ANALYTICAL CHEMISTRI‘. 129 Analytical Chemistry. ass-generating Apparatus for Analytical Purposes. EUCIEN M~~LLER (Chem. Zeit. 1907 31 1257).-A modification of J. M. Sanders’s appargtus (ibid. 1184). As now constructed i t consists of n U-tube with two or only one ground stoppers and a gas-delivery tube. The advantages gained are saving of material obtainment of a pure (washed or dried) gas and proper regulation of the gFs- current. L. DE K. Estimation of Organically-combined Iodine and Chlorine in (‘ Erythrosine.” FERDINAND JEAN (Ann. Chim. anccl. 1908 13 12-14 *).-Two portions of 0.6 gram each are burnt to ash one with addition of 0.5 gram of sodium carbonate 0.5 gram of potassium carbonate and 5 grams of ma.gneFium oxide. The ashes are treated with water the solutions are carefully neutralised with acetic acid made up to 100 c.c.and the iodine and chlorine are estimated. The difference between the results of the two ashes is then due t o organically-combined iodine or chlorine. The estimation of the halogen is carried out as follows 50 C.C. of each solution are titrated for iodine only by Pisani’s method (#/I0 silver in presence of iodide of starch). The other 50 C.C. are then titra.ted with N,lO silver for joint iodine and chlorine using potassium chromate as indicator. L. DE K. Estimation of Total Sulphur in Urine. RUGO SCHULZ (Pfiiger’s Archiv 1907 121 114-116).-Ten C.C. of urine are heated in a round-bottomed flask with 10 C.C. of fuming nitric acid; the * And Bztll. Xoc. CII&IL Belg 1908 22 45-46.130 ABSTRACTS OF CHEMICAL PAPERS.heating is continued for about fifteen minutes that is until all violent action has ceased and the contents of the flask are then cooled. Water and hydrochIoric acid are now added the solution is trans- ferred to a beaker and the sulphuric acid is precipitated in the usual way as barium sulphate. w. P. s. Assay of Highly Concentrated Sulphuric Acid. ERNST BUCHWALD (Chein. Zeit. 1907 31 1256).-About 5 grams of the sample are weighed accurately in a small weighing bottle and then made np t o a litre. The solution is placed in a burette and run into a flask containing 50 C.C. of accurately prepared N/lO sodium hydr- oxide free from carbon dioxide and coloured with phenolphthalein until the liquid is decolorised. Good results are also obtained by K jeldahl’s iodometric process (liberation of iodine from a mixture of potassium iodide and iodate).L. DE K. Titration of Sodium Thiosulphate. THADDEUS MILOBENDSKI (J. Russ. Phys. Chem. Soc. 1907’30 Chem. 1404-1411).-The three methods of titrating sodium thiosulphate namely those of Zulkowsky and of Volhard and direct titration with iodine have been found to be equally good ; but in cases where the iodine is only very slowly liberated the thiosulphate should be titrated by Zulkowsky’s method without however allowing as usual half an hour for the completion of the reaction. Z. K. Modification of Kj eldahl’s Process for Estimating Nitrogen in Foods. REMO CORRADI (Boll. d i m . farm. 1907 46 861-864). -2.5 Grams or if the amount of the nitrogen is less than 3% 5 grams of the food are weighed dried in a steam-oven and heated for five t o six hours in a Kjeldahl flask with 30 C.C.of fuming sulphuric acid. The colourless or pale yellow liquid is intro- duced into a 250 C.C. flask containing about 40 C.C. of water the cooled solution being then rendered faintly acid by means of 20% sodium hydroxide solution and finally made up t o volume. Twenty- five C.C. of the liquid are then treated with 25 C.C. of sodium hypo- bromite solution prepared by mixing 25 C.C. of bromine 350 C.C. of 30% sodium hydroxide solution and 275 C.C. of water. The nitrogen evolved is measured in a slightly modified DuprB’s azotometer (compare Abstr. 1906 ii 505) the volume being increased by 1.7%. This method gives numbers differing insensibly from those obtained by the ordinary Kjeldahl proceEs.Employment of Nitron- for Determining Nitrates in Soils and Plants. JAKOB LITZENDORFF (Zeitsch. angew. Chem. 1907 11 2209-223 I).-Correct results are obtained with soils containing 20-30 parts per million of nitric nitrogen. When however smaller amounts are present the extracts have to be evaporated down and in this process substances are produced which hinder the separation of the nitrate. This difficulty can be overcowe by repeatedly adding hydrogen peroxide to the extract during evaporation and by heating T. H. P.ANALYTICAL CHEMISTRY. 