ANALYTICAL CHEMISTRY. 89 An a1 y tic a1 Chemistry I A New Electrical Method for the Continuous Analysis of Gas Mixtures and its Application to the Measurement of the Velocity of Gas Currents. ADOLF KOEPSEL (Ber. deut. physikal. Ges. 1908 6 814-827).-The method depends on the variation of the thermal conductivity of a gas mixture with its composition. In its application the resistance of a wire suspended in the gas to be analysed is compared with that of a second wire contained in air or a gas mixture of constant composition. Both wires are traversed by a constant current but the difference between the thermal conductivi- ties of the surrounding gases results in temperature differences and consequently in differences in the resistances of the two wires which are measured. In the apparatus which has been constructed for the continuous electrical analysis two wires are suspended in the gas to be analysed and two in the standard gas and these are arranged so as to form the arms of a Wheatstone bridge. The method is recom- mended for the analysis of generator gas flue gases and mine gases.In the actual apparatus elaborate precautions are taken to avoid disturbances resulting from the action of the gas currents on the heated wires. The heated wires are so sensitive to the movements of gases that the arrangement may be utilised for the measurement of very slowly moving air-currents. Some observations relating t o the sensitiveness of the electrically- heated wires for different loads are communicated. For the wires used by the author all the gases and gas mixtures examined mere found to exhibit maximum sensitiveness when currents ranging from 0.30 to 0.34 ampere were passed through them.This variation in the current corresponded with a temperature range of 180' to 320'. H. M. D. The Mercury Bath a Too-little Known Useful Expedient in Gas-analytical Work. ALFRED STOCK (Ber. 1908 41 3834-3849).-The original contains a detailed illustrated descrip tion of a mercury-bath specially adapted for gas analysis together with directions for working with it. Gasometric Work by means of V. Meyer's Vapour Density Principle. JULIUS MAI (Ber. 1908 41 3897-3904. Compare Abstr. 1903 ii 98).-The method has now been simplified and the W. H. G.90 ABSTRACTS OF CHEMICAL PAPERS. sources of error reduced so that it may be employed in actual practice.The bulb is placed near the outlet tube and the constant temperature maintained by boiling xylene toluene aniline or water The asbestos cartridge has been displaced in favour of a glass tube 27-35 mm. x 6 mm. provided with one to two holes 4 to 6 mm. from the jop. The results obtained from calcium strontium and barium carbonates are in accordance with the theoretical ; magnesium carbonate gives low results. W. R. BERNIIARD TOLMACZ (Zeitsch. angew. Chem. 1908 21 2551).-The pipette is surmounted by a tap so bored that the interior of the pipette may be in communication with the outside air or with the suction tube or with neither according as it is turned. Tap pipettes with more than one bulb are found advantageous where several small quantities of the same liquid are to be measured out ; thus the contents of the upper bulb between two graduations on the stem may be accurately measured out and then the contents of the second bulb.J. V. E. Tap Pipettes. Burette for Calibrating Measuring Flasks. A. MULDER (Chem. Weebblad 1908 5 830-S31).-The burette consists of a fairly narrow tube with a large bulb and below it a smaller bulb a aide-tube being sealed on below the smaller bulb and a glass stopcock below the side-tube. Above the larger bulb is a zero mark between the two bulbs a 100 C.C. mark and below the smaller bulb a 110 C.C. mark. Above and below the last two marks there are a number of divisions corresponding t o tenths of a C.C. The portion of the tube below the stopcock is capillary to admit of the introduction of the water into the flask without wetting the neck of the latter. Water is introduced into the burette from a bottle at a higher level through the side-tube and a tube fitted with a pinchcock.The apparatus is suitable for the calibration of measuring flasks intended to contain any multiple of 100 or 10 C.C. An Indicator Highly Sensitive towards Alkali and Suitable for Titrations with Centinormal Solutions. ERWIN RUPP and R. LOOSE (Ber. 1908 41 3905-3908).-The authors find that p dimetlq Zaminoaao benzene-o - car boxv Z ic acid C02H*C,H;N N*C,H,*EMe surpasses other indicators and especially haematoxylin and iodoeosin when used in the titration of very dilute solutions (up to N/lOO) of weak bases (especially vegetable alkaloids).The indicator to which is given the name " methyl-red,'' is prepared by combining a diazotised solution of o-aminobenzoic acid with dimethylaniline in alcoholic solu- tion and crystallises from acetic acid in glistening violet needles. I t F colour in alkaline or neutral solution is pale yellow and in acid solution violet-red. J. C. C. Burette for the Analysis of High-Grade Oxygen. H. MURSCH- HAUSER (Zeitsch. angew. Chem. 1908,21 2503-2504).-A combination of two gas burettes of the Zuntz-Geppert type surrounded by a water jacket; one serves as a thermobarometer whilst into the other is A. J. W.ANALYTICAL CHEMISTRY. 91 introduced in the usual manner about 100 C.C. of the sample which may be measured to 1/50 C.C. By means of Hempel pipettes containing aqueous potassium hydroxide and alkaline sodium hyposulphite the carbon dioxide and oxygen are absorbed and the remaining gas consists of nitrogen.L. DE K. Detection of Hydrogen Peroxide Formaldehyde and Persulphates. S. ROTEIENFUSSER (Zeitsch. Nahr. Genussrn. 1908 16 589-591).