ANALYTICAL CHEMISTRY. Analytical Chemistry. ii. 121 Application of X-Rays to Quantitative Chemical Analysis. R. GLOCKER (Portschr. Geb. Rontgenstr. 1923 31 90-92; from Chem. Zentr. 1923 iv 763).-The author proposes as a measure of the amount of an element present in a substance the intensity ratio of the discontinuity of the K-absorption (Absorptionssprung) when the substance is submitted to a continuous X-ray pencil. For the K-discontinuity f photometrically measured the relation- ship f = e p c is given where p is the amount of the element present and c is a constant which is characteristic for the element and apparently an exponential function of the atomic number. The method has hitherto only been used for the K-absorption. G. W. R. Benzidine as a Reagent and as an Indicator for Specific Oxidation-potential.I. M. KOLTHOFF (Chem. Weekblad 1924 21 24).-The work of Feigl and others (A. 1920 ji 709; 1921 ii 278; 1922 ii 865; 2. anal. Chem. 1923,62,369) and of Olszewski (Chem. Ztg. 1923 47 273) on the detection of metals and phosphates has been examined and confirmed but the sensitiveness in most cases is not found so high as has been stated. An examin- ation of the behaviour of oxidising agents towards benzidine shows that the colour reaction is not dependent on the concentration of the former but on the oxidation-potential. Thus ferric salts give a coloration after one hour in a concentration as low as 1 mg. per litre ferrous salts not a t all; mixtures of ferrous and ferric salts give the same result independently of concentration the effect depending on the proportion of ferrous to ferric compound; no coloration is obtained with a mixture of 1 part of ferric and 9 parts of ferrous salt even in concentration of 10 g.per litre. Similarly iodates give a definite coloration but the result is negative in presence of iodides. If the potential is above the specific point the benzidine is oxidised; if below it remains unaltered but at the specific point the appropriate coloration is developed. Qualitative Reaction of Halogens in Organic Compounds. J. PICCARD and F. DE MONTMOLLIN (Helv. Chim. Acta 1923 6 1020).-The usual copper wire test for halogen in organic compounds sometimes fails because the substance volatilises before it decom- poses. It is recommended to hold a copper gauze in the flame about 1 cm.above the heated substance ; the copper gauze is then attacked by the halogen in the flame giving the green colour. If the substance gives a luminous flame two flames may be used a small one to heat the substance and a larger one above in which the copper gauze is held. To detect halogen in a volatile liquid a piece of filter-paper saturated with the liquid is placed a t the top of a vertical glass tube. The vapours fall down the tube at the lower end of which is attached a bent tube A flame containing copper gauze S. I. L.ii. 122 ABSTRACTS OF CHEMICAL PAPERS. is applied to the cnd of the latter tube and an intense green coloration is obtained. Determination of Chloride in Bleaching Powder. MATSUWO NAKAMURA ( J . Chem. I92d. Japan 1923,26,984-985).-The sodium arsenite-silver nitrate method for the determination of chloride gives no distinct end-point and the manipulation is too complex.In the method proposed the hypochlorite is converted into chloride with hydrogen peroxide solution and titrated with silver nitrate. Fifty C.C. of bleaching powder solution (4.0-6.0 g. per litre) is treated with 3-5 C.C. of 3:4 hydrogen peroxide and then boiled for several minutes for decomposition of the excess of the peroxide. The solution is neutralised with dilute nitric acid and titrated with an 0.1N silver nitrate solution using potassium chromate solution as indicator. The percent,age of the chloride is obtained by sub- tracting that of the available chlorhe from the total. -4 small amount of hydrogen peroxide remaining in the solution is not objectionable in the titration.K. K. S. OCHI ( J . Chem. Ind. Japan 1923 26 1152-1154).-The water in bleaching powder is determined by the combustion method by heating the tube quickly a t 200-250" (cf. ibid. 1923 26 349) but the amor- phous form of bleaching powder produces much chlorine according to the equation Ca(OG1)C1,€I,O=Ca(O€€),+Cl2. By absorption of the chlorine thus produced in a washing bottle containing potass- ium iodide solution next t o the absorption tube the amount of water fixed in the residue can be determined and thus the total water content. K. K. Determination of Bromine in Organic Compounds. TETSUO MAZUME and KIICHIRO KIKO ( J . Chem. Ind. Japan 1923 26 1133-1 139).