ANALYTICAL CHEMISTRY. Analytieal Chemistry. ii. 77 The Conditions for the Maximum Precipitation of an Amphoteric Electrolyte. ADA PRINS (Chem WeekbZad 1921 18 657-658).-The minimum solubility occurs a t a definite hydroxyl-ion concentration depending for each amphoteric electro- lyte on its solubility product as base and as acid. The concentrations of the positive and negative ions are inversely proportional to their charges. s. I. L. Use of Caesium Chloride in Microchemistry. ENRIQUE HERRERO DUCLOUX (Anal. Asoc. Quim. Argentina 1921 9 215- 227).-Caesium chloride may be used as a reagent in micro- chemistry by reason of the well-defined double chlorides which it forms with different metals. Crystallographic descriptions with photomicrographs are given of the double salts thus formed with silver mercury lead platinum gold palladium arsenic antimony cadmium tin copper aluminium iron zinc nickel cobalt man- ganese calcium magnesium thallium cerium and indium.G. W. R. Use of the Zeiss Water Interferometer (Rayleigh-Lowe) for the Analysis of Non-aqueous Solutions. ERNST COHEN and H. R. BRUINS (Proc. K. Akad. Wetemch. Amsterdam 1921 24 114-122).-A description of the Rayleigh-Lowe water inter- ferometer and the method of use for estimating the concentration of aqueous solutions are given. The accuracy obtainable with this instrument is about 99.9998% for aqueous solutions. To use this instrument for non-aqueous solutions greater precautions in the fixing of the temperature are required. The thermostat liquid should also be chosen so as to have a refractive index of the same order as that of the solvent used.When these precautions are taken an accuracy equal to that obtained for water solutims may be obtained with solutions in organic solvents. J. F. S,ii. 78 ABSTRACTS OF CHEMICAL PAPERS. A Buffer Solution €or Colorimetric Comparison. T. C. MCILVAINE ( J . BioZ. Chem. 1921 49 183-186).-The author covers the whole range from PII 2.2 to PH 8.0 by mixing two solu- tions only viz. a O.2M-disodium phosphate solution with 0-1M-citric acid. A table to obtain any desired PH is given and also a graph. G. B. A Simplified Form of Apparatus for Air Analysis. CHARLES CLAUDE GUTHRIE (J. BioZ. Chem. 1921 48 365-371).-The apparatus which is illustrated in the original “ differs from the well known forms in dimensions rather than in principles or in design.” E.S. Estimation of the Gases of the Blood. DONALD D. VAN SLYKE and WILLIAM C. STADIE ( J . BioZ. Chem. 1921 49 1-42).- Improvements in the technique of using the apparatus previously described for the extraction from the blood and measuring of the carbon dioxide (A. 1917 ii 422423) oxygen (A. 1918 ii 82) and carbon monoxide (A.? 1920 ii 53). All the gases can now be estimated in 1 C.C. of blood by a modified form of the original apparatus with narrow measuring tube enclosed in a water-jacket. In the oxygen estimation the blood is now laked with water and the amount of ferricyanide has been greatly reduced; the results are about 5% higher than those obtained by Haldane’s method. Particularly in the case of the carbon dioxide estimation the errors are fully analysed and examples of calculations are given. G.B Mechanical Shaker and other Devices €or Use with the Van Slyke Blood Gas Apparatus. WILLIAM C. STADIE (J. Bid. Chem 1921 49 43-46 ; cf. preceding abstract).-A motor- driven shaker a levelling scale and tonometer rotator are de- scribed with the help of figures. SCHIMMEL & Co. (Bey. Xchimmel & Co. 1921 56-61; from Chem. Zentr. 1921 iv 771).-Benzaldehyde is burnt in a small lamp so constructed that the amount of liquid burnt may be found by weighing a t the beginning and end of the experiment. The products of com- bustion are passed through two U-tubes containing glass beads moistened with O-OBN-potassium hydroxide solution foaming being prevented by the use of a few drops of petroleum. The washings from the absorption apparatus are united and after addition of an equivalent amount of O-OQN-sulphuric acid con- centrated in a closed flask.The solution is then made alkaline to phenolphthalein and the colour discharged by one or two drops of 0-02N-sulphuric acid. Five drops of 10% potassium chromate solution are added and the chlorine is estimated in the usual way with 0-02N-silver nitrate solution. Action of Potassium Ferrocyanide on Silver Haloids. G. B. BONINO (Gaxzetta 1921 51 ii 261-265).-Potassium ferro- cyanide does not appear to interact with silver bromide or iodide G. B. Estimation of Chlorine in Benzaldehyde. G. W. R.ANAL'YTICAt CHEMISTRY. ii. 79 buti with the chloride it reacts according to the equation 3AgC1+ K,Pe(CN),=Ag,KFe(CN) ,+3KC1.The chlorine ion may there fore be estimated in presence of the iodine ion by precipitating and washing the mixed silver haloids treating the latter with a known volume of 0.W-potassium ferrocyanide solution removing the insoluble silver potassium ferrocyanide by filtration and determining the amount of the residual potassium ferrocyanide by titration with standard permanganate solution. The reaction has not yet been studied quantitatively in presence of bromides. T. H. P. Estimation of Small Quantities of Bromides and Chlorides in Iodides. I. M. KOLTHOFF (Pharm. Weekblad 1921 58 1568-1569).-The iodide in solution is treated with excess of sodium nitrite in presence of sulphuric acid. After filtering and boiling to remove iodine and excess of nitrous acid chlorides and bromides are estimated by Volhard's method.The method is quantitative but does not distinguish between chlorides and bromides. s. I. L. Bromine Normally Present in Animal Tissues. A. DAMIENS (Bull. Sci. Pharmacol. 1921 28 85-93; from Chem. Zentr. 192 1 iv 847) .-Thirty grams of finely-divided tissue are extracted by heating with dilute potassium hydroxide solution. The dried residue is incinerated with a mixture of potassium nitrate and sodium carbonate and the ash dissolved in water. Iodine is estimated in a portion of the solution as follows the silver haloid precipitate obtained by addit'ion of silver nitrate solution in the presence of nitric acid is suspended in water and a stream of chlorine is passed first in the cold and then in the presence of 1 C.C. of sulphuric acid with warming.Air is passed through and the precipitate removed by centrifuging. The iodine is then estimated by a colorimetric method. Another portion of the solution is precipitated with silver nitrate and nitric acid. The precipitate is removed and the filtrate is treated with zinc and sulphuric acid. When less than a milligram of iodine is present bromine may be estimated colorirnetrically in the filtrate. In the presence of larger amounts of iodine the filtrate after the reduction of the silver haloid precipitate is neutralised with ammonia diluted to 40 C.C. and after addition of 1 gram of iron animonium sulphate concentrated to 10 C.C. Bromine may then be estimated as before (cf. A. 1921 i 476).Estimation of Sulphur in Pyrites. L. GADAIS ( A m . Chim. Analyt. 1921 3 330-335).