ORGANIC ANALYSIS 145 INORGANIC ANALYSIS. Detection of Bromine by Fluoresceln. H. Baubigny. (Bull. Soc. Chim., 1912, 11, 13.)-The author’s test (Comptes rend., 1897, 125, 654) has been stated by Pribram, and later by Labat, to be inconclusive as applied to the detection of bromine in animal organs, on the ground that chlorine or iodine would give the- same reaction. This is denied: the chlorine compound of fluorescein is almost colourless, and iodine should be converted into iodate by oxidation with permanganate, as indicated in the original paper. 0. E. M.146 ABSTRACTS OF CHEMICAL PAPERS Estimation of Alkalis in Silicates by Fusion with Calcium Chloride. E. Makinen. (Zeitsch. anorg. Chem, 1912, 74, 74-78.)-Instead of the calcium carbonate and ammonium chloride of the Lawrence Smith method (Ann.Chern. u. Pharrn., 1871, 159, 82), anhydrous calcium chloride is recommended. The calcium chloride should be prepared in the laboratory, as the purest commercial salt may contain as much as 0.02 grm. of alkali chloride in 7 grms., the amount of flux recommended for use with 0.5 grm. of the silicate. Intimate admixture of the silicate with the flux is unnecessary, as the latter, unlike the mixture of Lawrence Smith, really melts. The crucible is supported on asbestos millboard, so that its upper portion and the cover may remain relatively cool and so serve to prevent volatilisation of alkali chloride, The crucible is heated gently at first until the calciuw chloride is fused, but is then raised to a full red heat by means of a Teclu burner. After twenty-five to thirty minutes fusion the flame is withdrawn, the crucible cooled suddenly, and the analysis completed in the usual manner, except that the use of ammonium carbonate in-place of ammonium oxalate is recommended for removing the last traces of calcium.Ammonium carbonate is said to separate the last traces of calcium as well as oxalate, and is preferred because of the high temperature which is necessary to expel oxalic acid from the residue of alkali chlorides. When the results obtained diEer from those yielded by the Lawrence Smith method, they are higher, and the author thinks they should be preferred. G. C. J. The me of a blowpipe-flame leads to volatilisation of alkali. Estimation of Ferrous Iron in Silicate Bocks. M. Dittrich and A.Leon- hard. (Zeitsch. anorg. Chem., 1912, 74, 21-32.)-The paper contains a detailed description of the experiments which led Dittrich to publish the statement (ANALYST, 1911, 36, 602) that the uncertain end-point in the titration of ferrous iron in certain rocks, after decomposition by means of hydrofluoric acid, was due to the presence in those rocks of Ti,O,, and not to the presence of manganese. The following method of conducting the estimation of ferrous iron avoids any difficulty due to the presence of Ti,O,. From 0.8 to 1 grm. of the not too finely powdered sample is mixed with 0.5 grm. quartz sand (ANALYST, 1911, 36, 372) and 2 C.C. sulphuric acid, and decom- posed by means of 8 C.C. of hydrofluoric acid. The crucible is placed on a boiling water-bath and covered with an inverted fmnel, which is sufficient to prevent any appreciable oxidation during the few minutes required for decomposition of the rock.The contents of the crucible are poured, without previous cooling, into a mixture of 100 C.C. water, 10 grms. precipitated silica, 20 grms. potassium sulphate, and 2 C.C. sulphuric acid, and the mixture is titrated at once with permaaganate. G. C. J. Estimation of Chromium and its Separation from Vanadium Steels. J. R. Cain. ( J . Ind. Eng. Chem., 1912, 4, 17-19.)-After drawing attention to the difficulties which attend the estimation of chromium in chrome-vanadium steels, the author recommends the following method, which depends on the fact established by him (ANALYST, 1911, 36, 475) that chromium can be precipitated completely by boiling the nearly neutralised (ferrous) solution with barium carbonate, and on the observation of Noyes and Bray (Technology Quarterly, 1908, 21, 14), thatINORGANIC ANALYSIS I47 chromate may be separated from vanadate by precipitation as lead chromate if certain precautions are observed.The drillings, which should preferably contain not more than 0.06 or 0.07 grm. of chromium, are