-The authors have experimentally compared the method of Baubigny and Chavanne (A.1904 ii 203) and that of Stepanow (A 1905 i 335; 1907 ii 50) and Bacon (A. 1909 ii 179) for the determination of bromine in organic compounds finding that the former gives lower results than the theoretical whilst the latter is accurate and simple in manipulation. In the former case nitrous acid is formed and part of the alkali bromide is decom- posed the bromine being lost. In the latter method the titration with ammonium thiocyanate is conducted safely after coagulating the silver bromide without filtering it. Determination of Fluorine in Organic Compounds. J. PICCARD and C. BUFFAT (Nelv. Chim. dcta 1923 6 1047-1048). -The method consists in converting the fluorine into potassium fluoride by heating the compound with potassium a t 400".The compound (fluorobenzene 0.1-0-2 g.) is weighed in a thin-walled sealed glass tube which is introduced into a bomb-tube containing 20 C.C. of absolute ether. About 0.5 g. of potassium is added and the tube is drawn out exhausted and sealed. The ampoule is broken and the tube heated at first gently then a t 400". The E. H. R. Determination of Water in Bleaching Powder. I<. K.ANALYTICAL CHEMISTRY ii. 133 tube is then opened and the potassium fluoride determined by conductometric titration with calcium chloride. E. H. R. Volumetric Determination of [Sulphur in] Sodium Sulphide. V. HASSREIDTER (Chem. Ztg. 1923 47 891-892).- The iodometric method of determining sodium sulphide in solution (A. 1923 ii 790) suffers from the disadvantage that sulphites and thiosulphates which may be present in commercial sodium sulphide will also decolorise the iodine solution. These salts do not affect the determination of sodium sulphide if the latter is carried out by allowing the solution to run into a standard ammoniacal zinc solution until a drop of the latter just stains lead acetate paper.Alternatively the standard zinc solut im may be run into a known volume of sodiuni sulphide solution until the latter just fails to stain the test-paper. Somewhat more accurate results are obtained by allowing a known volume of the sulphide solution to flow into excess of alkaline copper sulphate solution separating qnd washing the copper sulphide dissolving it in nitric acid and depositing the copper electrolytically.One mg. Cu=1.227 mg. Na,S. The copper solution is prepared in a similar way to Fehling’s solution and should contain about 50 g. of copper sulphate crystals per 1. Microchemical Detection of Sulphuric Acid as Silver Sulphate. L. ROSEXTHALER (Mikrochemie 1923 1 47).- Characteristic crystals (rhombic pyramids) are obtained when a sulphate solution is treated with silver nitrate in nitric acid solution ; the crystals of silver sulphate may be obtained from a drop of 1% sodium sulphate solution; oxalates and chromntes do not interfere with the reaction. but chlorides must be removed me- H. C. R. viously by adding the requisite amount of silver nitrate. I w. P. s. Micro-volumetric Determination of Arsenic Anthony and Iron. A. BRUKL (Mikrochemie 1923 1 54-57).-The iodometric method is recommended for the determination of small quantities (less than 1 mg.) of arsenic or antimony N/500 iodine solution is used for the titration with starch solution as the indicator.In the case of ferrous salts the titration is made with titanium trichloride solution; a small combined container and burette for the latter is described means being provided for filling the apparatus New Light Filter. [Detection of Potassium.] LER. W. MCCAY ( J . Amer. Chem. SOC. 1923 45 2958).-A solution of 31 g. of crystallised chrome alum in 100 C.C. of water forms an excellent light filter for use in the detection of potassium. When this filter is used potassium may be detected by the flame test in the presence of sodium lithium strontium barium and calcium. The rays from rubidium and cesium however are not cut off.A Volumetric Micro-method for the Estimation of Sodium. HANS MULLER (Helv. Chim. Acta 1923 6 1152-1161).