VEGETABLE PHYSIOLOGY AND AGRICULTURE. 47 Analytical Chemistry. New Burette Holder. AZARIAH T. LINCOLN (J. Arner. Chrn. Xoc. 1905 27 1442-1443).-A new form of burette holder is described which has the advantages that the burettes are always vertical and parallel can be readily put into or removed from the clamp and can be easily moved up or down. For details the description and diagram in the original must be consulted. E. G.48 ABSTRACTS OF CHEMICAL PAPERS. Ignition in a Vacuum by means of the Electric Furnace. ERNST HAAGN (Chem. Zeit. 1905 29 1209).-The author recommends a special form or" electrical furnace for crucibles constructed by Heraeus for use with a vacuum pump. With this apparatus powdered marble may be ignited to constant weight under a pressure of 40 mm.in three minutes. P. H. Modification of Winkler's Process for the Estimation of [Dissolved] Oxygen in Water. HERMANN NOLL (Zeit. angew. Chem. 1905 18 1767-1768).-1n order to allow for the presence of organic matters in waters when using his manganese method for the estimation of dissolved oxygen Winkler takes equal portions of the sample and of distilled water adds to each a certain amount of man- ganic chloride solution and then titrates both with tbiosulphate after addition of potassium iodide. The difference in the iodine is calcu- lated into oxygen and added to the amount found. Two C.C. of a 50 per cent. manganous chloride 2 C.C. of a 40 per cent. sodium hydroxide and 20 C.C. of water are shaken in a flask until the mass has become brown; 50 C.C. of hydrochloric acid are then added and the whole diluted to 300 C.C.To 100 C.C. of distilled water and also to 100 C.C. of the sample are now added 10 C.C. of a 5 per cent. solution of potassium iodide and then 25 C.C. of the above manganic chloride solution and after five minutes the iodine in both solutions is titrated. A number of experiments is given showing however that the correc- tion may as a rule be neglected as the action of iodine on organic The author has slightly modified this procedure. matters usually occurring in waters is but trifling. L. DE K. Volumetric Estimation of Sulphates with Beneidine Hydro- chloride in Presence of Thiosulphates Sulphites and Sulphides. OTTO HUBER (Chem. Zeit. 1905 29 1227-1229).-It is found that Muller and Diirke's method (Abstr.1903 ii 751) of titrating sulphates by means of benzidine hydrochloride is not directly applicable in the presence of thiosulphates sulphites and sulphides ; the following modified procedure is however recommended. By heating the mix- ture with a few drops of bromine the thiosulphate sulphite and sulphide are oxidised to sulphate; after boiling off the excess of bromine and making neutral with sodium hydroxide the hot solution is treated with benzidine hydrochloride ; this gives the total sulphate. A second portion of the solution is run into N/10 iodine acidified with hydrochloric acid the excess of iodine being titrated back by means of N/lO thiosulphate. A third portion of the solution after boiling with cadmium carbonate to precipitate the sodium sulphide is cooled made up t o 250 c.c.filtered and divided into two portions A and B of 100 C.C. each. A is again run into iodine solution and titrated back with thiosulphate which gives the amount of thiosulphate and sulphite. B after boiling for twenty-five minutes with 10 C.C. of 2N-acetic acid to destroy the sulphite and 30 C.C. of 2N-sodium acetate is treated with 1 gram of potassium chlorate and allowed to cool; titration with iodine then gives the amount of thiosulphate present ; the addition ofANALYTICAL CHEMISTRY. 49 sodium acetate prevents the t hiosulphate from being destroyed during these operations. P. H. Determination of the Strength of the Solutions employed in Nitrogen Determinations. FELIX MACH (Lundw. Veel.suchs-Stat. 1905 63 71-80).-The sodium hydroxide solution employed is rather less than N/3 and is prepared by diluting 225 C.C.of a solution of 1 part of sodium hydroxide in 2 parts of water to 10 litres and adding a little solid barium hydroxide. The sulphuric acid is prepared by diluting 200 grams of strong acid to 10 litres. The indicator is 2 C.C. of a solution of congo-red (2 grams) in 25 per cent. alcohol (1 litre). The exact strength of the solution is determined by distilling a weighed amount of amuionium sulphate (about 0.4 gram) with mag- nesium oxide and absorbing the ammonia in 20 C.C. of the acid. Hydrogen sodium carbonate (0-6-0.7 gram) is gently heated until the weight remains constant washed into 20 C.C. of the acid heated to expel the carbon dioxide and the excess of acid titrated with congo- red.Potassium tetraoxalate (0.5 gram) is dissolved in about 100 C.C. of water and directly titrated with phenolphthalein as indicator. The nitrogen value is then calculated in the usual manner. The ammonium sulphate is prepared by adding a concentrated solu- tion of the salt (previously recrystallised several times) to absolute alcohol (2 parts) ; it is then filtered washed successively with alcohol and cooled distilled water partially freed from water with filter paper and stored in a moist condition. The calcium hydrogen carbonate is prepared by passing carbon dioxide through a cold saturated solution of the carbonate. The washed precipitate is kept in a moist condition. N. H. J. M. Estimation of Mixtures of S u l p h u r i c and Nitric Acids. GEORG LUNGE and E.EERL (Zeit. angew. Chern. 