ii. 66 ABSTRACTS OF CHEMICAL PAPERS. An alg t ical Chemistry . Improved Method of Collecting Gases from the Mercury Pump. FREDERICK G. KEYES (J. Amer. Chem. Xoc. 1909 31 1271-1273).-1n the ordinary metbod of collecting gases from the mercury pump in an inverted tube and transferring them to the gas analysis apparatus minute bubbles of air are liable to be trapped even when great care is exercised. The following arrangement has been devised in order to avoid this source of error. A wide tube for collecting the gases is sealed on to the up-turned end of the capillary tube of the pump The wide tube is provided a t the lower end with a side-tube which is connected to a reservoir of mercury by means of rubber tubing. The upper end of the wide tube bears a three-way stopcock furnished with capillary leads.Towards tbe end of an ordinary exhaustion the air in the wide tube can be expelled by raising the reservoir. On closing the stopcock and lowering the reservoir a vacuum is formed in the tube into which the bubbles passing down the pump capillary will be discharged. The method of employing the apparatus for the collection of gases is very simple and is described with the aid of a diagram. E. G. Automatic Filling Burette. R. FRAILONG (Bull. Assoc. Chim. sucr. dist. 1909 27 470-471).-A description of two kinds of automatic burettes one intended to deliver a definite quantity of reagent at once and the other for delivering the same drop by drop. The main parts of the apparatus are a reservoir with a tube supplying the burette a specially-constructed Mohr’s pinchcock which serves to open the outlet tube while closing simultaneously the supply tube and a capillary tube passing through the rubber cork of the burette and dipping slightly into the liquid; i t serves t o obtain the zero point automatically.L. DE I(. Apparatus for the Rapid Elec tro-analytical Separation of Metals. HENRY J. S. SAND (Trans. Furaday SOC. 1909,5,159-164). -A description of some developments made in the author’s apparatus for the rapid electro-analysis and separation of metals (compare Trans. 1907 91 373; 1908 93 1572). A special screw-cap has been provided which may be screwed down when the apparatus is not in use making it possible to transport it without taking out the mercury which is employed to make connexion between the stationary and moving parts in the electrolytic stand.A clutch arrangement has also been added which enables the operator to start or stop the rotation of the anode without stopping the motor. All the apparatus required for the measurement of the electrode potential is now fitted into a single box. The arrangement is so designed that by depressing a keg it will also allow the potential difference between the anode and the cathode to be read directly. TheANALYTICAL CHEMISTRY. ii. 67 capillary electrometer is retained as a zero instrument and a special portable form which is a modification of the Ostwald horizontal capillary electrometer has been designed. It is provided with an enclosed scale and will readily indicate one millivolt when observed with a lens T.S. P. Aluminium Dishes and other Appliances in Quantitative Analysis. JAROSLAV FORMANEE and FRANZ PEG (Chem. Zeit. 1909 33 1282-1283).-Dishes of aluminium may be substituted for those of platinum in electrolytical separations but it is necessary to have the inside of the dish plated with n very thin sheet of copper This may then in turn be coated electrolytically with tin or silver etc. according to circumstances A number of successful experiments are communicated. Several improved appliances for quick electrolytic estimations are also described. L. DE I(. Electrolytic Estimation of Chlorine in Hydrochloric Acid with the Use of a Silver Anode. FRANK A. GOOCH and H. L. READ (Amer. J. Xci. 1909 [iv] 28 544-552; Zeitsch. anorg. Cham. 1909 64 287-297).-The authors have shown by a large number of experiments that when using anodes of silver or silver-plated platinum gauze and currents of various strengths the results are not so satisfac- tory as might have been expected.This is caused by the fixing of oxygen as well as chlorine on the anode the removal of silver from the anode to the cathode and the formation of hypochlorous acid. Even when the anode is ignited to decompose silver oxides the results are still irregular and always low. L. DE I(. The Beilatein Reaction [for Halogens]. Hvao MILBOTH (Cham. Zeit. 1909 33 1249).-The author has shown that the Beilstein copper oxide flame test for halogens is not absolutely trustworthy as owing t o the formation of cupric cyanide a green coloration is occasionally noticed which although not quite 80 characteristic may be mistaken for the halogen colour.The reaction is given by picolinic acid quinolinic acid 2-methyl- picolinic acid dipicolinic acid methyl 2-hydroxynicotinate quinaldinic acid and nicotinic acid whereas no green flame was observed with quinoline isoquinoline isonicotinic acid cinchoninic acid cinchomeronic acid methylanthranilic acid apophylenic acid and papaveric acid. In the case of hydroxyquinoline the reaction is obtained even without the use of a copper oxide bead. L. DE SE(. Analysis of Mixtures of Halogen Acids. 11. WILLIAM M. DEHN (J. Amer. Chem. Xoc. 1909 31 1273-1275. Compare Abstr. 1909 ii 612).-The following method is recommended for the aaalysis of mixtures containing two halide ions. Silver nitrate is added in excess and the precipitated silver salts are collected and weighed.The solution containing the excess of silver nitrate is titrated with thiocyanate and ferric nitrate by the Volhard method. It is shown that if a is the weightl of silver nitrate required to precipitate both the halogens present and b is the weight of silver VOL. XCVIII 6ii. 68 ABSTRACTS OF CHEMICAL PAPERS. halides precipitated the quantities of the halogens can be calculated by the following equations For mixtures containing chlorine and bromine Cl=0"317a - 0.79763; Br= 1.79763 - 1.5166a.' For mixtures containing chlorine and iodine C1= 0 5 3 5 8 ~ ~ - 0.38773 ; I= 1.38773 - 1.1706a. For mixtures containing bromine and iodine Rr = 2 . 3 5 0 1 ~ ~ - 1.70043 ; I = 2.70073 - 2-9S51a.Mixtures containing three halide ions can be analysed by the same method if one of the halogens is estimated separately. Thus if iodine is estimated separately we have C1 = 0.S817a - 0 79763 (+ 0.2954 I) ; Br = 1.79763 - 1 . 5 1 6 6 ~ ~ ( - 1.2951 I). If the bromine is estimated separately C1 = 0 . 5 3 5 s ~ ~ - 0.38773 ( - 0 2280 Br); I = 1.38773 - 1-1706a (0.7720 Br). If the chlorine is estimated separately Br = 2.3501tc - 1-7004b ( - 0,4386 C1) ; I = 2.70073 - 2.9851a ( + 3.3557 Cl). E. G. Estimation of Perchlorates by means of Titanous Salts. VICTOR ROTHNUND and A. BURGSTALLER (Chem. Zeit. 1909 33 L245). -Contrary to Stahler's statement (Abstr. 