-The method described has been worked out especially for the deter- with an atmosphere of dry carbon dioxide. w. P. s. J. F. S.ii. 124 ABSTRACTS OF CHEMICAL PAPERS. mination of sodium in serum. It can be used for ordinary or micro- estimations For the latter 0.1 C.C. of serum is diluted to 1 C.C. with water in a cehtrifuging tube and there are added 1 C.C. of potassium pyroantinionate solution (2 g. per litre) and drop by drop 0.4 C.C. of 95% alcohol. After two hours the tube is centrifuged the supernatant solution removed and the precipitate washed three times with 30:/ alcohol. The precipitate of sodium pyro- antimonate is then treated with 1 C.C.of 2% potassium iodide solution 1 C.C. of concentrated hydrochloric acid solution and 2 C.C. of water and after ten minutes the free iodine is titrated with 0.01N thiosulphate. The method depends on the reduction of the antimonic acid in the sodium pyroantimonate Na,H,Sb,O 7 according to the equation Sb,0,+4HI=Sb20,+2H,0-t-21,. To obtain a sharp end-point in the titration the concentration of hydrochloric acid must not exceed the limits 7-9% and not more than four times the necessary potassium iodide should be used. Magnesium calcium and phosphate in amounts exceeding those likely to be present in serum do not affect the titration. E. H. R. Determination of Silver in Molecular Solution in Presence of Colloidal Silver. K.VON NEERGAARD (Arch. expt. Path. Pharm. 1923 100 162-189).-The silver-ions are determined in presence of colloidal silver by means of electrometric titration. Among the results are given the concentration of silver-ions found in a number of therapeutic preparations. Use of Picrolonic Acid €or the Micro- and Histological Detection of Calcium. J. KISSER (Mikrochemie 1923,1,25-31). -Characteristic crystals are obtained when a dilute solution of a calcium salt is treated with picrolonic acid solution ; the reaction may be obtained with 0.01 mg. of calcium. The reaction may be used to detect the presence of calcium in tissues but in this case the form of the crystals obtained is more varied. Calcium oxalate Determination of Calcium in Organic Material by de Waard’s Method. G.HECHT (Biochem. Z . 1923 143 342- 346).-With quantities as low as 0.1 mg. de Waard’s micro- method for determining calcium (A. 1920 ii 53) gives results accurate within 5%. The determination is not affected within these limits by the presence of phosphates magnesium or iron. C. R. H. does not react with picrolonic acid. w. P. s. T D Diphenylcarbazide as a Qualitative Reagent €or Metals. I. M. KOLTHOFF (Chem. W-eekblad 1924 21 20-22).-The form- ation of intensely coloured compounds between diphenylcarbazide and metals (Cazeneuve A. 1900 ii 627) is not satisfactory for general identification since the solutions must always be neutral but the conditions may be so chosen as to make the reaction specific for particular metals. The bluish-violet colour given with mercury salts in dilutions down to 0.1 mg.of metal per litre is very sensitive to the presence of halogen-ions but cupric salts give a colorationANALYTICAL CHEMISTRY. ii. 125 (reddish-violet 0.1 mg. per litre of copper) not affected by halogen- ions so that copper may readily be detected in presence of mercury (5 mg. in presence of 2,000 mg. of mercury per litre). Cadmium may also be detected in presence of mercury and copper if the latter be reduced and halides are added. Magnesium salts give no coloration but the hydroxide gives an intense red by which the metal may be identified in presence of the alkaline-earth metals. Reaction of Potassium Permanganate with Cuprous Oxide in Dilute Sulphuric Acid and the Determination of Metallic Copper and Cuprous and Cupric Oxides in their Mixtures.DENGORO NISHIDA and KEN-ICHI HIRABAYASHI ( J . Chem. Ind. Japan 1923 26 1123-1133).-Powdered pure metallic copper is practically insoluble in 3-6N sulphuric acid at room temperature by contact for twenty minutes; the addition of potassium per- manganate solution to the acid has no influence on the solubility. Cupric oxide is however completely dissolved in 6N sulphuric acid at room temperature; potassium permanganate in the acid has also no influence on the solubility. When cuprous oxide is stirred in 3-6N sulphuric acid a t room temperature it is com- pletely dissolved and potassium permanganate if added into the acid is consumed according to the equation 2KMn0,+4H2S04= 2KHSO4+2MnSO,+3H2O+50 and 5Cu2O +50 + 10H2S0 = lOCuSO,+ 10H20.One C.C. of 0.1N potassium permanganate solution corresponds with 0.007157 g. of the oxide. For analysis 0.5 g. of a mixture of metallic copper cuprous and cupric oxides is dissolved in nitric acid and the total copper content is determined by iodometry. Another weighed sample is stirred with 100 C.C. of 6N sulphuric acid and 1 C.C. of 0-1N potassium permanganate solution to which permanganate solution is gradually added until the pink colour of the solution no longer disappears; the percentage of the cuprous oxide is thus determined. After titration the residue is dissolved in nitric acid and the metallic copper is determined; the cupric oxide can then be calculated. s. I. L. K. K. Experiments Relative to the Determination of Uranium by means of Cupferron.J. A. HOLLADAY and T. R. CUNNING- HAM (Tram. Amer. Electrochem. SOC. 1923 43 329-339).