-Hydrogen peroxide may be detected in milk by adding about 10 drops of 2% alcoholic benzidine solution and a few drops of acetic acid to 10 C.C. of the milk; a blue coloration is obtained if hydrogen peroxide is present. For the detection of hydrogen peroxide in liquids other than milk it is necessary t o add a little milk serum as well as the reagents as the coloration is obtained only in the presence of the protein. Persulphates give a blue coloration with the benzidine solution alone and 1 part of persulphate in 1,000,000 may be detected by this means.Formaldehyde may be detected in milk by warming a portion of the milk with twice its volume of concen- trated hydrochloric acid to which has been added 3 drops of a 2% solution of a molybdenum uranium cobalt silver mercury or copper salt; a violet coloration being produced in the presence of form- aldehyde. w. P. s. Analysis of a Solution containing a Mixture of Sulphides Hydrogen Sulphides Polysulphides and Hyposulphites. F. DHUIQUE-MAYER (Chem. Zent?'. 1908 ii 1124-1125 ; from Rev. gen. Chim. pure. appl. 1908 11 273-274).-Ten C.C. of the solution are diluted to 100 c.c. and 10 C.C. of this again diluted to 200 C.C. and titrated with N/lO sulphuric acid using phenolphthalein as indicator (A). The liquid is next titrated with A'/lO iodine using starch as indicator (I).After decolorising the liquid with a small drop of the solution it is again titrated with N,/lO sodium hydroxide until pink (R). Another 10 C.C. of the solution are diluted and shaken with 2 grams of lead carbonate and the filtrate is titrated with N/10 iodine which now only gives hyposulphite (23). A = sulphides R = total sodium hydrogen sulphide R - A =true sodium hydrogen sulphide I - 2 R == polysulphide and hyposulphite H = hyposulphite ( I - 2R) - H = polysulphide. The coefficients used in the calculation are the tenth- part of the molecular weights; in the case of polysulphides the titre is divided by 2 and the result is expressed as sodium disulphide. L. DE I(. Estimation of [ Ammoniacal] Nitrogen.FRITZ TAURKE (Chem. Zeil. 1908 32 1176).-The usual plan is to distil 50 C.C. of the usual solution (=1 gram of the sample) with magnesia; the distillate is collected in a receiver containing 50 C.C. of N/2 acid and the excess of this is titrated with N/4 alkali. The amount of nitrogen is then calculated from the formula (50 - a)/2 x 0.007 x 100 = percentage. I n order to facilitate the operation the author starts with 35 C.C. of the solution ( = O T gram of the sample) and titrates from a burette graduated ti0 0.1 C.C. from 25' at the top t o 0' at the bottom. The92 ABSTRACTS OF CHEMICAL PAPERS. number read off then at once gives the percentage of ammoniacal C.*ALBERTE GARCIA (Bull. SOC. chim. 1908 [iv] 3 111 1-1114).-The point for which novelty is claimed in this apparatus is the device for eliminating all the air from the burette before the reaction is started.The nitrometer consists of a burette provided with a stopcock A communicating with a cylindrical reservoir of the usual pattern at the upper end and terminating below in a cylindrical bulb of 40 C.C. capacity which communicates laterally at its upper end with a glass funnel serving as a water reservoir and at its lower end with a reservoir of mercury. The water and mercury reservoirs connect with the burette by flexible tubing and can be used for levelling purposes and their connexions with the burette are controlled by stopcocks B and C respectively. I n using the apparatus the water reservoir and its connexion are filled with boiled distilled water of which a little is allowed to run into the burette.The stopcock B is then closed and the burette filled with mercury by opening stopcocks A and C. Stopcock A is then closed. By lowering the mercury reservoir a reduced pressure space is created in the burette so that solutions of reacting ingredients can be sucked in through A . A t the termination of the reaction water can be added through B and the resulting gas saturated with water vapour measured in the burette in the usual manner. nitrogen. L. DE K. New Mercury Nitrometer. T. A. H. Volumetric Estimation of Water-Soluble Phosphoric Acid in Superphosphates. LUDWIG SCHUCHT (Chem. Zeit. 1908 32 1201-1202).-Five hundred C.C. of the aqueous solution of the super- phosphate (20 grams t o 1 litre) are mixed with N/-potassium oxalate in order to remove the calcium but'as an excess must be strictly avoided it is best t o ascertain by a special experiment how much oxalnte is really required.After heating at 70' and subsequent cooling the solution is diluted to 550 c.c. and 110 C.C. of the filtrate (= 2 grams) are titrated in presence of methyl-orange with N,/2 alkali. Another 110 C.C. are mixed with 25 C.C. of the oxalate solution and then titrated in presence of phenolphthalein. The following example is given. I n the analysis of a superphosphate the first titration required 2.7 C.C. alkali; the second 11.2 c.c.; 11.2-2.7 C.C. = 8.5 C.C. x 0.0358 x 100/2 = 15.1% soluble P,O,. The amount of free acid = 2.7 C.C. x 0.0355 x 100/2 = 4.8% of P,O,. L. DE K. Estimation of Boric Acid in Insoluble Silicates.EDGAR T. WHERRY and WILLIAM H. CHAPIN (J. Amer. Chem. Soc. 1908 30 1687-1704). -Volumetric Method (WHERRY). - 0.5 Gram of the sample is fused with 3 grams of sodium carbonate for fifteen minutes. The mass is treated with 20-30 C.C. of dilute hydrochloric acid and a few drops of nitric acid and heated nearly to boiling in a 250 C.C. round-bottomed flask. After adding a moderate excess of pure calcium carbonate a refiux condenser is attached and the contents are boiled vigorously for ten minutes. The precipitate is collected on a BiichnerANALYTICAL CHEMISTRY. 93 funnel and washed with hot water. The filtrate :is then returned to the flask and after adding a pinch of calcium carbonate it is freed from carbon dioxide by boiling under reduced pressure j in this manner there is no fear of loss of boric acid by volatilisation. The liquid is filtered and rendered decidedly alkaline t o phenol- phthalein when 1 gram of mannitol is added.The solution is now titrated for boric acid with standard sodium hydroxide until pink when another gram of mannitol is added and the titration continued if necessary. The aodirim hydroxide must be titrated accurately as to its true hydroxide content; in that case it matters not i f a little carbonate should be present as this is neutralised by the calcium chloride present. Distillation Method [CHAPIN].