ii. 86 ABSTRACTS OF CHEMICAL PAPERS An aly t ical C he mist r y . A Modifled Explosion Eudiometer. F. H. CAMPBELL (Chem. News 1911 104 235-236).-A Chree-way capillary tap is fitted at the top of the eudiometer; this arrangement is convenient when it is desired to demonstrate the proportions in which various gases combine,ii. 87 ANALYTICAL CHEMISTRY as the latter may be introduced separately through the tap into the apparatus. w. P. s. Simplification of Gravimetric Analysis. ERNST MURMANN (Zeitsch. anal. CIbem. 191 1 50 742-747).-Instead of an ordinary filter a small disk of wet filter paper is placed over a perforated porcelain disk and when filtering suction is applied. A special apparatus is recommended which is figured in the original. Introduction of a minute quantity of filter-paper pulp is recom- mended.I n cases where the precipitate has t o be ignited the precipitation as sulphide is much assisted by the addition of mercuric chloride or i n the case of stsnnic oxide some mercuric sulphide may be added in order to get rz clear filtrate. The weight of the filter disk being seldom more than 0.02 gram the process is very suitable for the direct weighing of dried precipitates as it is not necessary to employ a weighing tube. Of course no pulp is then added. L. DE K. Standardising of Acids W i t h o u t the Aid of Alkali Solutions. FRIEDRICH KLINKERFUES (Chem. Zeit. 19 1 1 35 1274). -A definite volume of the acid intended for titrations is plac9d in a suitable receiver in which is then collected the ammonia evolved from an accurately known weight of ammonium sulphate.C u e should be taken that there shall be a slight excess of ammonia at the end of the distillation. After adding a suitable indicator bhe free ammonia present is titrated with the same acid and the strength of the latter is then found by an easy calculation L. DE K. The Application of Methyl-red to the Colorimetric Measure- ment of Hydrogen Ion Concentrations. SVEN PALITZSCH (Biochem. Zeitsch. 1911 37 131-138).-The substance p-dimethyl- aminoazobanzene-o-carboxylic acid is specially adapted to the measure- ment of the hydrogen ion concentrations between the zlmes pLL= 4.2 and pa= 6.3 in which the colours vary from violet-red to yellow. The author gives a table of corrections to be applied in the pre,-,encc of sdtrs and proteins in which cases the [HI*-concentratiuns were determined by him electrometrically. The corrections t o be applied are only small S.B. S. The Analysis of Chlorates. VON BuTTrAR (Chcrn. Zeit. 1911 35 137P).-Calcium carbonate is generdly used in the neutralisation of reduced chlorato solutions before reduction. Time is saved by using magnesium hydroxide made into a cream with water. Neutralisation is complete when a permanent turbidity appears and filtration is unnecessary C. H. D. Estimation of Perchloric Acid in Certain Perchlorates. HENRYK GOLBLUM (Zeitsch. and. Chem. 1911 50 741-742).-The author estimates perchloric acid in the nickel or cobalt compounds as follows About 0.2 gram of the salt is dissolved in 300 C.C. of water and electrolysed at the ordinary temperature for some seventy hours,ii.88 ABSTRACTS OF CHEMICAL PAPERS. using as cathode a platinum gauze and a very weak current up t o 0.1 ampere. The potential is rather high on account of the strong resistance and varies between 10 and 12 volts. The liquid containing the liberated perchloric acid is then titrnted Detection of Fluorine. ERWIN RUPP (Zeitsch. Nahr. Geimssna. 1911 22 496-497).-The ash of the material to be tested for the presence of fluorine (fluorides) is placed in a platinuni or lead crucible 2nd moistened with 3 drops of water and 1 C.C. of sulphuric acid. The crucible is then closed with tt rubber stopper through which passes a glass rod the end of the rod below the stopper being moistened with a drop of water. The crucible and its contents are now heated on a water-bath for about twenty minutes; during this time the water on the glass rod evaporates and should fluorides be present the rod becomes coated with a film consisting of silica sodium silicofluoride and calcium silicofluoride. When only traces of fluoride are present the portion of the glass rod below the stopper may be covered with a piece of rubber tubing leaving only the end of the rod exposed ; the action of the hydrofluoric acid is thus concentrated on a small surface G.B. VAN KAMPEN (Chem. Weekblad 1911 8 856-860).