-A number of experiments are described dealing with the estimation of uranium by means of cupferron which show that quadrivalent uranium or uranium in a lower state of oxidation can be pre- cipitated completely by means of a freshly prepared solution of cupferron from solutions containing 2-8 C.C. of sulphuric acid (d 1-84) in 100 C.C. On ignition the precipitate is converted quantitatively into U,O,. If the amount of sulphuric acid is less than 4 .c.c. per 100 c.c. aluminium and possibly phosphorus will be carried down with the precipitate whilst if the acidity is greater than 8 C.C. per 100 c.c. tlhe uranium Will not be completely precipitated. If the acidity is kept between 4 and 8”/ preferably at 6% a sharp separation of uranium from aluminium zmc calcium,ii.126 ABSTRACTS OF CHEMICAL PAPERS. magnesium and phosphorus can b3 obtained by a single precipit- ation. Uranium and vanadium can be separated and estimated with a sufficient degree of accuracy in tlhe presence of widely vary- ing quantities of iron aluminium. calcium magnesium and phos- phorus by the following method. The iron and vanadium are precipitated by cupferron from a 12yo sulphuric acid solution in which the uranium vanadium and iron are present respectively in the sexavalent quinquevalent and tervalent condition. Uranium aluminium calcium magnesium and phosphorus pass quantitatively into the filtrate. The vanadium in the precipitate is then estimated by any of the usual methods.The filtrate is evaporated with nitric acid to destroy Che cupferron and the solution then passed through a Jones's reductor to reduce tlhe uranium. The solution is then diluted t o 60,; sulphuric acid content and the uranium precipitated with cupferron filtered washed and ignited to u,o,. J. F. S. Determination of Iridium in Platinum by the Method of Fusion with Lead. R. Gir.cHItIs'r ( J . Amer. Chem. Xoc. 1923 45 2820-2828).-The analytical details of the Deville and Stas method for the estimation of iridium in platinum alloys containing 0-1-200,& of iridium have been investigated. It is Eound that the concentration of nitric acid t,he concentration of aqua regia the proportion of lead and the time and temperature of the lead fusion can be varied over a wide range without any efiect on the determination.The observations of Ueville and Stas to the effect that palladium and rhodium have no effect on thc determination and that ruthenium separates quantitatively with the iridium are confirmed. Gold is also found not to interfere with the determin- ation. Iron separates almost quantitatively with the iridium as observed by Deville and Stas. Iron can be separated satisfactorily from iridium by the following method. The alloy is fused with zinc and the excess of zinc removed by treatment with hydrochloric acid. The resulting alloy is then fused with potassium pyrosulphate. The fusion after cooling is extracted with dilute sulphuric acid. The iron all dissolves and the insoluble residue contains the iridium and silica.The loss in weight of crystalline iridium during the ignition periods is insignificant and thc weight of the crystalline iridium is not affected by heating and cooling in an atmosphere of hydrogen. Spectrographic examination of samples of iridium from the analyses showed that neither lead nor platinum is present in quantities sufficient to afiect the determination. The iridium results tend to be low by a small but variable amount. One factor in this error is a slight dissolution of iridium by aqua regia. A modified method is presented which combines the optimum of speed and accuracy. The modified method is as follows The alloy is fused with ten times its weight of granular lead for one hour a t about 1,000" in a graphite crucible and then allowed to cool.The button is removed and brushed to free it from carbon placed in a beaker containing 1 volume of nitric acid (d 1-42) to 4 volumes of water using 1 C.C. of acid per Q . of lead. The beaker is keptANALYTICAL CHEMISTRY. ii. 127 a t 85" for two hours when the alloy will be disintegrated. The solution is diluted to twice its volume