1909 ii 699) that the estimation of perchloric acid by boiling with excess of titanous sulphate and then titrating with permanganate is unsatisfactory the authors find that the results are trustworthy provided the operation takes place in a flask fitted with a cooling arrangement and in an atmosphere of carbon dioxide. The chloride formed in the reaction may be estimated by way of a check but the use of the nephelometer cannot be recommended in this case.L. DE K. Simple Process for the Estimation of Iodine. ERNST WINTER- STEIN and E. EERZFELD (Zeitsch. physiol. Chem. 1909 63 49-57),- The apparatus consists of a 250 cc. round-bottomed flask with a narrow neck which is closed by means of a ground joint through which pass two tubes like those of an ordinary washbottle. The exit tube is also furnished with a ground joint by means of which it is connected with a short condenser.The end of the condensicg tube is connected with two absorption flasks filled half way with a 10% solution of potassium iodide. Fifty C.C. of the solution to be tested for small quantities of iodides (bromides or chlorides do not interlere) are placed in the flask together with 5 C.C. of phosphoric acid and 10-20 C.C. of commercial hydrogen peroxide. A current of air is drawn through the apparatus and the solution is heated gradually to boiling. After thirty minutes the iodine is expelled and absorbed by the potassium iodide and may be titrated. When applying the process to urine the results are only qualitative but the full amount of iodine present may be obtained by rendering the urine alkaline with sodium hydroxide evaporating to dryness and destroying the iodised organic compounds by ignition.L. DE I(. Estimation of Iodine in Organic Substances. VINCENZO PAOLINI (Mon. Xci. 1909 [iv] 23 648).-The following method is recommended for the estimation of iodine in such substances as blood,ANALYTICAL CHEMISTRY. ii. 69 urine brain-substance sea-weed thyroid glands etc. A particular instance is given of the estimation of iodine in iodoform-gauze. Ten grams of the gauze cut into small pieces are treated in a flask with 40 grams of zinc dust and 60 C.C. of 25% sulphuric acid. The mixture is heated by means of a water-bath in a reflux apparatus for about three hours a further 50 C.C. of sulphuric acid being then added and the heating continued for some hours. The iodine present is thus converted into hydriodic acid. The gauze is then mashed several times by decantation with water and the solution made up to a volume of 1 litre.One hundred C.C. of this solution are treated with a few c.0 of potassium nitrite solution and extracted with 100 C.C. of carbon disulphide. The carbon disulphide is then separated washed with water and the iodine dissolved in i t is titrated with standard thio- sulphate solution in the presence of sodium hydrogen carbonate. w. P. s. Estimation of Iodine in the Thyroid. ATHERTON SEIDELL (J. Amer. Ciiem. Xoc. 1909 31 1326-1329).-Riggs (Abstr. 1909 ii 6 0 9 ) has stated t h a t the estimation of iodine in the thyroid by Baumann’s method (Abstr. 1896 ii 487) is liable to error owing t o the formatiou of iodate during the fusion and he has therefore suggested a modification involving a reduction process.I n the present paper an account is given of experiments which indicate that Baumann’s method is satisfactory and that Riggs’ modification is not only unnecessary but actually leads to greater error than the original method since it is difficult to remove the whole of the iodine from the aqueous layer before applying the reduction process and any iodine remaining in solution renders the results inaccurate. E. G. Estimation of Sulphur Dioxide and Sulphuric Acid in the Gases of Sulphur Furnaces. HENRI PELLET (Bull. Assoc. Clhn. SUCT. dist. 1909 27 468-469).-The author described his sulphi- carbonimeter some ten years ago in a pamphlet issued for private circulation and now communicates some results obtained. The pro- cess is based on the fact that on passing the gases through water the sulphuric acid is removed the sulphur dioxide being absorbed by a Estimation of Dithionic Acid and Dithionates.HENRI BAUBIGNY (Compt. rend. 1909 149 1069-1071. Compare Ashley Abstr. 1906 ii 800).-Preliminary experiments showed that it was not possible to oxidise dithionic acid completely to sulphuric acid by heating with aqua regia either in open or closed vessels. The solution containing the dithionate is mixed with a sufficient amount (12 t o 15 times the quantity of dithionate present) of a mixture of sodium carbonate aad nitrate. If a precipitate is formed it is filtered off and washed; the filtrate and washings are then evaporated to dryness and the residue carefully fused during from ten to fifteen minutes. The sodium sulphate so formed is estimated in the usual way with barium nitrate in presence of nitric acid.suitable reagent contained in a second washbottle. L. DE K. T. A. H. 6-2ii 70 ABSTRACTS OF CHEMICAL PAPERS. Estimation of Nitrogen in Soil Extracts. DENSCH (Chem. Zeit. 1909 33 1249-1251).-A reply to Mitscherlich (Abstr. 1909 ii 935).-The author upholds the accuracy of his process and prefers it to the more scientific method of Mitscherlich in the case of analysis of the aqueous extract of soils. For the reduction of the nitrates (any nitrite is first oxidised by means of permanganate) in acid solution a mixture of iron and zinc powder is recommended although iron alone Method of Estimating very Small Amounts of Nitrogen. T. ZELLEB (Landw.Verszcchs.-Stat. 1909 71 437-440).-The method described by Mitscherlich Herz and Merres (Abstr. 1909 ii 614) is considered impracticable owing t o the relativelx large and variable amounts of nitrogen unavoidably introduced in the reagents. may be used. L. DE K. N. H. J. M. New Distillation Arrangement for Ammonia Estimation. ADOLF BERTHOLD (Chem. I A%. 1909 33 1292).-By using this apparatus all loss of ammonia is avoided. The substance is placed in the flask d and the centre-piece b is put on. The receiver f which contains the standard acid is then raised by iz wooden block g until the delivery tube dips into the acid. The funnel c is now filled with alkali and by gently lifting the tube e the liquid is allowed to run into d but a little should be left in t h e funnel so as to form a seal; the remainin5 alkali is then removed by rinsing a few times with water still retaining the seal.Alter connecting the condenser a with the water supply the contents of d are heated over a Bunsen burner. When the dis- tillation is practically finished the wooden block is removed and the distillation con- tinued for a short time. When dealing with liquids contain- ing free ammonia these should be intro- duced through the funnel tube c after placing on the centre-piece and raising the receiver. Into e a drop catcher has been sealed. L. DE I(. Estimation of very Small Amounts of Ammonia in Large Quantities of Air. PAUL LIECRTI and ERNST &m!ER (Chem. Eeit. 1909 33 1265-1266).