1905 18 1681-1687)- The investigation of the commercial mixed acids involves the deter- mination of nitric nitrous and sulphuric acids. It is shown that the method previously recommended for the direct estimation of the sulphuric acid according to which the acid mixture is diluted with water evaporated on the water-bath and the residual acid titrated yields results which are too low in consequence of the loss of sulphuric acid in the process. The value for the nitric acid estimated by difference is consequently too high. The method of examination now recommended consists in determining (a) the total acidity ( b ) the nitrous acid from the permangsnate titre ( c ) the nitric and nitrous acids from the total nitrogen evolved in the nitrometer.The difference (c) - ( b ) represents the nitric acid ( a ) - ( c ) the sulphuric acid. An alternative method is to determine the sulphuric acid gravimetrically the nitrous acid by means of permanganate and the nitrogen by the " nitron " method of Busch (Abstr. 1905 ii 282). The nitric acid is then given by the difference in the values of the third and second determinations. H. IU D. Estimation of Nitric and Nitrous Acids. VOL. XC. ii. 4 JACOB MEISENHEIMER and FRIEDRICH HEIM (Bey. 1905,38 3834-3837).-The nitrite is de-60 ABSTRACTS OF CHEMICAL PAPERS. composed by hydrogen iodide and the nitrate subsequently by ferric chloride and hydrochloricacid and the nitric oxide gas evolved in each stage of the operation measured. The operation is carried out in an atmosphere of carbon dioxide in a suitable flask and does not take longer than an hour and a half.E. F. A. Estimation of Nitrous Acid. FRITZ RASCHIG (Bey. 1905 38 391 1-3914 Compare Meisenheimer and Heim preceding abstract).- I n solutions with which the sulphanilic acid and permanganate methods fail as for example those containing hydroxylamine the author estimates nitrous acid by addition of potassium iodide and sulphuric acid and after the lapse of at least two minutes titration of the liberated iodine with sodium thiosulphate in an atmosphere of carbon dioxide. Jf the thiosulphate titration is carried out too soon results even 10 per cent. too high are obtained probably owing to the inter- mediate formation of nitrosyl iodide NOI which decomposes slowly into nitric oxide and iodine.The estimation of nitrous acid by means of permanganate is carried out best by adding an excess of this reagent acidifying with sulphuric acid and after two minutes determining the unused permanganate with potassium iodide and thiosulphate G . Y. Estimation of Uitrate-soluble and Total Phosphoric Aoid in Basic Slag. FELIX MACH (Landw. Versuchs-Stat 1905,63,81-91).- With regard to the estimation of citrate-soluble phosphoric acid it is shown that whilst Wagner’s method generally yields satisfactory results this is not invariably the case. It is preferable to separate the silica from the extract. In determining total phosphoric acid the higher results obtained by Shenke’s modification is not due to the different relation between the citric acid and ammonia but probably to impurities present in the phosphoric acid precipitate which separates much more quickly than in the ordinary method.N. H. J. M. Estimation of Arsenious Oxide. CHARLES E. CASPARI and LEO R A. SUPPAN (Pharm Rev. 1905 23 334-338). -Three different methods are given for bringing arsenious oxide into solution previously to titration all giving equally trustworthy results. First the arsenious oxide is dissolved in warm dilute hydrochloric acid; to the solution a slight excess of sodium hydroxide is added the excess i u neutralised with N-sulphuric acid and the solution titrated with standard iodine solution after adding an excess of saturated sodium hydrogen carbonate solution. Second the arsenious oxide is dissolved in dilute sodium hydroxide solution the solution is cooled neutralised with N-sulphuric acid an excess of saturated sodium hydrogen carbonate solution is added and the titration proceeded with as usual.Third the arsenious oxide is treated with saturated sodium hydrogen carbonate solution at a boiling temperature the solution is cooled and any carbonate which may have formed is converted into hydrogen carbonate by the addition of a little N-sulphuric acid an excess of saturatedANALYTICAL CHEMISTRY. 51 sodium hydrogen carbonate solution is added and the whole titrated with iodine solution. w. P. s. Simplified Method of Elementary Analysis for Scientiflc Purposes. MAXIMILIANO DENNSTEDT (Ber. 1905 38 3729-3733. Compare Abstr. 