-A critical review of Lunge's method and certain suggested modifications. For accurate work the original method without alteration is to be preferred care being taken to adhere strictly to all details of the operations. A varia- tion of this method in which the insoluble gangue is not filtered off before the precipitation of the iron with ammonia is not recom- mended as this gangue contains substances such as the sulphates of barium strontium calcium and lead which may be partly G. W. R.ii. 80 ABSTRACTS OF CREMICAL PAPERS. dissolved by ammonia and reprecipitated on subsequent acidifica- tion and addition of barium chloride. A second modification of Lunge’s procedure which obviates the washing of the gangue consists in making up the solution in aqua regia to 100 c.c.filtering off 50 c.c. and proceeding with this aliquot portion as in the original method. This is free from serious objection provided the insoluble residue is comparatively small in amount. Finally a rapid control method not suitable for accurate work consists in diluting the original solution to about 800 c.c. adding ammonia directly to this digesting for two hours a t a moderate temperature cooling making up the volume to 1000 c.c. and filtering off 500 C.C. in which sulphate is estimated in the usual way. Volumetric Estimation of Sulphide by Oxidation to Sulphate. H. H. WILLARD and W. E. CAKE (J. Amer. Chem. XOC. 1921,43 1610-1614).-Sulphide is rapidly and quantitatively oxidised to sulphate by an excess of hypobromite in 26N-sodium hydroxide or by hypochlorite in 4N-sodium hydroxide.The excess of oxidising agent is determined iodometrically. The method gives accurate results for the estimation of sulphur in steels and in sulphides if care is taken to exclude all other reducing agents. I n the case of steels the method is used as follows 5 grams of steel are placed in a flask through which hydrogen may be passed which is connected with a 10-bulb tube for absorbing the hydrogen sulphide. Air is removed from the apparatus by hydro- gen and 100 C.C. of hydrochloric acid (d 1.1) are added. After the reaction has moderated the solution is heated just to boiling for five minutes after the steel has dissolved. The solution in the absorption tube should contain 6-7 grams of sodium hydroxide in 60 C.C. of solution.The contents of the bulbs are washed with as little water as possible into a flask containing 10 C.C. of 0-3h7- hypobromite solution and kept for three or four minutes then 2 or 3 grams of potassium iodide are added and the solution is diluted to 150 C.C. It is then exactly neutralised with concentrated hydrochloric acid and 5 C.C. excess added and titrated with 0.1N- sodium thiosulphate. If hypochlorite has been used instead of hypobromite more sodium hydroxide must be used. Sulphides which are soluble in hydrochloric acid are estimated in the same way as steels. Insoluble sulphides are ignited with powdered iron in an atmosphere of hydrogen or carbon dioxide thus producing ferrous sulphide.The estimation is then carried to completion as above. Estimation of Sulphurous Acid. VICTOR COPPETTI (Ann. Chim. Analyt. 1921 3 327-330).-The gravimetric method of Haas for the estimation of sulphurous acid which consists in expelling the sulphur dioxide from the solution under examination by distillation in an atmosphere of carbon dioxide absorbing the gm in a solution of iodine and weighing the resulting sulphuric acid as barium sulphate gives accurate results volumetrically if means are taken to prevent loss of iodine by volatilisation in the G. P. M. The error of the method is about 0.1%. J. F. S.ANALYTICAL CHEMISTRY. .ii. 81 current of carbon dioxide. For this purpose an apparatus is described consisting essentially of a 300 C.C. flask to contain the iodine solution to the bottom of which extends the gas delivery tube from the distillation flask.Surmounting the flask is a spherical absorption vessel containing N/10-thiosulphate solution through which the carbon dioxide and iodine vapours leaving the flask must pass. When distillation is complete the thiosulphate solution containing all the volatilised iodine is allowed to run back into the flask and the excess of iodine in the latter is titrated back with standard thiosulphate solution. The Estimation of Sulphates by means of a Suspension of Barium Chromate. I. M. KOLTHOFF (Rec. trav. chirn. 1921 40 686-699; cf. Andrews A. 1890 414).-Although it would appear on theoretical grounds that an estimation of sulphate by means of barium chromate was not practicable the method is valid since the reaction BaSO,+CrOi’ zz BaCrO,tSO,” proceeds very slowly from left to right.The solution acidified with hydrochloric acid should be heated during neutralisation with ammonia to avoid the loss of chromic acid which accompanies the precipitation of barium chromate at ordinary temperatures and having regard t o the hydrolysis of ammonium chromate a slight excess of ammonia should be added. Barium chromate is adsorbed by ferric aluminium and zinc hydroxides so that when these metals are present the method gives results which are too low but trustworthy results can be obtained when ions which affect the reacting substances are absent. I n presence of calcium the results are too low the error becoming smaller with increasing acidity of the solution and decreasing concentration of calcium.Practical details are given. G . F. M. H. J. E. Catalytic Action of Copper in the Oxidation of Ammonia by means of Persulphate. G. SCAGLIARINI and G. TORELLI (G‘axxetta 1921 51 ii 277-280).-Quantitative investigation of the oxidation of ammonia by means of either potassium or ammo- nium persulphate in presence of copper sulphate shows that the oxygen of the persulphate first oxidises the ammonia to nitrous acid which is converted into ammonium nitrite this being decom- posed with liberation of nitrogen by the heat developed 2NH3+ 30,=2HN02+2H20 2NH3+2HN0,=2NH4*N02 and 2NH4*N02 =4H20+2N,. The catalytic effect of copper salts appears to be due to oxidation of the copper to a more highly oxidised com- pound which passes on its surplus oxygen to the ammonia.The estimation of persulphate by means of ferrous salts may be replaced advantageously by the following method in which use is made of a Schultze and Tiemann’s apparatus furnished with a mercury valve (2. anal. Chem. 1870 9 401; Ber. 1873 6 1041) 40 C.C. of water and 0.4-0.6 gram of copper sulphate are boiled in the flask for about ten minutes the caoutchouc tube being then clipped and the boiling continued for five minutes to expel the air. The apparatus is allowed to cool the mercury rising inii. 82 ABSTRACTS OF CHEMICAL PAPERS. the barometer tube and a vacuum becoming established in the flask. A known volume of the persulphate solution and after- wards about 100 C.C. of ammonia solution are drawn into the cold flask through the clipped tube the flask being then heated and the evolved gas collected over water and measured.The results obtained in this way agree exactly with those yielded by the ferrous sulphate method. [Method for the Estimation of Tri- Tetra- and Penta- thionates present together in Solution with Sulphite Thio- sulphate and Sulphate.] E. H. RIESENFELD and G . W. Fmn (Z. anorg. Chem. 119 225-270). The Accuracy of Dumas's Method for the Estimation of Nitrogen in the Cases of Substances rich in Nitrogen. ERNST Mom (Ber. 1921 54 [B] 2758-2767).