-The apparatus consists of a U-tube contain- ing diluted N-sulphuric acid one limb of which is attached to a vertical spindle-shaped tube which is connected to the first of two pipettes inclined at about 45' t o the vertical; the top pipette is attached to a second U-tube.By means of a ventilating fan the air is drawn very rapidly through the apparatus. The acid is sweptANALYTICAL CHEMISTRY. ii. 71 mainly into the spindle-shaped tube where it is churned violently by the air current and a good deal of the acid is converted into spray; this however settles on the walls of the two pipettes and flows back; it is found that practically every trace of ammonia is removed. L. DE K. New Arrangement for the Estimation of Nitric Osmpounds in Sulphuric Acid. K. LEO (Chem. Zed. 1909 33 1218-1220).- Instead of shaking the sulphuric acid with mercury (Lunge’s process) 50 C.C.of the sample are placed in a specially constructed bulb the remaining space being taken up with mercury. By means of a tube furnished with three openings and connected with a mercury pressure flask mercury is forced into the acid like a spray and the action soon starts. The mercury displaced runs into another reservoir and from there into a beaker and is afterwards emptied into the pressure flask. The nitric oxide liberated is then transferred to a gas volumeter and measured. For working details the illustrations in the original article should be consulted L. DE K. Reduction of Nitric Nitrogen to Ammonia; New Procem for the Estimation of Nitrates. M. XMMANUEL POZZI-ESCOT (Bull. Aseoc. China sucr. dist. 1909 27 457-4459 ; Ann.Chim. aml. 1909 14 445-446).-The nitrate is placed in a distilling flask fitted with a funnel safety tube 5 to 6 grams of aluminium cuttings and 2 C.C. of a saturated solution of mercuric chloride are added and also 150 to 200 C.C. of water. After connecting the flask with the condensing apparatus a brisk evolution of hydrogen will be noticed; a solution of sodium hydroxide is added and the ammonia formed is distilled off and collected as usual in standard acid. I n order to decompose traces of ammoniated mercury compounds a few C.C. of sodium bypophosphite solution are added towards the end. L. DE K. Modifbation of -the Gtrandval and Lajoux Process for the Estimation of Nitrates in Waters Charged with Chlorides. L. PAECY (Bull. SOC. chim.1909 [iv] 5 1088-1090).-The author stated previously that in the presence of chlorides the process used by Graodval and Lajoux failed t o give correct results and he con- structed correction tables. * By operating as follows and using an improved reagent the use of these tables may be avoided. One volume of a mixture of 37 parts of sulphuric acid and 3 parts of phenol is mixed with 1 volume and a-half of hydrochloric acid and 1 volume and a-half of water and heated for half an hour on the water-bath. The residue left on evaporation of the sample is moistened with 1 C.C. of the reagent and heated on the water-bath for fifteen minutes. The residue is then rendered alkalirie with ammonia and the coloration compared with that of a standard watar similarly treated. L.DE K.ii. 72 ARSTRACTS OF CHEMtCAL PAPERS. Influence of Nitrites on the Estimation of Nitrates by Grandval and Lajoux’s Process. L. FARCY (Bull. SOC. chim. 1909 [iv] 5. 1090-1091).-The author states that although nitrites do not themselves affect Grandval and Lajoux’s reagent the least trace of nitrate added gives a coloration far more powerful than would be expected. The nitrite should therefore be estimated separately by means of Griees’ reagent and before estimating the nitrate it should be oxidised with permanganate or else destroyed by means of carbamide. L. DE K. Influence of Chlorides on the Estimation of Nitrates. L. FARCY (Bull. SOC. chim. 1909 [iv] 5 1091).-The loss in nitrate observed when applying the Grandval and Lajoux colorimetric test for nitrates in presence of chlorine is due to the formation of nitro-hydro- chloric acid; this may be proved by passing the gaseous products of the reaction through a solution of potassium iodide when the amount of iodine liberated corresponds with the loss observed.A slight mechanical loss owing to effervescence is also unavoidable. L. DE K. Coloured Substances Produced in Grandval and Lajoux’s Reaction. MAURICE LOMBARD (Bull. SOC. chim. 1909 [iv] 5 1092-1096).-The coloration is mainly caused by a mixture of o-nitrophenol-p-sulphonic acid and o-nitrophenol. The latter substance gives a more intense red coloration with ammonia. L. DE K. Estimation of Mineral Constituents in Vegetable Substances. L. VUAFLART (Bull. Assoc china. sucr. dist 1909 27 454-456).- Polemical. A reply t o Pellet (Abstr.1909 ii 755) on the subject of the volatilisation of phosphoric compounds on incinerating vegetable substances. The errors are avoided by using the moist combustion process with suiphuric acid. The use of alcohol when estimating nitrates is also discussed also the possibility of some magnesium ammonium phosphate remaining on the filter when dissolving the yellow molybdate precipitate in dilute ammonia containing some citrate solution. L. DE K. Estimation of Mineral Constituents in Vegetable Substances. HENRI PELLET. (Bull. Assoc. Chim. sucr. dist. 1‘309 27 456-457).- Polemical. A further reply co Vuaflart (preceding abstract). The author did use citrate solution and not nitrate as reported erroneously. Attention is called to the fact that t h e application of alcohol in the eRtimatiorr of nitrates has already been advocated by Bouesingault Estimation of the Free Acid in Superphosphates.FREDERICK B. GUTHRIE and A. ALEXANDER RAMSAY (J. Roy. SOC. New South WaZes 1909 43 G9-74).-The methods recommended for the estimation of free phosphoric acid have been examined. Herzfelder’s method i u which the free acid is extracted with ether and titrated in aqueous solution with methyl-orange as indicator gives the best results but in 1865. L. DE I(.ANALYTICAL CHEMISTRY. ii. 73 the values obtained are rather low. A modification of this method in which sodium alizarinsulphonate is used as indicator will probably prove to be the best method of estimating free phosphoric acid in superphosphates. E. M. D. Estimation of Total Phosphoric Acid in Basic Slags and Native Phosphates by the '' Citro-mechanic Method." E.GUERRY and E. TOUSSAINT (Bull. SOC. chirn. Belg. 1909 23 454-457).