and decanted through a double filter-paper. The residue is washed with hot water and the washings passed through the filters. The filter-papers are returned to the beaker without ignition. To the beaker is now added 15 C.C. of water 6 C.C. of hydrochloric acid (d 1.18) and 0.8 C.C. of nitric acid (d 1-42) for each gram of sample taken and the mixture heated a t 85'; in about one and a half hours the lead platinum alloy will be dissolved.The solution is diluted with twice its weight of water and filtered through a double filter-paper in which the iridium collects in fine crystals. The precipitate and filter-paper are washed with hot water and finally with hot dilute hydrochloric acid (1 100). The iridium and filter-papers are placed in a Rose crucible and dried the filter-paper is then destroyed and the residue heated with the full heat of a Tirrill burner. The ignited metal is then heated in hydrogen for five minutes and allowed to cool in the gas. The Colorimetric Determination of Cholesterol in Blood- serum. s. KRASTELEVSKY (Biochem. Z. 1923 143,403407),- Of the colour reactions of cholesterol the Salkowski reaction is found to be best suited for determining the substance in blood by colorimetric methods.The colour tone is not affected by the presence of other plasma constituents if water is excluded. One C.C. of the serum from 5-10 C.C. of blood is added to 5 C.C. of absolute alcohol shaken and kept in a thermostat a t 55-66' for twelve to twenty-four hours. To the dried powdered residue is added 5 C.C. of chloroform and after fifteen to twenty minutes 5 C.C. of pure sulphuric acid (d 143388). After shaking the mixture is left in an ice-chest for six to twelve hours and the chloroform layer is removed and compared with the standard colours of various known concentrations of pure cholesterol. Solutions of Bayer's Benzobrown and Benzoazurin in aqueous alcohol may be used as standards. The normal cholesterol content of blood-serum is 0~0005-0~0012 g.per C.C. J. P. Determination of Dextrose by Oxidation with Iodine. F. AUERBACH and E. BODLANDER (2. angew. Chern. 1923 36 602-607) .-Under the conditions described by Willstatter and Schudel (A. 1918 ii 337) the oxidation of dextrose by iodine proceeds beyond the gluconic acid stage and laevulose and sucrose are also to some extent attacked. By adjusting the alkalinity of the reaction medium however and the time of reaction a quantitative oxidation of dextrose is effected without any appreci- able reaction taking place with lzvulose and sucrose. To the sugar solution which should preferably contain not more than 200 mg. of dextrose in about 25 c.c. a quantity of an N/10 solution of iodine in potassium iodide is added such that a t least one-third to one-half of the iodine remains unchanged.To the solution is then added 100 C.C. of a mixture of equal volumes of one-fifth-molar solutions of sodium carbonate and sodium hydrogen carbonate (having pH 10.1-10.2) and this reaction mixture is kept in the J. F". S.ii. 128 ABSTRACTS OF CHEMICAL PAPERS. dark for one and a half to two hours. It is then acidified with 25y0 sulphuric acid (12 c.c.) and the unchanged iodine is titrated back with N/10 thiosulphate and starch. The quantity of thio- sulphate required is subtracted from that required in a blank experiment carried out simultaneously and of the difference 1 C.C. of the N/10 solution is equivalent t~ 9.005 mg. of dextrose. Under these conditions lactose also is quantitatively oxidised 1 mol. requiring two equivalents of iodine and the same applies a t any rate approximately to maltose.The method may be used to distinguish between natural and artificial honey and to detect adulteration of honey with artificial invert-sugar or dextrose. [Cf. R. 107.1 W. T. K. R. Comparison of the Action of Chlorine and Chlorine Dioxide on Wood. E. HEUSER and 0. MERLAU (Cellulosechem. 1923 4 101-109) .