-For estimating calcium fluoride in the basic slag of the Talbot process the author recommends Penfield’s method (Abstr. 1879 829) the conversion into silicon tetrafluoride the decomposition of the fluoride with water in accordance with the equation 3SiF + 4H20 + 4KC1= H,SiO + 2K2SiF6 + 4HC1 and the titration of the hydrochloric acid formed using lacmoid as indicator.with N / 1 0-alkali with phenolphthalein as indicator. L. DE K. of the glass. w. P. s. Estimation of Fluorides. A. J. \V The Estimation of Oxygen in Iron and Steel. ALLERTON S. CUSHMAN (J. Ind. Eng. Chcm. 1911 3 372-374).-The three important methods for estimating oxygen in steel are (1) heating the sample in a stream of dry chlorine ; (2) dissolving in special solvent. such as copper sulphate or bromiDe; (3) combustion in a stream of hydrogen and the paper contains a sketch of the apparatus employed with exact details of how the author conducts the latter method The hydrogen is generated from drillings of pure iron or Lcmossy ” zinc with dilute hydrochloric acid and is passed succersively through potassium hydroxide solution concentrated sulphuric acid over a roll of platinum gauze in a strongly heated silica tube and finally over phosphoric oxide.The finely divided borings (20-30 grams) are weighed into a platinum (or silica) boat introduced into the silica combustion tube and after passing hydrogen until all air has been removed the tube is rapidly heated to about 850” maintained a t this temperature during about thirty minutes whilst the purified dry hydrogen passes at TS rate of about 100 C.C. per minute ; the apparatus is cooled in the stream of gas and the tared absorption tube which is charged with phosphoric oxide re-weighed. A correction obtained by blank experiments must be employed to ensure accurate results.F. M. G. M.ANALYTICAL CHEMISTRY. ii. 89 Estimation of Sulphur in Pyrites. ARNOLD HECZRO (Zeitsch. anal. Chem. 19 11 50 748-753).-The following conclusions are arrived at. For a rapid and safe estimation of available sulphnr in pyrites Dennstedt’s process (combustion in oxygen absorption of the gases in sodium carbonate heated a t 450’ Abstr. 1905 ii 761) is very suitable. Tubes made of quartz instead of Jenn glass are however preferable as they do not retain traces of sulphuric acid; this saves t h e trouble of rinsing the tube with hot water. When precipi tnting the sulphate with barium chloride double the theoretical quantity of the latter should be added. The results agree very well with those obtained by Lunge’s process.L. DE K. The Phenolcilulphonic Acid Method for the Estimation of Nitrates in Water. ALFRED E. JOHNSON (Chern. News 1911 104 235)-Attention is drawn to the necessiLy of heating the sulphuric acid and phenol together for eight hours (compare Abstr. 1890 ii 832) in order to obtain a reagent which will yield a red coloration with nitrates without any development of a green coloration. w. P. s. Estimation of Nitrogen in Drainage Water and Rain Water by Schloesing’s Method. (Miss) ALIDA HUIZINGA (Chm. Weekblad 1911 8 882-895).-The usual method of carrying out Schloesing’s process gives inaccurate results arising from the presence of air in the solutions employed and of nitrites and carbonates in the samples examined. The author recommends evaporation of the water with potassium hydroxide instead of magnesium oxide elimination of carbonic acid by boiling with acetic acid and oxidation of the nitrite with acidified potassium permanganate.With these precautions the error in the estimation of the nitrogen in an ordinary sample does not exceed 0.1 mg. of nitrogen A. J. W. Bacteriological Methods for E s t i m a t i n g t h e Available Nitrogen in Fertilizers. JACOB G. LIPMAN (J. Ind. Eng. Chem. 1910 2 146-:48).