INORGANIC ANALYSTS 171 INORGANIC ANALYSIS. Standardisation of Alkalimetric Solutions. F. D. Dodge. (J. Ind. and Eryg. Chem., 1915, 7, 29-30.)-The use of potassium hydrogen phthalate, KHC,H,04, is recommended for standardising alkali solutions employed in volumetric analysis. It may be prepared by dissolving 50 grms. of resublimed phthalic anhydride in about 200 C.C. of water, exactly neutralising the solution by the addition of 50 per cent.potassium hydroxide solution (about 120 grms. of this solution will be required), then adding a further 50 grms. of the anhydride, and heating the solution until a11 the crystals have dissolved. The solution is now diluted with water to about 550 grms., filtered if necessary, and stirred while cooling. The crystals are collected, recrystallised from 300 C.C.of hot water, and dried at 110" C. The salt is anhydrous, Don-hygroscopic, stable, and colourless; it is soluble in 10 to 11 parts of cold water and in about 400 parts of alcohol. w. P. s. Detection of Bromides in the Presence of Thiocyanates, Cyanides, and Ferrocyanides. L. J. Curtman and A. G. Wikoff. (J. Amer. Chem. SOC., 1915, 37, 298-301.)-Thiocyanates, cyanides, and ferrocyanides interfere with the libera- tion of bromine from bromides by chlorine water, apparently owing to the latter having a selective oxidising action upon them.The following method of pre- cipitating them as cuprous salts in the presence of sulphurous acid enables 2 mgrms. of bromine to be detected in the presence of 500 mgrms. of each of the inter- fering substances : The solution, which should be neutral or slightly acidified with sulphuric acid, is mixed with 15 C.C.of a, saturated solution of sulphur dioxide, heated just to boiling, and treated while hot, and with coutinual stirring, with172 ABSTRACTS OF CHEMICAL PAPERS 2N-copper sulphate solution, which is added drop by drop until the supernatant liquid is blue.(A green colour in the case of thiocyanates indicates the addition of insufficient sulphurous acid.) The liquid is filtered while hot, and the filtrate and wash- ings evaporated to 5 to 10 c.c., any slight white precipitate formed in the concentra- fion being disregarded. I t is then tested with 1 C.C. of 3N-sulphuric acid and 1 C.C. of 1 per cent. potassium permanganate (or dilute hydrochloric acid and chlorine water) and shaken with carbon bisulphide.If permangaaate be used, care must be taken not to leave the carbon bisulphide too long in contact with it, since they react and form a small amount of manganese dioxide, which may dissolve and give a colora- tion similar to that given by small quantities of bromine. C. A. M. Comparison of the Gunning-Copper Method with the Kjeldahl-Gunning- Arnold Method for the Estimation of Nitrogen.0. F. Jensen. (J. Ind. and Eng. Chem., 1915, 7, 38-39.)-Estimations of nitrogen in bone meal, dried blood, cyanamide, and linseed-meal by the Gunning-copper method (use of 0.3 grm. of copper sulphate in addition to potassium sulphate) and the Kjeldahl-Gunning- Arnold method (use of metallic mercury in place of copper sulphate) showed that with either method a digestion of from one to one and a half hours gives a quantitative yield of ammonia, except in the case of dried blood, when three hours’ digestion is necessary in the Gunning-copper method.The latter method possesses the advan- tage that the use of sodium sulphide is unnecessary in the distillation part of the process; the copper sulphate also acts as an indicator in adding the sodium hydroxide before distilling, so that a large excess of alkali may be avoided.w. P. s. Rapid Method of Estimating Nitrates. E. Knecht. (J. SOC. Chem. Id., 1915, 34, 126-127.)-The method is a modification of one previously proposed by the author (Ber., 1903, 36, 166) depending on the fact that titanous hydroxide is capable of effecting the complete reduction of nitrates (and nitrites) to ammonia.When caustic soda is added to a solution of a, titanous salt, black titanous hydroxide is precipitated, and this begins to decompose almost at once, yielding nascent hydrogen and the white titanic hydroxide, probably according to the following equations : Ti,( SO,), + 6NaOH = Ti,(OH), + 3Na2S0,. Ti2(OH)6 + 2H,O = STi(OH), + H,.A convenient amount of nitrate for a determination is about the equivalent of 0.1 grm. of potassium nitrate. Ten C.C. of a 1 per cent. solution are measured into a copper flask, excess of caustic soda is added, and then about 20 C.C. of commercial titanous sulphate or chloride. The solution is at once distilled into standard acid, and the liberated ammonia estimated.The reagents must be added in the order stated, or there will be a loss of nitrous or nitric oxide. Pure potassium nitrate showed 99.80 and 99.83 per cent. as the result of duplicate determinations, the cost of which for materials is considerably less than one penny. H. F. E, H.INORGANIC ANALYSIS 173 Comparison of Some Methods for Total Phosphoric Acid in Super- phosphate.C. A. Peters and A. G. Weigel. ( J . Ind. and Eng. Chem., 1915, 7, 39-40.)