placed in a 300 C.C. Erlenmeyer flask and dissolved in fuming hydrochloric acid of which about 10 C.C. is taken for each grm. of steel. The solution is diluted to 100 or 150 C.C. with hot water and nearly neutralised with saturated sodium carbonate solution. A slight excess of an emulsion of barium carbonate is added and the mixture boiled vigorously for ten or fifteen minutes with small additions of barium carbonate every two or three minutes, the flask being kept covered to exclude air as completely as possible.The excess of barium carbonate used should not be more than 2 grms. or there may be dificulty in extracting the chromium by a single fusion. The precipitateis allowed to settleand is then filtered off and washed twice with hot water, these operations being conducted as quickly as possible. The filter and precipitate are transferred to a platinum crucible, the paper is burned away, 2 grms. sodium carbonate and about 0.25 grm. potassium nitrate are mixed in, and the mixture is maintained in a state of fusion for twenty minutes. The product of the fusion is digested with boiling water and the solution filtered.Hydrogen peroxide (1 or 2 C.C. of a 3 per cent. solution) is added to destroy any nitrite, and the solution is boiled for five minutes to decompose the excess of peroxide. The solution is cooled and transferred to a separator, rendered slightly acid with nitric acid and vigorously shaken to expel carbon dioxide. It is then transferred to a beaker, just neutralised with sodium hydroxide, and then made acid by addition of 2 C.C. nitric acid (1 : 1) for each 100 C.C. of solution. Lead nitrate (20 C.C. of a 20 per cent. solution) is added to the cold solution with vigorous stirring, and the lead chromate is filtered off on asbestos and washed three or four times with cold water. The asbestos mat is transferred to a beaker or flask and the lead chromate decomposed with hot hydrochloric acid (1 : 4).The solution iscooled, madeup to 150 or 200 c.c., and titrated against -& ferrous sulphate with ferricyanide as an outside indicator. The amount of sodium carbonate permissible in the fusion is limited by the fact that for a successful separation of chromium from vanadium not too much sodium nitrate should be present (Noyes and Bray, Zoc. cit.), and it is this necessity for restricting the amount of sodium carbonate which makes it advisable to take for analysis not more of the sample than corresponds to 6 or 7 cgrms. of chromium. Larger amounts are not readily extracted by 2 grm3. of sodium carbonate, especially if too great an excess of barium is present. The shaking of the solution to eliminate carbon dioxide deter- mines the subsequent formation of a precipitate which settles and filters well, and equally cogent grounds underlie other directions as to manipulative detail.Test numbers, obtained with solutions of a vanadium and chromium free steel to which known amounts of bichromate and vanadate were added, show maximum errors of +_ 0*0004 grm. with from 0.02 to 0.3 grm. chromium present. The filtrates from the barium carbonate precipitation were tested for chromium with negative results in every case, and in only a few cases was any trace of vanadium found in the chromium solution after titration, and in those cases the amount was negligibly small. On the other hand, it is well to examine the insoluble matter from the fusion by fusing it again with sodium carbonate and potassium nitrate.With much chromium in the original148 ABSTRACTS OF CHEMICAL PAPERS weight of drillings, a yellow-coloured solution may result. it may be estimated with sufficient exactness colorimetrically. If so, the chromium in G. C. J. Analysis of Higher Oxides of Lead. A. Chwala and E. Colle. (Gazz. Chim. Ital., 1911, 41, 551-579.)