ii. 146 ABSTRACTS OF CHEMICAL PAPERS Analytical Chemistry. Use of Centrifugal Apparatus in Chemical Analysip. OTTO NOLTE (Landw. Versuchs-Stat. 1915 87 449-457).-The author gives many instances in which the centrifugal apparatus possesses advantages over the filter f o r the separation of precipitates from liquids; in many cases precipitates can be much more rapidly and accurately separated by centrifugal action than by filtration. A centrifugal apparatus is described in which the separations are effected in small silica beakers o r other vessels suitable mountings being provided f o r these in the centrifugal drum. New Form of Gas Burette. HARXRIERMANX ( C ' h e m Zezt. 1916 40 84).-A gas-measuring burette) and a levelling tube are con- nected at their lower ends with a Y-tube the stem of which is fitted into one neck of a three-necked Woulfe's bottle; the centre opening of the latter carries a three-way tap connected with a caoutchouc pressure bulb and a thermometer is fitted into the other neck. The Woulfe's bottle is filled with the absorption solution; this is forced up into the burette and levelling tube by means of the pressure bulb; the tap atl the top of the burette is then closed and the level of the solution is adjusted by means of the tap on the centre neck of the bottle( this also serving for the escape of air from the bottle when the gas under examination is admitted t o the burette.W. P. S. A tubulure is provided a t the bottom of the bottle. w. P. s. Potassium Hydrogen Carbonate as a Staadard in Acidi- metry and Alkalimetry. G.INCZE ( Z c i t . anal. Chem. 1915 54 585-602) .-Potassium hydrogen carbonate is most readily pre- pared by passing a current of carbon dioxide into an alcoholic solution of potassium hydroxide ; the salt is then recrystallised by treating its aqueous solution with alcohol. It should be stored in well-closed perfectly dry bottles and under these conditions does not undergo any change; in the presence of moisture how- ever it loses carbon dioxide. On account of its high molecular weight and thk readiness with which i t may be obtained in a pure condition the salt forms a suitable basis for standardising acid solutions. w. P. s. Estimation of Oxygen (in Water) by Winkler's Method. 11. G. BRUHNS (Chem. Zeit. 1916 40 45-46 71-73. Compare this vol. ii 47).-The following modifications of this method are recom- mended Manganous sulphate solution (1 1) is used in place of manganous chloride solution and potassium hydroxide-sodium hydroxide solution (1 1 2) in place of sodium hydroxidepotassium iodide solution.After the water has been treated with 10 drops of each of these solutions and the precipitate has settled potassium iodide is added either in the form of crystals o r as a concentratedANALYTICAL CHEMISTRY. ii. 147 solution. The mixture is then acidified with dilute hydrochloric acid (1 1) o r sulphuric acid (1 3) a de'finite volume of the water is withdrawn from the flask by means of a pipette and the remain- der is titrated with thiosulphate solution in the usual way. The influence of nitrites etc.may be eliminated by adding solid potassium hydrogen carbonate after the manganese oxide precipi- tate has settled collecting the precipitate on a filter and t'hen treating i t with potassium iodide and acid. If the water contains ferric salts phosphoric acid should be used instead of hydrochloric acid f o r acidifying the mixture; the water should be treated previ- ously with permanganate if it contains ferrous salts. W. P. S. Estimation of Sulphuric Acid in the Presence of Phosphoric Acid. TH. VON FELLENBERG (Chenz. Zeizt7*. 1915 jj 1262 ; from M i t t . Lebensmittelunters. Hyg. 1915 6 191-195).-The author confirms a statement by Baragiola and Godet that the results obtained in the estimation of sulphuric acid are too high if phos- phoric acid is also present.The error may be reduced to a mini- mum if the barium chloride solution is added drop by drop in not too great excess t o the boiling sulphuric acid or sulphate solution which should contain about 0.13 C.C. of concentrated hydrochloric acid per 100 C.C. w. P. s. The Volumetric Estimation of Total Sulphur and Sulphates in Small Quantities of Urine. JACK CECIL DRUMMOND (Bi0~7zem. ,7. 1915 9 492-507).-The benzidine method of Rosenheim and Drummond (A. 1914 ii 455) for the estimation of urinary sul- phat,es can be carried out with as little as 2 C.C. of urine by titrating the precipitated benzidine sulphate with 0.02N-alkali provided certain details of technique especially in the' matter of filtration are followed. The method is also applicable to the estimation of total sulphur in small quantities of biological material by applying the precipitation of the benzidine sulphate to the solution of salts obtained after oxidation of the material by Benedict's method.W. D. H. The Analysis of a Mixture of Alkali Sulphides Thio- sulphates and Dithionates. J. A. MULLER (B2tl.Z. SOC. chinz.. 1916 [iv] 19 8-9).-Any sulphate present is first estimated by precipitation with barium chloride in a very dilute solution slightly acidified with acetic acid. The sulphur present as sulphide and thiosulphate is estimated by titrating an aliquot portion suitably diluted and feebly acidified with acetic acid with a standard iodine solution. Another portion is acidified with acetic acid and the hydrogen sulphide removed by evacuating the flask for ten minutes by means of a water-pump the residual liquid being titrated with the standard iodine solution.From a fourth portion of the liquid the sulphide is removed as described and the liquid then made alkaline with potassium hydroxide transferred to a silver capsule and evaporated to dryness with a little potassium nitrate the residue being gradually raised to a dull red heat. I n this residueii. 148 ABSTRACTS OF CHEMICAL PAPERS. the sulphate is estimated in the usual way. Owing t o the presence of an excess of alkali sulphides the only thionates likely t o be presentl are the dithionates and thus from the above data the neces- sary percentages can be calculated. Simple Method for the Estimation of Nitrogen in Urine. VICTOR C. MYERS (Chem. Zentr.1915 ii 858; from Munch. med. Tt'och. 1915 62 1076-1077).-A modification of Folin and Farmer's method (A. 1912 ii 702) is described. The colorations are compared in an Aut,enrieth-Konigsberger colorimeter (A. 1910 W. G. ii 910). w. P. s. A Modification of the Lunge Nitrometer. W. SENFTEN ( Chenz. Zeit. 1916 40 39-40).