1903 ii 103; 1905 ii 202 651).-Further details of the author’s method are given.The substance to be analgsed is placed in a boat which is weighed in a glass tube of such dimensions that it can be readily pushed into the combustion tube. This inner tube may be closed at the one end if the substance is readily volatile and also when the substance is difficult to volatilise provided carbon is not deposited during the volatilisation. An open inner tube may be used for substances such as sugar and proteids which deposit much carbon on heating. Another arrangement which gives very good results permits of oxygen being passed through the outer and inner tubes at the same time (see figure). i 0 The combustion is proceeding regularly when the contact substance glows brightly or when a small flame appears at the end of the inner tube which is pushed close up to the contact material This flame should remain in the inner tube that is it should not burn outside the rim of the inner tube.When this happens the burner under the capillary tube is to be removed and the oxygen supply increased. This double supply method may be used for most substances and permits of the regulation of the volatilisation with great ease. J. J S. Analysis of Silicates. 11. EDUARD JORDIS and WILHELM LUDEWICI (Zeit. anorg. Chm. 1905 47 180-189. Compare Abstr. 1905 ii 317 610).-The authors find that considerable loss by spirting takes place when ignited silica is treated with hydrochloric acid and water probably owing to the heat always developed when a fine powder is moistened with a liquid; this source of error is avoided by allowing the silica to cool and then moistening it with a few drops of cold water before the bulk of acid and water is added.The amount of silicic acid which passes into solution during the washing of the ignited silica on the filter depends largely on the base with which the silica has been combined; it is very small with barium and strontium silicates much greater with calcium silicate and is greatly increased by the presence of alkali metals. The colloidal silicic acid remains in solution in the filtrate even in the presence of a large excess of an alkali chloride. Kehrmann and Fliirscheim’s suggestion (Abstr. 1905 ii 411) as to the volatility of silicic acid is inaccurate ; the discrepancies observed are due to the foregoing sources of error and it is shown that when proper precautions are taken there is no appreciable loss.G. s. 4-252 ABSTRACTS OF CHEMICAL PAPERS. Simplified Method f o r Determining Potassium Sodium Calcium Magnesium and Phosphoric Acid in Hydrochloric Acid Soil Extracts. HUGO NEUBAUER (Landw. Versuchs-Stat. 1905 63 141-149).-A portion of the extract corresponding with 25 grams of soil is evaporated to dryness after adding calcium carbonate (0.5 gram) in the case of soils which do not effervesce with acid heated over a mushroom burner rubbed with a glass pestle and again heated until the organic matter is destroyed. It is then boiled with about 60 C.C. of water for half an hour diluted to 125 c.c. and filtered. The filtrate should be quite colourless and slightly alkaline.Potassium is determined in 100 c.c. as previously described (Abstr. 1900 ii 481) and also the sodium (Abstr. 1904 ii 209). The phosphoric acid is determined in the insoluble residue which is boiled for half an hour with dilute sulphuric acid (containing 5 C.C. of strong acid). The solution is diluted to 100 C.C. and precipitated with molybdate solution (100 c.c.) prepared without ammonium nitrate (Abstr. 1903 ii 102). Or the phosphoric acid may be determined preferably by von Lorenz’ method (Abstr. 1901 ii 278) in 25 C.C. of the solution (= 5 grams of soil) after adding 25 C.C. of nitric acid of sp. gr. 1.2. Calcium and magnesium are estimated in another portion of the original solution (corresponding with 25 grams of soil). After evaporating to dryness and heating as before the residue is heated with water and 2-5 grams of ammonium chloride (according to the amount of calcium) until no more ammonia is given off washed into a 125 C.C.flask and boiled with a few drops of ammonia. The cold solution is diluted to 125 c.c. filtered and the calcium and magnesium estimated in 100 C.C. by the usual methods. The method is suitable for all soils except those which contain very large amounts of sulphates. N. H. J. M. The Use of the Rotating Anode f o r the Estimation of Cadmium taken as the Chloride. CHARLES P. FLORA (Amer. J. Sci. 1905 [iv] 20 392-396).-When substituting cadmium chloride for the sulphate (Abstr. 