ORGANIC ANALYSIS 147 INORGANIC ANALYSIS. Behaviour of Antimony Stains or Deposits towards Hypochlorite Solution. W. Vaubel and A. Knocke. (Chew. Zeit., 1916, 40, 209-210.)-The ordinary hypochlorite test for distinguishing between antimony and arsenic stains, deposits, or spots, such as are obtained in the Marsh apparatus, gives reliable results only when freshly-prepared hypochlorite solution is employed.Old hypochlorite solution contains chlorite, and the latter dissolves the antimony mirror. Antimony stains on a porcelain surface have a varying composition; the velvet black stain is soluble in chlorite solution, whilst the brown-black or grey stains are insoluble. After twenty-four hours, the velvet-black stain becomes more or less insoluble. W. P. S. Detection of Cadmium. R.Salvadori. (AnnaZi Chim. AppZZc., 1916, 5, 25-26.)-An ammoniacal solution of ammonium perchlorate (20 per cent. in ammonia solution of sp. gr. 0.90) affords a sensitive test for cadmium, which it precipitates as a148 ABSTRACTS OF CHEMICAL PAPERS white crystalline double perchlorate [Cd.(ClO,), . 4NHJ. I n applying the test to the mixed sulphides of copper, cadmium and bismuth as precipitated in the second group, the precipitate is dissolved in nitric acid, and the solution treated with ammonia in excess, The bismuth hydroxide is filtered off, and the blue filtrate treated with the reagent, which precipitates the cadmium, even in the presence of five times its amount of copper. C.A. M. Solubility of Mineral Phosphates in Citric Acid Solution.G. S. Robert- son. ( J . Soc. C‘henz. Ind., 1916,35, 217-220.)-Experiments described in detail show that 2 per cent. citric acid will completely dissolve mineral phosphates if a sufficient number of extractions be made. In the case of most of the phosphates examined from 90 to 100 per cent. of the phosphoric acid was dissolved. The presence of a small amount of free lime or calcium carbonate caused a pronounced reduction in the solubility, and when much calcium carbonate is present the citric acid test, as ordinarily applied, is a test of the calcium. On the other hand, the greater the pro- portion of lime in actual combination with the phosphoric acid the greater the solu- bility.The fineness to which the slag is ground affects the solubility to the extent of about 10 per cent. decrease for each of the grades “Passes 100,” “ Refuses 100,’’ ‘‘ Refuses 60,” and “ Refuses 30.” The solubility is progressively decreased by calcining.The general conclusion drawn is that mineral phosphahes are a6 valuable sources of phosphoric acid as basic slag for plants, and that the citric acid test is valueless as a criterion of the relative suitability of phosphatic manures for plant life.C. A. M. Influence of Fluorspar on the Solubility of Basic Slag in Citric Acid Solutions. G. S. Robertson. ( J . ,306. Chem. Ind., 1916, 35, 216-217.)-The solubility of basic slag produced by the use of fluorspar ranges from 20 to 50 per cent. instead of from 70 to 90 per cent. in the citric acid test. Two samples (5 grms. each) of “ fluorspar slag^," when extracted five consecutive times with 500 C.C.of 2 per cent. citric acid solution, showed fairly constant ratios between the amounts of phosphoric acid and lime dissolved, whereas the ratios between the amounts of silica and phosphoric acid dissolved (e.g., 1 : 0.53 and 1 : 1.209) were SO divergent as to indicate that silica was not in combination with the phosphate.I n the case of highly citric-soluble slag, the ratio between the dissolved phosphoric acid and silica is fairly constant in successive extraction (e.g., 1 : 1.63 to 1 : 1.76). Hence the conclusion is drawn that the citric acid test is no true criterion of the solubility of phosphates in ‘‘ fluorspar slags ” or of their value as manures. These slags are con- pletely soluble if suficient time be given.Moreover, the phosphates in these slags show a close resemblance in composition to those in mineral phosphates, such as are used in the United States, in preference to highly citric-soluble slags. C. A. M. Detection and Estimation of Halogens in Organic Compounds. I. Drogin and M. A. Rosanoff. (J. Amer. Chem. Xoc., 1916, 38, 711-716.)