-1n order t o obviate the necessity of dissolving the substance in the cup of the nitrometer in its usual form the apparatus is modified by connecting the upper stop-cock with the decomposition tube with a ground glass joint. The stop- cock itself is triple bored and thO nitric oxide is measured externally by transference t o a gas burette. The substance is dis- solved in sulphuric acid in the decomposition tube in a vacuum if necessary after adjusting the mercury level to the stop-cock a t the base of the tube which is then turned off.When solution is complete the mercury is slowly admitted into the tube thus expel- ling the air through the upper stop-cock and the decolmposition is then conducted in the usual way. A special reversible gas-measur- ing burette is also described having a tap a t each end an snlarge- ment a t one end and graduate'd through its entire length. I f after absosption of one constituent of a gas mixture the remaining volume is so small that it is contained entirely within the enlarged bulb the burett.e is simply reversed and the volume can then be read off on the second reversed scale. G. F. M. Detection of Arsenic in Beverages. L. VUAFLART ( A m . 'Fnlsif. 1915 8 414-415).-The author states t h a t 1 mg.of arsenic per litre of liquid can be detected by means of Bougault's reagent.. To detect arsenic in beer 250 C.C. of the sample are treated with a few drops of b,romine filtered after twenty-four hours and the filtrate then mixed with 1 C.C. of saturated sodium phosphate solution 5 C.C. of magnesia mixture and 80 C.C. of ammonia. After twenty-four hours the precipitate is collected dissolved in 20 C.C. of nitric acid (I 4) the solution treated with 2 C.C. of 20% magnesium nitrate solution and evaporated to dryness. The residue is then ignited and heated with 10 C.C. of t h e reagent A similar process is used for the detection of arsenic in wines. w. P. s. Estimation of Arsenic in Urine Blood and Animal Organs. WILHELM KOSIAN (Chem. Zentr.1915 ii 1311 ; from P71nrm. Post 1915 48 321).-One hundred and fifty grams of urine blood or serum are evaporated t o a syrup and 50 grams of this or the same quantity of finely divided animal substance etc. are warmed withANALYTICAL CHEMISTRY. ii. 149 20 C.C. of hydrochlmic acid and a quantity of potassium chlorate; ths excess of chlorine is then expelled 10 grams of ferrous sulphate are added the mixture is transferred t o a distillation flask and diluted to 200 C.C. with hydrochloric acid. The cold mixture is saturated with hydrogen chloride and then distilled in a current of the same gas until 80 C.C. of distillate have been collected in a receiver containing 50 C.C. of water. The distillate is now diluted to 200 c.c. and 10 C.C. of i t are treated with 3 C.C.of concentrated hydrochloric acid methyl-orangs is added and the mixture is titrated with. N / 10- or N / 100-potassium bromate solution until it Formulae for Use in the Indirect Analysis of Generator Gas. FRITZ HOFFMANN (Chew. Zeit. 1916 40 81-82).-The following formulze are given f o r calculating the proportions of carbon mon- oxide hydrogen and methane in generator gas after the carbon dioxide and oxygen have been removed. Letl V bel the total volume of the three gases K the volume of the carbon dioxide formed when they are burnt together C the contraction in volume S the volume of oxygen required for the combustion then Being given ; V K C. K C S. c s v. S V K . is colourless (compare Gyory A. 1893 ii 554). w. P. s. A. B. C. D. CO = 2K/3+ V-2C/3 4K/3+CI3-rS 4V/3+S/3- C K-22513- V/3 Hg z= V-K C-Is c-s V-K CH,= 2K/3+2C/3- V S-KJ3-CJ3 22513- V13 2S/3-V/3 The formulze A and I3 are principally used since the values C and h' are determined directly in the course of the analysis.W. P. S. Fractional Combustion of Gases over Copper Oxide. E. TERRES and E. MAUGUIN (Chem. Zentr. 1915 ii 1309; from J . Gasbeleucht. 1915 58 8-ll).-Hydrogen is oxidised completely when passed over copper oxide a t 250-300° but carbon monoxide is oxidised a t 305O only to the extent of 90-94%. A mixture of equal volumes of hydrogen and carbon monoxide is burnt com- pletely a t 300'. Acetylene and ethylene even when mixed with hydrogen do not busn quantitatively a t 300O; a quantity of carbon separates. Benzene behaves in a similar manner.Methane begins to burn atl 310° and the presence of hydrogen lowers this tempera- ture by a few degrees. Pure methane is oxidised completely a t a red-heat and the presence of hydrogen is without influence. A mixture of carbon monoxide and methane cannot be separated by fractional combustion but this is possible in the case of a mixture consisting of equal volumes of carbon monoxide and hydrogen on the one hand and methane on the other. I n using the Drehsclimidt- Ubbelohde apparatus notice should be taken of the fact that at high temperatures copper oxide loses oxygen and that this may lead to error. w. P. s. A New Method for the Simultaneous Estimation of Carbon Hydrogen and Mercury in Organo-mercuric Compounds. V. GRIGNARD and A. ABELMANN (Bzcll. SOC. chim. 1916 [iv] 19 25-27).-The ordinary combustion tube A is modifisd as shownii.150 ABSTRACTS OF CHEMICAL PAPERS. in the sketch a t the) exit end. It is slight'ly constricted a t e just before i t leaves the furnace and then is allowed t o project 20 cm. beyond this. I n this external portion is fitted the tube B held in position a t the open end by a rubber stopper and a t the constric- tion by a cone of calcined asbestos. B contains a spiral of gold wire and is weighed before and after the experiment and during the combustion and is kept slightly heated to prevent t'he condensation of water. Ro'und the tube B by means of the narrow tube t is kept passing a slow current of dry oxygen a t a pressure slightly higher than that existing in the major part of A in order to prevent any leakage of water or mercury vapours through the asbestos stlopper a t e.The carbon dioxide and water vapour are collected and weighed as usual. Can Carbon Dioxide in Sea Water be Directly Estimated b-y Titration ? SERGIUS MORGULIS and EVERETT W. FULLER (J. Biol. Chem. 1916 24 31-35).-1n recent investigations of the metabolism of marine organisms Moorre and his collaborators have estimated the amount of carbon dioxide produced in the water by titration with a centinormal solution of an alkali. The authors believe that this method is so inaccurate as t o invalidate the conclu- sions based on its results and quote results of experiments which in their opinion show that only approximately one-third of the carbon dioxide can be determined under such conditlions.They also suggest that an indirect method of estimation based on precipitation of the carbon dioxide with N / 5-barium hydroxide solution and the titration of the excess of this reapent with NIlO- W. G. or AT/ 5G-acid is susceptible' of development t o greacer accuracy.' D. F. T. A Copper Cathode and Iron Anode in the Electro-analysis of Brass. J. GUZMLN and J. M. FERNANDEZ LADREDA (A?tuZ. Fis. Quim. 1915 13 308-315).-An account of t.he results obtained in the electrolytic analysis of cartridge cases. A. J. W. Analysis of Copper-Aluminium-Zinc Alloys. H. GRAEFE (Chem. Zeit. 1916 40 lO2).