-From the results of an experimental study of the different methods proposed for the analysis of lead peroxide the author recommends the following modification of Lux’s method (Zeit. anal. Chem., 19, 153) as the most convenient and reliable : From 1 to 2 grms. of the sample are treated with 100 C.C. of oxalic acid solution, and the liquid boiled for ten to fifteen minutes, until solution is complete. The excess of oxalic acid is then titrated with standard permanganate solution, and the amount of lead peroxide calculated from the result- nitric acid and 25 C.C.of YbO, + C2H20, = PbO + H2O + 2C02 The duration of the boiling with nitric acid and the concentration of that acid have considerable influence upon the results. The results of experiments have shown that the differences between the results given by the oxalic acid method and the iodometric methods are entirely due to the occurrence of secondary reactions in the latter, leading to the formation of lead iodide and methyl acetate, so that part of the iodine escapes titration- In some instances the error thus produced may be considerable, and, in any case, should always be taken into account in iodometric estimations of oxidising substances in the presence of weak acids and heavy metals.Various samples of commercial lead peroxide analysed by the foregoing method contained from 43.5 to 95-56 per cent. of lead peroxide. Pb(CH,COO), + 21 = CH3COOCH3+ PbI2+ GO2 C. A. M. Bismuthate Method for Manganese. D. J. Demorest. ( J . Ind. Eng. Chem., 1912, 4, 19.)-The following modification of the bismuthate method is recommended for use with chrome and chrome-vanadium steels. The drillings are dissolved in nitric acid (sp. gr. l a l ) , boiled to expel nitrous fumes, and potassium bismuthate is added until permanganic acid or manganese dioxide persists on boiling. Potassium nitrite is added to dissolve the manganese dioxide, and the solution is boiled to expel nitrous fumes, after which it is cooled and the manganese oxidised by bismuthate in the cold as usual.The solution is then filtered rapidly through asbestos on glass wool and titrated with a standard solution of sodium arsenite, made by dissolving 2.25 grms. of arsenious acid in a hot solution of sodium carbonate and diluting to 2.5 litres. The test numbers, obtained in presence of three or four times as much chromium or vanadium as manganese, are good, the worst being 0.671 per cent. found with 0,666 present. G. C. J. Determination of the Density of Minerals by Means of Rohrbaoh Solution of Standard Refractive Index. H. E. Merwin. (Amer. J. Science, Silliman, 1911 [43, 32, 425 ; through Chem. Zentralbl., 1912, I., 524.)-The density of a solution contained in the cell of a refractometer is adjusted until a fragment of theINORGANIC AJSALYSIS 149 mineral under examination is just suspended in it ; its refractive index is determined, and from this its density, and hence the density of the sample, is found.For Rohrbach (barium mercury iodide) solution, the relation between D (density) and n (refractive index), at 20" C., is as follows : D = 3.449 3.396 3.246 3.180 3.046 2.980 n = 1.7686 117960 1.7312 1.7195 1.6944 1.6823 D = 2.748 2.649 2.648 2.367 2.163 2.067 For densities between 2-25 and 3.40 it may be expressed, with a maximum error of f 0-002, by the empirical formula : D20= 5.39 (n - 1.5467) + 2-25 =5*39n - 6.0865. For variations from 20" exceeding 3" a correction of +_ 0*001 per L- 2" is necessary. For densities between 2.0 and 2.25, D=5.7n-6.567; and between 3.4 and 3.5, D = 552n - 6.313.For approximations, the mineral is brought to suspension, and n then determined on a drop of the liquid. 12 = 1.6391 1.6207 1.6205 1.5685 1.5320 1.5148 0, E. M. Estimation of Nitrates in Sewage Effluents. H. Silvester. (J. Sot. Chem. Ind., 1912, 31, 95-96.)-Effluents derived from sewage containing wasto gas liquor may contain thiocyanates, and in presence of thiocyanates the phenol- sulphonic acid method for the estimation of nitrates is not trustworthy. The zinc- copper couple method is recommended as the best for sewage effluents. I n using this method it is convenient to have an approximate idea of the amount of nitrate present, and for this purpose the diphenylamine test serves well. To 2 C.C. of the effluent a few drops of the diphenylamine reagent and 10 C.C.of sulphuric acid (free from nitrate) are added. The intensity of the blue coloration which results enables the analyst to form a rough estimate of the amount of nitrate present. G. C. J. Detection of Nitrates in Presence of Nitrites. H. K. Sen and B. B. Dey. (Zeitsch. anorg. Chem., 1912, 74, 52-54.)