-Twenty to twenty-five C.C. of sulphuric acid are intro- duced into a 250 C.C. Jena flask by means of a funnel tube so as not t o moisten the neck. 2.5 Grams of the powdered sample are added and after gently shaking the whole is boiled for ten minutes. When cold water is added a little at a time and when cooled down the solution is diluted to the mark well shaken and filtered. Twenty- five C.C. of the filtrate are placed in a beaker and neutralised with ammonia and when cold 30 C.C. of ammonium citrate solution (= 10 grams of citric acid) are added followed by 15 C.C. of ammonia. After starting the stirring machine 35 C.C.of magnesium mixture are added drop by drop and after twenty-five minutes the precipitate is collected washed with dilute ammonia and ignited. No correction need be made for the volume of the insoluble matter as this contains relatively the same amount of phosphoric acid as the solution. L. DE K Estimation of Hypophosphoric Phosphorous and Hypo- phosphorous Acids in Presence of One Another and of Phosphoric Acid. ARTHGR ROSENHEIM and JAEOB PINSKER (Zeitscl.. anorg. Chent. 1909 64 327-341).-A method was required for the analysis of the acid obtained from the action of moist air on yellow phosphorus. Hypophosphoric acid (the pure sodium salt way used) may be estimated by titration with permanganate if certain precautions are observed.The slightly acid solution with the addition of a small quantity of permanganate is heated to 80-go" titrated rapidly with oxalic acid more perrnanganate added and so on until a red coloration persisting for five minutes is obtained. The excess is then titrated with oxalic acid 10H2P03 + 2KMn0 + 3H,SO + 2H20 = 10H3P0,+ 2MnS04 + K,SO,. Iodine is without action on hypophosphoric acid. Uranyl salts give a yellow precipitate according to the equation 2NaHP0 + U02(N03)5 = UO,( HPO,) + 2NaN0,. Phosphorous acid may be estimated by means of permanganate under the same conditions as hypophosphoric acid. It reacts quanti- tatively with iodine when heated in a closed flask to form phosphoric acid and may thus be estimated in the presence of hypophosphoric acid. Hypophosphorous acid may be titrated with permanganate as above 5E3P0 + 4KMn0 + 6H2S04 = 5H3P0 + 4MnS0 + 6H,O + 2K,SO and also reacts quantitatively with iodine when heated in acid solution in a closed flask.In all cases the oxidation of any ono acid is unaffected by the presence of one or more of the other acids. A.ii. 74 ABSTRACTS OF CHEMICAL PAPERS. mixture containing phosphoric phosphorous hypophosphoric and hypophosphorous acids may be analysed by (1) oxidising with nitric acid and precipitating with magnesia thus estimating total phosphorous ; (2) titrating with permanganate which oxidises all but the phosphoric acid ; (3) titrating with iodine which oxidises only phosphorous and hypophosphorous acids ; (4) titrating with uranyl nitrate which precipitates only phosphoric and hypophosphoric acids. An application of this method of analysis to the product obtained by the oxidation of phosphorus in moist air a t 5 O shows that the solution contains principally hypophosphoric acid.Equations for calculating the proportions are given. C. H. D. Apparatus for the Estimation of Carbon Dioxide in Milk. A. BAR ILL^ (J. Pharm. Chim. 1909 [vi] 30 452-453).-The apparatus which is figured in the original consists of a flask con- nected with (1) washbottles containing potassium hydroxide solution through which air freed from carbon dioxide may be introduced at will and (2) washbottles containing ammoniacal bnryta water for the absorption of the carbon dioxide liberated. Eight hundred C.C. of milk are introduced into the flask heated at 50° and the last traces of carbon dioxide swept by a current of purified air into the baryta solution where it is estimated. The ‘ I combined ” carbon dioxide is then estimated by adding excess of tartaric acid t o the milk and proceeding as before.The apparatus gives results within 1% of the theoretical. T. A. H. Estimation of Potassium in Potassium Silicates. AART VERWEY (Zeitsch. anal. Chem. 1909,48 760-762).-A slight modifica- tion of Lawrence Smith’s well-known method. One gram of the finely- powdered silicate is intimately mixed with 1 gram of pure ammonium chloride and 5 grams of alkali-free calcium carbonate and the whole is introduced into a platinum crucible 4.5 cm. in height and 3.5 cm. in width. After covering the mixture with a layer of 3 grams of calcium carbonate the crucible is heated until the ammonia is expelled and after putting the lid on a strong heat is applied for an hour with a Teclu o r a powerful Bunsen burner.The mass is then boiled with water for fifteen minutes and when cold the liquid is made up to 200 c.c. and in 50 or 100 C.C. of the filtrate the potassium is estimated by acidifying with hydrochloric acid and evaporating with excess of platinic chloride as usual. L. DEK. Estimation of Zinc and Analysis of Zinc Ores. K. VOIGT (Zeitsch. angew. Chem. 1909 22 2280-2285).-Complete analyses are given of a few zinc ores by known methods. The author prefers precipitating $he zinc as zinc ammonium phosphate after removing any calcium with ammonium carbonate in ammoniacal solution. Sulphur is best estimated after a preliminary fusion with a mixture of 5 parts of anhydrous sodium carbonate and 3 parts of potassium chlorate.If the ore contains an appreciable amount of matterANALYTICAL CHEMISTRY. ii. 75 insoluble in acid this should be fused with sodium carbonate and then again treated with hydrochloric acid. L. DE K. The Most Rapid W e t Lead Assay. JULIUS F. SACHER (Chem. Zeit. 1909 33 1257-1258).-A slight modification of Alexander’s process (separation of lead as sulphate dissolution in ammonium acetate and titration of the slightly acidified solution with standardised ammonium molybdate with tannin as external indicator). The author finds the previous separation of the lead as sulphate unnecessary and titrates the nitric acid solution at once after adding first a sufficiency of ammonia and then an excess of acetic acid.Supposing 1 C.C. of the molybdate solution=0*01 gram of lead a Correction must be made depending on the final volume of the liquid. For 50 c.c. 0.18 C.C. of molybdate should be deducted; for 100 c.c. 0.3 c.c.; for 125 c.c. 0.40 C.C. ; for 150 c.c. 0.45 C.C. ; for 175 c.c. 0.50 c.c.; for 200 c.c. 0.54 C.C. ; for 250 c.c. 0.60 C.C. ; and for 300 C.C. of liquid 0.75 C.C. of molybdate solution. If iron is present to any extent it should be removed from the solution by heating this for some time on the water-bath. L. DE K. Detection of Mercury in Urine. WILH. BECKER (Pharm. Zeit. 1909 54 987).-Polemical. A reply t o Stich (Abstr. 1909 ii 1055). The method given by Almdn only yields approximate results but the processes communicated by Farup and by Schumacher-Jung allow the determination of 0.1 mg.of mercury in urine. Bunge’s balances provided with a glass micrometer and indicating 1/20 mg. are recommended. L. DE I(. Detection of Mercury in Urine. FRITZ GLASER and A. ISENBURU (Chem. Zeit. 1909 33 1258).-To 250 C.C. of the sample are added 5 grams of pure aluminium sulphate and the solution is heated and precipitated with ammonia. The precipitate is then collected dissolved in hydrochloric acid and heated on the water-bath for forty-five minutes in presence of a copper spiral. The deposit is then as usual sublimed and converted into the red iodide. L. DE K. Electro-analysis of Mercury Compounds with a Gold Cathode. F. MOLLWO PERKIN (B-ans. Paraday Soc. 1909,5 45-48).