-T~~S method is briefly as follows 0.5 gram of the borosilicate is fused with 3 grams of potass- ium sodium carbonate. The mass is decomposed by 12 C.C. of dilute hydrochloric acid (1 l) and the solution put into a distilling flask together with 12 grams of dry calcium chloride.The flask is connected with a generating flask in which methyl alcohol is being boiled and when 25 C.C. of this have collected heat is applied and the vapours (methyl alcohol and methyl borate) are passed through a short condenser and collected in an Erlenmayer flask to which a water trap is attached. When 100 C.C. of distillate have collected another 100 C.C. are collected in a second receiver. The methyl alcohol is distilled off after addition of an amount OF sodium hydroxide sufficient t o retainall the boric acid The resulting solution is then titrated for boric acid as usual using either glycerol or mannitol as a medium. L. DE K. Combustion Analysis.JAMES WALKER and THOMAS BLACKADDER (Proc. Roy. Soc. Eldim 1908 28 708-?12).-The dimensions of the tube used for copper oxide combustions have been reduced so as to secure the advantages of the Dennstedt furnace and to make this method of analysis more suitable for students working at an ordinary labopatory bench A furnace of the Dennstedt pattern but only 60 cm. in length and a Jena glass combustion tube 66 cm. long and not more than 8 mm. internal diameter are employed. Tho tube is charged with coarsely- powdered copper oxide and the substance to be analysed also mixed mlth copper oxide is burnt in a moderately rapid current of oxygen at a dull red heat. Details are given for carrying out combustions of compounds containing nitrogen and also volatile substances such as benzene the results obtained in each case being very good.A combustion tube of the same dimensions may also be used for estimating nitrogen by the direct method with equally satisfactory results. J. V. E. The Use of Sodium Peroxide for the Quantitative Analysis of Organic Compounds. 111. HANS PRINGSHEIM (Ber. 1908 41 4267-4271. Compare Abstr. 1904 ii 447; 1905 ii 609).-A modification of the older method is described. Substances containing 75% or more of C + H + S require eighteen times their weight of peroxide and those with 50-75% C + H + S,94 ABSTRACTS OF CHEMICAL PAPERS. sixteen times. Substances containing 25450% C + H + S are mixed with half theirweight of acompound rich in C and H(for example,sugar); those with less than 25% C+H+S with their own weight of the compound rich in C and H.The first of these mixtures is used with pixteen times its weight of peroxide and the second with eighteen times its weight. The admixing is carried out in an iran crucible by means of an iron nail The lid is placed on the crucible which is introduced into a dish containing cold water. The iron nail is raised t o a red heat and introduced through the hole in the lid of tho crucible. If the mixture is in the right proportions the whole mass becomes molten and after cooling for a few minutes is dissolved in cold water and sulphur halogens arsenic or phosphorus estimated as already described. Excellent results can be obtained. Estimation of Potassium as Phosphomolybdate. A. SCHLICHT (Chem. Zeit. 1908 32 1125-1126 1138-1140.Compare Abstr. 1907 ii 130).-About 0.1 gram of potassium sulphate free from ammonium salts is mixed with 10 C.C. of the author's reagent and evaporated to dryness. The residue is taken up with 10 C.C. of hot 5% solution of magnesium sulphate and when cold the precipitate is collected on a weighed asbestos filter and washed first with 10 C.C. and then with 5 C.C. of 5% magnesium sulphate then again three times with 5 C.C. of 5% ammonium nitrate and finally with alcohol and ether. 0.1 Gram of pure potassium sulphate thus treated yields 0.8071 gram of dry and 0.7860 gram of ignited residue. The reagent is prepared as follows 100 grams of molybdic acid are dissolved in water with addition of 250 grams of crystallised sodium carbonate the whole is evaporated to dryness and the residue heated for two hours at 150'.It is then dissolved in water up to 500 C.C. Meanwhile 6 grams of magnesium pyrophosphate are dissolved in boiling nitrohydrochloric acid and then evaporated repeatedly with nitric acid to expel the chlorine. The residue is dissolved in a little water and nitric acid D 1.36 and diluted to 500 C.C. with acid of the same strength. To this is then added gradually and with stirring the molybdate solution also another 500 C.C. of water and after standingLfor two hours the liquid is J. J. S. filtei ed. L. DE I(. Volumetric Estimation of Potassium in Animal Fluids. W. A. DRUSHEL (Anzer. J . Xci. 1908 [iv] 26 555-662*).-The process has been described previously (Abstr. 1908 ii 66). The only modification is that the " cobalt-yellow " is washed with weak brine (1 1) instead of with water ; this permits the use of a coarser asbestos felt without danger of loss.It is then titrated with N/10 permanganate 8s usual. Potassium in Urine.-Ten to fifty C.C. of the sample are evaporated to dryness in a small platinum dish and the residue is heated with 6-10 C.C. of a mixture of nitric and sulphuric acids (9 1); when the first violent action is over the acid is expelled and the residue * and Zeztsch. nnorg. Chern. 1909 61 137-146.ANALYTICAL CHEMISTRY. 95 ignited. The residue is then dissolved in a little water and a few drops of acetic acid and without previous filtration the liquid is treated with the sodium cobalt nitrite reagent as directed. Potassium in Circulatory flhids.