-When nitrites are destroyed by means of urea, as in Piccini's method for the detection of nitrate in presence of nitrite, small quantities of nitrate are invariably formed by the decomposition of nitrous acid into water, nitric oxide, and nitric acid. I t is shown that the reaction of hydrazine sulphate with nitrites (ANALYST, 1911, 36, 468) is unattended by the formation of even such traces of nitrate as may be detected by the diphenylamine reaction, and hydrazine sulphate is therefore recommended for the preliminary destruction of nitrite when it is wished to test for nitrate in a liquid containing nitrites.G. C. J. Note by Abstractor.-Stutzer and Goy (ANALYST, 1911, 36, 524) have stated that the decomposition of nitrite by hydrazine sulphate is never complete. Estimation of Phosphoric Acid by von Lorenz's Method. H. Neubauer and F. Lucker. (Zeitsch. anal. Chem., 1912, 5lY161-175.)-The following procedure is recommended in estimating phosphoric acid by this method, the principal modi-150 ABSTRACTS OF CHEMICAI, PAPEE28 iication introduced being that acetone is employed instead of alcohol and ether for washing the phosphomolybdate precipitate. The reagents required are : (1) Xolybdic acid reagent.One hundred and fifty grms. of ammonium molybdate are dissolved in 400 C.C. of boiling water, and the solution, when cold, is added gradually to a solution of 50 grins. of ammonium sulphate in 450 C.C. of nitric acid of sp. gr. 1.4 ; the whole is then diluted to a volume of 1 litre, filtered, and stored in a dark place. (2) Nitric acid containing sulphuric acid. Thirty C.C. of sulphuric acid of sp. gr. 1-84 are added to 1 litre of nitric acid of sp. gr. 1.20. (4) A 2 per cent ammonium nitrate solution. A quantity of from 10 to 15 C.C. of the phosphoric acid solution under examination (for instance, the solutions obtained in the analysis of fertilisers for total and citric acid soluble phosphoric acid) is diluted to a volume of 50 c.c.; should sulphates not be present, the dilution is made with the nitric acid containing sulphuric acid, otherwise with the dilute nitric acid.The mixture is then heated nearly to boiling, and 50 C.C. of the molybdate reagent are added, the whole being stirred for five minutes. After the lapse of from two to eighteen hours (the longer period is necessary if less than 3 mgrms. of phosphoric be present), the precipitate is collected in a Gooch crucible, washed with the 2 per cent. ammonium nitrate solution about four times, and then twice with acetone. The crucible is now placed for at least thirty minutes in a vessel in which the pressure has been reduced below 150 mm., and then weighed. The precipitate thus obtained contains 3.295 per cent. of phosphoric anhydride (P,O,).I t is shown that the acetone removes the ammonium nitrate completely from the precipitate and filter, and results of estimations are recorded which show that the method is accurate . w. P. s. (3) Nitric acid of sp. gr. 1.20. (5) Neutral acetone. Determination of the Penetration of Salts in Treated Wood. (Forest Service CiwuLa?; 190, U.S. Dept. of Agriculture ; through Chem Tyade JozLrizal, 1912, 50, 12.)-If zinc chloride is the preservative, a representative disk of the wood is dipped first, for not more than ten seconds, into 1 per cent. potassium ferrocyanide, and then, after removal of the excess with filter paper, into 10 per cent. uraniuiii acetate; it is then allowed to dry. The untreated portion will now be dark red, the treated portion somewhat lighter than the natural wood.The method allows 0.2 pound per cubic foot to be detected, Copper and iron sa1t.s are determined by the dark-red and deep blue coloration, respectively, of the treated zone on immersion in potassium ferrocyanide solution ; mercuric chloride by the blackening in hydrogen sulp hide sola t ion. 0. E. M. Rapid Method for Determination of Sulphur in Roasted Blende. C. C. Nitchie. ( J , Ind. Eng. Chern. 1912, 4, 30-32.)