-Using a gold flag electrode a large number of estimations of mercury were made using mercuric chloride bromide and sulphate. The electrolytes employed were nitric acid sulphuric acid potassium cyanide and sodium sulphide. The results obtained were always high-from 0.5 to 2.5%. This was at first attributed t o occluded hydrogen but it was not found that the gold electrode increased in weight when made the cathode in dilute sulphuric acid and the current passed for twenty hours. When the electrode coated with mercury was treated in a similar manner it also showed no increase in weight. Since the electrode was cathode the increase in weight could not be due to oxidation and when a platinum electrode was run in series with it ehs mercury deposited on the platinum was always slightly less thanii.76 ABSTRAC rS OF CHEMICAL PAPERS. that theoretically required. The author cannot explain the results which show however that a gold electrode cannot be satisfactorily used for analytical purposes. A rotating silver electrode showed similar resirlts to the gold electrode. Usually the time required to deposit the mercury was from five to six hours but on stirring the electrolyte by means of a powerful magnetic field the time of deposition was reduced to fifty minutes ; the results were still too high. The author comes to the conclusion that the best electrode on which to deposit mercury is one of mercury the :deposition being very rapid if a rotatinganode is used (compare Smith and Kollocb Abstr. 1905 ii 859 ; 1906 ii 194).Two quartz vessels one of them fitted with a siphon side tube are described for use with a mercury cathode. T. S. P. Estimation of Manganese by Volhard and WOES Method. WALDEMAR M. FISCHER (Zeitsch. anal. Chem. 1909,48 751-760).-A slight modification of the Volhard-Wolff permanganate titration pro- cess. The sulphuric or hydrochloric acid solution (containing about 0.35 gram of manganese) is diluted t o 500 c.c. and aqueous sodium hydroxide is added until a slight precipitate forms which is then at once redissolved by a few drops of dilate sulphuric acid. One gram of freshly ignited zinc oxide and 10 grams of zinc sulphate are added and after heating the solution to boiling it is titrated with .AT/10 permanganate. One C.C. of pure glacial acetic acid is added and the liquid is again heated to boiling.This causes the precipitate t o coagulate and to settle rapidly and the pink coloration to disappear ; the titration is then continued until the pink colour is restored. L. DE I(. Separation of Manganese and Chromium. FEEDINAND FALCO (Arch. Yharm. 1909 247 431-436).-A modification of Kassner’s process by means of which good results are obtained when the amount of manganese does not exceed 0.015 gram The solution is placed in a covered beaker and sodium peroxide added in small portions. The liquid is then heated until the precipi- tate has separated in brown flocks. The liquid is decanted through a filter and the precipitate wached with boiling water and finally collected on the filter ; the filtrate contains the chromium as chromate.The manganese peroxide is dissolved off the filter by means of hot 20% hydrochloric acid and returned to the beaker. Excess of bromine water is added followed by an excess of ammonia and the liquid is heated to boilbg. The precipitate is then collected washed ignited and weighed as manganosomanganic oxide. The filtrate which may still contain traces of chromium is added to the main solution. L. DE I(. Quantitative Separations by Means of Ammonium ‘( Cup- ferron ’’ (Nitrosophenylhydroxylamine). OSCAR BAUDISCH (Chem. Zeit. 1909 33 1298-1 300).-Analysis of By*own Iron Ore.-Five grams of the ore are dissolved in 60 c.c of Btrong hydrochloric acid ; theAXALPTICAL CHEMISTRY. ii. 77 iron is fully oxidised by boiling with potassium chlorate and when cold the whole is diluted to 500 C.C.Twenty-five C.C. of the liquid are placed in a beaker and 20 C.C. of hydrochloric acid and 100 C.C. of cold distilled water are added. While stirring a solution of 3 grams of ‘‘ cupferron ” in 50 C.C. of water is added. The iron is completely pre- cipitated when a white precipitate begins to form. The precipitate is now collected and washed with cold water until the washings are no longer acid; the filtrate is used for the estimation of manganese. The filter is then washed with dilute ammonia (1 1) to remove the excess of ‘‘ cupferion,” and the precipitate is then washed once with water and finally converted by ignition into oxide and weighed as such. Andysis of Nickel Ore.-The iron and copper present are precipitaked jointly with ‘6 cupferron,” and in the liltrate the nickel is estimated by a suitable method.The copper is extracted from the washed precipitate with strong ammonia and the copper is then obtained as oxide by evaporating the solution and igniting the residue; the oxide may be reduced to metal and weighed as such. The precipitate insoluble in ammonia is then ignited and weighed as ferric oxide. The use of ‘6 cupferron ” in a number of other cases is suggested. L. DE I(. Rrtpid Electrolytic Estimation of Cobalt. PIERRE BRUYLANTS (Bull. SOC. chim. Belg. 1909 23 383-400).-A lengthy paper containing a large number of experiments recorded in sixteen tables. The electrolytic estimation of cobalt in ammoniacal solution gives good results with a cathodic potential of 1.35 volts.If the same process as for the rapid estimation of uickel is applied a spongy and partly oxidised metallic deposit is obtained In any case the metal is very readily oxidised and redissolved so that it must be washed without interrupting the current. Under proper conditions cobalt may like lead and manganese be deposited as peroxide at the anode. Electro- lysis from an oxalic acid solution cannot be recommended as the deposit then contains carbon. An approximate separation of cobalt from zinc may be obtained when operating in ammoniacal solution as directed above. If sodium sulphite is also added the cobalt although free from zinc contains an appreciable amount of sulphur. L. DE I(. Gravimetric Estimation of Chromium ; Quantitative Hydro- lysis of Sesquioxides.WALTER SCHOELLER and WALTER SCHBAUTH (Chem. Zed. 1909 33 1237).-The neutralised solution of the chromic salt which contains 0.1-0-2 gram of metal and is diluted to about 300 c.c. is heated to boiling and 3 C.C. of aniline are added in portions of 1 C.C. After boiling for five minutes the chromium has been precipitated quantitatively as hydroxide in a form which may be readily washed by decantation. Salts of aluminium zinc and ferric iron are also precipitated by aniline but whether the process is sufficiently accurate for these metals has not as yet beep decided The process may be used in the presence of manganous salts. L. DE &ii. 78 ABSTRACTS OF CHEMICAL PAPERS. Elstirnation of Titanic Acid in Ilmenite. ELISE ROEB (Chem. Zeit. 1909 33 1225-1226).