead, the product formed containing rather more oxygen than corresponds with the mixture YbO + PbO,.I n slightly alkaline solution nickel is oxidised to Ni,,O,,, whilst in strongly alkaline solution the oxidation proceeds approximately to the formation of Ni,O,. An oxide, containing somewhat more oxygen than corresponds with the formula Co203, is formed when cobalt is oxidised with permanganate in strongly alkaline solution.w. P. s. Separation of Metals of the Tin Group in Qualitative Analysis. J. M. Weleh and H. C. P. Weber. ( J . Ind. and Eng, Chem., 1916,38,1011-1016.)- The reduction of stannic compounds to stannous compounds by means of metallic lead can be carried out with sufficient ease and accuracy to make it available as a method for the detection of tin.These two observations are the basis of the method of analysis which follows. The sulphides of the tin group are separated from those of the copper group by treatment with ammonium polysulphide to which about 5 per cent. of sodium hydroxide has been added. This mixture does not dissolve copper sulphide as ammonium polysulphide alone does, nor does it dissolve mercury sulphide as sodium sulphide alone does. The sulphides of the tin group are dried superficially by suction or by pressing between filter-paper, and are then heated with 10 C.C.of hydrochloric acid for ten minutes on the water-bath. The arsenic sulphide is filtered off, and the filtrate is diluted to about 70 c.c., heated, and saturated with hydrogen sulphide. Antimony, if present, begins to separate before tin, and can be recognised by its red colour.If this gives place to brown with further addition of hydrogen sulphide, tin is present, and with practice the approximate ratio of tin to antimony can be judged by the shade. The mixture is next boiled to expel excess of hydrogen sulphide, 5 C.C. of 3 per cent. hydrogen psroxide solution are added, and the solution heated until theINORGANIC ANALYSIS 259 precipitate is redissolved.Oxalic acid (5 to 10 grms.) is added and hydrogen sulphide again passed through the hot solution, which is allowed to cool while it is being saturated. The antimony sulphide is filtered off, and a portion of the filtrate is boiled with 1-2 grms. of granulated test lead for two or three minutes. The solution is then chilled and filtered into a solution of mercuric chloride.In presence of tin, white mercurous chloride is formed. The method will detect as little as 0.5 mgrm. of tin. G. C. J. Estimation of Tin in Tin Ashes. N. Welwart. (Chem. Zeit., 1916, 40, 458- 459.)-The method described consists essentially in separating the tin as metastannic acid, converting this into the sulphide, dissolving it in hydrochloric acid, reducing the solution with metallic antimony, and titrating the resulting stannous chloride with standard iodine solution.From 3 to 10 grms. of the sample are boiled with 30 to 100 C.C. of nitric acid (1:l) until nitrous fumes are no longer given off, the solution is then diluted with three times its volume of water, boiled for a further five minutes, and filtered.The insoluble portion is washed first with dilute nitric acid, then with water, dried, ignited in a porcelain crucible, and weighed. A weighed portion of about 0-3 grm. of this ignited residue is fused with a mixture of sodium carbonate and sulphur, the melt dissolved in dilute hydrochloric acid, the solution boiled until all hydrogen sulphide has been expelled, and then diluted with three times its volume of water.To the hot solution are added 25 C.C. of concentrated hydrochloric acid and 2.5 grms. of antimony powder, and the mixture is boiled for thirty minutes while a current of carbon dioxide is passed into it; the mixture is then cooled, the current of carbon dioxide being maintained, and titrated with TG iodine solution which has been standardised against pure tin.Lead, copper, antimony, and zinc, do not interfere with the estimation, but if large quantities of iron are present the sulphide melt must be dissolved in water and the solution filtered before hydrochloric acid is added. w. P. s. Volumetric Estimation of Tin by Potassium Xodate. G. S. Jamieson. (J. Ind. and Eng. Chem., 1916,8,500-502.)-1n the absence of interfering substances stannous solutions can be titrated satisfactorily with potassium iodate.Among the substances that interfere are ferrous, cuprous, and antimonious salts and precipitated metals. Pure stannous solutions are made 80 strongly acid with hydrochloric acid that the fully titrated solution will not contain less than 10 per cent. of hydrogen chloride.To the cooled solution, contained together with 6 C.C. chloroform in a stoppered bottle, a 0.9 per cent. solution of potassium iodate (1 C.C. =0.01 grm. tin) is added rapidly while shaking the bottle so as to give the contents a gentle circular motion until the iodine colour, which gradually appears, has increased to the maximum amount. The stopper is then inserted and the solution well shaken.The titration is continued with thorough shaking of the closed bottle after each addition until the end-point (disappearance of the violet colour of the chloroform) is obtained. With