-A mathematical treat- ment of the influence of the errors in measuring weight of substance volume of nitrogen temperature and pressure on the accuracy of the process. The usual procedure of estimating the volume accurately tlo within 0.05 or 0.1 c.c.and the temperature and pressure to degrees Centigrade and millimetres of mercury is sufficient for substances containing 20-25 yo of nitrogen but involves considerable error when more than this amount is present. The errors due to inaccurate reading of pressure and temperature cannot be minimised by increasing the weight of substance taken. On the other hand the errors due to volume and weight of substance become con- siderable when a small quantity of substance is taken and can be diminished by increasing the amount. The practice of using small weights of material when dealing with substances rich in nitrogen by Dumas's method is to be deprecated; Pregl's method should be used in preference. The error involved in the measurement of pressure does not depend to an appreciable extent on whether the gas is moist or dry (above 50% potassium hydroxide solution) but the error involved in measurement of temperature is lower in the latter case.In spite of this fact the measurement is generally made in preference over water by reason of the customary large diameter of the Schiff's nitrometer and the formation of foam over the potassium hydroxide solution. The temperature is first brought to whole degrees by addition or subtraction of x" in the direction of smallest change and the pressure is then changed by 3x mm. in the same sense as the alteration of temperature. A plea is entered for the more uniform recording of analytical data in the literature and more precise statement of the exact condition of the nitrogen with respect to moisture as also for the reduction of pressures to 0".The Kjeldahl Nitrogen Method and its Modifications. A. E. PAUL and E. H. BERRY ( J . Assoc. O#. Agric. Chem. 1921 5 108-1 32) .-Investigations are described into the most suitable T. H. P. See this vol. ii 45. A simple and accurate method of calculation is as follows. H. W.ANALYTICAL UHEMISTRY. ii. 83 apparatus and method for the nitrogen estimation with particular reference to the case of cotton-seed meal as presenting special difficulty. The most suitable type of bulb-trap for use in the distillation is the one in which both inlet and outlet tubes enter the bulb and are bent in opposite directions. It is advisable although not absolutely necessary to have enough acid in the receiver to neutralise all the ammonia distilled.Practically all the ammonia appears in the first 75 C.C. of the distillate and all is in the first 100 C.C. During the initial digestion of the material with acid the flame should never touch the flask above the surface of the liquid; the flask should be protected by a ring of asbestos. The volume of the digesting liquid should a t no time be less than 10 C.C. If mercury is used to aid the digestion there is a loss of from 2 to 15% of ammonia unless enough potassium sulphide is added to precipitate all the mercury before distillation. The use of copper sulphate during digestion does not necessitate the sub- sequent addition of potassium sulphide. The use of perman- ganate is unnecessary.Digestion with sulphuric acid alone never gives maximum results. Mercury gives a much more rapid digestion than copper salts and potassium sulphate is more efficient than sodium sulphate. The most rapid and efficient digestion is given by the use of 0.7 gram of mercuric oxide and 10 grams of potassium sulphate. In this way the liquid becomes clear in one to one and a half hours and further heating for three hours completes the digestion. The amount of copper used in the digestion has little effect on the result. A 2 gram sample of cotton-seed meal and similar substances is preferable. Digestion should not be carried out in an atmosphere containing nitrous fumes. Nitrates in the digestion mixture are not only entirely lost but they also bring about a considerable reduction in the amount of ammonia recovered.A. G. P. Modification in the Kossel-Neumann Method for the Estimation of Phosphorus in Organic Substances. MARIO A. MANCINI (Biochem. ter. sper. 1921 8 4-7; cf. Falk and Sugiura A 1915 ii 577).-@5 Gram of the substance is heated in a 200 C.C. Kjeldahl flask with 6-8 C.C. of sulphuric acid (d 1*184) the boiling being continued for one hour. Nitric acid (d 1.4) having been carefully added drop by drop boiling is con- tinued until the evolution of oxides of nitrogen ceases. The opera- tion is repeated four or five times. The clear light yellow liquid is washed into a beaker with hot water and ammonium hydroxide solution is added in slight excess followed by 30-40 C.C. of 40% ammonium nitrate solution and 20 C.C. of .%yo nitric acid.The liquid is mixed a t loo" with 120-140 C.C. of a boiling 3% solution of ammonium molybdate. After being kept for a t least two hours it is filtered and the precipitate washed with a hot solution containing 50 C.C. of nitric acid and 50 grams of ammonium nitrate per litre. Hot 25% ammonium hydroxide solution is poured over the filter which is washed with hot water until the washings give no reaction with Nessler's reagent. To the filtrate is addedii. 84 ABSTRACTS OF CHEMICAL PAPERS. 50 C.C. of ammoniacal " magnesia mixture," and after being kept for twenty-four hours the precipitate is filtered washed and ignited in the usual manner. The Composition and Preparation of a Neutral Solution of Ammonium Citrate. C. S. ROBINSON ( J . Assoc. Ofl.Agric. Chern. 1921 5 93-97).-To overcome the confusion attached to the term " neutral ammonium citrate solution," it is recom- mended that such a solution should be one showing a pH value of 7.0. The solution contains 45.33 grams of ammonia and 172.00 grams of anhydrous citric acid per litre a t 20" and has d 1.09. To prepare a litre of solution 172.00 grams of anhydrous citric acid are dissolved in 700 C.C. of water nearly neutralised cooled and made up to a convenient volume maintaining the density above 1.09. Of this 5 C.C. are diluted to about 20 c.c. and standard ammonia solution is added until the colour produced with phenol-red indicator matches that produced with the same quantity of indicator by an equal volume of neutral standard phosphate solution (50 C.C.of M/5-dihydrogen potassium phosphate +29*63 C.C. of M/5 sodium hydroxide in 200 c.c.). The calculated amount of ammonia is then added to the bulk of the solution. A. G. P. CHEMICAL ABSTRACTS. A Modified Method for the Estimation of Phosphoric Acid. A. W. CLARK and R. F. KEELER ( J . Assoc. 08. Aqric. Chern. 1921 5 103-105).-Two grams of the sample are dissolved in 30 C.C. of concentrated nitric acid and 10 C.C. of hydrochloric acid. The solution is diluted to 200 C.C. and filtered through a dry filter. A portion equivalent to about 0.25 gram is neutralised with ammonia and acidified with nitric acid. Fifty C.C. of 20% ammon- ium nitrate solution are added and then sufficient ammonium molybdate solution. After remaining over-night the precipitate is collected on a Gooch crucible washed eight times with 2% nitric acid (12-13 C.C.each time) then twice with cold water and dried for two hours a t 120". Precipitation in the cold and drying a t 120" give a less variable precipitate than the usual method. The conversion factor of ammonium phosphomolybdate to phosphoric acid is 0.03723. Nephelometric Method for the Estimation of Phosphoric Acid and its Compounds in Small Quantities of Blood. W. R. BLOOR (Bull. SOC. Chim. Biol. 1921 3 451475).