131 the sufficiently concentrated extract with hydrogen peroxide in a flask in boiling water for several hours until decolorised. N. H. J. M. Estimation of Phosphorus in Calcium Carbides F.WILLY HINRICHSEN (Chem. Zentr. 1907 ii 1356-1357; from Mitt. K. MateriaZpriiyg.-Amt. Gyoss. Lichterfeelde Vest 25,110-1 12)-On pass- ing impure acetylene through a solution of sodium hypochlorite ex- plosions occur now and then and the phosphorus is not completely absorbed as it occurs in part as an organic phosphorus compound. The process recommended by Lidholm (Ahstr. 1904 ii 776) gives satisfactory results. L. DE K. A New Modiflcation of Petermann’s Method for Estimating Citrate-Soluble Phosphoric Acid in Precipitated Calcium Phosphate (Futterkalk). GUSTAV FINaERLTNG and ADOLF GROM- BACH (Zeitsch. anal. Chm. 1907 46 756-761).-0ne gram of the sample is put into a 2OOc.c. flask and moistened with 5 C.C. of alcohol 100 C.C. of Petermann’s alkaline citrate solution are added and the whole is rotated for half an hour.The liquid is then heated a t 40° for another hour with constant shaking and when cold diluted with water to 200 C.C. and filtered. One hundred C.C. of the filtrate are mixed with 20 C.C. of strong nitric acid evaporated t o half the bulk and when cold neutralised with ammonia and mixed with 50 C.C. of Hallenser’s solution. To the cold solution are then added drop by drop 20 C.C. of magnesium mixture and after shaking for half an hour the precipitate may be collected. L. DE K. Detection of Arsenic by Means of the Marsh Apparatus. HEINRICH STRUVE (Zeitsch. and. Chem. 1907 46 761-764).-The author recommends the use of sheet zinc which almost invariably he finds to be free from arsenic. Addition of activating agents are Estimation of Commercial Silicon ; Separation of Silica and Silicon.FRITZ LIMMER (Chem. Zeit. 1908 32,42).-0-25-0.50 Gram of the finely-powdered sample is heated in a slow current of dry chlorine care being taken not to apply heat until every trace of air has been expelled from the apparatus. The silicon (also aluminium iron &c.) is volatilised whilst the silica remains in the residue which is then tested in the usual way by fusion with alkali carbonate &c. The silica is deducted from the amount of total silica obtained from a separate portion of the sample and the difference calculated into superfluous. L. DE K. silicon. L. DE K. Estimation of Carbon in Pig-Iron and Steel MAX ORTHEY (Chem. Zeit. 1908 33 31-33).-A criticism of some of the methods used for the estimation of carbon in iron or steel.Sarnstrom’s modified chromic acid process and the copper ammonium chloride method are quite satisfactory for the ordinary kinds of iron. The method of burning in a current of oxygen in presence of bismuth oxide gives good results with all kinds of iron. I n the case of samplcs132 ABSTRACTS OF CHEMICAT PAPERS. oxidised with great difficulty the chlorine process may be used as a check. Some other recognised methods were tried but with less satisfactory results L. DE K. The Use of Silver in the Combustion of Nitrogenous Substances. F. EPSTEIN and RICHARD DOHT (Zeitsch. anal. Chew. 1907 46 771-773).-A silver spiral 10 cm. long heated to intense redness is again recommended instead of metallic copper in order to decompose the nitrogen oxides formed i n the combustion of nitrogenous organic substances (C and H estimation).Silver also effectually retains halogens. L. DE K. Dennstedt's Method of Elementary Analysis Applied t o the Hzemoglobin Derivatives. JEAN ZALESKI (Bull. Acad. Sci. Cracow 1907 ii 646-651).-The elementary analysis of substances such as acetonehaemin or iodohaemin (Merunowicz and Zateski ibid. 633) is attended with great difficulty when ordinary methods are employed but is rendered simple by the use of Dennstedt's method (compare Abstr. 1906 ii 51). G. B. Estimation of Carbon Dioxide in Electrolytic Chlorine. PETER PHILOSOPHOFF (Chent. Zeit. 1907 31 1256-1257).-The author (Abstr. 1907 ii 908) now dispenses with the use of brine in order to facilitate the readings and reads off the volume of the gas before and after absorption by means of a graduated levelling tube attached to the Bunte burette and filled with mercury.L. DE K. Quantitative Estimation" of Metals in Organic Substances. J. ROTRE (Chem. Zentr. 