pure stannous solutions the results are exact. Stannic solutions are reduced by boiling with sheet nickel and at least half their260 ABSTRACTS OF CHEMICAL PAPERS bulk of concentrated hydrochloric acid for about forty-five minutes.Towards the end a current of carbon dioxide is introduced into the flask, and under these conditions the solution is cooled almost to Oa C. The cooled solution is filtered through absorbent cotton, sand (1 inch), and more cotton supported on a perforated disc in a, calcium chloride tube, the filtrate being received in a stoppered bottle containing 6 C.C.of chloroform. Other methods of filtration, which were tried, failed to remove finely divided nickel which interferes with the subsequent titration. The flask and filter are washed with dilute (1:l) hydrochloric acid, and the solution titrated as above described. From the consumption of iodate, a deduction has to be made before calculating the tin. This deduction, which approximates 0-5 c.c., is determined by means of a blank determination, a fresh blank being run whenever a new sheet of nickel is taken into use.With solders and type metals, a 0.5 grm. sample is heated with 15 C.C. of sulphuric acid, until sulphur dioxide is no longer evolved. After cooling, 20 C.C. of water and 15 C.C. of hydrochloric acid are added, after which the tin is reduced and titrated as above described. Bronzes are decomposed with nitric acid, and the metastannic acid filtered off through a Gooch crucible.The washed precipitate and asbestos are boiled with 15 C.C. of sulphuric acid for two or three minutes, the solution cooled, 20 C.C. water and 15 C.C. hydrochloric acid added, the asbestos filtered off on a Gooch crucible, and the filtrate reduced and titrated as above described.G. C. J. Use of Titanium Trichloride in Volumetric Analysis. A. Monnier. (Ann. Chim. anal., 1916, 21, 109-113.)-Titanium trichloride solution may be used for the titration of ferric and cupric salts and chromates, methylene blue being used as the indicator except in the case of cupric salts where safranine or induline is used for the purpose.In the titration of ferric salts, the solution is acidified with hydrochloric acid, 4 drops of a 0.2 per cent. methylene blue solution are added, the mixture is boiled and titrated, in an atmosphere of carbon dioxide, with titanium trichloride solution which has been standardised against a known quantity of iron. The dis- appearance of the blue cqlour indicates the end-point of the titration.The small quantity of titanium trichloride required to decolorise the added indicator is estimated separately and deducted from the total quantity. A similar procedure is adopted in the case of cupric salts. The reason for using safranine or induline in this titration is that these substances are not reduced until all the copper has been reduced, whilst methylene blue is reduced before the copper.The end-point in the titration of chromates is quite sharp, provided that the solution is sufficiently dilute. Copper and iron in the same solution may be estimated as follows: An aliquot portion is titrated in the presence of methylene blue, whilst a, second aliquot portion is titrated with safranine or induline as the indicator.The first titration gives the quantity of ferric iron, the second the ferric iron and copper together. For the estimation of titanium in ores, etc., the hydrochloric acid solution of the metals is reduced with zinc, filtered, the filtrate treated with an excess of standard ferric chloride solution, and this excess then titrated with titanium trichloride solution. w. P. s.INORGANIC ANALYSIS 261 Many of these details are to be found in the monograph entitled ‘I New Reduc- Longmans, tion Methods in Volumetric Analysis.” Green and Go.London, EDITOR. By E. Knecht and E. Hibbert. Estimation of Vanadium by Cwpferron. W. A. Turner. (Qhem. News, 1916, 113, 284-285.)-The vanadium in a metavanadate solution can be precipitated quantitatively by means of Baudisch’s ‘‘ cupferron ” reagent (the ammonium salt of nitrosophenylhydroxylamine ; cf. ANALYST, 1910, 35, 78, 32’7, 453). A necessary condition is that the solution be acid with a mineral acid, but not more than 1 per cent. of hydrochloric or sulphuric acid is needed. The 6 per cent. solution of the reagent, prepared as directed by Baudisch (ANALYST, Zoc. cit.), is added to the solution until the appearance of white nitrosophenylhydroxylamine shows that excess has been added, the vanadium compound being of a mahogany colour. A few C.C. of the reagent are added in excess, the precipitate is filtered off on paper without delay and washed with cold 1 per cent. sulphuric acid containing 1.5 grms. of ‘‘ cupferron” per litre. The precipitate and paper are dried, then heated in a platinum crucible with a small flame to expel volatile substances, the paper incinerated at as low a tempera- ture as possible, and the residue finally heated for some time in the uncovered crucible to insure its being in the highest state of oxidation (V,O,). The test numbers given in the paper are exact, but no example of separation of vanadium from other metals is given. G. C. J.