-An investigation of the method proposed by Schmidt and Graumann (A. 1921 ii 912) for the removal of incrustive substances from plants for example lignin from wood by means of chlorine dioxide. It is shown that the method is serviceable as a means for the rapid separation of raw+ellulose so-called " skeleton-substance " from wood and that under certain con- ditions it is a useful alternative method for the quantitative analysis of wood.When wood is treated by the Schmidt and Graumann method hexosans remain in the " skeleton-substance " whereas pentosans are found in both the cellulose and the lignin portions. For instance spruce wood freed from ash and resin and containing 10+30~0 of pentosans and 6.38% of mannose yielded 62.71 yo of " skeleton-substance," which contained 4.4% of pentosans 5.4374 of mannose and 52.84% of pure cellulose. By dialysis and subsequent evaporation of the lignin extract obtained by Creatment of wood with chlorine dioxide and aqueous sodium sulphite 70% of the lignin originally present in wood was obtained in a dry state and this contained 8-16y0 of pentosans. Hence in the determination of pure cellulose and lignin in wood by means of the method of Schmidt and Granmann allowance must be made for the presence of mannose and pentosans in thc " skeleton-substance " and the lignin extract.The pure cellulose- content of wood as determined by the methods of Schmidt and Graumann and of Cross and Bevan was 52.2% and 52.8% re- spectively and the lignin content as determined by the former method and by Willstatter and Zechmeister was 30.3% and 2S.01~0 respectively. Chlorine dioxide has an appreciable action on xylan but not on pure cellulose and cotton. Pentosans present in wood are partly removed by treatment with hot water* and hot dilute solutions of sodium sulphite (cf. Euler following abstract) but this solubility is not dependent on simultaneous processes of oxidatioii or hydrolysis.Methods for carrying out the analysis of wood by the method of Schmidt and Graumann are fully described and also ail improved form of Gooch crucible for use with these methods. A. J. H.ANALYTICAL CHEMISTRY. ii. 129 Determination of Lignin in Wood by the Method of Schmidt and Graumann. A. C. \TON EULER (CeZZuZosechem. 1923 4 109-1 13).-In view of the considerable difTerence (9.37 %) between the lignin-content of pine wood as determined by Willstatter and Zechmeister and by the method of Schmidt and Graumann (A. 1921 ii 912) the latter method has been critically examined. Since albumin is not affected by treatment with an aqueous solution of chlorine dioxide it is probable that about O4y0 of wood-gum containing xylose mannose and dextrose will remain in the so- called carbohydrate " akeleton-substance " ; the latter also retains about 0.2% of mineral substances so that an error of l"/b is possible in the determination of the carbohydrate content.A more serious error arises from the fact that when wood after successive extrac- tion with benzene and alcohol whereby fats resins and turpentine are removed is subjected to the various processes of washing by means of hot and cold water and aqueous solutions of sodium sulphide as described in the Schmidt and Graumann method it suffers a considerable loss of weight due to the solubility under such conditions of wood-gum and lignin. For instance one sample of pine wood freed from fats resins and turpentine was subjected to thirteen successive treatments comprising steeping in water for twenty-four hours washing with hot and cold water extraction with a 2% aqueous solution of neutral sodium sulphite and subsequent washing with water and the respective losses in weight were 6.5 2.8 0;7 1-5 1.5 0.9 0.7 1.2 1.7 1-2 0.6 0.4 and 0.7% (total=20.4%).The Schmidt and Graumann method is therefore likely to give a low carbohydrate-content and a corre- spondingly high lignin-content. Attention is directed to the apparent solubility of lignin in a neutral aqueous 2% solution of sodium sulphite but this solubility may be due to the formation of a small amount of sodium hydrogen sulphite during hydrolysis of the wood complex. A. J. H. Detection and Determination of Tartaric Acid. M. FRAN~OIS and C. LORMAND ( J . Pharrn. Chirn. 1923 [vii] 28 433-442).-In Kling's method for determining tartaric acid in wine (cf.A. 1910 ii 359) which depends on tJhe precipitation of the very insoluble calcium racemate by addition of Z-tartaric acid in excess and calcium acetate incorrect results are liable to be obtained on account of the sparing solubility of calcium Z-tartrate. The present authors find the solubility of the calcium tartrates in water a t 20" to be calcium d-tartrate and calcium Z-tartrate both C,H40,Ca,4H,0 0.232 and 0.249 g. per l. respectively ; calcium racemate 0.0493 g. per 1. The solubility of the d-tartrate in alcohol of 32 vol. yo strength is however sufficientlysmall (0-038 g. per l. a t 20') to allow of the direct determination of tartaric acid by precipitation of this salt. The sample for analysis should contain about 0.4-0-6 g.of tartaric acid and is made up to 100 c:c. with water. To this solution is added 20 C.C. of a solution of calcium acetate (32 g. of pure calcium carbonate and 120 C.C. of glacial acetic acid made up to 1 1. with water) and then after stirring,ii. 130 ABSTRACTS OF CHEMICAL PAPERS. 