1907 ii 1362-1364; from Mitt. K. &fate&Iprufs.-Amt. Gross. Lichterfdde West 25 105-1 06) .-The dried substance is heated in a round-bottomed flask with 15 C.C. of fuming nitric acid and 2 C.C. of sulphuric acid for every gram taken at first gently afterwards,more strongly on a sand-bath. Substances which are readily inflammable are first moistened with four times the weight of nitric acid D 1.4. When sulphuric fumes begin to appear the contents are allowed to cool and another portion of nitric acid is added and t h i s operation may be repeated if necessary. Finally the bulk of the sulphuric acid is driven off and any metal will be found in the residual mass.L. DE K. Analysis of Sodium Peroxide. R. NIEMEYER (Chem. Zeit. 1907 31 1257)-With care the permanganate method gives trustworthy results. 0.15-0.2 Gram of the peroxide is weighed out in a little dish placed in a weighing tube. This is then seized with a pair of tongs and quickly immersed in a beaker containing 500 C.C. of water; care should be taken that none of the peroxide floats on the surface as otherwise the experiment fails. A sufficiency of dilute sulphuric acid is now added and the liquid titrated at once with iV/ 10 permanganate. L. D E K ,ANALYTICAL CHEMISTRY. 133 Quantitative Separation of Barium from Strontium.ZELDA KAHAN (Analyst 1908 33 12-14).-The method proposed depends on the precipitation of the barium as chromate in ammonium acetate solution. The solution containing the barium and strontium salts is treated with ammonium dichromate solution until precipitation is ne;trly complete ; the dichromate solution is added drop by drop with constant stirring. Ammonium acetate solution is then added until the solution becomes colourless a few more drops of the dichromate solution are added so that the solution remains pale yellow and next a little more ammonium acetate solution. After the lapse of about three hours the precipitate of barium chromate is collected in a Gooch crucible washed with ammonium acetate solution until the filtrate gives only a slight coloration with silver nitrate solution and is then w.P. x. dried at a temperature of 180’ and weighed. Estimation of Mercuric Chloride in Pastilles. UMBERTO SAPORETTI (Boll. chinz. farm. 1907 46 S65-867).-The amount of mercuric chloride contained in pastilles used for antiseptic purposes may be readily estimated as follows. Five of the pastilles are dissolved in water in a porcelain basin or crystallising dish with tall sides and t o the solution decolorised with chlorine water or dilute hydrochloric acid sodium hypophosphite is added gradually as long as any action takes place. The finely-divided mercury is then caused t o collect into a globule by gently boiling the solution and is washed with dilute hydrochloric acid. The amount of mercury is estimated by weighing or measuring.For the latter estimation the author has devised a narrow burette or (‘ hydrargyrometer,” by means of which the weight of mercuric chloride per five pastilles can be read off directly. T. H. P. Estimation of Manganese in Potable Water. EDMUND ERNYEI (C‘iiem Zeit. 1908 32 41-42).-The sample is acidified with sulphuric acid and any iron removed by shaking with a slight excess of zinc oxide. One hundred C.C. of the filtrate are mixed with 5 C.C. of 3076 sulphuric acid the chlorine is removed with slight excess of silver sulphate and the whole is boiled for twenty minutes after adding 1-2 grams of potassium persulphate. When cold the liquid is made up t o the original volume; the permanganate formed is evtimated by adding potassium iodide and titrating with N/100 thiosulphate with starch as indicator.The process may be used also for the estimation Detection of Nickel as Double Nickel Ammonium Molybdate. 31. EMMANUEL POZZI-ESCOT (Ann. Chim. and. 1908 13 16. Compare Abstr. 1907 ii 818).