-The following method is suggested for the analysis of alloys consisting approximately of copper 25-27% aluminium 14-18% and zinc 54-57% together with traces of silicon cadmium tin lead and iron.A quantity of 0.5 gram of the turnings is dissolved in 15 C.C. of nitric acid (D 1-4) the solution is diluted and metastannic acid is separatd by filtration. The sodution is then dilubd t o 150 c.c. heated toANALYTICAL CHEMISTRY. ii. 151 60-70° and the copper and lead deposited electrolytically; with a current of 0.2-0.4 ampere and 2-3 volts these metals are deposited within forty-five minutes whilst the zinc remains in solution. The solution is next partly neutralised with solid potass- ium hydroxide aluminium zinc and iron are precipitated by the addition of Eodium carbonate the precipitate is collected washed with hot water and dissolved in a very small quantity of dilute (I :4) sulphuric acid. The solution is treated with an excess of potassium hydroxide and the cleas liquid (except for a small precipitate of ferric hydroxide) measuring about 150 C.C.is sub- mitted to electrolysis. The copper-coated cathode should be rotated a t 600-800 revolutions per minute and the current increased from 1 to 5 amperes in steps of 1 a.mpere every ten minutes. The zinc is deposited in forty minutes and without interrupting the current is washed dried with alcohol and weighed. The electrolyte con- tains the iron and aluminium and these two metals are then separated in the usual way. w. P. s. Sensitive Clinical Method for the Detection of Mercury in Urine. 31. PERELSTEIN and J. ABELIN (Chem. Zmzlr. 1915 ii. 1027; from Munch. med. Woch. 1915 62 1181-1183).-Five hundred C.C. of the urine are treated with 10 C.C.of concentrated hydrochloric acid boile'd f o r one minute cooled 6 C.C. of ammonia 25 grams of sodium acetate and 10 C.C. of 10% ferric chloride solution are added and the mixture is again boiled. The precipi- tate formed is collected on a filter while the solution is still hot washed with a small quantity of hot water then dissolved in con- cenhrated hydrochloric acid and the mercury deposited on copper in the usual way. The deposited mercury may be identified by Qualitative Analysis of the Iron Group in the Presence of Phosphates. ROBERT GILMOUR (Chem. News 1916 113 1-3 ; 13-15).-The solution from which the metals of the silver and copper groups have been removed is boiled to expel hydrogen sulphide and the alkaline earths are precipitated as sulphates by means of sulphuric acid and alcohol.The precipitated sulphates are converted into carbonates and analysed according t o a method described previously (A. 1915 ii 486). The filtrate from the sulphates is evaporated t o expel alcohol sodium hydroxide is added followed by an excess of sodium peroxide the mixture is boiled for some minutes and filtered. The filtrate contains aluminium chromium and zinc a-nd most of the phosphate; the precipitate consists of hydroxides and phosphates of iron nickel cobalt man- ganese and magnesium. After the precipitate has been dissolved in hydrochloric acid the solution is evaporated t o dryness the residue dissolved in water and iron and phosphate separated by the basic acetate process. The filtrate from this precipitate contains nickel cobalt manganese and magnesium ; it is boiled treated with hydrogen sulphide and the nickel and cobalt sulphides are separated bj filtration.The filtrate is rendered ammoniacal the converting i t into mercuric iodide. w. P. s.ii. 152 ABSTRACTS OF CHEMICAL PAPERS. manganese is separated as sulphide and tl10 magnesium subse- quently as snimonium magnesium phosphate. w. P. s. Rapid Estimation of Iron in Presence of Organic Substances F. FERRARI (Ann. Chim. Applicata 1915 4 341-343).-The pre- liminary destruction of the organic matter usually effected when iron is t o be estimated in solutions containing hydroxylic organic acids polyhydric alcohols sugars etc. may be avoided by the following melhod. To the solution which is made up to a t least 150 C.C.and contains hydrochloric acid in excess (10.15 C.C. of the concentrated acid) chlorine water is added in small po,rtions until a faint odour of the halogen persists. A cold clear 6% solution of " cupferron " is then added drop by drop in amount a t least one-fifth greater than that necessary for the complete precipitation of the iron the liquid being kept in motion meanwhile. After fifteen t o twenty minut'es when the pulverulent precipitate has mostly collected into large friable crystalline masses these are broken up with a glass rod and the liquid filtered by aid of a filter pump the suction at first gentle being continued until the precipitate is dry when i t is washed twice with cold water acidified with hydrochloric acid and afterwards with water alone until the filtrate loses its acidity; i t is then treated several times with ammonia solution and finally washe$ again with water.After thorough drying and cautious heating t o redness in a covered plakinum crucible the precipitate is calcined in the air and the residual ferric oxide weighed. The whole of the opesations require about one and a-half hours. The method gives exact results but is not applicable in presence of copper silver lead mercury tin bismuth titanium or zirconium all these giving insoluble salts with " cupferron." T. H. P. Separation and Estimation of Nickel. T. L. WALKER (Tyans. Roy. SOC. Canada 1915 [iii] 9 93-97).-The following method has been investigated far the separation of iron from nickel and cobalt The hot solution containing iron nickel cobalt copper and lead chlorides is oxidised with a few drops of hydrogen peroxide in order t o convert all the iron into the ferric condition.After nearly neutralising with sodium carbonate xed lead is added gradually to1 the boiling solution until i t is seen t o be in excess by the distinct re'd colour of the precipitate which is heavy and granular. The separation of the iron is complete in a few minutes. After collecting the precipitate the lead and other metals are precipitated f8rom the filtrate as sulphides after which the fresh filtrate is concentrated tlo 100 c.c. 3 C.C. of concentrated sulphuric acid and 30 C.C. of strong ammonia added and the nickel and cobalt deposited on a revolving cathode. It is found that nickel and iron can be separated satisfactorily in this way but not cobalt and iron there being about a 10% loss with cobalt. Litharge does not give1 any better result with cobalt than does red lead.ANALYTICAL CHEMISTRY.ii. 153 Neumann's method (Chem. Zeit. 1901 25 731) fw the separa- tion of nickel from iron in steel and matte is found to be unsatiS- factory. Cobalt may be deposited successfully from chloride solutions under the following conditions To the solution of cobalt chloride containing 0.13 gram of cobalt are added 3 C.C. of concentrated sulphuric acid 30 C.C. of strong ammonia solution and water to 150 C.C. Deposition is complete in twenty minutes a t a revolving cathode using 4 amps. and 5 volts. The accuracy of the dimethylglyoxime method for nickel is about the same as the red lead method.T. S. P. Rapid Analysis of Commercial Antimony Sulphide. A. HUTIN (,4nn. ChZm. aital. 