30 C.C. of alcohol of 95 vol. ?h strength. After keeping the mixture for twenty-four hours a further 30 C.C. of the alcohol is added. After a further twenty-four hours the crystals of calcium d-tartrate are collected preferably in a Gooch crucible dried in the air a t room temperature and weighed. The results are accurate to about 1%. The characteristic appearance of the crystals of calcium d-tartrate may be used for the detection of the acid.W. T. K. B. Determination of Malic Acid in Fruit Syrups and other Fruit Products. F. AUERBACH and D. KRUGER (2. Unters. Nahr. Genussm. 1923 46 177-217).-A 25 C.C. sample of the material is acidified with h7/2-hydrochloric acid allowing an excess of 0.5 C.C. over the ash-alkalinity. The latter is determined by titration to Methyl-orange. In the case of ( a ) natural juices lemonade etc. (with small sugar content) 100 C.C. of 96% alcohol is added and ( b ) syrups marmalade etc. (containing much sugar) 400 C.C. of alcohol is needed. The solution is filtered through a dry filter and if the volume of the precipitate is considerable a correction is made for its volume. Ten C.C. of the filtrate is titrated with NjlO sodium hydroxide solution (litmus) and the amount of barium carbonate required to neutralise is calculated.Three hundred and fifty C.C. of filtrate ( b ) or 75 C.C. of filtrate ( a ) to which 250 C.C. of alcohol is added is neutralised with the requisite amount of barium carbonate. Two C.C. of 50% barium acetate solution is added and the liquid heated on a water-bath until precipitation is complete. The precipitate is filtered after twenty-four hours and washed with 95y0 alcohol and dried in a steam oven. Pre- cipitate and paper are placed in a 50 C.C. graduated flask 0.1 g. of barium citrate and 0.05 g. of barium tartrate are added and the whole is diluted with water nearly to 50 C.C. After several hours' shaking the volume of liquid is made ul) to the mark with a saturat'ed solution of limium citratc and tartrate. The liquid is filtered and if necessary decolorised with animal charcoal.The malate in the solution is determined polarimetrically . For this 20 C.C. of solution is sliakcn for four hours with 3.5 g. of uranyl acetate and 2.5 C.C. of dljl0 dkodiuni citrate solution and diluted to 25 c.c.; or 10 C.C. of solution is mixed with 2 C.C. of glacial acetic acid and 10 C . C . of satnmted aqueous ainmonium inolybdatc solution decolorised if necessary with animal charcoal and diluted to 25 e.c. The necessary corrections are described and the effect of various sugars amino-acids p r e s ( ~ v r ~ t i v ~ s chtc. on the accuracy of the method detailed. A. G. P. Determination of Hydpoxyl Values [of Oils and Fats] by Normann's Method. W. MEIGEN and 0. RAMGE (Cheni.Umschau 1024 31 3-4).-Norinann's method of determining the hydroxyl values of oils and fats was tested on specially purified samples of mono- and di-hydroxystearic acids and on mixtures of known proportions of these acids with cacao butter and olive oil. In all cases the values obtained were somewhat higher thanANALYTICAL CHEMISTRY. ii. 131 the theoretical the discrepancies ranging froni 1% to 4%. The fact that high values were obtained by this method was confirmed by parallel determinations of the hydroxyl values of castor and rape oils and of the methyl esters of the mixed fatty acids prepared from these oils. In each case the value obtained for the oils themselves was higher than that calculated from the hydroxyl value of the methyl esters of the mixed fatty acids.According to Griin (Uel & Fett. Ind. 1919 Nos. 13 14) the theoretical value is obtained by applying Normann's method to the methyl esters. When however the hydroxyl value is not too large the error introduced falls within the experimental error and in such cases Normann's method may be considered applicable in view of its convenience. H. C. R. A. EIBNER and K. SCHMIDINGER (Chem. Umschau 1923 30 293-302).