-A detailed description of the author's adaptation (A. 1918 ii 452) of Kober and Egerer's method (A. 1915 ii 794). Toxicology of Arsenic. N. TARTJGI (Boll. Chim. Farm. 1921 60 569-576) .-The results of experiments with rabbits and guinea-pigs show that arsenic either organically combined or mixed mechanically with organic matter always yields gases containing arsenic when putrefaction occurs. Such evolution of gas may occur quickly and must be borne in mind in cases of suspected arsenical poisoning.In the putref action of animal matter the mechanism by which gaseous arsenic compounds are formed A. G. P. E. S.ANALYTICAL CHEMISTRY. ii. 85 is more complicated than the action of the " arsenic moulds," since such gases are not formed in the initial stages of the putre- faction where aerobiosis appears to predominate and since also the gases contain not only alkylarsines but hydrogen arsenide as well. T. H. P. f The Use of Silica Crucibles for the Estimation of Potassium in Soils. J. S. JONES and J. C. REEDER (Soi2 Sci. 1921 12 419-432).-For the estimation of potassium in soils by the fusion method silica crucibles may be used instead of platinum provided that certain limits of temperature are observed.To ensure perfect fusion the muffle must reach a temperature of 812" and to avoid loss of potassium by volatilisation the tem- perature must not exceed 855". For this purpose an electrically heated muffle is preferable. An electrical arrangement for heating a silica crucible of the J. L. Smith type 10 cm. long 2 cm. in diameter at the top and 1.8 cm. at the bottom is described and is very satisfactory for this type of work. Detection of Magnesium in Presence of Manganese and Phosphoric Acid. A. PURGOTTI (Guxxetta 1921 51 ii 265- 266).-In presence of phosphoric acid manganese is precipitated almost completely as tertiary manganous phosphate even from solutions containing large proportions of ammonium chloride ; the precipitate undergoes gradual or in the hot rapid trans- formation into pale pink crystals of manganous ammonium phosphate analogous in composition and properties to magnesium ammonium phosphate.The formation of this manganese pre- cipitate and hence the danger of mistaking it for the magnesium precipitate may be avoided by treating the hydrochloric acid solution not only with ammonia solution to remove the kations accompanying magnesium but also a t the same time with ammonium sulphide which eliminates the whole of the manganese as sulphide; the filtrate is then tested for magnesium. If the metals of the third group are precipitated in the usual way with ammonia solution manganese phosphate is precipitated in considerable quantity in addition to certain amounts of calcium barium and strontium phosphates a little manganese and much calcium barium and strontium passing into solution.Similar behaviour is shown by a mixture of magnesium phosphate with a manganese salt but less magnesium goes into solution and still less manganese is precipitated as phosphate. If the quantity of manganese salt is very considerably greater than that of the calcium barium and strontium phosphates the equilibrium Ca3(P0,)2+ 3Mn(OH) s lHn3(POp),+3Ca( OH) is displaced towards the right hand side and the calcium barium and strontium pass into solution almost completely. Apparatus for the Gasometric Estimation of Zinc in Zinc Powder.EDGAR BEYNE (Ann. Chim. Amlyt. 1921 3 360).- The apparatus consists of a gas burette the upper part of which below the glass stop-cock is expanded into a bulb of 300 C.C. W. G. T. H. P.ii. SG ABSTRACTS OF CHEMICAL PAPERS capacity and the lower part is graduated in 0.5 C.C. from 300 C.C. to 375 C.C. The lower end of the burette is connected with a rubber tube to a bulb for adjusting the liquid level and gas pressure in the burette and the upper end is connected through a small refrigerating spiral to a Koninck apparatus which is charged with the zinc dust under examination and a mineral acid. In a simpler form of apparatus the latter is replaced by an ordinary flask con- nected by means of a glass tube with the stop-cock of the burettc. G. F. M. The Direct Iodometric Estimation of Lead Peroxide.SAMUEL GLASSTONE (T. 1921,119 1997-2001). A Very Sensitive Reagent for Copper the Kastle-Meyer Reagent. PIERRE THOMAS and GEORGES CARPENTIER (Cmpt. rend. 1921 173 1082-1085).-The Iiastle-Meyer reagent which is a 2% solution of phenolphthalein in 20% potassium hydroxide solution decolorised by boiling with zinc powder gives a pink coloration with copper salts and is capable of detecting 1 part of copper in 100,000,000 parts of water. Four drops of the reagent are added to 10 C.C. of the solution to be tested and then one drop of hydrogen peroxide ( 5 - 4 vols.). The Iodometric Estimation of Copper and Arsenic present together especially in Paris and Schweinfiirth Green. I. M. KOLTHOFF and C. J. CREMER (Pharm. WeeEbZad 1921 58 1620-1624).-Arsenic trioxide can be estimated by means of iodine if the cupric-ion concentration is very much reduced which can be effected by addition of excess of pyrophosphate or tartrate with which the copper forms complex salts.O*& 0.8 Gram of the pigment is boiled with 25 C.C. of water and 5 grams of sodium pyrophosphate to LL clear solution. After cooling N/lO-iodine solution is added until the deep blue solution becomes green; the end-point is very sharp. The iodine required is equivalent to the arsenious oxide present. Ten C.C. of QN-sul- phuric acid and 2 grams of potassium iodide are then added and after ten minutes the iodine liberated by reduction of the cupric compound is titrated with N/lO-thiosulphate starch being added towards the end.The results were checked by analysis of the pigments by the standard Lunge-Berl method This method was found unsatis- factory for copper the results being always high. Addition of hydrazine sulphate before the sodium hydroxide reduces the copper compound the element being precipitated and weighed as metal ; this modification is much more accurate than the accepted method. The iodine titration is rapid and simple a'nd gives accurate results for both elements. Separation and Estimation of Copper Lead Antimony and Tin. Analysis of White Metals. A. &ING and A. LASSIEUR (Compt. rend. 1921 173 1081-1082).-05--1~0 Gram of the alloy is dissolved in 10 C.C. of hydrochloric acid in the presence of potassium chlorate. The solution is diluted to 100 c.c. W. G. s. I. L.ANALYTICAL CHEMISTRY.ii. 87 and neutralised with sodium hydroxide. Any precipitate formed is redissolved by the addition of 4 - 5 grams of tartaric acid. The liquid is transferred to a conical flask coated inside with wax and to it are added 10 C.C. of concentrated hydrofluoric acid and then after half an hour 10 grams of sodium acetate 1 C.C. of glacial acetic acid and water to bring the volume to 300 C . C . A white precipitate of lead fluoride is formed but to the liquid 20 C.C. of a 10% solution of sodium sulphide are added and after a time the precipitate of the sulphides of copper lead and antimony is filtered off. In the filtrate the tin may be estimated either by precipitation with cupferron (cf. A 1920 ii 452) or electrolytically after decomposition of the complex fluoro-compound by the addition of boric acid and redissolving the tin sulphide by boiling with hydrogen peroxide.The precipitate of the mixed sulphides is extracted with 80 C.C. of sodium sulphide solution (d 1.14) and the antimony estimated electrolytically in the extract after the addition of potassium cyanide. The copper and lead sulphides are dissolved in nitric acid and the two metals estimated simul- taneously by electrolysis. W. G. Estimation of Mercury in the Mercurial Pills of the [French] Codex. MAURICE FRANQOIS ( J . Pharm. Chim. 1921,24 369-379; Ann. Palsif. 