-In this case the mass (blood serum lymph) is oxidised with nitric acid alone and the residue is heated but not strongly enough to cause an explosive decomposition.The mass is then heated with sulphuric acid and finally ignited. In the case of defibrinated blood this may be treated with about 2 C.C. of bromine and after an hour it is evaporated to dryness and charred. The mass is then extracted with hot water and the solution evaporated t o dryness after adding a little sulphuric acid. Potassium in Milk-The milk is evaporated to dryness and the residue treated with nitric acid. The whole is evaporated to dryness and ignited to burn off most of the organic matter. The residue is then moistened with sulphuric acid and once more ignited. L. DE I(. The Determination of Calcium Carbonate in Soils by the Methods of Bernard and of Treite and its Significance in the Selection of Soils [for Vineyards].KARL VOTRUBA (Chem. Zentr. 1908 ii 1065 ; from Allgem. Wein-Zeit. 1908 No. 30).-Bernard’s method consists in gently powdering the soil passing it through a 0.3 mm. sieve and determining the carbon dioxide evolved in a calcimeter from the fine soil. I n Treitz’s method the soil is stirred with water allowed to settle for half a minute decanted alid allowed to settle for half an hour and the carbon dioxide given off from the deposit on treatment with acid determined. The two methods did not give concordant results in the author’s tests but in general itlappears that the amount of carbonate in the fine soil is more important than that present in coarse lumps.Other properties of the soil however have t o be taken into account in selecting soils for vineyards. E. J. R. Detection of Poisonous Metals. GE-ORGE D. LANDER and H. W. WINTER (Analyst 1908 33 450-452).-The following methods are suggested for a rapid and simple qualitative search for poisonous metals in organic substances such as viscera etc. A Reinsch test is first made for arsenic antimony bismuth and mercury; if a deposit is obtained on the copper foil numerous methods are avail- able for its further examination. A portion of about 50 grams of the sample is then treated with 50% nitric acid (in quantity sufficient to cover the substance) from 5 to 10 C.C. of copcentrated bulphuric acid being also added The mixture is heated until dense brown fumes are evolved diluted and filtered.The filtrate is rendered strongly alka- line with ammonia a slight excess of ammonium sulphide is added and the precipitated phosphates chromium hydroxide and iron lead mercury copper bismuth and zinc sulphides are collected on a filter and washed. The metals are thus obtained practically free from organic matter and may then be separated and identified by known methods. w. P. s.96 ABSTRACTS OF CHEMICAL PAPERS. Microchemical Analysis. IV. Mercury Bismuth Lead Copper Cadmium. NICHOLAS SCHOORL (Zeitsch. anal. Chem. 1908 47 729-754).-The sulphides of the metals are treated with nitric acid and the insoluble mercuric sulphide is separated ; a drop of the nitric acid solution is then heated on an object glass with a drop of sulphuric acid If on diluting with water a white insoluble residue remains it is probably lead sulphate and its presence is confirmed by other reactions.The sulphates of bismuth copper and cadmium are treated with water and if basic bismuth sulphate remains insoluble it is identified by the appearance of the microscopic crystals when it is converted into potassium bismuth sulphate. The solution containing the copper and cadmium is next evaporated the residue is heated to dull redness and treated with water ; cadmium sulphate dissolves whilst copper oxide remains undissolved. Cadmium may be identified by the shape of the crystals when it is converted into either cadmium oxalate or cadmium tetrarubidium chloride. Copper may be identified by the reaction described by Meerburg and Filippo (Abstr.1906 ii 52). w. P. s. Volumetric Estimation of Lead in Ores. JOSEPH A. MULLER (Bull. SOC. chirn. 1908 [iv] 3,1131-1133).-The author has examined the Schwatz volumetric process both in the case of lead nitrate solu- tions of known strength and of a mixture of known composition in- cluding the usual constituents of a lead ore and finds that it gives low results. These are due (1) to the adhesion of some lead sulphate (even in presence of excess of sodium acetate) to the lead chromate when this is precipitated by the standard dichromate used and (2) in the case of the mineral mixture to the difficulty of extracting the whole of the lead especially when the mixture contains ferric oxide as this is liable to form with lead oxide lead ferrate which is attacked with diEculty by sodium hydroxide solution.Weighing the precipitated lead chromate gives better results but it is preferable to weigh as the sulphide as recommended previously (Abstr. 1905 ii 118 119). T. A. H. Decinormal Solution of Potassium Permanganate. BERNARD COLLITT (Pharm. J. 1908 27 724).-Experiments showing that a standard solution of potassium permanganate suffers no loss in strength for at least twelve months if kept in the dark even although the bottle JOSEPH A. MULLER (Bull. Xoc. chim. 1908 [iv] 3 1133-1136).-The author has examined the volumetric methods described by Carnot (Abstr. 1889 311) and by Zulkowki. He finds that the former which depends on the use of a standard solution of hydrogen peroxide gives results which vary with the dilution of the solution titrated Zulkowki's process depending on the liberation from potassium iodide of iodine and titration of the latter with sodium thioaulphate gives trustworthy results if no other oxidising agent is present.En the latter case titration with a standard solution of a lead salt can generally be resorted to but as cshown in a may have to be frequently opened. L. DE K. Estimation of Chromic Acid.ANALYTICAL CHEMISTRY. 97 previous paper (see preceding page) it is safer to weigh the precipitated lead chromate. T. A. H. Electrolytic Estimation of Nickel. AUGUST SCHUMANN (Zeitsch. angem. Chm. 1908 21 2579-2583).