-The estimation of phosphoric acid by precipitation first as phospho- molybdate and then as ammonium magnesium phosphate gives high results unless the phosphate solution has been previously evaporated to dryness to remove silica. Trustworthy results are obtained by other methods-namely, the volumetric method of Pemberton, and the direct precipitation of the phosphoric acid as ammonium magnesium phosphate in the presence of ammonium citrate or citric acid, and direct weighing of the yellow phosphomolybdate precipitate.w. P. s. Volumetric Estimation of Silver. C. Debrun. (Ann. Falsific., 1914, 7, 407-409.)-1n the Gay-Lussac method for the volumetric estimation of silver, the strength of the standard chloride sodium solution employed is so adjusted that 100 c.c., measured at 1 5 O C., are equivalent to 1 grm.of silver. If measured at other temperatures, a correction must be applied for the contraction or expansion of the solution, rand the author gives the correction for each degree between 0' and 35' C. Thus, 100 c.c., measured at 15' C., will measure 99.953 C.C.at 0' C.; 99.938 C.C. at 5" C. ; 99.956 C.C. at 10' C. ; 100.092 c.0. at 20' C . ; 100.205 C.C. at 25' C. ; 100.328 C.C. at 30' C.; and 100-484 C.C. at 35' C. w. P. s. Analysis of Mine Waters. J. H. Capps and 0. W. Boies. (J. P 7 q ~ Chem., 1915,19, 65-75 ; through J. Soc. Chem. Ind., 1915, 34, 175.)-Free sulphuric acid in mine water cannot be determined directly, owing to gradual hydrolysis of the ferric salt during the titration.Electrolytic reduction of the ferric salt invariably resulted in the liberation of measurable amounts of hydrogen. This source of error was avoided by reducing with cadmium amalgam in an atmosphere of hydrogen. A portion of the mine water (50 c.c.) was first titrated with potassium permanganate (1 C.C. =0*005258 grm.Fe) for ferrous iron. A small separating funnel and a three- way tap were sealed to the central tube and side tube respectively of a Drexel gas-washing bottle, and after introducing a convenient amount of the mine water and exhausting, the bottle was well shaken ; hydrogen was admitted, and the bottle again shaken till equilibrium resulted. The process was repeated twice, the amalgam run in, and the bottle shaken vigorously until the attainment of complete reduction --i.e., till a drop of the solution gave no colour with potassium thiocyanate (about two minutes).One aliquot part of the solution (50 c.c.) was then titrated with sodium carbonate for free acid and another portion (50 c.c.) with permanganate for total iron. Methyl orange indicator was employed in the former operation, the end-point being ascertained by comparison with a standard neutral tint.The absolute end-point of the standard was ascertained by conductivity measurements made during titration of a similar solution in presence of methyl orange. The result- ing tint was permanent when the solution was enclosed in a sealed bulb of clean glass. An artificial '' mine water " containing 0.429 grm.of ferric iron, 0.774 grm. of ferrous iron, and 3.85 grme. of sulphuric acid per litre, also traces of aluminium, calcium, magnesium (as sulphates), and sodium (as chloride) gave, respectively, OD174 ABSTRACTS OF CHEMICAL PAPERS analysis by the above method, 0.427 grm., 0.769 grm., and 3.825 grms. Similar solutions of ferric iron in sulphuric or hydrochloric acids, after reduction (four to five minutes) in the same way with cadminm amalgam, gave extremely concordant results upon subsequent titration with permanganate and bichromate, respectively.In these cases the preliminary expulsion of oxygen from the solutions was unneces- sary, but the air in the apparatus was displaced by carbon dioxide. Zinc amalgam effected a rapid reduction (0.5 minute), but the results were vitiated by the simultaneous displacement of appreciable amounts of iron from the solutions. Separation of Tungsten from Molybdenum.E. E. Marbaker. (J. Amer. C3wm. Soc., 1915, 37, 86-95.)-The tungsten and molybdenum are first obtained a8 a solution of sodium tungstate and molybdate. To this boiling solution, which should not contain more than 0.3 grm.of either metal, a 25 per cent. solution of stannous chloride in concentrated hydrochloric acid is added in the proportion of about 20 C.C. for each 0.16 grm. of tungstic acid supposed to be present. Larger amounts of tin give trouble at a later stage, whilst the separation of tungsten is incomplete unless enough stannous chloride is added. The total volume of the solution should lie between 60 and 300 C.C.After boiling for a few minutes, the blue precipitate of W,Os is allowed to settle, the orange supernatant liquid is decanted through a filter and the precipitate washed by decantation with 5 per cent. hydrochloric acid until the washings give no reaction for molybdenum when tested by the zinc and thiocyanate method. The precipitate is then brought on to the filter, allowed to drain, partially or completely dried, according as it is small or large in amount, and ignited to WO,.The filtrate and washings are evaporated to small bulk, 5 to 10 grms. of 20-mesh zinc is added to throw tin out of solution, and after ten minutes the green solution