1916 21 3-7).-Moisture and calcium sulphate are estimated by the usual methods and free sulphur by extracting the sample with acetone. Commercial antimony sulphide consists chiefly of a mixture of the tri- and penta-sulphides; these are converted into antimony tetroxide when ignited j n the presence of mercuric oxide and their quantity can be calchlated from the weight of tetroxide obtained. The total quantity of antimony can also be estimated by reduction in a current of hydrogen. Samples which contain antimony tetroxide are not completely soluble in JAROSLAV M~LBAUER ( C h r n . Zeit. 1916 40 l08).-Antimony glass consists essentially cf antimony oxide (90.5%) and antimony sulphide (8-7%) and these two constituents may be estimated as follows 1 gram of the finely divided material is boiled f o r five minutes with 100 C.C.of dilute hydrochloric acid (1 :4) the liquid is filtered through asbestos the insoluble portion washed first with the dilute acid then with water dried and weighed. The insoluble portion is now washed with a mixture of concentrated t'artaric acid solution and dilute hydrochloric acid and after the total filtrate has been neutralised with sodium hydroxide sodium hydrogen carbonate is added t o the solution and the latter titrated with N/lO-iodine solution. The1 result gives the quantity of antimony oxide present. The insoluble residue of antimony sulphide is then dissolved in concentrated hydrochloric acid the solution boiled t o expel hydro- gen sulphide and titrated with / 10-iodine solution undm the same conditions. w.P. s. boiling hydrochloric acid. w. P. s. Rapid Analysis of the So-called Antimony Glass. Estimation of Titanium in Ferrotitanium. G. EOHL (Chem. Zeit. 1916 40 105-106).-Gooch7s method as modified by Thorn- ton (A 1912 ii 1000) was found to yield concordant results but these were lower by about 0.8% than those obtained by volumetric methods. Of the latter thO methylene-blue process (A. 1913 ii Estimation of Methyl Alcohol in Ethyl Alcohol. G. REIF (Chem. Zentr. 1915 ii 1056-1057 ; from Arbeit. H. Gesundh.- *4mt. 1915 50 50-56).-The method depends on the fact that VOL CX. ii. 7 1079; 1914 ii 73) is rapid and accurate. w. P. s.ii. 154 ABSTRACTS OF CHEMICAL PIPEKS.methyl iodide reacts with methyl sulphide in the cold to form tri- methylsulphonium iodide according to the equation Me1 + SMe = S1Lle31 wliilst the cosresponding ethyl compound is not formed a t the ordinary temperature if the quantity of methyl iodide in the mixtiire of methyl and ethyl iodides is not too small but in this case its formation can- be prevente'd by the addition of ether. The trimethylsulphonium iodide behaves as a salt of hydriodic acid and can be estimated volumetrically. The mixture of the alkyl iodides is obtained as described by Wirthle (A. 1912 ii 607) and the specific gravity of the mixture will indicate approximately the quantity of methyl iodide present. Two C.C. of the iodide mixture are placed in a 20 C.C.flask and if the quantity of methyl iodide exceeds 5% 2 C.C. of methyl sulphide are added; otherwise only 1.5 C.C. of methyl sulphide and 0.5 C.C. of dry ether are added. The flask is closed placed in a desiccator a t 20° for twenty-four hours the trimethylsulphonium iodide is then rinsed on t o a dry filter with dry ether and washed with the same. The crystals are now dissolved in water and titrated with silver nitrate solution. One mol. of silver nitrate corresponds with 1 mol. of methyl alcohol. F o r the estimation of methyl alcohol in brandy a preliminary test is first made as described by Wirthle (Zoc. czt.) t o detect the actual presence of this alcohol. One hundred C.C. of the sample are then treated with 10 C.C. of N/1-sodium hydroxide solution and distilled in an apparatus fitted with a Le Be1 still-head the quantity of distillate collected being 10 C.C.more than the amount of alcohol present. The quantity of total alcohol is now estimated in the distillate and from the result is calculated the number of C.C. of the distillate corresponding with 10 C.C. of alcohol. This quantity of the distillate is used for the preparation of the mixed iodides; the boiling point observed during the distillation will indicate the proportion of methyl alcohol present and consequently the quan- tities of iodine and phosphorus t o be used but an excess of 1 gram of iodine per 10 C.C. of the alcohols is without influence. When large quantities of methyl alcohol arel present i t is recommended that the boiling point of the brandy should be closely observed so that a distillate containing approximately 10% of methyl alcohol may be obtained.The mixed iodides are then treated as above described. w. P. s. Estimation of Methoxyl in Compounds Containing Sulphur. ALFRED KIRPAL and THEODOR BUHN (Monatsh. 1915 36 853-863. Compare A. 1914 ii 497).-The authors describe improvements in their modification of the Zeisel method for the estimation of methoxyl. They find that i t is not necessary to use hydrogen and t h a t a current of carbon dioxide gives satisfactory results. The pyridine to absorb the methyl iodide is now enclosed in two small bubbling tubes wholly composed of glass the only necessary con- nexion being for the attachment to the remainder of the apparatus. After evaporating off the pyridine in a dish the remaining meth- iodide is dissolved in w a k and titrated with N / 1 0 silver nitrate solutioa using sodium chromate as an indicator.ANALYTICAL CHEMISTRY.ii. 155 With this modification of the Zeisel process there is no need for a pure hydriodic acid because the common impurities for example hydrogen sulphide and phosphine do not affect the pyr- idine. The new method is also applicable t o the estimation of the methylimide group but is of no value for the estimation of ethoxyl as ethyl iodide is only partly absorbed by the pyridine. This method is also satisfactory f o r methoxyl estimations in sulphur compounds but here the advantage over the Zeisel process is not so great as a t first appears because the latter can be made t o give accurate results if the mixture of gas and vapour from the digestion flask is first passed through a slightly acidified solution of cadmium sulphate before reaching the silver nitrate solution.I n this case any hydrogen sulphide is precipitated as cadmium sulphide. The unf avourable and inaccurate results obtained by Benedikt and Bamberger (A. 1891 1296) who endeavoured to effect a similar result by the use of cadmium iodide solution is explained by the fact t h a t with this reagent the cadmium sulphide produced tends t o remain in part in colloidal solution in which form i t is more active and reacts with the methyl iodide giving methyl mercaptan. Experimental results are given indicating the applicability of the Zeisel method with the use of cadmium sulphate t o the estimation of methoxyl and ethoxyl groups in compounds containing sulphur in various states of combination.Detection of Glycerides by the Magenta-Sulphurous Acid Reagent. MAURICE FRANSOIS ( J . Phnrm. Chim. 191 6 [vii] 13 65-77).