-For determining whether a particular car-load of roasted blende is sufficiently roasted to be suitable for use in the spelter furnaces, or whether it should be returned to the kiln for additional roasting, a rapid inethod is essential. The following method is recommended as rapid, and sufficiently exact for the purpose. The sample is ground no niore finely than is necessary to get a 1-grm.sample which shall be fairly representative; finer grinding prolongs the time necessary for the subsequent roasting of the ore. OneINORGANIC ANALYSIS 151 grm. of the sample is weighed out into a boat, which is placed in a silica combustion- tube previously heated to 1000" C., and a current of air, freed from carbon dioxide, is led through the tube and through an absorption apparatus containing standard caustic alkali. X fume of zinc oxide, formed by reduction of zinc oxids by zinc sulphide, volatilisation of the zinc, and its subsequent oxidation in the air-current, passes through the absorption apparatus, only negligible traces being dissolved by the alkali solution. Experiment has shown that the roasting may be discontinued when this fume is no longer seen in the absorption apparatus.The only sulphur then remaining with the roasted ore is that combined as calcium sulphate. The whole of the "false " sulphur-sulphur which should have been eliminated in the kiln, but which remained as sulphide or sulphate of zinc or lead in the sample- will have passed into the absorption apparatus, mostly as sulphur dioxide with some trioxide. The excess of alkali hydroxide is titrated with standard acid, using phenol- phthalein as indicator, and the sulphur in 1 grin. of the sample is calculated from the alkali neutralised by the gases resulting from roasting the ore. The roasting process usually occupies only about six minutes. The maximum error of the method is -0.1 per cent.on the ore, which is quite near enough for the purpose, the percentage of '' false " sulphur in roasted blende being of the order of 1 per cent. Without; modification the method is not sufficiently exact for application to unroasted blende. G.. C. J. Rapid Volumetric Method for Estimation of Free Sulphur. C. Davis and J. L. Foucar. (J. SOC. Chem. Ind., 1912, 31, 100.)-The method depends on the quantitative conversion of the sulphur into thiocyanate by digestion with an alcoholic solution of sodium cyanide, and the subsequent titration of the thiocyanate. One grm. of the finely ground (60 mesh) spent oxide or other material is dried for an hour in the steam oven and then transferred, with 1.5 grm. sodium cyanide and 50 C.C. absolute alcohol, to a 250 C.C. flask.The contents of the flask are boiled under a reflux condenser for two hours, and the alcohol is then completely removed by distillation. The residue is dissolved in 100 C.C. of hot water, and the solution, when cool, made up to 250 C.C. To 25 C.C. of this solution, 75 C.C. of water and 5 C.C. of a saturated solution of iron alum are added and the mixture heated, with constant stirring, to 95" C., filtered, and the filter washed until free from thiocyanate. To the cool filtrate, 5 C.C. of nitric acid and enough water to bring the bulk up to aboutl 500 C.C. are added, and the mixture is titrated with silver nitrate until the red colour of ferric thiocyanate disappears, or excess of silver nitrate may be added, and the solution titrated back with TG thiocyanate.Should the sample contain thiocyanate, as is most likely in the case of spent oxide, a blank test must be con- ducted, the thiocyanate being extracted with water. Test numbers obtained on applying the method to sulphur crystals agree well with those obtained by oxidising the sulphur and weighing as barium sulphate. The test numbers for spent oxides are from 0.3 to 04 per cent. lower than those obtained by oxidation of the carbon bisulphide extract, but the latter method includes as free sulphur the sulphur of any compound soluble in carbon bisulphide. G. C. J.152 ABSTRACTS OF CHEMICAL PAPERS Estimation of Sulphur in Illuminating Gas. J. Niermeyer. ( J . f. Gas- beleuchtwq, 1911, 54, 1078-1079 ; through Chem. Zentralbl., 1912, I., 375-376.)