-That the formation of nitrates in organic substances bears a direct relationship to the proportion of nitrogen they contain and generally also t o the ease with which they ixndergo decomposition has been demonstrated previously. From the fact t h a t nitrate formation is preceded by the formation of ammonia i t would appear that a similar relationship should exist between ammonifica- tion and nitrification and the present paper contains an account of preliminary experiments for the purpose of testing this hypothesis. The nitrogenous material to be tested was mixed with 100 grams of soil and 1 gram of calcium carbonate and maintained at its optimum moisture content (18%) ; a further quantity of water was added in the ammonification experiments to ensure an organic matter containing 75% moieture; the ammonia determination was made after six days and the nitrate estimation after four weeks.The results showed that as a general rule easily ammonifiable substances are also readily nitrifiable the highest ammonia content being found in concentrated tankage ground fish solid and liquidii.90 ABSTRACTS OF CHEMICAL PAPERS. fresh manure dried blood and bone-meal whilst the nitrate content followed the order concentrated tankage cottonseed meal ground fish dried blood and solid and liquid fresh manure. The high position of cottonseed meal in the nitrate series whilst possessing low availability in the ammonia series is considered (from the result of other experimenh) to be due to the depressing effect of soluble carbohydrates on ammonification. F. M. G. M. A Bacteriological Method for Estimating Available Organic Nitrogen. J. M. MCCAUDLESS and F. C. ATHINSON (J. Ind. Eng. Chem. 1911 3 174-175).-An account of experiments to determine the available ammonia in two samples of fertilisers prepared from leather by a special process which rendered it soluble; cottonseed meal was employed for comparison. Tho total nitrogen content in the initial materials also in their filtered boiling water extracts was estimated the extracts shaken up with soil the filtered solutions incubated at 38-40' under both aerobic and anwobic conditions (the surface of the liquid in the latter being protected by a layer of oil) and the ammonia estimated a t intervals.After two hundred and ten hours the experiments were stopped by exhaustion of the flasks ; in the anaerobic experiments practically no reaction had taken place but in the others considerable quantities of soluble organic nitrogen had been converted into ammonia ; i n one case (3.35% ammonia) the action was retarded probably by the toxic effect of excess of ammonia on the organism producing it.The insoluble residues were treated with soil extract in the presence of a nutritive mixture (devoid of nitrogen) and after twenty-six days the extracts and residues separately examined when i t was found that a quantity of insoluble proteid matter had been converted into ammonia. The author concludes that two distinct processes are involved ( 1 ) the conversion of insoluble proteids into a soluble form and (2) their subsequent change into available ammonia but whether both these changes are the work of the same or of separate organisms has yet to be determined. F. M. G. M. Estimation of Phosphorus in Iron and Steel. C. REICHARD (Pharm. 2entr.-h. 1911 52 1314-1315).-0wiug to the high molecular weight of the ammonium phosphomolybdate (" yellow ") precipitate and the fact that even minute quantities of phosphorus in iron or steel still yield a precipitate that can be weighed with great accuracy it is proposed to limit the amount of iron taken for analysis to 1 gram.L. DE K. Estimation of Inorganic Phosphorus in Animal Tissues. A. C. WHITTIER (J. Ind. Eng. Chern. 1911,3,248-250).-An account of various methods employed for tbe estimation of phosphorus in animal tissue with a discussion on the difficulty (owing to the rapidity of subsequent chemical change) of determining the relative proportions in organic and inorganic combination at the time of death. The cnncliisions drawn are (1) that the determination of organicANALYTICAL CHEMISTRY. ii. 91 phosphorus by the barium method gives high results with uncoagulated extracts owing to barium phosphate passing through the filter.(2) A t the boiling temperature water has very little hydrolysing action on organic cornpourids of phosphorus in animal tissue. (3) Enzymes and bacteria have a greater hydrolytic action on organic pho3phorus compou~ids than boiling. (4) Coagulation of the proteins by boiling clears the solution giving more complete precipita.tion and also arrests the action of enzymes and bacteria. F. n!t. G. M. The Quantitative Separation of Organic Phosphorus and Phosphates iu Fodders. GUSTAV FINGERLING and A. H ECRING (Biochent. Zeitsch. 