- Stepanoff (ANALYST, 1907, 32, 52) described a simple method for the estimation of halogens in organic compounds, depending on solution of the compound in absoluteINORGANIC ANALYSIS 149 alcohol, reduction with sodium: and subsequent estimation of the halogen acid by Volhard’s method.Bacon (ANALYST, 1909, 34, 111) failed to get satisfactory results when following Stepanoff’s directions, but described a modification which he found satisfactory.Maryott also (ANALYST, 1911, 36, 287) devised a modification of the method which he found satisfactory. On the other hand, Walker and McRae ( J . Amer. Chem. SOC., 1911, 33, 598) obtained poor results when following Bacon’s directions. I n view of the convenience of the method, as compared with that of Carius, the authors have reinvestigated it, have failed to get satisfactory results by either of the modifications above referred to, but have found conditions under which every organic substance tested by them yields the whole of its halogen as halogen acid, which can be readily estimated by Volhard’s method.Since the history of the method shows that apparently trifling details may affect the results, the authors describe their modification at some length.The reagents required are absolute alcohol, sodium, nitric acid, -& silver nitrate, Tz ammonium thiocyanate, and a solution of ferric ammonium alum. The alcohol is redistilled from sodium. The sodium must be free from chlorine, or its chlorine content known. The ferric indicator is made up by adding concentrated nitric acid to a saturated solution of iron alum until the red colour changes to straw yellow. The following quantities of alcohol and sodium are necessary and suf3ficient for the complete extraction of the halogen from apparently any organic combination. If W denotes the weight of substance taken (between 0.2 and 0.3 grm.), and if the halogen present is chlorine, then the number of C.C.of alcohol required is 156 W, and the number of grms. of sodium 21.5 W. If the halogen is bromine, then 68 W C.C. of alcohol and 9.4 W grm. of sodium are required. If the halogen is iodine, 44 W C.C. alcohol and 6.1 W grm. sodium are required. The substance is dissolved in the requisite quantity of alcohol in a 300 C.C. Kjeldahl flask fitted with a reflux condenser, using a free flame to warm the alcohol if necessary.After cooling, small pieces of sodium are introduced through the condenser-a few at a time-so that the addition occupies about thirty minutes, and during at least the second half of this time the dissolution of the sodium is aided by a small flame under the flask. The mixture is next boiled for an hour with occa- sional shaking, cooled and diluted with water, the first portions being added 1 C.C.at a time so long as vigorous action ensues. When finally diluted to about 250 c.c., the mixture is acidified by addition of dilute (1 : 3) nitric acid, transferred to a beaker, mixed with a measured and more than sufficient volume of silver nitrate solution, and filtered. The filtrate, mixed with 10 C.C. (no less) of the ferric indicator, is then titrated with thiocyanate.As a qualitative test, a few mgrms. of the substance are dissolved in 1 C.C. of alcohol, small fragments of sodium are added one at a time, and the mixture warmed. When cool the solution is acidified with 1 C.C. of dilute nitric acid, filtered if necessary, and tested with 10 drops of approximately TT silver nitrate. G. C.J. Use of Hydrofluoric Acid in Separation of some Heavy Metals from Tin, Antimony, Tungsten, and Molybdenum by Means of the Electric Current. L. W. McCay and N. H. Furman. (J. Amer. Chem. SOC., 1916, 38, 640-652.)-It has already been shown that from a nitsohydrofluoric acid solution of copper, lead,150 ABSTRACTS OF CHEMICAL PAPERS tin, and antimony, the last two metals being in the higher state of oxidation, the electric current precipitates the copper and lead only, the former at the cathode, the latter as peroxide at the anode (&Cay, J.A~ner. Chem SOC., 1914, 36, 2375.) I t is now shown that silver and mercury can be separated from tin and antimony in a similar manner, and further that neither tungsten nor molybdenum interferes. The silver deposits contain no weighable amounts of platinum, but the mercury deposits usually do.The amount of platinum present varies, but seems to increase with the quantities of tin and hydrofluoric acid, and with the time during which the current acts. When the current strength is as high as 3 amperes, and a Classen rotating anode is employed, the amount of platinum found in the mercury deposits is about 0.5 mgrm.On treating the deposits with nitric acid, the platinum remains in the form of a black powder, or dark and very thin film, and can be rapidly and accurately determined, and the necessary correction made. G. C. J. Estimation of Sulphur in Spent Oxide. E. V. Espenhahn. ( J . SOC. Clzern. Ind., 1916,35,292-293.)