1905 ii S59) some important differences were noted. Good results were obtained in solutions containing sulphuric acid cyanides and pyrophosphates but with phosphates such care is required that the phosphate method cannot be recommended when other methods are available.With great care fair results may be got in tho presence of urea formaldehyde and acetaldehyde. I n presence of acetates the method proved unworkable with cadmium chloride although the results had been most satisfactory in the case of the sulphate. L. DE K. Microchemical Reaction for Copper in Presence of Lead and Bismuth. PIETER A. MEERBURCI and H. FILIPPO (Chem. Centr. 1905 ii 1466; from Chem. Weekblad. 2 641-643).-The presence of 0.0001 mg. of copper can be detected by adding cesium chloride to th0 solution of the copper salt in hydrochloric acid. When a small quantity of cEsium chloride is added to a solution of a copper salt red needle-shaped or six-sided prismatic crystals (CuCl,,CsCl 4) are formed together with other colourless compounds ; the addition of a furtherANALYTl CAL CHEMISTRY.53 quantity of czesium chloride causes the separation of yellow crystals (CuCl,,ZCsCl ?) from which the red crystals may be regenerated by again increasing the quantity of cupric chloride. The long prismatic crystals cause a marked polarisation of light but the shorter ones have as a rule no such action. The presence of hydrochloric acid appears to accelerate the reaction for copper but cobalt and iron may interfere even in the presence of ammonium chloride. The iron may be removed as ferric oxide by repeatedly evaporating the solution but the cobalt cannot be separated.When a small quantity of cssium chloride is added to a solution containing cobalt reddish-brown crystals are formed and yellow or yellowish-green crystals separate on the addition of a larger quantity of cwsium chloride. Lead and bismuth have no effect on the reaction and in mixtures of these metals with copper bismuth may be detected simultaneously by Behren’s reaction and the lead finally obtained as cmium lead copper nitrite. When a large quantity of lead is present colourless highly refractive apparently rhombohedra1 crystals are sometimes formed. E. W. W. Gasometric Estimation of Copper with Hydrazine Salts. ERICH EBLER (Zeit. unorg. Chem. 1905 47 371-376).-The author finds that the oxidation of hydrazine salts by solutions of cupram- moni.um salts proceeds strictly in accordance with the equation 4(CuS0,,4NH3) + N2H,,H,S0 = 2(Cu2S0,,4NH3) + N + 2NH + 2(NH,),S04 under all conditions and on this fact has been based a method for the estimation of copper and of hydrazine compounds by measurement of the evolved nitrogen.The reacting substances are heated together without access of air in an apparatus described by Hempel and the nitrogen measured over mercury. Test analyses gave satisfactory results. Metallic salts which themselves reduce hydrazine exert a disturbing action and must be removed before the experiment ; this can be effected satisfactorily by precipitation as metal with a hydroxylamine or hydrazine salt the copper remaining in solution as the cuprammonium compound. G. S. Gasometric and Volumetric Estimation of Mercury by Hydrazine Salts and Gasometric Estimation of Hydrazine by Mercury Salts.ERICH EBLER (Zeit. nnorg. Chenz. 1905 4’7 377-385). -When mercuric salts dissolved in acetic acid containing sodium acetate are gently heated with hydrazine salts they are reduced quantitatively according t o the equation BHgCI + N2H = 4HC1+2Hg+N2. This reaction may be made use of for the estima- tion both of hydrazine and of mercuric salts by measurement of the evolved nitrogen. For this purpose the reacting substances are heated in an atmosphere of carbon dioxide the nitrogen is driven over into a Schiff’s nitrometer containing potassium hydroxide and subsequently measured over mercury. Mercuric salts in ftmmoniacal solution are also reduced completely by hydrazine salts and the reaction may be employed for the quantitative estimation of mercuric salts either by measuring the54 ABSTRACTS OF CHEMICAL PAPERS.volume of nitrogen evolved as previously described for copper (see preceding abstract) or by titrating the excess of hydrazine salt. According to the latter method ammonia is added to a weighed quantity of the mercuric salt dissolved in hydrochloric acid until t h e solution becomes clear a known excess of 1/40 normal hydrazine sulphate is added the mixture is then heated on the water-bath until the precipitate has settled allowed to cool made up to a definite volume filtered and an aliquot part titrated with iodine by the method given by Stolle (Abstr. 1903 ii 100). These methods for the estimation of mercury are also applicable t o mixtures except when substances such as silver and copper wbich are themselves reduced by hydrazine are present ; n method of removing these metals by means of hydrazine is described.G. S. Influence of the Presence of Titanium on the Estimation of Aluminium in Presence of Iron and Phosphoric Acid. HENRI PELLET and CH. FRIBOURG (Ann. Cl~inz. n n d . 1905 10 41 6-420).