-0*5 Gram of the finely-powdered ore is heated with 10 grams of anhydrous sodium carbonate finally over the blowpipe for half an hour.The liquid mass is poured into a platinum dish placed in cold water and when cold it is put together with the crucible into a spacious beaker containing 50 C.C. of cold water. The portion left undissolved is washed on a filter then transferred to an Erlenmeyer flask of one litre capacity and heated with dilute hydro- chloric or sulphuric acid until quite dissolved. When cold the solution is rendered slightly alkaline with sodium hydroxide and then again slightly acidified with sulphuric acid. One hundred C.C. of strong sulphur dioxide solution are added and enough water t o obtain a volume of 700 C.C. The whole is now boiled on an asbestos plate for one hour and another 20 C.C.of sulphur dioxide solution are added. After remaining for a few hours in a slanting position the precipitate is collected washed with boiling water ignited and weighed as titanium peroxide. L. DE K. Estimation of Thorium in Monazite. Colorimetric Estima- tion of Small Amounts of Platinum. JOHN C. H. MINGAYE (Records Geol. Survey 8. S. Wales 1909 8 276-286).-Estimatiom of Thorium irt Monccxite.-Samples of Carolina monazite have been analysed by several methods. The precipitate obtained with sodium thiosulphate must be redissolved precipitated with oxalic acid and extracted with ammonium oxalate. Precipitation with moist lead carbonate (Giles Abstr. 1905 ii 615) gives good results if the precipi- tation is repeated. Precipitation as basic acetate (Haber Abstr.1898 ii 295) is rapid and sufficiently accurate. Several Australian monazites have also been examined the richest being a sand from Black Swamp Torrington New South Wales contbining 4.1 2% of thoria. Colorimetric Estimution of Platinum-The small quantities of platinum occurring in beach sands or alluvial deposits may be estimated by fluxing with litharge and charcoal to obtain a lead button cupelling with the addition of silver and parting with nitric acid D 1.28 which dissolves platinum and silver leaving gold and iridosmine. The silver i E precipitated as chloride and the filtrate evaporated twice to dryness with hydrochloric acid. The hydrochloric acid solution filtered from traces of lead and silver chlorides is treated with stannous chloride or potassium iodide and the coloration obtained compared with that of a standard solution containing 0.065 gram platinum per litre. Quantities of platinum as small as 0.06 gram per ton may be estimated in this way.C. H. D. Methods for the Quantitative Chemical Analysis of Animal Tissues. I. General Principles. WALDENAB KOCH (J. Amer. Chem. Soc. 1909 31 1329-1 335).-A discussion of the general principles on which accurate methods of analysis of animal tissues depend and of the manner in which the analytical data should be applied in order to yield results of value t o the physiologist. The difficulties attending such work are pointed out. It is suggested that the constituents of the cell may be conveniently arranged in theANALYTICAL CHEMISTRY. ii.79 following classes (1) lipoids including phosphatides cerebrins and cholesterol ; (2) extractives such as creatine taurine and hypo- xanthine ; (3) inorganic constituents represented by the ash ; (4) proteins ; (5) fats ; (6) carbohydrates namely! glycogen. Very few of the methods at present employed for estimating these various groups are at all satisfactory and it is therefore considered more useful to study the distribution of the three elements nitrogen phosphorus and sulphur which are of special interest as they are invblved in somewhat different phases of cell activity. From a physiological standpoint it is desirable t o study the same tissue under different conditions of which the three most important are (1) the period of growth or development ; (2) conditions of pathological change; and (3) conditions of starvation the first of these being the most favourable for a study of the chemical transformations in the cell.As an illustration analyses are given of (1) a very young brain (2) an adult brain and (3) a brain from a case of dementia prscox. The young brain contained a larger quantity of extractives than the adult brain whilst the latter contained a greater proportion of lipoids. The brain from the case of dementia prscox resembled the normal brain but contained a smaller quantity of partly oxidised sulphur compounds soluble in water. E. G. Methods for the Quantitative Chemical Analysis of Animal Tissues. 11. Collection and Preservation of Material. WALDEMAR KOCH and SIDNEY A. MANX (J.Amer. Chem. Xoc. 1909 31 1335-1341. Compare preceding abstract).-An account is given of the precautions to be observed in order to obtain uniform and representative samples for analysis and of the methods of preserving the material. The chief methods of preservation are (1) immersion in alcohol (2) drying and (3) the addition of a dehydrating agent such ag sodium sulphate or gypsum The relative advantages and disadvantages of these three methods are discussed. E. G. Methods for the Quantitative Chemical Analysis of Animal Tissues. 111. Estimation of the Proximate Constituents. WALDEMAR KOCH and EMMA P. CARR (J. Amer. Chem. Soc. 1909,31 1341-1 355. Compare preceding 'abstracts).-Methods are described for effecting the separation and estimation of the six classes of constituents referred to by Eoch and the results of the analysis of a sample of meat by these methods are tabulated.Methods for the Quantitative Chemical Analysis of Animal Tissues. IV. Estimation of the Elements with Special Reference to Sulphur. WALDEMAR KOCH and FRED. W. UPSON (J. Amer. Chem. Soc. 1909 31 1355-1364. Compare preceding abstracts).-Methods are described for estimating the total sulphur in animal tissues and its distribution in the lipoids extractives proteins and inorganic sulphates. I n the lipoids extractives and proteins the sulphur exists in both the oxidised and non-oxidised condition the former being represented by the ethereal or R*SO,*OH type and the E. G.ii. 80 ABSTRACTS OF CHEMICAL PAPERS. latter by the cystine or R*S*H type.Methods are given for the estimation of the sulphur in each of these states of combination. The results of a study by these methods of the distribution of sulphur in brains of three different ages are tabulated. The per- centage of total sulphur does not show any great variation. The results indicate however that the lipoid sulphur increases at first with age and decreases later and the organic sulphur compounds soluble in water decrease with age. E. G. Volumetric Estimation of Phenol by Lloyd’s Method. Tribromophenol Bromide and Hexabromophenoquinone. S. C. J. OLIVIER (Rec. trav. chim. 1909 28 354-367).