-A modification of a test for glycerol described recently by Franqois and Boismenu (A. 1915 ii 110) may be used for the detection of glycerides in the presence of paraffins waxes resins etc. The substance mixed with sand is heated in a test-tube for ten minutes nearly t o redness and the vapours produced are con- ducted into another test-tube containing a quantity of the magenta reagent (sulphurous acid 220 c.c. 0.1% magenta solution 30 c.c. and concentrated sulphuric acid 3 c.c.) ; white fumes produced a t the commencement of the heating should be prevented from coming into contact with the reagent by removing the source of heat for a few moments a t a time.The tesbtube containing the reagent is then heated in a boiling water-bath for fifteen minutes. If glycerides are present in the substance tested the reagent is coloured red while cold and the coloration changes to blue on heating; paraffins waxes resins and caoutchouc yield vapours which give a red coloration with the cold reagent but the latter becomes colourless when heated. The author also discusses the mechanism of the formation of the red and blue colorations. G. ADANTI (Boll. Chin2. Farm. 1916 55 33-35).-The method described by the author is identical with Bertrand’s method (A. 1907 ii 136). D. F. T. w. P. s. Indirect Estimation of Sugars.T. H. P. The Fehling-Soxhlet Method of Estimating Sugar. RUOSS (Zeitsch. anal. Chem. 1916 56 1-23).-The author gives 7-2ii. 156 ABSTRACTS OF CHEMICAL PAPERS. tables sliowiiig the quantities of sugar solutions (containing from 0.1 t o 1% of reducjiig sugar) required t o seduce 10 C.C. of Fehling’s solution. I n the titration of Fehling’s solution with the sugar solution the end-point is found by filtering the solution after the bluish-green coloration has disappeared acidifying the filtrate with acetic acid and testing with potassium ferrocyanide solution. If a brown coloration is obtained the titration is repeated adding more of the sugar solution until the filtered solution ceases t o give a coloration with f errocyanide. If the sugar solution contains nitro- genous substances the end-point is ascertained by acidifying the solution with acetic acid containing sodium chloride and adding a small quantity of potassium thiocyanate. The cuprous salt forms a precipit’ate whilst any cupric salt remains in solution and gives a coloration on the addition of ferrocyanide solution.I n solutions containing from 0.01 t o 0.1% of reducing sugar 10 mg. of sodium potassium tartrate should be added to the sugar solution in order to prevent the precipitation of cupric oxide from the Fehling’s solution. Iodometric estimation of the excess of cupric salt remain- ing after Fehling’s solution has been partly reduced by a sugar solution yields untrustworthy results owing t o the action of oxidised sugar compounds on the liberated iodine.w. P. s. Comparative Action of Sucrose and Invert-sugar on the Cupropotassic Solution. L. MAQUENNE (Coin@. !~-eiad. 19 16 162 145-149. Compare this vol. ii 56).-A study of the varia- tion of the reducing powers of invert-sugar and sucrose with varia- tions in temperature and duration of heating. Starting from 70Q there is a very slight steady rise in the reducing power of invest- sugar with rise in temperature but f o r sucrose the increase is much more marked particularly from 90° to looo. Working with mixtures of these sugars the most satisfactory temperature is 70° where under the given conditions the reducing power of invert- sugar is eight to ten thousand times greater than that of sucrose. For temperatures above 6 5 O the time of heating has very little effect on thel reducing power of invert-sugsr but in the case cf sucrose the duration of heating has a very marked influence on the reducing power which f o r temperatures in the neighbourhood of 70° is almost exactly proportional to the time of heating.W. G. The Estimation of Reducing Sugars In the Presence of an Excess of Sucrose. L. MAQUENNE (Compt. rend. 1916 162 207-213. Compare Saillard this vol. ii 55).-The author has studied the effect of temperature and duration of heating on the estimation his conclusions being in agreement with those of Saillard (7oc. cit.). He has also examined the effect of varying the amount of sucrose present on the estimation of small quantities of reducing sugars. The results show that in the) case of products poor in invert- sugar 20 grams of sugar should be used when working with an external temperature of 6 5 O o r looo and 10 grams when working with boiling solutions the total volumes of liquid used being 42.5 C.C.and 36.2 C.C. respectively. The method of estimationANALYTICAL CHEMISTRY. ii. 157 should be adapted to the case under examination using the metho,d based on the separation of the cuprous oxide f o r products rich in invert-sugar and the thiosulphate method f o r products poor in invert-sugar. It is necessary to carry out the reduction as soon as possible after the sugar is dissolved and t o titrate immediately after the reduction and cooling. W. G. Use of Yeast for tbe Invereion of Sucrose and the Estima- tion of the Latter Polarimetricallg or Chemically.H. PELLET (BuZZ. A ssoc. C'him. Sucr. Dist. 1915 33 29-30).-The addition of sodium salicylate prevents alcoholic fermentation when sucrose is inverted by yeast and since larger proportions of yeast may b0 used the time required for the inversion is shortened. The results obtained in the estimation of sucrose in molasses when the sugar lias been inverted with yeast treated with sodium salicylate agree with those found by the ordinary inversion method. W. P. S. Preparation of Very Active Salioylated Yeast for the Inversion of Sucrose. H. PELLET (Bdl. Assoc. Chinz. Suer. Dist. 19 15 33 30-33).-Yeast is liquefied immediately when treated with 756 of its weight of sodium salicylate; the solution thus obtained retains its activity (inverting power on sucrose) f o r several months and a quantity of it corresponding with 3 grams of yeast is capable of inverting 3.5 grams of sucroge within twenty minutes. w.P. s. Uee of Different Substances in the Preparation of Invertase Solution. Possible Eources of Error. H. PELLET (Bzdl. Assoc. Chim. Sucr. D i d . 1915 33 33-34).-Thymol chloroform and formaldehyde have been used in the preparation of invertase solution from brewers' yeast; whilst the former is optically inactive and does not reduce Fehling's solution the last two in dilute solution alt'hough optically inactive reduce1 Fehling's solution. An error may therefore be1 introduced in the estimation of the cupric- reducing power of sugar solutions t o which invertase solution con- taining either of these substances has been added.Sodium sali- cylate is to be preferred t o any of the above substances in the w. P. s. preparation of invertase solutions. Estimation of Raffinose by Means of Different Yeasts. H. PELLET (Bit!!. ,4ssoc. Chim. Suer. Dist. 1915 33 41).