-The method depends on the fact that sulphur, when heated with iron, forms ferrous sulphide, which gives off its sulphur as hydrogen sulphide on treating with a strong acid.About 0.5 grm. of the oxide, finely powdered, is washed by decantation, and finally on a filter with hot water, to remove sulphates and thiocyanates. After drying, the oxide is scraped into a hard-glass crucible about 30 mm. long and 10 mm. in diameter, mixed with 4 grms.reduced iron and covered with a layer of the same. The crucible is fitted into a hole in an asbestos board, and the air displaced by carbon dioxide led in through a tube, the end of which is just above the layer of iron. A small flame is first applied, and the temperature raised until ths crucible reaches dull redness, which is maintained for ten minutes. After cooling in an atmosphere of carbon dioxide, the contents of the crucible are allowed to fall into a small Erlenmeyer flask which is connected to three wash bottles, each containing 150 C.C.of & iodine acidified with hydrochloric acid, and a fourth with a, few C.C. of The mass is covered with water and, after passing carbon dioxide for a few minutes, 50 C.C. of concentrated hydrochloric acid are run in through a funnel.As the reaction slackens, the solution is heated to boiling, and a, steady stream of carbon dioxide passed through the flask for ten minutes. The excess of iodine is titrated back with thiosulphate, and the free sulphur calculated. For the estimation of total sulphur, 0-5 grm. of the unwashed sample is reduced with iron and treated as described.For the estimation of tar, a weighed portion of the sample is extracted with carbon bisulphide in a Soxhlet extractor. The difference between the percentage of extract so obtained and the percentage of free sulphur gives the percentage of tar. G. C. J. thiosulphate. Decomposition of Tetrathionates in Alkaline Solution as a Source of Error in Certain Iodine Titrations. R. M.Chapin. (J. Anzer. Chem. Xoc., 1916, 38, 625-626.)-1f acid solutions containing tetrathionates are to be titrated with iodine, or subjected to any treatment involving assumption that the tetrathionate present has remained unaffected, they should never be neutralised by any substance ofINORGANIC ANALYSIS 151 distinctly alkaline properties. I t has long been known that tetrathionstes react with hot, concentrated caustic alkali to form thiosulphate and sulphite, but that the reaction might progress with significant rapidity in cold and dilute solutions does not appear to have been suspected.On investigation it was found that in FQ cold, caustic alkaline solution the reaction is almost complete in fifteen minutes, whilst even in ria solution it proceeds to the extent of about 40 per cent.in the same time. & alkali carbonate is about as active as & caustic alkali, whilst even & sodium carbonate or & ammonia are active enough to bring about very serious changes in a solution in fifteen minutes. If it is necessary temporarily to make alkaline a solution containing tetrathionate (e.g., a solution to which iodine and excess of thiosulphate have been successively added), sodium bicarbonate seems to be the only tolerable alkali, and even when this reagent is used it is necessary to keep the temperature low and the cencentration of alkali down to g, whilst it is a good precaution to saturate the solution with carbon dioxide.G. C. J. Water Analysis. L. W. Winkler. (Zeitsch. auyew. Chem., 1916, 29, 44-46).-Methods are described for the estimation of dissolved oxygen in waters containing much nitrite and organic matter, and for the estimation of alkalis and of zinc. DissoZved Oxygen.-The nitrites and organic matter are first destroyed by treating the water with an excess of chlorine, and the oxygen estimation is then carried out as usual; the result is a measure of the dissolved oxygen plus the excess of chlorine.The second portion of the water is now treated with chlorine, and the excess of the latter titrated; the difference between the results of the two estimations gives the quantity of dissolved oxygen. The details of the method are as follows : A bottle of 250 C.C. capacity is filled with water, 10 drops of 50 per cent. sulphuric-acid are added, and 1 C.C.of hypochlorite solution (a filtered solution containing 0.5 grm. of calcium hypochlorite and 25 grms. of crystallised sodium sulphate per 100 c.c.) is run in by means of a pipette reaching to the bottom of the bottle. The latter is closed with