-1n the author’s experience Koppeschaar’s process for the estimation of phenol does not as Lloyd states (Abstr. 1905 ii 209) give inaccurate results and on the contrary the modified form described by Lloyd (Zoc.cit.) gives results less trustworthy than those given by the original process. In this investigation the author prepared tribromophenol bromide and hevabromophenoquinone and records certain new data regarding these substances. Tribromophenol bromide when heated in an atmosphere of carbon dioxide begins to evolve bromine even below looo and it is to this decomposition that the different melting points ascribed to this substance are due. Mixed with excess of potassium iodide and hydrochloric acid the bromide liberates 99.6% of the theoretical quantity of iodine so that the anomalous results recorded by Lloyd cannot be due as he suggests to the formation of this substance in the titration of phenol by Koppeschaar’s process.Hexabromophenoquinone was prepared by Benedikt’q process ( A bstr. 1879 717) but could nob be obtained crystalline. This substance is not formed when excess of bromine is allowed to react with phenol under the conditions prescribed by Koppeschaar so that Lloyd’s anomalous results with this process are not due to this cause. In using Koppeschani’s process the time of action of the bromine on the phenol may be reduced to five minutes if not more than 0.09 gram of phenol is used and if the solution of sodium bromide and bromine is about 0.8 decinormal. The addition of sodium bromide inhibits the formation of t r i bromophenol bromide. The end reading is clearer if chloroform is added as suggested by Lloyd (Zoc. cit.). T. A. H. Application of Arsenious Acid in Volumetric Analysis.I. FRANZ M. LITTERSCHEID and J. BORNEMANN (Zeztsch. angew. Chem. 1909 22 2423-2427).-Estimu,tion of Dextrose [in Urine].-Fifty C.C. of copper sulphate solution (49.848 grams per litre) are placed in a 200 C.C. measuring Bask 20 C.C. of the usual alkaline tartrate solution are added and the liquid is heated to boiling. Twenty-five C.C. of the dextrose solution (urine) not exceeding 1% in strength are added and the boiling is continued for two minutes. To the hot liquid are then added 50 C.C. of arsenious acid solution (9.9 grams per litre; this is equivalent t o the copper solution) and also 30 C.C. of 96% acetic acid. When cold 7 grams of potassium iodide are added i n six t o eightANALYTICAL CHEMISTRY. ii. 81 portions. When all the copper inchding the reduced oxide has passed into cuprous iodide the liquid is diluted to the mark and filtered the first portion of 10 C.C.being rejected. One hundred C.C. are theu nearly neutralised with 20% sodium hydroxide and after adding 5 grams of sodium hydrogen carbonate and diluting t o 400-500 c.c. the free arsenious acid is titrated with N/lO-iodine as usual. Or the liquid may be nearly neutralised with pure ammonia and diluted to 400-500 C.C. without addition of hydrogen carbonate. The arsenious acid found is equivalent to the copper reduced by the dextrose. A table is given showing the amount of dextrose corresponding with the C.C. of iodine used. L. DE K. Simple Distilling Apparatus for the Estimation of Pentosans by Tollens’ Method. JOHANN TISCHTSCHENKO (J.Landw. 1909 5’7 229-230).-The usual apparatus consists of a flask fitted with a tap funnel and a Kjeldahl distilling bulb to which the condenser is attached. This is now simplified by making the funnel and bulb in one piece the funnel passing down through the middle of the bulb ; the double boring of the cock is thus avoided. A sketch ir given. N. H. J. M. [Estimation of Glycogen.] KARL GRUBE (PfCgep’s Archiu 1909 130 322-3 24). EDUARD PFLUUER (ibid. 3 5-32?) .-Polemical. Quantitative Estimation of Lactic Acid in Cheese. SHIGEHIRO SUZUKII and EDWIN B. HART (J. Amel*. Chem. Soc. 1909 31 1364-1367).-1n connexion with a study of the chemical changes which take place during the ripening of cheese i t was necessary to find a trustworthy method for estimating lactic acid.Palm’s method (Abstr. 1887 307) depending on the formation of basic lead lactate 3Pb0,2C,H603,-has been found to be inaccurate. Partheil’e method (Abstr. 1903 11 189) in which the lactic acid is distilled in a current of superheated steam when applied to i-lactic acid must not be carried out below 1304 and the distillation must be continued until 750 C.C. of distillate have been collected. This method however cannot be used in presence of other acids :such as malic citric tartaric oxalic or succinic and is therefore unsuitable for cheese analysis. The usual method of estimating lactic acid by adding dilute sulphuric ;acid t o the cheese extracting with^ ether and afterwards separating the acid in the form of its zinc salt gives fairly satisfactory results when carefully carried out.E. G. Estimation of Cinnamic and Benzoic Acids in Mixtures of the Two Acids. ANNE W. K. DE JONG (Roc. trav. chim. 1909 28 342-348. Compare Abstr. 1908 ii 993).-The methods already described such as those of de Jong (Abstr. 1906 ii 315) and Scberinga (Abstr. 1907 ii 823) give good results for the volumetric estimation of these two acids but are unsuitable in cases where other acids are present or where the constituents must also be isolated and identified. I n the process now described cinuamic acid is converted intoii. 82 ABSTRACTS OF CHEMICAL PAPERS. dibromophenyIpropionic acid whioh can be separated quantitatively from benzoic acid and identified. Preliminary experiments with mixtures of the two acids by (1) fractional crystallisation ; (2) extraction by solvents ; (3) sublimation and (4) precipitation of metallic salts did not give promising results.The process eventually discovered consists in dissolving the mixture under examination in carbon disulphide and determining either by titration or weighing the bromine necessary to convert the cinnamic acid present into dibromophenyl propionic acid. This may then be separated quantitatively from benzoic acid by extracting the latter with carbon tetrachloride or by removing it by sublimation. T. A. H. Sodium Phosphotungstate as a Reagent for Uric Acid and other Reducing Substances. CARLO CERVELLO (Arch. exp. Puth. Pharm. 1909 81 434-437).-The reaction in urine with sodium phosphotungstate described by Richaud and Bidot depends on the presence of uric acid which reduces the reagent and forms a blue colour.Other reducing substances (dextrose hydroxylamine morphine etc.) give the same reaction. The blue colour which ferrous salts give with the tungatate passes in the presence of atmos- pheric oxygen into a green and then a red colour as lower tungsten oxides are formed. Reducing agents produce analogous changes in molybdic salts. W. D. H. LUDWIG KRAUSS (Biochem. Zeitsch. 1909 22 131).-A question of priority. The author claims that the reaction with iodic acid described by Frankel and Allers (hbstr. 