-Experiments still in progress showing that the processes now in use for estimating aluminium in presence of iron or phosphoric acid all give an alumina contaminated with titanium oxide should this be present in the ashes of plants &c. operated upon. The weighed alumina compound should therefore be fused with potassium hydrogen sulphate and the resulting solution tested colori- metrically with hydrogen peroxide.It is still doubtful whether the titanium occurs in the precipitate as oxide or phosphate L DE I(. Estimation of Metallic Iron in Reduced Iron. H. CORMIMB~UFF and 1;. GROSMAN (Amn. Chim. ccnal. 1905 10 430-422).-0ne gram of the sample is treated for six hours with 25 C.C. of 2iV iodine with occasional shaking. Two hundred and fifty C.C. of water are then added and the uncombined iodine titrated with 2Nsodium thiosulphate. One C.C. of iodine solution = 0.056 gram of metallic iron. L. DE K. Separation of Iron from Zinc by means of Ammonia. W. FUNK (Zeit. angew. Chem. 1905 18 1687-16S9).-The ferric iron solution is mixed with excess of ammonium chloride and then neutralised with ammonia. Sufficient ammonia is now added to fully precipitate the iron and then an excess of about thirty times the amount already used.The whole is heated to 70-SOo and the precipitate is collected and washed with a hot 5 per cent. solution of ammonium chloride. To remove the zinc completely the precipitate must be redissolved in hot dilute hydrochloric acid and reprecipitated in the same manner. The iron precipitate may then be washed ignited and weighed. Allowance must be made for any silica present. As this renders the estimation troublesome the author prefers the acetate method for the estimation The zinc may be titrated in the filtrate by the usual process. of iron in presence of zinc. L. DE K.ANALYTICAL CHEMISTRY. 55 Estimation of Tin in Copper-Tin Alloys. ARTHUR G. LEW (Analyst 1905 30 361-364).-The alloy is distilled in a current of chlorine and the volatile stannic chloride collected in receivers con- taining water.About 0- 5 gram of the finely-divided alloy is placed in a 30 C.C. distillation flask the side-tube of which is bent first up for a short distance and then down again so that any drops of liquid projected into the side-tube drain back into the flask. The flask is connected with two Volhard receivers containing water india-rubber stoppers and connections being used. The entry tube for the chlorine reaches about half-way down the bulb part of the flask ; around it and below the side-tube is wrapped a fairly tight plug of glass wool t o retain the fine spray of other chlorides carried up by the stannic chloride. The chlorine is passed through a wash-bottle containing concentrated hydrochloric acid and is then thoroughly dried before it enters the distillation flask.The latter is gently warmed when it has become filled with chlorine. The last traces of stannic chloride are removed from the flask by adding about 10 C.C. of hydrochloric acid and continuing the distillation until the bulk of the acid has been distilled off. The stannic chloride in the receivers is precipitated with hydrogen sulphide and weighed as the oxide. If antimony is present in the alloy it will distil with the tin and may be separated by the usual methods. w. P. s. Estimation of Bismuth. Separation from Copper Cadmium Mercury and Silver. ARTHUR STAEHLER and WILHELM SCHARFEN- BERG (Be?. 1905 38 3862-3869).-The precipitation of bismuth as phosphate in presence of dilute nitric acid does not yield satisfactory results as the bismuth is liable to dissolve slightly,. especially if hydrochloric acid is present.This difficulty can be overcome by precipitating in presence of phosphoric acid which dissolves the phosphates of copper cadmium silver and mercury ; lead phosphate is too sparingly soluble to be easily separated in this way. The actual precipitation is carried out by adding hot sodium phosphate to a boiling solution of the bismuth salt acidified with nitric acid in order to prevent the separation of basic salt. Analyses are given to illus- trate the separation of bismuth from chlorides from copper from cadmium from mercury and from silver. Cadmium is separated most exactly by electrolysis the electrolyte being prepared by dissolving the sulphide in nitric acid.Mercury can be separated quickly and exactly from a hot ammoniacal solution as sulphide entirely free from excess of sulphur. T. M. L. Estimation of Bismuth and Separation from the Heavy Metals as Phosphate. BEINRICH SALKOWSKI [with BERNHARD SENDHOFF] (Ber. 1905 38 3943-3944. Compare J. pr. Chem. 1868 [i] 104 172).-Bismuth is precipitated and separated quantitatively from copper cadmium mercury silver lead iron manganese cobalt nickel zinc chromium and aluminium as the phosphate in nitric acid solution which must be free from hydrochloric acid or chlorides. The precipi- tate which can be ignited without risk of reduction is weighed as56 ABSTRACTS OF CHEMICAL PAPERS. BiPO,. The insolubility of bismuth phosphate in dilute nitric acid solution may be utilised for the detection of the metal.a. Y. Temperature of Combustion of Methane in the Presence of Palladiumised Asbestos. H. G. DENHAM (J. SOC. Chem. Ind. 1905 24 (xxiii) 1202-1 205).