-The linseed oil examined was of Dutch origin and was of iodine value 173.6 and acid valuc 2-3. Its composition was as follows or-linolenic acid 20.1% ; isolinolenic acid 2.7% ; a-linoleic acid 17.0% ; p-linoleic acid 41.8% ; oleic acid 4.5% ; hydroxy-acids 0.5% ; glycerol 4.1% ; saturated acids 8.3% ; phytosterol 1.0%. By treating the oil with nitrous acid 0.6% of an elaidin was obtained which was identified as di-elaidopalmitin.Bromination of the oil in alcoholic solution gave 54.2% of crude solid bromide which on purification was identified as brominated di-a-linolenic-a-linoleic glyceride. The oil contained about 25% of this mixed glyceride. The form in which the remaining unsaturated acids were combined was not determined. [Cf. B. Feb.] H. C. R. G. WALKER (J. Xoc. Chem. Ind. 1923 42 497-498~).-1t is suggested that the percentage of cineole in essential oils may be determined from the freezing point and density of a mixture of the oil with a-naphthol. To make such determinations possible measurements have been made of freezing points and densities a t 20"/4" of mixtures of 3.6 g. of a-naphthol and 3.85 g.of artificial eucalyptus oil con- taining from 65 to 95% of cineole. These oils were prepared by adding pure cineole to various eucalyptus oils free from cineole. Curves are given showing the change of density with the percentage of cineole for each variety of oil used namely pinene Eucalyptus mdiata E. nora angelica E. citriodora E. dityes and E. cineritolia residues. E. H. R. Quantitative Analysis of Linseed Oil. The Determination of Cineole in Essential Oils. Microchemical Identification of a-Monoamino-Acids. 0. WERNER (Mikrochemie 1923,1 3346).-A microchemical method for the separation and identification of the more important a-mono- amino-acids is described and also outlined in tabular form; it consists essentially in the sublimation of certain of the substances in a special apparatus.The operation is carried out under reduced pressure and means are provided for cooling the receiver. The character of the crystalline sublimates (in those cases where the substance sublimes) and of the crystals etc. obtained when theii. 132 ABSTRACTS OF CHEMICAL PAPERS. sublimates and residues are treated with cunric livdroxide reagent or phosphotungstic acid affords a means of ;dentaying the amino- acids. w. P. s. Amylolytic Enzymes. R. FABRE and H. P ~ N A U ( J . Phamz. Chim. 1923 [vii] 28 280-304 341-348).-The causes of thc variations observed in determining the amyloclastic activity of pancreatin and diastase preparations with commercial potato starches are critically examined and they are shown to be principally influenced by the character of the water employed in the manufacture of the starch (cf.A. 1923 ii 512). Water with,a pH on the alkaline side activates the pancreatic diastase and inhibits malt diastase whilst the reverse is the case with water having a pII on the acid side. I n order therefore to obtain comparable results it is recommended to prepare the starch employed for such determinations under standard conditions of maceration to avoid influences which tend to produce an extensive demineralisation of the starch granules and that the water employed in its preparation should be absolutely neutral and to adopt as analytical method either that of Bertrand or that of Grimbert’s modification of Lehmann’s method (A. 1913 ii 254) both of which are more exact than the method of direct reduction. Twenty-five to 30% of the sugar produced during the saccharification under the conditions required by the pharmacopoeia is soluble in absolute alcohol. With malt diastase maltose is the only sugar produced whilst in the case of pancreatic dimtase which contains maltase small quantities of dextrose are also formed [Cf. B. 27.1 D. R. N. The Inorganic Elements of Blood Plasma. A. P. BRIGGS ( J . Biol. Chem. 1923 57 351-357).-A system of analysis of the inorganic constituents of b,looci has been developed by means of which determinations may be made of the sodium potassium calcium magnesium chloride and phosphate content of blood using one small sample. The important point is the choice of an anti-coagulant which will not interfere with the subsequent deter- minations. Lithium citrate lias been selected as the most suitable for this purpose. Nine to 15 C.C. of plasma obtained from blood treated with this anti-coagulant is sufficient for the complete analysis. Proteins are precipitated with 20% trichloroacetic acid and the potassium is determined in a portion of the filtrate by a colorimetric method based on its precipitation as cobaltinitrite and the determination of the latter by Griess’s nitrite method. Separate portions of the filtrate are used for determining the remaining constituents slight modifications of existing methods being used for this purpose. The method has been applied to a number of pathological bloods. E. S.