1921 14 340-347).-The pills are heated with concentrated nitric acid for two hours on a water-bath and after a further two hours the solution is filtered and the destruction of the organic matter is completed by adding a small quantity of a 10% solution of bromine.After ten minutes the liquid is rendered strongly alkaline with sodium hydroxide and after adding potassium iodide the mercury is precipitated in the form of a grey powder by means of formaldehyde solution. After two hours the supernatant liquid is decanted through a filter the mercury is washed with dilute sodium hydroxide solution and is then treated together with the filter-paper and its contentp with diluted acetic acid and AV/lO-iocline solutiou. After ten minutes’ agitation the mercury is completely transformed into mercuric iodide which dis- solves in the potassium iodide present and the excess of iodine is titrated back with tlhiosulphate. The method is of general application to all the mercurial pills of the French Codex without modification except in the case of opiated mercurous iodide pills with which after treatment with nitric acid a crystalline precipitate of mercuric iodonitrate HgI,,Hg(NO,) is formed.All that is necessary in this case however is t o dissolve the crystals in the liquor by addition of potassium iodide and omitting the bromine treatment to proceed with the precipitation of the mercury as*above. In pills containing soap although the f a t acids are not destroyed by the nitric acid treatmen@ they do not retain any mercury in combination and form a layer on the surface of the liquor which solidifies on cooling and is therefore easily separated. ERNST JOSEF KRAUS (Chem. Zeit. 1921 45 1173).-The neutral or faintly acid solution containing the aluminium in the form of sulphate and free from G.F. M. Volumetric Estimation of Aluminium.ii. 88 ABSTRACTS OE' CHEMICAL PAPERS. other interfering metals is titrated with standard disodium hydro- gen phosphate solution a few drops of silver nitrate solution being used as indicator as yellow silver phosphate only commences to form after all the aluminium has been precipitated as phosphate according to the equation A12(S04),+2Na,HP04~2A1P0,+ 2Na,S0,+H2S0,. The titration is preferably carried out in a boiling solution as the silver phosphate is more pronouncedly yellow in colour and therefore more easily noticeable under these conditions. In presence of other metals such as iron etc. the aluminium should first be separated by addition of excess of sodium hydroxide for example and finally precipitated as hydroxide with ammonia.The precipitate after washing is dissolved in a slight excess of dilute sulphuric acid to form a solution suitable for the titration. The method gives good results even when only very small quantities of the metal are to be estimated. Improved Method for the Separation of Iron and Man- ganese. M. CARUS (Chem. Zeit. 1921 & 1194).-In the usual method for the separation of manganese from the metals of the iron group by precipitating the latter as basic acetates the con- tamination of the precipitate with manganese is not due to the co- precipitation of the basic acetate of this metal but to the formation of insoluble higher oxidation products owing to the action of dissolved oxygen.A perfect separation of the manganese in one operation even when a large excess is present is obtained by causing the precipitation with sodium acetate to occur in presence of a small quantity of hydrogen peroxide in which case no oxida- tion of the manganese salts can occur. The precipitate obtained after washing with dilute acetic acid containing a small quantity of sodium acetate and hydrogen peroxide and fhally with hot water is then completely free from manganese. G. F. M. I. &I. KOLT- HOFF and E. H. VOGELENZANG (Rec. truv. chim. 1921 40 681- 685) .-By reason of the hydrolysis of potassium chromate solution neither chromic acid nor dichromate can be neutralised using phenolphthalein as indicator. A study of the neutralisation of dichromate with sodium hydroxide leads to the following con- clusions :-Thymolphthalein should be used as indicator ; if phenolphthalein is used the solution must have been saturated previously with sodium chloride or barium chloride must be used to precipitate the chromate formed.The barium chloride may only be added when the solution is already yellow otherwise the precipitate of barium chromate contains chromic acid. It is claimed that the results are correct to within 0*2% but the method G. P. M. The Acidimetric Estimation of Dichromate. is not recommended owing to difficulGes with the indicator. H. J. E. Modified Method for the Estimation of Iron and Vanadium after Reduction by Hydrogen Sulphide. G. E. F. LUNDELL and H. B. KNOWLES ( J . Amer. Chem. Xoc.1921 43 1560- 1568).-The estimation of iron or of vanadium by reduction withANALYTICAL CHEMISTRY. ii 89 hydrogen sulphide followed by titration with potassium permah- ganate or dichromate ordinarily leads to high values. These high values are not due to the presence of sulphur and persist in spite of the complete expulsion of hydrogen sulphide and the avoidance of organic matter extracted from filter-paper. They may be ascribed to polythionic acids which are not destroyed by boiling in moderately concentrated sulphuric acid solution and are volatilised slowly from dilute and more rapidly from concentrated solutions. Reasonably accurate estimations may be made by restricting the volume of the original solution to 100 C.C. and pro- ceeding by the following modified process.The solution is acidified with sulphuric acid until it contains 2.5% sulphuric acid by volume reduced by hydrogen sulphide for thirty minutes in cold solution and for a further fifteen minutes during which the solution is raised to the boiling point. The hot solution is treated with 15 C.C. of 1 1-sulphuric acid and boiled for thirty to sixty minutes during which a stream of carbon dioxide is passed through until the total volume of liquid is about 50 C.C. The solution is cooled the current of carbon dioxide being maintained diluted to 200 c.c. and titrated with standard permanganate solution. In some cases it is advisable to filter off the separated sulphur after the hydrogen sulphide treatment. The filtrate is treated for a further ten minutes with hydrogen sulphide and the process completed as above.The average errors found by this method are for iron o'5y0 and for vanadium 0.1%. For accurate estimations of iron and vanadium in solutions containing platinum (as in rock analysis) a preliminary separation of the hydrogen sulphide group followed by expulsion of the gas complete oxidation with permanganate and reduction with sulphur dioxide is recommended. J. F. S. Estimation of Vanadium and Chromium in Ferrovanadium by Electrometric Titration. G. L. KELLEY J. A. WILEY R. T. BOHN and W. C. WRIGHT ( J . Ind. Eng. Chem. 1921 13 939-941) .