-A criticism of the methods proposed from time to time. Fresenius-Bergmann's ammonia method is free from sources of error and contrary to general belief i t is not interfered with by the presence of nitrates or chlorides.Brand's ammonium carbonate method also gives excellent results but the author has not as yet tried this in the presence of nitrates and chlorides. Addition of sodium pyrophosphate is unnecessary. Classen's oxalic acid method cannot be recommended as the nickel deposit is contaminated with carbon. The electrolysis succeeds best by the use of a platinum-gauze cathode a platinum spiral as anode and a weak current. A sufficiency of ammonium sulphate should be present. L. DE K,. Volumetric Estimation of Nickel with Potassium Cyanide. HERMANN GROSSMANN (Chenz. Zeit. 1908 33 1223).-The author has applied this process to mixtures of nickel and cobalt.To the solution is added some sucrose to prevent precipitation of cobalt and the nickel is then precipitated by means of dicyanodiamidine sulphate ammonia and aqueous potassium hydroxide. The precipitate is collected on a filter and re-dissolved in warm dilute hydrochloric acid. When cold a slight excess of ammoria is added and the nickel 'is titrated with potassium cyanide as usual. Further experiments including the estimation of nickel in the presence of zinc are in progress. L. DE K. Separation of Tin Arsenic and Antimony; Analysis of Bronzes. DINAM (Chew,. Zentr. 1908 ii 1207-1208 ; from Mon. Xci. 1908 [iv] 11 600-602).-Three grams of the bronze are heated with 30 C.C. of nitric acid D 1.4 ; the filtrate contains zinc lead and the greater portion of the copper ; the precipitate contains arsenic anti- mony tin and a little copper and lead.It is dissolved in a boiling solution of 10 grams of oxalic acid and 5 grams of ammonium oxalate and after adding a little hydrochloric acid a current of hydrogen sulphide is passed through the hot liquid for two hours. The tin remains in solution and may be deposited electrolytically whilst the arsenic antimony copper and lead are precipitated as sulphides. On treating with boiling aqueous potassium hydroxide the copper and lead are left undissolved ; these are then dissolved.in nitric acid and the solution added to the main acid liquid. The alkaline solution is heated with excess of hydrochloric acid the filtrate is oxidised with potassium chlorate and after expelling the excess of chlorine the antimony is estimated iodometriccllly with potassium iodide and sodium thiosulphate.The precipitated arsenious sulphide is dissolved in potassium hydroxide and also estimated iodometrically in the same manner as the antimony. The acid liquid is concentrated t o about 15 c.c. 3 C.C. of saturated sodium acetate are added and the copper and lead are deposited by electrolysis; the liquid contains the zinc VOL. XCVI. ii. 798 ABSTRACTS OF CHEMICAL PAPERS. which is precipitated by boiling with sodium carbonate and finally weighed as oxide. If the bronze should contain phosphorus this will be found in the liquid from which the tin has been deposited electrolytically and it may be precipitated as ammonium magnesium phosphate or ammonium phosphomolybdate. If iron should be present in the alloy a portion of it will cont,aminate the zinc and another portion remains in the liquid from the tin deposit from which it may be precipitated with ammonium sulphide.L. DE K. Estimation of Antimony and Arsenic in Lead-Antimony Alloys. GEO. M. HOWARD (J. Arrzer. Chern. Soc. 1908 30 1789-1 790).-A slight modification of the author’s process (Abstr. 1908 ii 429). The finely-divided alloy is heated with hydrochloric acid until action ceases the flask is removed from the hot plate and 0.5 C.C. of nitric acid is added. As soon as the arsenic and antimony have dissolved the liquid is boiled for some five minutes and then treated as directed. If the antimony does not exceed 0.1 gram 5 grams of tartaric acid will be sufficient to prevent precipitation.If there is a large excess of lead chloride a larger amount of sodium hydrogen carbonate should be used when titrating the arsenic; the blue colour should be per- manent for at least one minute. In oxidising the arsenic 10 C.C. of 3% hydrogen peroxide generally suffices and the solution must be distinctly alkaline; if this is not so as shown by separation of sulphur a little more sodium hydroxide should be added. L. DE K. Identification of Dihydric Phenols. A Delicate Reaction for Resorcinol. T. SILBERMANN and N. OZOROVITZ (Chern. Zentr. 1908 ii 1022; from BuZ. SOC. Sci. BUCUYBSC~ 19OS 17 41-42).-With formaldehyde in acid solution the dihydric phenols give characteristic resinous insoluble condensation products. When resorcinol is heated with formaldehyde in presence of hydrogen chloride or sulphuric acid a very voluminous flocculent precipitate is slowly formed.The colour of this substance is very pale red but when it is added to con- centrated sulphuric acid o r heated with concentrated hydrogen chloride solution i t changes to a brilliant carmine-red and on the addition of water or when neutralised this colour changes to pale orange. The carmine-red or orange-coloured substance in the presence of free alkali changes to Bordeaux-red. An acid solution of resorcinol (1 100,000) gives a white turbidity with formaldehyde which becomes rose-pink on addition of hydrogen chloride and yields a red solution on the addition of excess of sodinm hydroxide. Catechol with formaldehyde in acid solution gives a very pale yellow flocculent precipitate which changes when boiled with concentrated hydrogen chloride solution to a violet-brown ; dilution with water or neutralisation of the acid changes it to a flesh-pink colour; excess of alkali produces a greenish-brown colour.Quinol requires more acid for condensation with formaldehydeANALYTICAL CHEMISTRY. 99 to take place; a very pale violet flocculent precipitate is obtained which changes to pale brown when boiled with strong hydrogen chloride solution diluted with water or when neutralised and t o olive-green in presence of excess of alkali. J. V. E. The Sodium Nitroprueside Reaction for Acetone. A. C. H. ROTHERA (J. Physiol. 