-Different observers give widely-varying estimates of the maximum temperature at which the hydrogen in a gaseous mixture can be burnt over palladiumised asbestos without the simultaneous combustion of methane (compare Phillips Abstr. 1894 ii 393; Richardt Abstr. 1904 ii 167). The author finds that a mixture of methane and oxygen in the proportions for complete com- bustion passing over palladiumisedasbostos a t the rate of 1 C.C. per 34 sec. begins to burn a t 514-546'; a t twice the speed the tempera- ture is about 50' higher.The temperature of initial combustion rises when the proportiori of either gas is increased and is not lowered by the addition of hydrogen. Carefully prepared palladiumised asbestos gives consistent results in such combustion provided tho temperature does riot rise above 500-5509 That the action of the catalyst is not due to superficial oxidation was proved by passing oxygen over the palladiumised asbestos displacing the oxygen by nitrogen and the latter by metham; carbon dioxide was not produced a t 600'. c. s. The Iodine Value of Mineral Oils. EDMUND GRAEFE (Chem. Rev. Fett. Hum. Ind. 1905 12 296-299).-The quantity of un- saturated compounds in commercial mineral oils may be approximately estimated from the iodine value of the oil but allowance must be made for the sulphur compounds also existing in crude oils (Abstr. 1904 ii 514).In conjunction with the bromine and sulphuric acid tests tho iodine value is of use in detecting coal tar products in mineral oils. Petroleum oil remains uncoloured when carefully poured on the surface of warm sulphuric acid whilst other oils are coloured yellow or brown. American petroleum absorbs considerably more bromine than do oils from other countries. w. P. s. Detection of Methyl Alcohol in Liquids containing Ethyl Alcohol. FRANZ UTZ (Chem. Centr. 1905 ii 1 4 6 7 ; from Phccrm. Centr.-H. 46 736-737).-1n order to detect the presence of methyl alcohol in liquids which contain ethyl alcohol Kalm (D.-Am. Apoth. Zeit.) recommends the following method. 0*5-1*0 C.C. of the liquid to be tested is diluted to 5-1U C.C.with water and a red-hot copper spiral is repeatedly plunged into the solution. When 5 C.C. of milk and a solution of ferric chloride are then added and the mixture poured on to concentrated sulphuric acid a violet-blue line is formed at the zone of contact of the two layers if methyl alcohol is present in the original sample. The author finds that even when the quantity of methyl alcohol is small the formaldehyde which is formed by the oxidising action of the cupric oxide is not lost by volatilisation. Acetone gives a similar reaction but the colour is not so intense as that formed in the case of methyl alcohol. E. W. W.ANALYTICAL CHEMISTRY. 57 Detection of Methylated Spirit in Tinctures &c. ERNST SCHMIDT and RUDOLF GAZE (Arch.Pharrn. 1905 243 555-558).- Detailed directions are given for the testing of various spirits and tinctures in order to find whether they have been made with alcohol to every 100 litres of which 5 of methyl alcohol containing 30 per cent. of acetone have been added. Preparations made with methylated alcohol of this description have been put on the market lately. C. F. B. Estimation of Sugars. JULES WOLFF ( A m . Chirn. anal. 1905 10 427-431).-The amount of the precipitated cuprous oxide is estimated by dissolving it in an acid solution of ferric sulphate (free from ferrous iron) and then titrating the ferrous iron formed with Calculation of the Proportion of Lactose Hydrolysed in a Solution of this Sugar submitted to the Action of Lactase. Measurement of the Activity of Lactase.CR. PORCHER (Bull. Xoc. chirn. 1905 [iii] 33 1285-1295. Compare Porcher Abstr. 1905 ii 540 and Brachin Abstr. 1904 i 1069).-A critical dsumk of the polarimetric phenylosazone iodometric and cupric reduction methods of estimating galactose and dextrose in presence of lactose is given and it is shown that the method based on the reduction of Fehling's solution gives the best results when only small quantities of the two hexoses are present. For this reason this method is best adapted for the measurement of the hydrolytic activity of lactase towards lactose. A detailed description of the method adopted by the author in carrying out such estimations and a graphic method of calculating the results are given in the original. Polarimefric Estimation of Starch. ERICH EWERS (Zeit.ofentl. Chem. 1905 11 407-415).