191 1 3’7 452-456).-In Stutzer’g process the inorganic phosphates are extracted from the material by dilute hydro- chloric acid and precipitated from the solution thus obtained by ammonia and calcium chloride.The precipitated phosphates are theu dissolved in nitric acid aud precipitated from this acid solution as phosphomolgbdate. The authors find that this process is in most cases satisfactory but it is not applicable to those materials which contain phytin which is precipitated with inorganic phosphates. As this substance is present in a large number of fodders a new process is necessary for estimating phosphoric acid in materials of this character. S. B. S. Comparison of Methods for the Estimation of Phosphates in Vegetable Substances. A. PONTE (Chenz. Zentr. 1911 ii 900; from Stax. sperim. ayrar. ital. 1911 44 459-461).-1ncineratiou of the substance with the addition of calcium oxide yielded results which agreed with those obtained by electrolytic oxidation whilst from 2 to 11% less phosphoric oxide was found by the ordinary method of incineration.w. P. s. A Possible Fallacy in Fleitmann’s Test [for Arsenic]. WALTER J. DILLING (Phacrm. J. 1911 [iv] 33 811).-Arsenic having been found in a urine by the Fleitmann test when its presence was thought to be highly improbable the author investigated the cause and found that when applying this test the test-tube containing the zinc and sodium hydroxide with the suspected material should be heated until a stream of hydrogen bubbles is being evolved and no longer; the tube should be removed from the flame and the silver nitrate paper then placed over the mouth of the test-tube. If as might be done unintentionally the mixture is boiled the steam escaping (which perhaps also contains volatile organic products) exerts a more or less strong reducing action and the spot obtained may be mistaken for an arsenical one.L. DE 3. The Estimation of Arsenic in Insecticides. E. B. HOLLAND (J. Ind. Eng. Chcm. 1911 3 168-171).-A review of numerous methods employed for the estimation of arsenic with a discussion as t o their adaptability to the analyses of insecticides in which arsenic tri- or pent-oxides are the active constituents. The Avory-Haywoodii. 92 ABSTRACTS OF CHEMICAL PAPERS. method (Proc. Assoc. 08. Agr. C’hm. 1905 22 27) is considered most trustworthy and from it the author carries out the following modified procedure. About 2 grams of the sample are boiled during five minutes with 50 C.C.of sodium acetate solution (1 2) cooled shaken with about 60 C.C. of dilute hydrochloric acid (1 3) filtered from insoluble matter and the volume made up to 500 C.C. About 25-50 C.C. are neutralised with dry sodium hydrogen carbonate and treated with 25 C.C. of sodium potassium tartrate (10%) (to redissolve the copper) and titrated with N/20-iodine in the presence of starch. For samples containing lead arsenate the following method is recom- mended Twograms of the sample are boiled with 66 C.C. of nitric acid (1 3) cooled filtered and made up to 500 C.C. From 50 to 100 C.C. are removed evaporated with 10 C.C. of diluted sulphuric acid (2 l) and finally heated to 150-200°; the residue is dissolved in water treated with 10 C.C.of potassium iodide (1 65 lOOO) and boiled until free iodine is expelled ; the solution is rapidly cooled neutralised with sodium hydroxide re-acidified with sulphuric acid starch paste added any remaining free iodine dostroyed with sodium thiosulphate the liquid rendered alkaline with sodium hydrogen carbonate and titrated with N/20-iodine solution. F. N. G. M. Elementary Analysis. RICHARD DOHT (Zeitsch. anal. Chem. 1911 50 756).-The author thinks that it would have been better if Kurtenacker (Abstr. 1911 ii 823) had given analytical figures t o prove why silver and iron are not suitable for the reduction of nitric oxides in the combustion of nitrogenous substances for experiments conducted by himself and Epstein (Abstr. 1908 ii 132) led to a different view.L. DE K. Detection of Carbon Disulphide in Oils. E. MILLIAU (Compt. rend. 19 11 153 1021-1023).