1908 37 491-494).-The following modifica- tion of the test is recommended creatinine not reacting in the manner described. Five to ten C.C.of a solution of acetone in water or urine are taken and a solid ammonium salt (chloride bromide acetate or carbonate) added; then 2 or 3 drops of a freshly-made 5% solution of sodium nitroprusside and 1-2 C.C. of strong ammonia. A character- istic permanganate colour is produced. W. D. H. Detection and Estimation of Formaldehyde in Wine Stored in Barrels which have been Disinfected with Formaldehyde. FRIEDRICH SCEAFFER (Zeitsch. Nahr. Genussrn. 1908,16 674-676).- Barrels which have been treated with formaldehyde and subsequently steamed may still contain sufficient of the aldehyde to contaminate wine which is stored in them. The author has detected traces of formaldehyde in such mines by means of the test described by Arnold and Mentzel (Abstr. 1902 ii 480).Before being tested the wine should be distilled with phosphoric acid and the distillate rendered alkaline with sodium hydroxide in order to decompose the aldehyde- sulphurous acid compound which distils over ; after the lapse of fifteen minutes the mixture is acidified with sulphuric acid and the test is applied. For the estimation of formaldehyde in mine the latter is distilled with phopphoric acid until the volume of the distillate is about three-fourths of that of the quantity of wine taken. The distillate is rendered alkaline with sodium hydroxide and after fifteen minutes is exactly neutralised with sulphuric acid using rosolic acid as indicator. The neutral solution is then shaken with a slight excess of N/1 ammonia (about 2 C.C.are usually sufficient) and at the end of three hours the excess of animonia is titrated with N/10 sulDhuric acid. The estimation der>enda on the combination of the amionia with the aldehyde accordiig to the equation 6CH,O + 4NH = C,H,,N + 6H,O. w. Y. s. Colorimetric Method for the Estimation of Formaldehyde in Milk. EDWARD W. T. JONES (Chem. News l908,98,247),-The method depends on the coloration produced when a mixture of milk formalde- hyde and hydrochloric acid containing a small quantity of ferric chloride is heated. The reagent is prepared by dissolving 0.25 gram of pure iron wire in hydrochloric acid oxidising the ferrous salt with nitric acid and precipitating the iron as ferric hydroxide by the addition of am- monia; the precipitate is collected on a filter washed and then dis- solved in 500 c .~ . of concentrated hydrochloric acid. Ten C.C. of the sample of milk containing formaldehyde and 10 C.C. of the reagent are heated together for twenty-five minutes a t a temperature of about 80' ; 30 C.C. of water is next added and when cold the mixture is passed through 7-2100 ABSTRACTS OF CHEMICAL PAPERS. a filter the precipitate washed with a little water and the filtrate is diluted t o a volume of 100 C.C. The blue colour of the filtrate is then compared with the colorations obtained by similarly treating quantities of pure milk t o which have been added known amounts of formaldehyde. w. P. s. Characteristic Reaction of Uric Acid. DOMENICO GANASSINI (Boll. chint. farm. 1908 47 715-726).-The author gives the follow- ing test for the presence of uric acid or a urate which is not disturbed by the presence of proteins and is not given by xanthine hypoxanthine caffeine theobromine theophylline heteroxanthine pnraxanthine 1 7 - dimethylhypoxanthine 3-methylxanthine carbamide alloxan alloxantin allantoin creatine creatinine glycine tyro sine leucine hippuric acid nuclein nucleic acid or the bile acids.To a small quantity of uric acid or a urate dissolved or suspended in water 10% sodium hydroxide solution is added until the whole is dissolved giving a distinctly alkaline solution. Subsequent saturation of the liquid with potassium persulphate and gradual addition of 10% aqueous zinc sulphate until the precipitate formed no longer redissolves causes the precipitate to assume a bluish-green colour.The reaction is shown distinctly by 10-2Oc.c. of a 0.01% solution of uric acid and is still more sensitive if carried out directly on solid uric acid or a solid urate or the precipitated double urate of silver and magnesium. The reaction may be applied to the detection of uric acid in urine in the following manner Three C.C. of 10% sodium hydroxide solution are added to 10% of urine and the liquid filtered 5 C.C. of the filtrate being treated with 2 C.C. of 10% aqueous zinc sulphate. The gelatinous white precipitate formed when mixed on the filter with a little potassium persulphate will assume a bluish-green colour i f uric acid 1s present. I n blood traces of uric acid may be detected as follows A few C.C. of blood-serum are well mixed with one-half the volume of freshly- prepared Salkowski-Ludwig reagent and left for about twelve hours in the dark. The faint white or grey precipitate is collected and washed five or six times with water on a small filter being kept as far as possible away from the light. A little powdered potassium persulphate and 2 or 3 drops of 10% sodium hydroxide solution are then applied to the precipitate in which a bluish-green colour indicates the presence of uric acid.By this means the author has detected although not in all cases traces of uric acid in the blood of the ox and horse. The blood-serum of birds is also found to contain uric acid which is further met with in samples of blood from healthy and diseased human beings. T. H. P. Iodometric Estimation of Uric Acid [in Urine].FERNAND REPITON (Chem. Zentr. 1908 ii 1139 ; from Rev. gem. Chim. pure appZ. l908,11,285-286).-Twenty C.C. of urine are mixed with 5 C.C. of acetic acid and 20 C.C. of copper tartrate solution which has been bleached by means of a saturated solution of ammonium sulphite areANALYTICAL CHEMISTRY. 