-Three grams of the ferrovanadium are dissolved in a mixture of nitric acid and hydrochloric acid sulphuric acid is then added the mixture evaporated to expel all nitric acid and hydrochloric acid cooled and the solution diluted to 1000 C.C.One hundred C.C. of this solution are treated with 25 C.C. of sulph- uric acid (d 1-58) diluted to 300 c.c. boiled and 20 C.C. of 10% ammonium persulphate solution and 10 C.C. of o.2570 silver nitrate solution are added; the mixture is boiled for ten minutes 5 C.C. of hydrochloric acid (1 3) are then added the boiling is* continued for a further ten minutes the mixture treated with 26 C.C. of sulphuric acid cooled a t 5" and titrated with ferrous ammonium sulphate solution the end-point of the titration being determined electrometrically. This titration is a measure of the vanadium and chromium together. The vanadium is estimated by boiling 100 C.C. of the original solution with a few C.C. of ferrous sulphate solution adding 20 C.C.of sulphuric acid d 1.58 and 40 C.C. of nitric acid (d 1-40) diluting the mixture to 200 c.c. and boiling it a t such a rate that the volume is reduced to 100 c,c in one hour,ii. 90 ABSTRACTS OF CHEMICAL PAPERS. The solution is then cooled and titrated as before. The difference between the two titrations is a measure of the amount of chromium present. w. P. s. [Estimation of Antimony as Sodium Antimonate.] E. S. TOMULA (2. unorg. Chern. 1921 118 81-92).-See this vol. ii 74. Estimation of Bismuth. 0. A. CRITCIIETT (Eng. and Min. J. 1921 112 58).-A solution of the ore in a mixture of nitric and hydrochloric acids having been evaporated with sulphuric acid until fumes appear is diluted treated with a drop of hydrochloric acid filtered and the filtrate boiled with sodium thiosulphate and aluminium foil.The precipitate is collected returned to the beaker with hot water potassium hydroxide added and the liquid after having been boiled is filtered through the original paper. The latter with the precipitate is repeatedly evaporated with nitric and sulphuric acids until the paper is completely destroyed. After dilution boiling filtering addition of a slight excess of ammonium hydroxide and again boiling the precipitate is collected washed and dissolved in hot dilute nitric acid. The bismuth may then be estimated (a) as oxide after treatment with ammonium carbonate ( b ) as oxychloride after neutralising with ammonium hydroxide adding a little hydrochloric acid diluting and boiling or (c) by titration with permanganate after dilution treatment with ammonium oxalate boiling and separation with subsequent washing by decantation with hot water of the precipitate of bismuth oxalate.CHEMICAL ABSTRACTS. Analytical Chemistry of Tantalum Columbium and their Mineral Associates. I. The Use of Tartaric Acid in the Analysis of Natural Tantalocolwnbates. 11. The Separa- tion of Zirconium from Tantalum and from Columbium. WALTER RAYMOND SCHOELLER and ALAN RICHARD POWELL (T. 1921,119 1927-1935). Evaluation of the Degree of Unsaturation of Mineral Oils in the Bergius Process. H. I. WATERMAN and J. N. J. PERQUIN (Rec. truv. chim. 1921 40 677-680; cf. Dean and Hill Technical Paper 181 Bureau of Mines 1917).-Determination of the iodine number of an unrefined mineral oil before and after treatment by the Bergius process shows that it is a little greater after treatment.1%. J. E. Estimation of Phenanthrene. ARTHUR G . WILLIAMS ( J . Amer. Chem. Soc. 1921 43 1911-1919).-The phenanthrene is oxidised by iodic acid to phenanthraquinone which is precipitated as toluphenanthrazine by means of 3 4-tolylenediamine and weighed as such the procedure being as follows For materials containing 30% or more of phenanthrene 0.25 gram is weighed into a 50 C.C. conical flask 0.75 gram of iodic acid and 20 C.C. of glacial acetic acid are added and the mixture is boiled €or two and a half hours under an air condenser. After cooling forANALYTICAL CHEMISTRY. ii. 91 several hours any anthraquinone formed from anthracene present in the original material is filtered off on a Gooch crucible and washed with the minimum amount of glacial acetic acid.The filtrate and washings are evaporated to slightly less than 25 C.C. and then the volume is made exactly to 25 c.c. the mixture is cooled and 1 gram of 3 4-tolylenediamine is added and the flask left in run- ning water a t 20" over-night. The toluphenanthrazine is collected in a Gooch crucible and washed first with 25 C.C. of 50% acetic acid saturated with the phenanthrazine and then with 200 C.C. of cold water. The precipitate is dried and weighed and to the weight is added 0.053 gram to allow for the toluphenanthrazine remaining in solution in the 25 C.C. of glacial acetic acid. The factor for conversion into phenanthrene is 0.6052. Carbazole if present in amounts exceeding 10 yo interferes in the determination of phenanthrene and some of the high-boiling coal tar constituents interfere ; a crude anthracene cannot be directly analysed.For the detection of phenanthrene the material is oxidised as described above the mixture being cooled and filtered. The filtrate is poured into water and the precipitate collected and washed with water. The precipitate is warmed with concentrated sodium hydrogen sulphite solution and any residue is filtered off. The filtrate is washed in a separating funnel with one or two portions of carbon tetrachloride and then after the addition of a fresh portion of carbon tetrachloride is acidified with hydro- chloric acid containing ferric chloride. The carbon tetrachloride layer which contains the phenanthraquinone is separated and tested by Hilpert and Wolf's reaction (cf.A. 1913 ii 733) using a solution of antimony pentachloride in carbon tetrachloride a purplish-red precipitate being obtained on boiling if phenanthrene was originally present. W. G. Chemical Analysis of Caoutchouc Articles. ANDFA DUBOSC (Ann. Chim. Analyt. 1921,3 335-344).-A resume of the methods employed and the estimations necessary for a complete chemical analysis of caoutchouc and ebonite articles. (1) The acetone extract contains the natural resins added resins free sulphur oils and waxes and is evaporated to dryness and quantitatively examined for these substances by the usual methods. (2) The chloroform extract should be almost colourless. Brown coloration indicates the presence of tar or asphaltic adulterants.(3) The extract in alcoholic potash of the insoluble residue of the chloro- form extraction should not exceed 15% from a material con- taining about 50% of caoutchouc a larger proportion indicating added oils or fats. (4) The aqueous extract reveals the presence of starch or dextrins. (5) The estimation of total sulphur is best carried out by Henriquez's method as modified by the Bureau of Standards. (6) The ash is determined on the residue from the acetone extraction and a rough estimate of the caoutchouc content can be obtained by subtracting ash and total sulphur from 100. (7) None of the various methods which have been proposed for the estimation of caoutchouc itself give entire satisfaction. Theii. 92 ABSTRACTS OF CHEMICAL PAPERS.total foreign matter may be obtained however by solution in boiling nitrobenzene diluting with chloroform filtering through a tared filter washing with acetone drying and weighing. Estimation of Oxalic Acid in Urine. E. SALIIOWSKI (Biochem. Z . 