-The different' actions of top and bottom fermentation yeasts on raffinose noted by Hudson (A. 1914 i 1147) have been investigated by the author and found t o be as stated; the method of estimating raffinose described by Hudson and Harding (A. 1915 ii 803) may be used far the analysis of molasses containing a small quantity of raffinose. It is mentioned that 1 gram of anhydrous raffinose when inverted with hydrochloric acid yields 0.765 gram of reducing sugar; inverted with top fermentation yeast i t gives 0.745 gram and with bottom fermentation yeast 0.965 gram of reducing sugar all expressed in terms of invert-sugar.w. 9. s.ii. 158 ABSTRACTS OF CHEMICAL PAPERS. Detection of Small Quantities of Oxalic Acid in Wine. HANS KREIS and W. I. BARAGIOLA (Chem. Zentr. 1915 ii 810-811 ; from Schweiz. Apoth. Zeit. 1915 53,397-400).-0xalic acid is added occasionally t o wine to decrease the quantity of calcium salts present and t o increase the " residual acidity"; even when the quantity of oxalic acid added is insufficient t o precipitate the whole of the calcium a small quantity of oxalic acid remains in solution owing t o the solubility of calcium oxalate in the presence of tartaric acid. To detect a small quantity of oxalic acid in wine 50 C.C. of the sample are heateld to boiling 3 C.C. of 5% calcium chloride solution are added and the mixture is rendered ammonia- cal; acetic acid is now added t o the hot solution until the latter is just' acid in reaction (an excess of acid must be avoided) the mixture is cooled submitted t o centrifugal action and the sediment examined under the microscope.Characteristic crystals of calcium oxalat'e will be observed if tlhe wine contained as little as 0.02 gram of oxalic acid per litre. Still smaller quantities (0.01 gram per litre) of oxalic acid may be de€ected by treating 50 C.C. of the wine (cold) with 2.5 C.C. of 5% calcium chloride solution 2.5 C.C. of acetic acid and 5 C.C. of saturated sodium acetate solution. After twenty-four .hours the mixture is shaken in order to disturb the sediment of calcium oxalate (if present); the greater part of the precipitated calcium tartrate adheres t o the bottom of the vessel.The liquid is now submitted to centrifugal action the sediment transferred to a small drawn-out tube again submitted t o centri- f ugal action and examined microscopically. The calcium oxalate contained in the cell tissue of grapes does not find its way into the) must o r wine. w. P. s. Acidimetric Estimation of Oxalic Acid Using Methyl- orange aa Indicator and its Estimation by t h e Iodide-Iodate Method. G. BRUHNS (Zeitsch. a?znZ. Chewz. 1916 55 23-51).- Oxalic acid may be titrated accurately with alkali solutions using methyl-orange as indicator provided that an equivalent quantity of calcium chloride1 solution is added when the acid has be'en almost completely neutralised. Without the addition of calcium chloride the end-point of the titration is not shasp owing to the tendency of the added alkali t o form hydrogen salts with the acid.The iodide-iodate method is not satisfactory when applied to oxalic acid; the liberation of iodine is extremely slow towards the end of the reaction and if calcium chloride is added the results obtained are slightly too low. w. P. s. Modification of the So-called Mohler's Reaction for Benzoic Acid. J. GROSSFIELD (Chem. Zentr. 191 5 ii 1313-1314 ; from Zeitsch. Nahr. Genussm. 1915 30 271-273).-Von der Heide and Jacob's method (A. 1910 ii 359) of carrying out Mohler's test (compare Robin A. 1908 ii 1078) is trustworthy but somewhat hdious and the reduction of t'he dinitrobenzoic acid with ammon- ium sulphide has little advantage.The author modifies the method as follows The dry alkali benzoate residue is mixed with 0.1 gramANALYTICAL CHEMISTRY. ii. 159 of potassium nitrate and 1 C.C. of concentrated sulphuric acid the mixture is heated for twenty minutes on a boiling water-bath then cooled diluted with 2 C.C. of water again cooled and treated with 10 C.C. of 15% ammonia and 2 C.C. of 2% hydroxylamine hydro- chlmide solution. A red coloration devedops more o r less rapidly according to the amount of benzoic acid present; the coloration develops more quickly if the mixture is heated and reaches it6 greatest intensity when the mixture is cooled. I n the colorimetric estimation a standard iron solution (containing 1 mg. of Fe,03 per c.c.) mixed with thiocyanate solution may be used for the com- parison.w. P. s. Analysigl of Medicated Drinks. B. H. ST. JOHN ( J . SOC. Chem. Itzd. 1916 35 68-69; from J . Assoc. Off. Agric. Chem. 1915 1 343-345).-The acidified sample is extracted with a mixture of chloroform and alcohol and the acids present are transferred to aqueous solution by washing the extract with sodium carbonate solution. Caffeine if present is then estimated by evaporating the washed chlcrof o m solution to dryness. I n the sodium carbonate extra& salicylic acid is estimated by adding an excess of iodine and after warming for an hour collecting and weighing the precipitate; its weight multiplied by 0.4657 give6 the amount of sodium salicylate present. Saccharin ” and bennoic acid which are also present in the sodium carbonate extract are estimated the one by conversion into its silver salt and the other by eteam distillation.Phosphoric acid is precipitated directly from the beverage as ammonium magnesium phosphate which is converted into phosphomolybdate and back again to ammonium magnesium phosphate and weighed as pyrophosphate. Glycerol is estimated by the American official method f o r the estimation of this substance in wines but since the residue of glycero’l obtained may contain caffeine and other substances it is oxidised with permanganate and the glycerol calculated from the amount of oxalic acid pro- duced. G. F. M. Melting Point of Salicylic Acid and a Test for the Presence of p-Hydroxybenzoic Acid. HENRY L. SMITH (Analyst 1916 41 3-5) .-Salicylic acid prepared from different sources and dried over sulphuric acid under reduced pressure had m.p. 158.5O; p-hydroxybenzoic acid melted a t 213.5O the same melting point being obtained after crystallisation and drying over sulphuric acid. For the detection of phydroxybenzoic acid in salicylic acid the formation of basic calcium salicylate which is leas soluble than the calcium salt! of p-hydroxybenzoic acid may be employed to concen- trate the latter acid. The salicylic acid to be tested is mixed with an excess of calcium hydroxide and water evaporated to dryness and the residue extracted with water. The aqueous extra& is acidified extracted with ether and the ethereal extract evaporated. When an ethereal solution of this residue is allowed to evaporate on a microscope slide the phydroxybenzoic acid crys-ii.160 ABSTRACTS OF CHEMICAL PAPERS. tallises in tufts which are quite different in appearance from the crystals of salicylic acid seen a t the same time. w. P. s. New Reaction of Picric Acid and its Application. J. CASTETS ( J . Pkarm. Chim. 1916 [vii] 13 46-49).