101 added. The copper solution should be standardised iodometrically. After the cuprous urate has settled the liquid is made up to 100 c.c. and an aliquot portion of the filtrate is taken to ascertain the amount of copper left in solution. In order to obtain it in the higher degree of oxidation 4 C.C. of sulphuric acid are added and the liquid is boiled t o expel the sulphur dioxide. When cold sodium carbonate is added until a precipitate forms which is then redissolved by means of acetic acid.Potassium iodide is now added and the iodine liberated titrated with standard thiosulphate. The difference in the titrations represents the uric acid L. DE I(. Detection of Salicylic Acid in Wine Milk Butter and Preserved Tomatoes. UMBERTO SAPORETTI (Boll. chim. farm 1908 47 751-755).-The following modified methods based on the coloration given by dilute ferric chloride solution with salicylic acid are more rapid and convenient than those usually employed and are capable of detecting salicylic acid in the dilution 1 80,000. Wine.-Ten C.C. of the mine are shaken in a test-tube with a mixture of 3 C.C. of ether and 1 C.C. of light petroleum previously acidified with 10% hydrochloric acid. The ethereal solution after separation is allowed to flow down the side of another tube containing a very dilute ferric chloride solution as also is a further quantity of the solvent after shaking with the wine.On evaporating the ether and light petroleum by gentle heating the lower liquid gradually assumes the characteristic violet colour. As certain substances in pure wines may give a somewhat similar coloration the presence of salicylic acid may be confirmed by adding a little of the coloured ferric chloride solution drop by drop to water which then turns violet. NiZk-Ten C.C. of the milk are heated to 60-70' with 3-4 drops of acetic acid and the clear liquid removed by a pipette is run t o the bottom of a beaker containing dilute ferric chloride solution; the appearance of a violet ring at the surface of contact of the two liquids indicates the presence of salicylic acid.Butter.-Five to ten grams of the butter are melted in a porcelain capsule and poured into a test-tube the capsule being then rinsed out with 7-8 C.C. of aqueous alcohol (1 part 95% alcohol to 4 parts of water) which is added to the test-tube. After being rendered faintly acid with 10% sulphuric acid the butter is shaken repeatedly warmed slightly and then allowed to set but not solidify. The supernatant liquid is then removed in a pipette and run to the bottom of a beaker containing dilute ferric chloride solution. If salicylic acid is present a violet ring appears where the two liquids meet. Preserved Tomatoes.-To 10 grams of the material,.placed in a porce- lain capsule are added a few drops of 10% sulphuric acid and about 5 C.C.of ether the whole beihg well mixed. The ethereal liquid is then poured into dilute ferric chloride solution and the operation repeated with a few C.C. of light petroleum which is also subsequently added to the ferric chloride solution. On gradually heating the latter if salicylic acid is present the lower liquid slowly assumes the characteristic violet colour. T. H. P.102 ABSTRACTS OF CHEMICAL PAPERS. Bang’s Method of Sugar Estimation and itB Application in the Analysis of Urine. A. C. ANDERSEN (Biochem. Zeitsch. 1908 15 76-94).-The experiments indicate that Bang’s method (Abstr. 1907 ii 136) is applicable to sugar estimations in urine. The error due to the reducing substance normally present in urine is small and this even can be almost entirely eliminated by treatment of the urine with lead acetate or mercuric nitrate.Bang’s method can also be applied to the estimation of sugar in molasses if the solutions are first clarified by the mercuric nitrate method. S. B. S. Estimation of Reducing Sugars. PERCY H. WALKER (BUZZ. Assoc. chim. sucr. dist. 1908 26 341-342).-Results of experiments are given showing that slightly higher figures are obtained when the precipitated cuprous oxide is weighed as such than when it is estimated electrolytically or converted into cupric oxide before weighing (Abstr. 19OS ii 902). The disadvantage of weighing the copper as cupric oxide is that the tables published by the author (Abstr. 1907 ii 585) cannot be used.w. P. s Estimation of Dextrose in Urine with the Weidenhaff‘s Fermentation Saccharometer. GEORG GREGOR (Chem. Zeiztv. 1908 ii 987-988 ; from Zeitsch. ally. Oesterr. Apoth. Ver. 46 41 9-420).-The author who has successfully employed this apparatus of precision recommends that the weight of the mercury should always be taken before each experiment as small globules are easily lost Analysis of Camphorated Oil for Camphor Substitutes. FREDERIC W. RICHARDSON and W. WALTON (Analyst 1908 33 463 -466).-The following process is proposed for the analysis of camphorated oil in order to detect arid estimate synthetic camphor oil of turpentine and essential oil of camphor any or all of which may b e present in addition to natural camphor. Ten C.C. of the camphorated oil are placed in a small retort and heated to a tempera- ture of 200° by means of an oil-bath; a current of dry carbon dioxide is passed through the retort during the heating and the neck of the retort is connected with a weighed two-necked Woulff’s bottle.A weighed small spiral glass condenser is fitted in the second hole of this bottle. The volatile substances collecting in the neck of the retort the Woulff’s bottle and condenser are weighed (the neck of the retort being cut off for this purpose) dissolved in carbon tetrachloride and the solutions are united. This solution is weighed and portions are used for the eqtimation of the specific gravity iodine number and polymerisation with sulphuric acid. The last-mentioned process is carried out by treating a weighed portion of the solution (about one-half) with one-third of its volume of sulphuric acid D 1.84 the acid being added slowly and the mixture cooled.The latter is then distilled in steam and the carbon tetrachloride solution separated from the distillate and weighed; the change in specific gravity and refractive index of the solution enables the weight of the tobal camphors 6s be calculated whilst the polarimetric reading in a when washing the apparatus. L. DE x.ANALYTICAL CHEMISTRY. 