its stopper, the contents mixed, and, after ten minutes, 1 C.C. of manganous chloride solution and 2 C.C. sodium hydroxide- potassium iodide solution are added, and the mixture is titrated in the usual way with thiosulphate solution.The bottle is then filled a second time with the water, the same quantities as before of sulphuric acid and hypochlorite solution are added, and, after ten minutes, 3 C.C. of the contents of the bottle are withdrawn (to compensate for the 3 C.C. of manganous chloride, etc., added in the first estimation), a, crystal of potassium iodide is added, and the mixture titrated with thiosulphate solution.Experiments with distilled water showed that the same result was obtained whether the oxygen was estimated directly or after treatment with chlorine and allowing for the excess of the latter. Alkalis.-The method described depends, essentially, on the relative insolubility of sodium and potassium chlorides in isobutyl alcohol and on the solubility of mag- nesium chloride in this solvent (cf.ANALYST, 1913, 38, 531); the potassium is then separated from the mixed chlorides as potassium hydrogen tartrate and estimated by titration. From 200 to 2,000 C.C. of the water, according to the quantity of chlorides and nitrates present, are evaporated to dryness with the addition of 1 grm.of barium152 ABSTRACTS OF CHEMICAL PAPERS carbonate, the dry residue is extracted with five successive quantities of 10 C.C. of hot water, the extracts are filtered, the filtrate is breated with a few C.C. of hydrochloric acid and 1 C.C. of nitric acid, and evaporated in a small glass basin; a few C.C. of hydrochloric acid are then added to the residue and the evaporation is repeated.The residue, while still warm, is rubbed down with 1 C.C. of absolute alcohol and, after some minutes, when the greater part of the alcohol has evaporated, a further 1 C.C. of the same is added, followed by 2 C.C. of freshly-distilled isobutyl alcohol. The alcoholic liquid is decanted through a very small filter, and the residue extracted several times with isobutyl alcohol, using about 10 C.C.of this solvent altogether. The residue in the basin and on the filter is now dissolved in water, the solution treated with a drop of hydrochloric acid, evaporated to dryness, dried at 150" C., and weighed. The residue still contains traces of sulphate and of magnesia,, but this error is counterbalanced by the very small quantities of the alkali chlorides which are dissolved by the isobutyl alcohol.To estimate the potassium, the residue of mixed chlorides thus obtained is dissolved in lithium hydrogen tartrate solution (0.5 grm. of lithium carbonate and 2 grms. of tartaric acid are dissolved in 100 C.C. of water, 50 C.C. of 95 per cent. alcohol are added, the solution then shaken with 1 grm.of potassium hydrogen tartrate, and allowed to settle), 10 C.C. of this solution being used for each 0.1 grm. of mixed alkali chlorides; the mixture is stirred and placed aside for two hours, the vessel containing it being covered to prevent evaporation of the alcohol. The precipitated potassium hydrogen t'artrate is then collected on a small filter which has been moistened with 50 per cent.alcohol (by vol.) saturated previously with potassium hydrogen tartrate, and this alcoholic solution is used for washing the precipitate; the latter is now dissolved in hot water and the solution titrated with FG sodium hydroxide solution. Zinc.-Water which has been passed through galvanised iron pipes frequently contains a small quantity of zinc. The presence of this metal may be detected by the following test, which also affords some indication of the amount of the metal contained in the water : Two quantities of 100 C.C. of the water are each treated with 1 C.C. of 10 per cent. hydrochloric acid and 1 C.C. of hydrogen sulphide solution; to one portion 5 C.C. of a 15 per cent. ammonium acetate solution are then added. If the water contains 10 mgrms., or more, of zine per litre, a turbidity appears almost at once after the addition of the ammonium acetate; with 5 mgrms. per litre the turbidity is seen after the lapse of thirty seconds, whilst 2 mgrms. of the metal per litre produce a very faint turbidity after one minute. The other portion of the water, which serves for comparison, remains clear for at least five minutes even when the water contains 10 mgrms. of ferrous iron per litre. Ferric salts vitiate the test, since a turbidity, due to the separation of sulphur, forms in each portion of the water. w. P. s.