-The reaction depends on the formation of 2-bromo-4 6-dinitrophenol and the red coloration which this substance gives with ammonia vapour or potassium cyanide. A solution containing picric acid is treated with bromine water heated t o boiling cooled and extracted with ether. The ethereal solution is separated and divided into two portions one of which is evaporated and the residue treated with ammonia vapour; a red coloration is obtained which is destroyed by an excess of ammonia o r by the addition of water.The other portion is evaporated drop by drop on a piece of filter p a p r ; a red coloration is obtained if the paper is then treated with ammonia vapour or with a drop of potassium cyanide solution and again dried. For the detection of picric acid in urine 100 C.C. of the sample are acidified with 2 C.C. of hydrochloric acid and extracted with chloroform. The chloroform solution is evaporated the residue dissolved in water treated with bromine and the test carried out as described. I n the case of beer the liquid is evapor- ated the1 residue treated with alcohol f o r twelve hours the alcoholic solution then decanted filtered evaporated the residue dissolved in water the solution acidified and treated as described for urine.The test will detect the presence of 5 mg. of picric acid per litre of urine. w. P. s. Estimation of Uric Acid in Blood. CARL MAASE and HERMANN ZONDEK (Cherri. Zentr. 1915 ii 858-859; from Miinch. med. T170ch. 1915 62 1110-llll).-Fivel C.C. of the blood are boiled with the addition of 25 C.C. of iV/lOO-acetic acid the hot mixture is filtered the coagulum is washed with 60 C.C. of boiling water containing 0.5 gram of sodium acetate the filtrate is acidified with 2.5 C.C. of 50% acetic acid and evaporated tlo a volume of 5 C.C. This residual solution is neutsalised with sodium carbonate solution transfe'rred t o a flask 2.5 C.C. of saturated sodium carbonate solution and 1 C.C. of Autenrieth's phosphotungstic acid reagent are added the mixture is diluted to 25 c.c.and after ten minutes the colora- tion is compared with a standard scale in the Autenrieth-Konigs- berger colorirneter (A. 1910 ii 910). To correct f o r the influence of other blood constituents (tyrosine and peptone) 0.5 mg. is deducted from the result obtained. w. P. s. Sensitive Method for the Examination of Oils. A. MAZZARON (Chem. Zentr. 1915 ii 1316-1317; from Staz. sperim. agrar. ital. 1915 48 583-594).-The method depends on the measurement of the quantity of sulphur dioxide evolved when an oil is treated with concentrated sulphuric acid. Twenty C.C. of the oil are placed in a flask connected with an absozption vessel containing a definite quantity of N/lO-iodine so'lution and 5 C.C. of sulphuric acid (D 1.8417) are added to the oil.A current of air is drawnANALYTICAL CHEMISTRY. ii. 161 through the apparatus the mixture in the flask is shaken for fifteen seconds and the current of air continued until all the sulphur dioxide has been absorbed by the iodine; the excess of iodine is then titrated with thiosulphate solution. The reaction should be carried out a t a temperature of 20°. The number of C.C. of N / 10-iodine solution required to oxidise the sulphur dioxide evolved from 20 C.C. of an oil is termed the sulphur dioxide number of the oil. This number f c r various oils was found t o be as follows Olive oil 2.4; sesame oil 49-5; cottonseed oil 137.5; maize oil 65; soja-bean oil 223; r a p oil 15; earthnut oil 7. There does not appear t o be any definite relation between the sulphur dioxide number and the sulphuric acid thermal number of the oils.w. P. s. EJtimation of Fat (in Animal Substancep etc.). HELMAN ROSENTI-IAL and P. F. TROWBRIDGE ( J . Biol. Chem. 1915 20 7 11-7 17).-Whilst Soxhlet’s method (simple extraction with ether) yields trustworthy results in the case of such substances as pork fat where! the fat is the only soluble substance present it is not trustworthy f o r the estimation of fat in blood liver etc. and for the latter ,purpose the authors prefer the following modification of a method proposed by Liebermann and Szekely (A. 1898 ii 655). The sample (20 grams of blood or 10 grams of liver) is heated on a water-bath f o r two hours with 30 C.C. of 20% sodium hydroxide solution the warm solution then transferred to a separ- ating funnel acidified with 35 C.C.of 20% hydrochloric acid (I) 1*1) cooled and extracted with ether. The ethereal solutions are evaporated the residue is dissolved in about 25 C.C. of light petroleum (b. p. 30-50°) 15 C.C. of 95% alcohol are added and the solution is titrated with AT/ 20-alkali solution using phenolphthalein as indicabor. The result is calculated into neutral glyceride. The authors also prefer this method t o that of Kumagawa and Suto (A. ,1908 ii 331) as in the latter there seem t o be many possibilities of e.rror in the separation of the unsaponifiable substances. w. P. s. Estimation of Formaldehyde in the Presence of Acetone. J. BOES and H. WEYLAND (Chem. Zentr. 1915 ii 982-983; from Hyg. Rundschau 1915 25 657-663).-A definite quantity of the f ormaldehycle solution containing acetone is treated in a stoppered flask with an exces of neutralised ammonium chloride solution ; litmus solution is added followed by an excess of N/1-sodium hydroxide solution and after one hour the mixture is titrated with N / 1-acid.The reaction proceeds according to the equation 4NH4C1+ 6CH20 + 4NaOH = C,H,,N + 4NaCl+ 10H,O. w. P. s. Analysis of Vanillin. J. F. SACHER (Chem. Zentr. 1915 ii 1027 ; from Deutsch. Parfumerie Z e i t . 1915 1 187-188).-The method propoeed depends on the fact that vanillin in alcoholic solution can be titrated with N/lO-mdium hydroxide solution usingii. 162 ABSTRACTS OF CHEMICAL PAPERS. phenolphthalein as indicator ; the end-point of the titration is sharp. The result of the titration will indicate whether or not small quantities of foreign substances are present. w.P. 8. Significance of the Strophanthin Reaction which is Obtained by the Action of Sulphuric Acid on Stropbauthus Seeds. ALESSANDRO BALDONI (Chem. Zentr. 1915 ii 984-985 ; from Arch. Farmacol. sperim. 1915 19 511-528).-Fresh strophanthus seeds (St. Komb6) give a green coloration when t,reated with con- centrated sulphuric acid but the reaction is not obtained with old seeds. I n some cases where a reaction is not obtaine'd with the concentrated acid a coloration is given when 80% acid is used; 50% sulphuric acid never yields a reaction. Exposure of the seeds to sunlight causes a decrease in the intensity of the coloration. The substance in the seeds which gives the reaction can be extracted partly by means of ether or 95% alcohol but not by light petroleum.Although the reaction will not distinguish between the various strophanthins it is useful in conjunction with other tests. w. P. s. Estimation of Urea and Extractives in Urine by the Xanthhydrol Method. L. MAESTRO (Chem. Zentr. 