-Starch is completely dissolved by successive treatment with glacial acetic acid dilute hydrochloric acid and hot water and the solution obtained may be examined in the polariscope. The following conditions are necessary for the attain- ment of correct results 10 grams of the starch or starchy substance are heated in a flask with 50 C.C. of glacial acetic acid the flask being immersed in a boiling water-bath for twenty minutes. About 130 C.C. of cold water are then added and the heating continued for one hour at a temperature of 45'. I n the case of potato starch the digestion with water a t 45' must be omitted. After cooling from 2 t o 3 C.C. of saturated potassium ferrocyanide solution are added the liquid is diluted to 200 c.c.filtered and the filtrate polarised in a 200 mm. tube at a temperature of ZOO. The reading which mill be quite small (generally + 0.2 degree of the Ventxke scale) gives the amount of soluble carbohydrate in the sample. A second quantity of 10 grams of the starch is then heated with 50 C.C. of glacial acetic acid for ten minutes in a boiling water-bath. Ten C.C. of dilute hydrochloric acid (1 10) are now added and the digestion continued for exactly six minutes. The mixture is then diluted with water to a volume of 180 c.c. and heated for a further fifteen minutes except in the case of potato starch. The aolution is finally '' cleared '' with potassium permangana t e. L. DE K. T. A. H.58 ABSTRACTS OF CHEMICAL PAPERS. ferrocyanide diluted to a volume of 200 c.c.filtered and polarised. From the experiments described it is seen that all starches give approximately the same reading when the latter is calculated on the pure dry starch. For instance the following readings were obtained with various starches when treated as above wheat starch + 52.7' ; wheat flour + 52.7' ; rice starch + 52.6' ; maize starch + 52.4' ; potato starch +53*8' Yentzke when the actual readings mere calculated on the quantity of pure starch known to be present in the samples. w. P. s. Direct Estimation of Acetyl and Benzoyl Groups. RICHARD MEYER and ERNST HARTMANN (Ber. 1905 38 3956-3958. Compare R. and H. Meyer Abstr. 1896 ii 226 ; A. G. Perkin Trans. 1905 87 107 ; Sudborough and Thomas ibid. 1752).-05-0*7 gram of the substance is heated in a reflux apparatus for one hour with 5 grams of pure sodium hydroxide and 50 C.C. of methyl alcohol which has been distilled over potassium carbonate.The solution is then cooled acidified with 50 C.C. of phosphoric acid of sp. gr. 1.104 then distilled with steam and the distillate titrated with 0.1N barium hydroxide using phenolphthalein as indicator. Benzoyl groups are estimated in the same way. Attention is drawn to the fact that acid is formed when either methyl or ethyl alcohol is boiled for some time with sodium hydroxide. J. J. S. Hydrolysis of Sodium Palmitate. ROBERT CORN (Ber. 1905 38 3781-3784. Compare Zed. o f l e d . Chem. 1905 11 58; Kanitz Abstr. 1903 ii 248).-In reply to Schwarz (this vol. ii 657) the author gives experimental details to show that sodium hydroxide can be estimated accurately in aqueous solution in presence of sodium palmitate by titration of the hot solution with aqueous N/2 hydro- chloric acid with phenolphthalein as indicator and taking as the neutral point the change of colour from deep red to a light rose.The total sodium present as hydroxide and as palmitate is deter- mined by titration of the hot aqueous solution with LV/Z hydrochloric acid using methyl-orange as indicator. G. Y. Detection of Foreign Golouring Matters in Fats. GEoRa FENDLER (Chem. Rev. Fett. Rarx. Ind. 1905 12 207-209 and 237-239).-Of all the methods proposed for the detection of added colouring matters in fats and particularly in butter and margarine not one serves as a general test for each or all of the colours usually employed. The hydrochloric acid test only gives positive indications in the case of a few azo-dyes.On treating butter with a freshly- prepared ethereal solution of nitrous acid the fat is completely decolorised even if saffron marigold Orleans-yellow or Martius-yellow is present whilst turmeric aniline-yellow tropaeolin dimethyl- aminoazobenzene and certain '' butter colours " are not affected. Ex- tracting the fat by shaking with alcohol affords useful information as to the presence of foreign colours especially if the alcohol becomes strongly tinted but as nearly ail fats are soluble to a slight extent inANALYTICAL CHEMISTRY. 59 alcohol the coloration obtained must be compared with that yielded by a similar f a t known to be uncoloured.w. P. s. Behaviour of Milk towards Magenta-Sulphurous Acid Solution and the Detection of Formaldehyde in Milk. EICHHOLZ (Jfilchw. Zents.. 1905 1 499-500).-The proteids of milk have the property of giving a red coloration when treated with a solution of magenta decolorised with sodium sulphite but this property is destroyed by the addition of a trace of acid or alkali. On the other hand a solution of magenta bleached by means of sulphurous acid (Schiff’s reagent) only gives a coloration with milk when the latter contains formaldehyde. As formaldehyde after a time combines with the albumin of the milk and no longer gives a coloration with Schiff’s reagent it is better to distil the milk and apply the test to the distillate. w. P. s. Detection of Formaldehyde in Witch Hazel.WILLIAM A. PUCKNER (Amer. J. Piharm. 