103 200 mm. tube multiplied by 0.22385 and divided by 0.95 gives the amount of natural camphor. This deducted frdm the total camphors leaves the quantity of synthetic camphor. When the amounts of tho two camphors are known the iodine number due to turpentine and oil of camphor enables the percentages of the last-named substances to be calculated.The.following iodine numbers are assumed in the calculation natural camphor 5 ; synthetic camphor 7 ; essential oil of camphor lS0 and oil of turpentine 370. w. P. s. Gasometric Estimation of Urea. A. RONCHBSE (J. Pharm. Chim. 1908 28 444-452 ; Bull. SOC. chirn. [iv] 3 1135-1141).- Ten C.C. of urine are mixed with 10 C.C. of dilute solution of basic lead acetate (100 C.C. officinal solution and 150 C.C. of water) and 10 C.C. of the filtrate which is now free from uric acid are treated in the usual manner with sodium hypobromite. A check experiment is then made by means of 5 C.C. of a 2% urea solution and the volumes of nitrogen evolved are compared. If the urine contains sugar a judicious amount of dextrose should be added to the urea solution.If ammonium salts are present these evolve nitrogen which would count as urea. I n accurate investigations it is therefore necessary to estimate the ammonia separately and to calculate the result into urea. The following method is recommended 10 C.C. of urine are diluted to 100 C.C. with water free from carbon dioxide and a few drops of phenolphthalein are added. AT/10 sodium hydroxide is slowly added until the liquid turns a pale rose colour. Twenty C.C. of neutralised formaldehyde (1 1) are now added and the solution is titrated with N/10 sodium hydroxide until the liquid turns pink. A correction is made by adding 0.1 C.C. to every 3 C.C. of alkali used. The sodium hydroxide represents the ammonia. L. DE K.Estimation of Urea in Urine. STANLEY R. BENEDICT and FRANK GEPHART (J. Anzer. C?Lem. Xoc. 19OS 30 1760-1764).-The following process is recommended five C.C. of urine are placed in a test-tube and mixed with 5 C.C. of dilute hydrochloric acid (1 4). The mouth of the tube is covered with a cap made of lead foil and it is then placed in a small autoclave which is heated for nn hour and a half at 150-155'. When cold the contents of the tube are transferred to a 800 C.C. Ejeldahl distilling flask 400 C.C. of water and 20 C.C. of 10% sodium hydroxide are added and the ammonia formed is distilled and titrated as usual. Allowance should be made for pre-existing ammonia. L. DE K. Application of the Carbamino-reaction V. HANS LIEBERMANN (Zeitsch. physiol. Chem. 1908 58 84-91).-Various basic substances in solution were treated in the cold with lime-water and carbon dioxide alternately in the presence of phenolphthalein the gas being passed into the solution until the indicator was only faintly coloured. The filtered liquid was afterwards warmed to 20' with excess of lime water (and in a few cases boiled) and the calcium carbonate thus precipitated was weighed and the nitrogen in the filtrate estimated.In this way the ratio N CO and the degree of completeness of the104 ABSTRACTS OF CHEMICAL PAPERS. carbamino-reaction are ascertained. The ratios found were as follows for piperazine 0.95 piperidine 0.93 coniine 1.56 tetrahydroquinoline 4.07 tryptophan 1 -55 glucobamine 0.99 taurine 0.93. Estimation of Vegetable Alkaloids by means of Mercuric Potassium Iodide [Mayer’s Solution].GUNNAR HEIKEL (Chenz. Zeit. 1908 32 1149-1151 1162-1163 1186-1187 1212-1213). -Mayer’s reagent (6.775 grams of mercuric chloride and 25 grams of potassium iodide per litre) h:ts been used for the estimation of alkaloids by simply adding it to the slightly acid solution so long as a pre- cipitate is formed. The author has sensibly improv’ed the method by adding a larger amount of the reagent than necessary and then titrating the excess with potassium cyanide the excess of which is in turn titrated with standard silver nitrate in presence of ammonia ; the potassium iodide serves here as an indicator. The author communicates a large number of expsriments showing the amount of Mayer’s reagent required to precipitate 0.1 gram of the more important pharmaceutical alkaloids. The results which in some cases depend somewhat on concentration and acidity are not scienti- fically accurate but sufficiently near the truth for pharmaceutical purposes. When applying the process t o drugs these are as a rule extracted with chloroform in presence of ammonia. The solvent is then evaporated and the residue taken up with water acidified with sul- phuric acid and the acid solution is titrated as directed. I n other cases they are simply treated with acid water. Plants are extracted with a solvent immiscible with water and this is lhen shaken with acid water. For exact details of working in the assay of the various drugs and their preparations the original paper should be consulted. S. B. S. L. DE K. Behaviour of‘ Gelatin and Proteose to Bromine Water. ERNST SALKOWSKI (Zeitsch. pl,pioZ. Chem. 1908 57 526-528).- The addition of bromine water to a solution of gelatin causes the formation of a sticky precipitate which has been described as characteristic. h solution of proteose (or gelatose) behaves however in the same way. W. D. H. Diazo-reaction of Atoxyl. PAUL EHRLICH and ALFRED BERTHEIM (Chem. Zeit. 1908 32 1059).-The authors call attention to the fact that the reaction recently described by Covelli (Abstr. 1908 ii 1000) had been described in the literature previously; compare Ehrlich and Bertheim (Abstr.. 1907. i. S12) Blumenthal and Herschmann (Abstr. 1908 (1 613),’and Bertheim’ (Abstr. 1908 i 591). J. V. E. 9