1921 118 259-266).-A reply to Bau’s criticism of the author’s method (A 1921 ii 356). T. S. HAMILTON W. B. NEVENS and H. S. GRINDLEY (J. Biol. Chem. 1921 48 249-272).-Further improvements are made in the application of Van Slyke’s method to the estimation of amino- acids in feeding stuffs (cf. Eckstein and Grindley A. 1919 ii 204). Non-protein nitrogen is first removed from the material by suc- cessive extractions with anhydrous ether cold absolute alcohol and cold 1% trichloroacetic acid any protein removed by the latter being recovered by precipitation with colloidal ferric hydroxide.The main portion of the protein is then extracted with dilute (0.2%) sodium hydroxide. Starch is removed from the residue by treatment with hot 2% trichloroacetic acid and the remaining protein extracted by treatment first with boiling 20y0 hydrochloric acid and then with cold 5% sodium hydroxide. A small quantity of protein extracted with the starch is recovered by precipitation of the latter by addition of alcohol. The various fractions of protein thus obtained are hydrolysed with concentrated hydrochloric acid united and submitted to the Van Slykc analysis. The method is applied to oats corn cotton-seed meal and lucerne. J. SNAPPER and E. LAQUEUR (Arch. Nierl. Physiol.1921 6 48-57).-To 100 C.C. of urine 25 grams of sodium chloride are added and a litt’lc con- centrated hydrochloric acid. An aliquot portion is ext~actcd six times with ethyl acetate and the total extract is washed one(’ with one-quarter of its volume of water. The wash water is washed with an equal volume of ethyl acetate which is added to the original extract. After evaporation of the ethyl acetate the urea is decomposed by sodium hypobromite and the residual hippuric acid estimated by Kjeldahl’s method. The Rotation of Dextrose in Solutions of Trisodium Phosphate. Mutarotation as an Analytical Method. HANS MURSCHHAUSER (Biochem. Z. 1921 117 215-225).-The muta- rotation of dextrose is accelerated by trisodium phosphate. It follows a unimolecular law the velocity constants being also linear functions of the concentration of sodium phosphate.As the mutarotation is a function of the hydroxyl ion its use is indicated for distinguishing salts ol different alkalinity. Conditions Affecting the Quantitative Estimation of Reducing Sugars by Fehling ’s solution. Elimination of certain Errors Involved in Current Methods. P. A. QUISUM- BING and A. W. THOMAS ( J . Amer. Chem. XOC. 1921 43 1503- 1526).-The various sources of error in the current methods of G. F. M. H. K. Estimation of the Amino-acids of Feeding Stuffs. E. S. Estimation of Hippuric Acid in Urine. G. B. H. K.ANALYTICAL CHEMISTRY. ii. 93 using Fehling's solution are discussed. The inter-relationship of temperature 60-llO" and time of heating ten minutes to two hours to determine the best time and optimum temperature for reduction has been investigated.Fehling's solution has been studied from the point of view of the nature and concentration of the alkali concentration of copper sulphate and potassium sodium tartrate to find the maximum and minimum concentrations of these constituents necessary to give the greatest yield of cuprous oxide and to ensure the formation of $he complex cupric tartrate ion Auto-reduction of Fehling's solution a t different times and temperatures of heating has been measured showing those con- ditions under which absolutely no " blaiik " reduction is obtained. Conditions affecting the physical properties of the precipitated cuprous oxide the photosensitiveness and the keeping quality of Fehling's solution are recorded. A study of surface oxidation involving different methods of heating and vessels of various sizes has been carried out in order to determine the loss of copper due to surface oxidation and how to avoid it.The catalytic effect of the walls of the container has been demonstrated. A modified method of procedure for the estimation of sugars by means of Fehling's solution is described. In this method the solutions required are (1) copper sulphate solution containing 82.4 grams of CuS0,,5H20 per litre (2) alkaline tartrate solution; 376 grams of crystallised potassium sodium tartrate dissolved in water in a litre flask and the calculated amount of sodium hydroxide solution added to make 1 litre of this solution containing 130 grams of sodium hydroxide.The sodium hydroxide is made from material purified by alcohol which is kept in concentrated solution for several days to allow carbonates and other insoluble impurities to separate. To make an estimation 25 C.C. of the copper sulphate solution 25 C.C. of the alkaline tartrate solution and 50 C.C. of the sugar solution are placed in a beaker and warmed on a water-bath a t 80". After exactly thirty minutes the cuprous oxide is filtered on a Gooch crucible and washed. This may either be dried and weighed as cuprous oxide or dissolved in nitric acid and estimated electrolytically. From the weight of copper the amount of sugar may be obtained from tables which are given in the paper or by means of the following equations in which x is the amount in milligrams of copper obtained from y mg.of sugar dextrose y=0.474~+0*0001 15x2 ; lzvulose y=0*526x+ 0.000078~~ ; lactose y=0*813~+0~000003x~ ; maltose y=O*99Ox+ 0*0000005~~ ; and invert-sug?r y=0~504~+0~0000870x~. This method is designed for use in the analysis of saccharine materials containing sucrose dextrose lavulose invert-sugar lactose and maltose. It is not claimed that the method can be used for the accurate quantitative estimation of 0.1 % or less of invert-sugar in approximately pure sucrose. J. F. S. Errors in the Estimation of Sugar by Fermentation with Yeast. CARL LANCE (Berlin Klin. Woch. 1921 58 957-959; from Chem. Zentr. 1921 iv 848).-The estimation of sugar inii. 94 ABSTRdCTS OF CHEMICAL PAPERS. urine by fermentation with yeast is subject to certain errors.The evolved gas is not entirely carbon dioxide but contains also nitro- gen from the reaction of carbamide ammonia or amino-acids with nitrites. The latter may be formed by reduction of nitrates in the urine or may occur in the yeast. There may also be an evolution of carbon dioxide from carbonates present in the urine. Errors may be diminished to some extent by sterilisation and precipitation of carbonates with calcium chloride. The estimation of sugar in urine however gives only an approximate value. G. W. R. Limitations of the Modified Lewis-Benedict Method of Blood Sugar Estimation. RUTH A. GUY (Biochem. J. 1921 15 575-576).-0*0001 Gram of acetone added to 2 C.C. of blood produces an observable colour change in Lewis and Benedict's picric acid method of sugar estimation (A.1915 ii lll) but no change is noted with similar quantities of acetoacetic acid. G. B. The Catalytic Power of Flour. 0. FERN~DEZ and A. PIZARROSO (Anal. Fib. Quim. 1921 19,. 265-268).-The catalytic activity of flour measured by its decomposition of hydrogen peroxide has been proposed as a mea'sure of its grade of extraction. The results obtained by the authors using samples of flour of different origin showed no correlation between catalytic activity and total- or phytin-phosphorus. The test is held to be of little value. G. W. R. Identification of Oxycellulose by means of the Barium Compound. ERNST BECKER (ZelZstoSf u. Papier 1921 1 5-7).