19 15 ii 984 ; from Arch. Farmacol. sperim. 1915 19 572-576).-A modifica- tion of this method (A. 1914 ii 506 593) is described. One C.C. of the urine is mixed with 9 C.C. of water and 20 C.C. of glacial acetic acid and 5 C.C. of at 6% solution of xanthhydrol in glacial acetic acid are added in five equal portions a t intervals of ten minutes. After one hour the precipitate is collectesd washed with methyl alcohol dried and weighed; its weight multiplied by 142.857 gives the quantity of urea per litre of the urine.The extractives may be estimated in the filtrate. w. P. s. Purification of the Residues of the Ether and Chloroform Extracts in the Forensic Detection of the Alkaloids. A. CARDOSO PEREIRA (Chem. Zeit. 1916 40 39).-The residues from the ethereal or chloroform extracts in forensic analysis may be completely freed from ptomaines by boiling the original tartaric acid extract after extraction with ether with a few C.C. of per- hydro1 for a few minutes and then proceeding with the analysis in the usual way. Strychnine nicotine and coniine are unaffected by t'his treatment but marphinel is converted into oxidation products. G. F. M. G. FENDLER and W. STUBER (Chem. Zentr. 1915 ii 1315; from Zeitsch.ATahr. Genussm. 1915 30 274-277. Compare A. 1914 ii 757).-A method described by Burmann (A. 1910 ii 465) was found to be untrustworthy. w. P. s. Estimation of Caffeine in Coffee. Detection of Safflower -in Saffron. G. VICARI (Chem. Zentr. 1915 ii 1318; from Mitt. Lebensmittelunters. Hyg. 1915 6 195-197).-The phoecpho'mdybdic acid test described by Verda (A,,ANALYTICAL CHEMISTRY. ii. 163 1914 ii 305) is useful for the detection of safflower (Carthamus tinctorius) in saffron; the latter gives a hlue coloration whilst Distinction between Erythrosin and Rose-Bengal. Estima- tion of their Commercial Value. ALEXAXDRE LEYS ( A m . Chi1)z. anal. 1916 21 25-32).-Erythrosin the sodium salt of tetra- iodofluorescein contains 57.7% of iodine and is free from chlorine; its solutions are practically free from fluorescence. It is however rarely met with in a pure state in commerce a portion of the iodine being replaced by chlorine ; these mixed halogen compounds may exhibit fluorescence particularly when the halogens are attached to both the phthalic and resorcinol nuclei.Rose-bengal the sodium salt of tetraiododichlorofluorescein contains 53.53% of iodine and 7.48% of chlorine and is less valuable than erythrosin. Estimations of the iodine and of the ratio of iodine t o chlorine in a sample will give information as t o whether it is pure erythrosin. The matesial should be extracted with alcohol t o separate mineral substances the alcoholic solution then evaporated and the residue dried a t looo. A weighed portion of this residue is fused with a mixture od potassium hydroxide and sodium carbonate cooled d i s solved in water and an aliquot portion of the solution titrated in the usual way with silver nitrate solution; the iodine is then estimated in another portion of the solution the potassium nitrite safflower yields a red coloration.w. P. s. method being used for the purpose. w. P. s. Simple Apparatus for the Estimation of Small Quantities of Albumin. RICHARD WEISS (Chenz. Zciztr. 1915 ii 81 1-812 ; from Miinch. med. Woch. 1915 62 1010).-Five C.C. of phospho- tungstic acid reagent (phosphotungstic acid 1.5 gram ; hydrochloric acid 5 c.c.; and 95% alcohol 95 c.c.) are placed in a specially con- structed albuminometer and filtered urine is added until a distinct turbidity is produced.The reagent will give a turbidity with as little as 0*0001 gram of albumin and the quantity of urine added t o produce a turbidity will consequently contain this amount of albumin. The apparatus is graduated so t h a t the quantity of urine used may be noted. w. P. s. Estimation of Albumin in Urine. W. AUTENRIETH and FRIEDA MINK (Chem. Zentr. 1915 ii 1265; from Miinch. med. Woch. 1915 62 1417-1421).-A colorimetric method based on the biuret reaction is described; the albumin is first precipitated by warming the urine with the addition of acetic acid. The Auten- rieth-Koenigsberger colorimeter (A. 1910 ii 910) is used for w. P. s. Rapid Methods for the Estimation of Albumin and Sugar in Urine. EMIL LENK (Chenz. Zentr. 1915 ii 1265 ; from Deutsch. med. Woch. 1915 41 1281).-For the estimation of albumin the urine is diluted so that it contains not more than 0.4% of albumin and is filled into an Esbach tube up to the mark comparing the colorations obtained.ii.164 ABSTRACTS OF CHEMICAL PAPERS. U ; Esbach’s reagent is theii added up t o the mark K a small quantity of powdered pumice-stone is introduced the tube is closed and inverted about ten times but not shaken. The mixture becomes clear within two minutes and the estimation can be com- pleted in ten minutes. Sugar is estimated by titration against Eehling’s solution potassium ferrocyanide solution acidified with acetic acid being used as an external indicator in determining the end-point of the titration. w. P. s. Test for Drobilin in Urine and Faeces. ADOLF EDELMANN (Chem. Zentr. 1915 ii 1059; from Wien. ?din. TVoch. 1915 28 978-979).-The test depends on the formation of a red compound when urobilin is treated with mercuric chloride. Five C.C. of a 10% alcoholic mercuric chloride solution are added to 10 C.C. of the urine and the mixture is shaken with amyl alcohol; the amyl alcohol layer is then transferred to1 a test-tube and mixed with a few C.C. of clear filtered 10% alcoholic zinc chloride solution. If large quantities of urobilin are present the amyl alcohol solution will exhibit a red colour and a green fluorescence appears on the addition of the zinc chloride; with a small quantity of urobilin the green fluorescence is seez only when a ray of light is projected through the clear liquid. The reactiotn is not affected by any of the other constituents of normal or pathological urine. F o r the detec- tion of urobilin in fzces a few grams of the latter are mixed with a \-ery small quantity of water an excess of alcoholic mercuric chloride solution is added the mixture filtereld and the filtrate is treated with a few C.C. of alcoholic zinc chloride solution; a green fluorescence is seen if urobilin is preselnt. w. P. s. Detection of Bile in Soaps. F. STEINITZER (Chem. Zentr. 1915 ii 810; from Chem. Rev. Fett. Harzdnd. 1915 22 69-70). -Five grams of the soap are dissolved in hot water the solution is filtered and the filtrate treated with dilute sulphuric acid. After cooling the cake of fatty acids is separated dried between filter- paper then heated a t 70° in a test-tube with 10 C.C. of sulphuric acid (1 l) and the hot mixture treated with a few drops of 10% sucrose solution. If biliary acids are present a red to violebred coloration develops on keeping thel mixture a t 70° f o r a few minutes (Pettenkofer’s reaction). w. P. s.