is decanted through a filter and the tin is washed several times with hot water, which is also filtered and collected.The filtrate and washings are heated to 60' C. and passed through a Jones reductor, through which 50 C.C. of dilute (1 : 40) hydro- chloric acid has first been passed. The filtrate and washings are followed through the reductor by 150 C.C. of dilute (1 : 40) hydrochloric acid, and finally by 150 C.C. of hot water, all the liquid being collected in the flask of the reductor, which has been previously charged with 20 C.C.of a 10 per cent. solution of iron alum and with 20 C.C. of Reinhardt's solution (90 grms. manganese sulphate, 175 C.C. sulphuric acid, 175 C.C. syrupy phosphoric acid, 650 C.C. water). When the green solution enters the ferric solution, the colour changes to red, This red solution is finally titrated with permanganate, an operation attended with no difficulty, as the red colour disappears as the titration proceeds, and the final change is from faint bluish green to pink, The molybdate is reduced in the reductor to Mo203, which reacts quan- titatively with the ferric solution to give molybdate and ferrous salt.The results, both for tungsten and molybdenum, are never more that 0.5 per cent.in error. The method for tungsten, if applied to a solution of pure sodium tungstste, gives low results ; but if a little (0.2 grm.) sodium chloride or potassium sulphate be added, the results are exact. In presence of notable amounts of molybdenum, it is unnecessary to make any such addition. G. C. J.APPARATUS, ETC. 175 Electrolytic Separation of Zino, Copper, and Iron from Arsenic.A. K. Balls and C. C. McDonnell. (J. Ind. and Eng. Clzem., 1915, 7, 26-29.)- For the electrolytic separation of zinc, copper, iron, and small amounts of lead from arsenic, it is essential that the latter should be in the pentavalent condition, since in the lower form of oxidation it will be partially precipitated. Another essential condition, especially when iron is present, is that there should be an excess of alkali (e.g., 10 to 20 grms.per 100 C.C. of electrolyte). Potassium hydroxide is more satisfactory than sodium hydroxide, and the precipitation of iron or copper hydroxides is prevented by the addition of tartaric acid. The addition to the electrolyte of glycerol, or a mixture of equal parts of glycerol and alcohol, causes better adherence of zinc deposits to the basin, which may be of nickel, roughened by etching or a sand-blast.To prevent deposition of carbon with the iron, the ourrent should be maintained below 1.7 amperes per 100 sq. cm., and preferably below 1 ampdre, while the addition of a few C.C. of alcohol will prevent iron oxide from depositing at the anode. In order to make copper deposits adhere firmly, rapid rotation (1,000 to 1,600 revolutions per minute) of the anode is necessary, and an addition of about 2 grms. of potassium nitrate is made.The arsenic may be estimated iodimetrioally in the liquid siphoned off the deposits. C. A. M. APPARATUS, ETC. Fractional Distillation with Regulated Still-Heads. M. A. Rosanoff and C. W. Bacon. (J. Amer. Chem.SOC., 1915, 37, 301-309.)-1t is stated by Brown (J. Chem. Soc., 1880, 37, 49 ; 1881, 39, 517), that in distillation with a still-head maintained at a constant temperature the composition of the distillate is constant, and is identical with that of the vapour evolved by a mixture whose boiling-point equals the temperature of the still-head. When, however, the boiling-point curve of a mixture passes through a maximum or a minimum, Brown's law requires modi- fication.If the boiling-point curve be plotted with the abscises representing the composition of liquid mixtures, the diagram may be divided into two parts by a straight line perpendicular to the composition axis and passing through the maxi- mum or the minimum of the curve, and each part may be regarded as an independent diagram.The composition of the vapour from the regulated still-head will then depend upon whether it fall within the right or left-hand side of the diagram. For example, in the case of a mixture of carbon bisulphide and acetone (where the boiling-point curve passes through a minimum), a mixture containing only a small percentage of acetone was distilled in an apparatus of the type previously described (J. Amer. Chem. SOC., 1909, 31, 448; and ANALYST, 1914, 39, 306). The still-head was kept constant at 40-02O C., and seven consecutive fractions of distillate contained 75.2 to 75.1 per cent. of carbon bisulphide-a composition falling to the right of the perpendicular. In a second experiment the mixture contained 29 per cent. of carbon bisulphide (a composition lying within the left part of the diagram), while the still- head was maintained at the same temperature. Fifteen consecutive fractions of the distillate then contained from 55.4 to 55.3 per cent, of carbon bisulphide. Analogous176 REPORTS results were obtained with mixtures of ethyl acetate and ethyl iodide (boiling-point curve passing through a, minimum) and of chloroform and acetone (boiling-point curve passing through a maximum). C. A. M.