1905 77 501-503).-The following test is proposed as an improvement on the United States official test for the detection of formaldehyde in ‘‘ witch-hazel extract,” a preparation resembling the Liquor Hamamelidis of the B.P. One C.C. of the solution is added to 5 C.C. of a freshly-prepared solution of 0.01 gram of salicylic acid in 100 C.C. of concentrated sulphuric acid. If as little as 1 gram of formaldehyde in 10,000 C.C. is present a red coloration is produced. The quantity of salicylic acid used must not exceed the amount stated or the sensitiveness of the test will be impaired. w. P. s. Sensitive Colour Reaction for Formaldehyde Oxygenated Compounds of Nitrogen or Proteid Matters. E. VOISENET (Bull.Xoc. chim. 1905 [iii] 33 1198-1214).-When 0.1 gram of albumin is dissolved in 2 or 3 C.C. of water and to this is added a drop of a solution containing 5 per cent. of formaldehyde then on the addition OF from 6 to 9 C.C. of hydrochloric acid (4N) containing 0.01 gram of nitrous acid per litre the liquid acquires immediately a rose tint which becomes rose-violet and finally passes into an intense bluish-violet. Investigation of the conditions under which this reaction takes place shows that the colour forms most readily at a temperature of 50° that it is not produced if any one of the three substances proteid formaldehyde or nitrous acid is in excess and that hydrochloric acid may be replaced by sulphuric acid. Feeble pink colorations are produced by the action on proteid matters of hydrochloric acid containing nitrous acid in the absence of formaldehyde but these are quite distinct from the intense colours produced in presence of the aldehyde.The formation of the colour seems to be due to the interaction of the formaldehyde with the oxidation products formed by the action of nitrous acid on the scatole indole or similar substances produced by the action of hydrochloric or sulphuric acid on the proteids thus the coloration is given by most proteids by the products of pancreatic digestion of proteid and by scatole and indole but not by the products of peptic digestion,60 ABSTRACTS OF CHEMICAL PAPERS. Similarly nitrous acid may be replaced by almost any oxidising agent. Formaldehyde is the only aliphatic aldehyde which gives the coloration though similar colorations are given by some of the aromatic aldehydes.A distinct coloration is produced by solutions containing one part of formaldehyde in 10,000,000 of water. It is suggested that the reaction may be utilised for the detection of formaldehyde in food-stuffs nitrous products in sulphuric or hydro- chloric acid nitrates in water added water (containing nitrates) in milk or albumin in urine. Precise directions for the use of the reaction in these special cases are given in the original. T. A. H. Estimation of Caffeine in the Presence of Acetanilide. WILLIAM A. PUCKNEK (Yl~ccrm. lieu. 1905 23 345-350).-I)oth caffeine and acetanilide may be estimated in headache remedies by the following method. About 2.5 grams of the sample are treated with 30 C.C.of water containing 2 C.C. of Nsulphuric acid. The solution is extracted five times with chloroform using 20 C.C. each time and the extracts filtered through asbestos into a flask. The chloroform is then distilled off and the residue dried a t 100' and weighed. The residue is now dissolved in 50 C.C. of boiling water cooled and the excess of acetanilide allowed to crystitllise out. The crystals are collected on a small cotton-wool filter and mashed with about 25 C.C. of water. The filtrate and washings are treated with 1 C.C. of dilute hydrochloric acid and 25 C.C. of Nj5 iodine solution and shaken occasionally for half an hour. The precipitate of caffeine periodide is collected on an asbestos filter washed with 25 C.C. of N/10 iodine solution containing about 0.25 C.C. of hydrochloric acid and then dissolved in a solution of 2 grams of sodium sulphite in 10 C.C. of water. The filter is washed with about 15 C.C. of water and the solution and washings extracted several times with chloroform. The residue obtained on evaporating the chloroform is dried for three hours at 100" and weighed. The method yields about 99 per cent. of the caffeine actually prisent in the sample. w. P. s. Choline Periodide and the Quantitative Precipitation of Choline by Potassium Tri-iodide. VLADIMIR STAN~K (Zeit. physiol. (?]hem. 1905 46 280-285).-Choline periodide which is obtained by the addition of iodo-potassium iodide t o a solution of choline was first described by Griess and Harrow (Trans. 1885 47 298); it can be obtained in crystalline form and may be used for the estimation of choline. It is suggested that this reaction may be employed to deter- mine the amount of choline in animal and vegetable tissues. The per- iodidc has the composition C,H,,ONI,T and it may therefore be spoken of as a n enneaiodide. W. D. H.