- According to Dennstedt and Ahrens, the ratio between sulphur dioxide and sulphur trioxide in the products of combustion of illuminating gas shows only small varia- tions: 92 to 94 per cent.of the total sulphur exists as the dioxide and 6 to 8 per cent. as the trioxide. On this observation the author has based a simple method for the estimation of sulphur. A known volume of gas is burnt, the sulphur dioxide is estimated in the products, the quantity of sulphur found is increased by 6 to 8 per cent. and calculated per cub. m. of gas. The sulphur dioxide is absorbed in g8 iodine solution, to which an excess of potassium iodide has been added to reduce the volatility of the iodine. The glass tube of a Drehschmidt sulphur-estimation apparatus is used, and the thin, downwardly inclined portion of the tube is utilised as a cooler for the products of combustion, which are then aspirated through the washing vessel containing the iodine.The gas-consumption is adjusted at 70 to 80 litres per hour ; 10 C.C. of & iodine solution, 10 C.C. of 10 per cent. potassium iodide solution, a few drops of starch solution, and 50 C.C. of water are placed in the washing vessel, and a current of air is aspirated. The burner is then placed under the glass tube, and, when the iodine solution is decolorised, the volume of gas burnt is read off on the meter. The quantity of iodine used represents 1.6 mgrm. of sulphur ; this is divided by the number of litres of gas burnt, and the result is increased by 6 per cent., giving mgrms. of total sulphur per litre. J.F. B. Estimation of Sulphur in Nitrocellulose. C. Kullgren. (Zeitsch. ges. Schiess- und Sprengstofwese?~, 1912, 7, 89.)-About 0.75 grm. of the substance, denitrated by heating for thirty minutes on the water-bath with 10 C.C. hydrochloric acid, is dried in a large porcelain boat. It is then burnt i n a stream of oxygen by the method of Bnrlow (J. Amer. Chenz. SOC., 1904, 26,341), the sulphur being absorbed as sulphuric acid in water or alkali. Hydrogen peroxide is added to the solution to ensure conversion of any sulphur dioxide to sulphuric acid, which is then precipitated with barium chloride solution. The results showed satisfactory agreement. 0. E. M. Estimation of Sulphuric Acid. K. K. Jarvinen. ( A m . acad. scient. Fennica, 1911, Serie A.2. Nr. 16; through Chem.Zentralbl., 1912, I., 526.)-By the method given, about -0025 grin.-molecule of sulphuric acid is titrated with a maximum accuracy of 0.2 per cent. The following ions may be present : sodium, potassium (magnesium j, ammonium, in almost unlimited quantity ; calcium, alu- minium, manganese, zinc, iron, to the extent of 4 equivalents ; copper, 1 ; chloride, 8 ; nitrate and borate, 4 ; phosphate, 2 ; oxalate, 1. Strontium, lead, chromium, and chromate ions must be absent. The solution, containing -002 to -003 grm.-molecule sulphate in 40 to 50 c.c., is neutralised if necessary until about 1 to 3 equivalents ( = 2.5 to 7.5 C.C. 2 N) hydrochloric acid are left free; 3 equivalents are only required if 2 of phosphoric acid or 1 of oxalic acid are present. Ferric salts, 1 to 4 equivalents, are reduced by boiling with 9 to 2 equivalents of hydroxylamine. The solution is boiled up and slowly precipitated with about 40 C.C. (1 to 5 equiv-INORGANIC ANALYSIS 153 slents) of hot benzidine hydrochloride solution ($+Fc HCI). The liquid is now evaporated to one-half, cooled, filtered, the precipitate washed with water contain- ing benzidine sulpbate, transferred to the original flask, and titrated hot with ; sodium hydroxide free from carbon dioxide, and phenolphthalein. The result is usually not more than 0.1 to 0-2 per cent. high. 0. E. M. Rapid Method of Estimating Zinc. K. Voigt. (Zeitsch. angew. Chem., 1912, 25, 205-206.)-V. Hassreidter has stated (Zeitsch. angew. Chew,., 1911, 24, 2471) that the author’s method (ANALYST, 1912,35) is open to objection on the ground that zinc may be underestimated owing to the re-formation of zinc silicate in presence of hydrated silioa. The author points to the test numbers in his earlier paper for proof of the accuracy of the method, and adds that many of those test numbers were obtained in presence of 30 to 40 times as much (‘ soluble I’ silica as zinc. He also points out that Hassreidter’s experiments were not conducted in accordance with his directions, since Hassreidter used the smallest possible excess of ammonia, as he wished finally to complete the analysis by titration with sodium sulphide, which is rendered difficult or impossible by excess of ammonia. The large excess of ammonia directed to be used by the author prevents the zinc passing into the precipitate as silicate. A still larger excess is without influence on the results, but serves no useful purpose (cf. ANALYST, 1912, 35). G. C. J.