-When 2 grams of oxycellulose are shaken with 50-60 C.C. of barium hydroxide solution for four hours and the insoluble portion is washed with water until the filtrate is free from barium compounds a barium compound of oxycellulose remains of which the content of barium varies according to the origin of the oxycellulose. CHEMICAL ABSTRACTS.Benzidine Hydrochloride as a Reagent for Wood Cells. C. VAN ZIJP (Pharm. Weekblud 1921 58 1539-1542).-A solution of 0.2 gram of benzidine in 19 C.C. of water to which 1 C.C. of 25% hydrochloric acid has been added gives an orange stain to ligneous matter in microscope sections. Sections of material containing tannins or acids which act on iron should be washed free from iron with a 1% solution of hydrochloric acid in alcohol after cutting. Starch can be stained with iodine in the same section without interfering. The reagent also detects diseased tissues in Hevea bark and for this reason as well as because oPits cheap- ness as compared with phloroglucinol is of great use for control on rubber plantations.Detection of Formaldehyde with Phenols. B. PFYL G. REIF and A. HANNER (C'hem. Zeit. 1921 45 1220-1221).- Phenol reactions for the detection of formaldehyde as previously carried out (cf. A. 1921 ii -663) are not sufficiently trustworthy s. I. L.ANALYTICAL CHEMISTRY. ii. 95 when the test is to be adapted to the detection of methyl alcohol in potable spirits and tinctures after distillation with permanganate as colour reactions are often simultaneously given by the traces of other aldehydes alcohols etc. which pass over into the dis- tillate and the mixed colours produced allow of no certain con- clusions being drawn.A solution of guaiacol or of upomorphine hydrochloride in concentrated sulphuric acid (0.02 gram in 10 c.c.) gives however a reagent with which a sharp distinction can always be observed if methyl alcohol is present in the original tincture. The reaction is best carried out by adding a few drops of the distillate to 0.5 C.C. of the reagent in a watch glass. With the guaiacol reagent a clear dark red colour is produced when formalde- hyde is present which is readily distinguished from the pale yellow coloration produced in its absence whilst with the apomorphine reagent a characteristic precipitate is formed in presence of form. aldehyde. As little as 025% of methyl alcohol in tinctures etc. can be detected with certainty by the new method.ED. JUSTIN-MUELLER (J. Pharm. Chim. 1921 [vii] 24 334-336).-As little as 1 part of furfuraldehyde in 600,000 parts of solution may be detected by the blue coloration which is obtained when 5 C.C. of the solution 5 C.C. of hydrochloric acid and 0.02 gram of orcinol are heated G. F. 11. Orcinol Reaction of Furfuraldehyde. together cGoled and shaken with a few-drops of amyl alcohol. w. P. s. Source of Error in Tests for Acetone. E. PITTARELLX (PoZicZinico 1921 28 621; J . Am. Med. ASSOC. 76 1803).- Caoutchouc under the influence of steam or boiling water yields a volatile substance which responds to all the tests most character- istic of acetone ; therefore conclusions based on technique involving distillation in the presence of caoutchouc may be misleading. CHEMICAL ABSTRACTS.Microchemical Reactions of “ Dulcin ” [p-Ethoxyphenyl- carbamide]. G. DENIGI~S and R. TOURROU (Compt. rend. 1921 173 1184-1186).-When a few particles of “ dulcin ” on a microscope slide are moistened with one drop of nitric acid (d 1.39) tlhey dissolve and on the addition of a drop of water microscopic orange OF brick-red crystals of p-ethoxynitrophenylcarbamide are obtained. These crystals are soluble in chloroform and on evaporation after the addition of a drop of acetic acid character- ist8ic crystals are obtained. If dilute nitric acid is used in place of the concentrated acid the “dulcin ” does not dissolve but the mass gradually becomes permeated with crystals of the nitro- compound and under the microscope there is the appearance of a marked effervescence.“ Dulcin ” is soluble in concentrated sulphuric acid or glacial acetic acid and is deposited as a micro- crystalline precipitate from such solutions on the addition of water or alkali. W. G. The Vitali Reaction for Cocaine. PAUL HARDY ( J . Pharm. Chini . 1921 [ vii] 24 325-330) .-Pure cocaine snd most sub-3. 96 ABSTRACTS OF CHEMfCAL PAPERS. stituted cocaines yield a yellow coloration with the Vitali test in the cold ; isoatropylcocaine however yields a violet coloration and the presence of a small quantity of this substance in certain specimens of cocaine causes these to give a violet coloration with the test. When the reaction mixture is heated pure cocaine Morphine Codeine and Narcotine in Indian Opium. JITENDRA NATH RAKSHIT (AnaZyst 1921 46 481-488).-The morphine content of the bulk of Indian opium is between 8.5 and 10.5y0.The B.P. process for estimating morphine is inefficient in India owing to the high temperature which prevails. The U.S.P. process is better and compares favourably with the author's polarimetric process (Analyst 1918 43 321). The solubility of codeine and narcotine has been determined in various solvents a t temperatures between 20" and 100". Methods have also been worked out for the estimation of codeine and narcotine for which the original should be consulted. Estimation of Tannin and Colouring Matters in Wines. W. FRESENIUS and L. GRUNHUT (2. anal. Chem. 1921 60 406- 417).-One hundred C.C. of the wine is evaporated to expel alcohol cooled diluted to 100 C.C. and 50 C.C. of this solution are treated with a few grams of pure animal charcoal; after some hours the mixture (a furthhr quantity of charcoal must be added if all the colour has not been removed) is diluted to 1 litre filtered and 400 C.C. are diluted to 1 litre treated with 10 C.C. of sulphuric acid (d l*ll) and 30 C.C. of indigo solution (3 grams of indigo dissolved in 20 C.C. of sulphuric acid and diluted to 1 litre) and the mixture is titrated with permanganate solution. Twenty C.C. of the de- alcoholised wine not treated with charcoal are titrated in a similar manner. The difference between the two titrations is a measure of the tannin and colouring substances in the wine. The permanganate solution should contain 1.33 grams of potassium permanganate per litre and be standardised against 10 C.C. of N/lO-oxalic acid solution ; the volume of permanganate solution used is divided into 0.0416 to obtain the tannin value of 1 C.C. of permanganate solution. Determination of the Size of Particles. Attempts to Explain the Formation of Layers in Clay Turbidities and their Use in Soil Analysis. ERNST UNGERER (KoZZoid Chem. Beihefte 1921 14 63-95).-The layers in clay suspensions are brought about solely by the size and weight of the suspended particles so that each layer corresponds with particles of a definite size. The turbidity between two layers contains particles uniformly dispersed. Layers are formed both in solutions containing electro- lytes and in solutions which do not contain electrolytes but a high concentration of a coagulating electrolyte will probably prevent the formation of layers. The individual layers either rise or fall with a uniform velocity and from the rate of falling or rising the size of the particles in a given4ayer may be calculated by means of Stokes's law. produces a t most a slight yellow coloration. w. P. s. H. K. w. P. s. J. F. S.