-A modification of the author’s previous test for nickel. The neutral or slightly acid solution is mixed with a small quantity of saturated solution of ammonium molybdate and then with a large excess of saturated solution of ammonium chloride. The mixture is heated gently when should nickel be present it becomes turbid-in a few minutes. of manganese in ferruginous water deposits. L. DE K. L. DL K.134 ABSTRACTS OF CITEMICAL PAPERS. Volumetric Estimation of Tin by Means of Potassium Dichromate.H. REYNOLDS (Chem. News 1908 97 13-1 5). - The process is based on the fact that sulphonated azobenzene is bleached by stannous chloride,rand that the red colour is restored by potassium dichromate. The metal is dissolved in hydrochloric acid inacurrent of carbon dioxide free from oxygen a definite amount of potassium dichromate (1 C.C. -0.01 gram of tin) very nearly sufficient to complete the oxidation is added and then sufficient of the indicator to fully mask the green colour of the reduced chromate. More dichromate is now added until the colour changes t o red. The process may be used for the assay of ferro-tin and Britannia metal L. DE K. New Method for Separating Titanium and Zirconium. MAX DITTRICH and 5. FREUND (Zeit.sch. anorg. Cl~m. 1907 56 344-345).-The method depends on the fact that a neutral solution of titanium nitrate gives in the cold with ammonium salicylate a precipitate (titanium salicylate) easily soluble in hot water whilst the corresponding precipitate with zirconium nitrate is insoluble in exces of the reagent.A solution of the mixed nitrates neutralised with sodium carbonate is added drop by drop to a boiling concentrated solution of ammonium salicylate (1 salt to 5 water). The solution is then boiled for some time concentrated filtered the precipitate ,washed with a boiling solution of ammonium salicylate until free from titanium and the zirconium then estimated as dioxide. The titanium is precipitated from the boiling filtrate with excess of ammonia and also estimated as the dioxide. G. 5.Separation of Titanium and Thorium by Means of Ammonium Salicylate. MAX DITTRICH and S. FREUND (Zeitsch. unorg. Chem. 1907 56 346-347).-The separation is effected exactly as described in the preceding abstract; thorium like zirconium being completely precipitated from a neutralised solution of the nitrate when the latter is slowly added t o a boiling concentrated solution of ammonium salicylate. G. 8. Simultaneous Precipitation of Titanium and Zirconium in the Presence of Iron. MAX DITTRICH and 8. FREUND (Zeitsch. anorg. Chen?. 1907 56 337-343. Compare Abstr. 1905 ii 287).- Four methods for the simultaneous precipitation of titanium and zirconium salts from a mixture also containing iron are given. A complete separation in t'he presence of sodium acetate is possible when the iron is in the ferrous st,ate.The mass obtained by fusing a mix- ture of the three salts with acid potassium sulphate is dissolved in water saturated with hydrogen sulphide filtered nearly neutralised with sodium carbonate and the hydrogen sulphide removed by passing a stream of carbon dioxide through the hot solution. Then whilst the solution is protected from oxidation by means of carbon dioxide excessANALYTICAL CHEMISTRY. 135 of sodium acetate is added the solution heated for one hour and the mixed precipitate of titanium and zirconium removed by filtration. According to another method which has some advantages ammonium sulphate is used inetead of acetate otherwise the procedure is the same as above. The titanium and zirconium are obtained as oxides.The separation in question may also be effected by means of sulphurous acid or sodium thiosulphat'e; full details are given in the paper. The titanium in the mixed precipitate is estimated colorimetrically with sulphuric acid and hydrogen peroxide and the zirconium by difference. G. S. Separation of Thorium Titanium and Zirconium from Iron. MAX DITTRICH and S. FREUND (Zeihch. anorg. Chem. 1907 56 348-352. Compare preceding abstracts).-It is shown that the com- plete separation from iron of (a) titanium and thorium ( b ) thorium and zirconium (c) titanium zirconium and thorium can be effected by the sodium acetate method described in the preceding abstract. I n order to separate the individual oxides from the mixture of the three oxides obtained in ( c ) they are converted into nitrates the zirconium and thorium are precipitated with ammonium salicylate as described above and are then separated by the ammonium oxalate method due to Jannasch.G. S. The Action of Methyl Sulphate on Oils of the Aromatic and Aliphatic Series. THOMAS WEATHERILL HARRISON and F. MOLLWO PERKIN (AnaZyst 1908 33 2-9).-The authors find t h a t Valenta's method (Abstr. 