1909 ii 628) mas previously described by himself (Apoth. Zeit. 1908 701). Estimation of Morphine in Opium; Extract of Opium and Tincture of Opium. HEINRICH PRERICHS (Chena. Zentr. 1909 ii 1499-1500 ; from Apoth.Zeit. 1909 24 592-596).-Helfen- berger’s process is recommended using 5 grams of dilute ammonia (17 + S3)and water saturatedwith ether. Aconvenient separating funnel for washing the morphine crystals is described. The process is also applicable to the extract and the tincture of opium ; the latter should be concentrated to about one-third. Opium may be reduced to a d e h i t e morphine content by addition of starch; for the extract P. W. GILL F. G. ALLISON and HARRY 8. GRINDLEY (J. Amer. Chenz. Xoc. 1909 31 1078-1093).- A modification of the hydrolysis method. The resulting solution after heating the urine with hydrochloric acid in an autoclave is mixed with an excess of sodium carbonate and the ammonia formed from the urea is then expelled in a current of air and absorbed in standard acid.Under these conditions no ammonia is yielded by the products of the hydrolysis of creatinine uric acid or hippuricacid. A The Iodine Reaction of Adrenaline. W. D. H. lactose may be used. L. DE K. Estimation of Urea in Urine. special aeration apparatus is described. L. DE I(.ANALYTICAL CHEMISTRY. ii. 83 Estimation of Urea Allantoin and Amino-acids in Urine. DOROTHY E. LINDSAY (Bio-Chem. J. 1909 4 448-454).-Nitrogen is estimated by (a) Bohlaad’s method ; this gives the nitrogen of amino- acids hippuric acid creatinine allantoin and urea. Estimation by Folin’s method (6) gives the nitrogen of urea allantoin and ammonia ; estimation by the Morner-Folin method (c) gives the nitrogenof urea and ammonia. Ammonia and creatinine are estimated separately by Folin’s method.The difference between (a) and (6) gives the nitrogen of amino-acids (including hippuric acid) and of creatinine ; the differ- ence between (b) and (c) gives the dlantoin nitrogen. This combina- tion of methods was found to give good results with artificial mixtures. W. D. H. Estimation of Purine Bases in Urine. ERNEST L. KENNAWAY (J. Physiol. 1909 39 296-310).-When urine is treated by the Camerer-Arnstein method for the estimation of total purines uric acid loses ammonia; this loss occurs when the precipitate of silver- magnesium urate is boiled. The purine bases do not undergo this loss. The method is therefore only correct when employed for the estima- tion of purine bases in urine from which the uric acid has been removed.W. D. H. Vanillin &s a Test for Antipyrine and Kryogenine ; Detection of Antipyrine in Pyramidone. CHAELES PRIMOT (Chem Zentr. 1909 ii 479 ; from BUZZ. Sci. Pharmacol. 1909 16 270).-If a small crystal of antipyrine is moistened with 2 C.C. of a reagent consisting of 1 gram of vanillin 6 grams of dilute hydrochloric acid (1 I) and 100 grams of 95% alcohol and the whole evaporated on the water-bath a dark orange-yellow ring is formed first and afterwards a similar stain on the bottom of the dish is noticed. The test shows the presence of even 0.00095 mg. of antipyrine. Pyramidone does not give the test and even 0*005 mg. of antipyrine may thus be detected in 0.1 gram of pyramidone. Rryogenine gives a distinct greenish- yellow coloration when this test is applied.L. DE I(. Attempt to Estimate Indoxyl in Urine. L. A. MENNECHET (Chem. Zentv. 1909 ii 1499; from Bull. Sci. Pharmacol. 1909 16 458-460).-The urine is defecated by 10 vol. % of basic lead amtate and a portion of the filtrate representing 50 C.C. of the urine is mixed with an equal voliime of pure hydrochloric acid. After shaking for five minutes 5 C.C. of chloroform are added and the shaking is continued; if the chloroform remains colourless 2 to 3 drops of hydrogen peroxide solution are added. The chloroform is drawn off and the liquid is shaken repeatedly with fresh portions of chloroform. The united chloroform extracts are shaken first with water containing a few drops of sodium hydroxide and then with water and the indoxyl is titrated by means of standard hypobromite.The chloroform does not become quite colourless at the end of the titration but retains a slight lilac colour. L. DE K.ii. 84 ABSTRACTS OF CHEMICAL PAPERS. Cause of the Vanillin Hydrochloric Acid Reaction for Camphor. 0. TUNMANN (Chem. Zentr. 1909 ii 1010 ; from Schwedz. Wocbschr. Chem. Pharm. 1909,4'7 617-519),-'J!he active substance contained in commercial camphor to which the reaction with vanillin and hydrochloric acid is due is only an impurity and is not derived from the oil cells of the camphor tree but exists in the parenchyma cells It is probably closely related to the phloroglucotannoids and it seems t o play an important physiological rSle. L. DE K. A Colour Reaction for Gelatin. R. ED. LIEBEGANG (Zeitsch. Chem. 2nd.Kdloids 1909 5 248).-RIixed aqueous solutions of cupric chloride and excess of tripotassium phosphate give in the presence of gelatin a violet coloration instead of the usual whitish-green precipitate. G. S. W. The VaIue of Benzidine for the Detection of Minute Traces of Blood. E. J. MCWEENEY (Sci. Proc. Roy. Dubl. Soc. 1909 12 216-223).-The reagent is made by dissolving a pinch of benzidine in 3-4 C.C. of glacial acetic acid just before use for the mixture itself turns blue after about a minute 0.5 C.C. is transferred to a test- tube and 2 C.C. of 10% hydrogen peroxide are added. To the white opaque mixture is added the material to be examined for blood and its presence is denoted by an immediate blue coloration. The material requiring investigation consists nearly always of suspected' stains and may be usually obtained as a fine dust by scraping. Should it be found impracticable to obtain a scraping the fabric may be treated with a drop of normal saline solution the stained fibres removed to another slide and treated between slide and cover-glass with a drop of the reagent when the presence of blood reveals itself bF the brilliant blue coloration of the affected fibre. The mass may then be treated on the slide with a drop of 32% potassium hydroxide so as to render visible the outlines of the individual red cells. None of the secretions and excretions hitherto tested by the author (saliva nasal mucus urine etc.) has been found to behave towards benzidine in the same way as blood but precisely similar results are given by many freshly-cut vegetables and fruits (potato and apple for instance) but these no longer give the reaction when boiled for some time whereas the reaction given by blood solutions is unimpaired by boiling for ten minutes. A positive result with the benzidine test should not be looked on as absolutely positive proof of the presence of Mood although i f the result is negative the absence of blood may be safely assumed. No regard should be paid to colour ,changes occurring after the lapse of one minute and a blank test should be made to make sure thst the ,paterials and the test-tubes to be used are satisfactory. L. DE I(.