1906 ii 310) is untrustworthy for the estiinationiof tar oils in mixtures of the same with mineral oils. Whilst t a r oils are soluble in all proportions in methyl sulphate mineral oils are not insoluble. The method is however of use qualitatively; methyl sulphate removes most of the tar oil together with only a small quantity of mineral oil when shaken with mixtures of these oils.If the methyl sulphate layer is then saponified with potassium hydroxide and the solution diluted and extracted with ether the tar oil mixed with only a small portion of mineral oil is recovered. It can then be tested to prove that it is a tar or aromatic oil. The fact that methyl sulphate dissolves a portion of an oil is no proof that the dissolved portion is an aromatic hydrocarbon. w. P. s. Estimation of Naphthalene in Coal Gas and in Spent Oxide of Iron. C. J. DICKENSON GAIR (J. Xoc. Chem. Ind. 1907 26 1263-1264).-The method described previously (Abstr. 1906 ii 201) has been modified as follows so as to be available either gravi- metrically or volumetrically. The naphthalene picrate obtained is collected on a filter and washed with about 400 C.C. of dilute picric acid solution to remove the acetic acid and is then washed into a beaker with a little cold water; 180 C.C.of water are added together with 25 C.C. of N/10 sodium hydroxide solution and the mixture is boiled for ten minutes with the result that the naphth-136 ABSTRACTS OF CHEMICAL PAPERS. alene is liberated. If on testing with lacmoid indicator the solution does not give a green coloration a further quantity of sodium hydroxide is added and the boiling continued. N/20 picric acid solution is now added in excess as is shown by the indicator becoming yellowish-brown and the solution is titrated with N/10 sodium hydroxide solution until the green coloration is per- manent. The quantity of sodium hydroxide solution required t o decompose the naphthalene picrate is thus found ; 1 C.C.of A710 sodium hydroxide solution corresponds with 0.01 28 gram or 0.197 grain of naphthalene. For the estimation of naphthalene in spent oxide of iron 10 grams of the undried oxide are placed in a closed flask and extracted with 100 C.C. of 80% alcohol. At the end of three hours the solution is poured through a filter the residue is washed with dilute alcohol and to the filtrate are added 300 C.C. of concentrated picric acid solution. After a short time the naphthalene picrate is collected on a filter and the naphthalene is then estimated volumetrically as described above except that there is no necessity to wash the picrate before decompos- Estimation of Alcohol in Wine. MARCEL DUBOUX and PAUL PUTOIT (Ann. Chim. a n d 1908 13,4-9).-Five volumes of redistilled aniline are mixed with 3 volumes of alcohol (95' French). A mixture of 1 volume of nitrobenzene and 9 volumes of alcohol may be used also. Fifteen C.C.of the mixture are placed in a test-tube 3.5 cm. in width and 15 cm. long fitted with a doubly-perforated cork through which pass a delicate thermometer and a glass stirrer bent upwards at the lower end. Ten C.C. of the alcoholic solution (wine- distillate) are added and t h e whole is heated with constant stirring until a clear solution is obtained The temperature at which this occurs is noted and referencemade to a table constructed from similar experiments made with alcoholic mixtures of known composition. ing i t with sodium hydroxide solution. w. P. s. L. DE K. Formation of Methylacetol (Acetylmethylcarbinol) in the Acid Fermentation of Wines. J. PASTUEREAU (J. Pharrn. Chcm. 1908 [vil 27 10- 12).-The author states that acetylmethylcarbinol is a normal constituent of wine-vinegars and that it is formed during the acetic fermentation of the wine. Its presence may be ascertained and its amount determined by subjecting the neutrdised vinegar t o distillation under reduced pressure or by passing a current of steam. The ketone is then estimated in the distillate either by Fehling's solution or by means of ammoniacal silver nitrate containing sodium hydroxide. 2.85 Grams of copper reduced = 1 gram of methylacetol ; 4 atoms of silver = 1 mol. of methylacefol. Methylacetol is also characterised by yielding with phenylhydrazine acetate an osazone melting a t 2 4 3 O of which the corresponding osotetrazone melts gt 151'. With semicarbazide a semicarbazone is obtained melting at 200'. L. DE E.