-Vegetable oils employed as lubricants should be free from carbon disulphide. This substance may be detected by heating 25 grams of the oil with 10 C.C. of a concentrated potass- ium hydroxide solution adding 150 C.C. of water and a little sodium hydrogeu carbonate. The mixture is then poured into 20 C.C. of hydrochloric acid and the evolved gas tested for hydrogen sulphide by means of lead acetate paper. A second method consists in distilling 50 grams of the oil with 10 C.C. of amyl alcohol. The first 4 C.C. of distillate are heated in a sealed tube on the water-bath with 1 C.C. ol cottonseed oil and a few centigrams of sulphur. The presence of traces of carbon disulphide is recognised by the appearance of a red coloration.w. 0. w. Estimation of Silica in Iron Ores. FR. MOLDENHAUER (Zeitsch. anal. Chem. 1911 50 754-755).-111 order to obtain accurate silica estimations it is advisable to use platinum dishes as i t is often impossible to remove the silica rendered insolu bIe by evaporation from porcelain basins. L. DE I(.ANALYTICAL CHEMISTRY. ii. 93 Rapid Estimation of Zinc. X. VOIQT (Zeitsch. angew. Chem. 19 1 1 24,2195-2198).-The process is briefly as follows One to three grams of the zinc ore are dissolved in a mixture of 10 C.C. of nitric acid (D 1.2) and 20-30 C.C. of hydrochloric acid (D 1.9) preferably in a porcelain dish so that the siliceous mass may be broken up with a pestle. After boiling for a while on a sand-bath the whole is rinsed into a 200 C.C.flask 50-60 C.C. of ammonia are added and the solution is again boiled for a short time. When cold dilute ammonia (1 3) is added up to the mark and the solution is well shaken and filtered. One hundred C.C. of the filtrate are mixed with a little bromine water to precipitate any manganese and the turbid liquid is then boiled with about one gram of aluminium scraping8 until the copper present is all precipitated. Five C.C. of sodium sulphite solution (1 4) and a few C.C. of dilute ammonia are added the liquid is again heated to boiling and at once filtered; the residue is washed with a hot dilute ammoniacal solution of sodium sulphite. The filtrato is neutralised with hydrochloric acid and an extra 10 C.C. of dilute acid (1 3) are then added After diluting t o 150 C.C.and heating to boiling the zinc is at once titrated with ferrocyanide (2 1.63 grams of potassium ferrocyanide and 14 grams of crystallised sodium sulphite per litre) using ammonium molybdate solution (9 grams per litre) as external indicator. The end reaction is shown by the formation of a reddish-brown spot. L. DE K. Estimation of Copper. Modiflcation of the Iodide Method. E. C. KENDALL (J. Amer. Chem. Xoc. 1911 33 1947-1952).-A modification of the iodide method of estimating copper is described in which the nitrous acid formed during the solution of the copper in nitric acid is destroyed by che addition of sodium hypochlorite the chlorine thus liberated being removed by treatment with phenol. The solution of copper of volume about 50 t o 60 c.c.is placed in a flask and the acidity rendexed equivalent to about 4 or 5 C.C. of con- centrated nitric acid. The temperature should not be above 25'. Solution of sodium hypochlorite is added and about two miuutes later 5% solution of phenol is quickly introduced. Sodium hydroxide is added in slight excess to prevent the formation of nitrophenol and tlle solution is afterwards acidified with acetic acid. Potassium iodide solution is now added and the iodine liberated is titrated with standard sodium thiosulphate. This modification gives accurate results is more rapid and requires less attention than the original method. E. G. The Estimation of Manganese by the Sodium Bismuthate Method. PAUL H. M. P. BRINTON (J. Ind. Eng. Chem.2911 3 337-239).-The author has repeated tbe work of Blair (Abstr. 1904 ii 683) and others on the estimation of manganese by the sodium bis- mutbate method and corroborates the statement that for gmall amounts of manganese it is the most accurate method known. The following modifications are suggested that pure Sorensen sodium oxalate should be used t o standardise the potassium permanganate and that the empirical fwtor 0.1 656 instead of the theoretical factor 0.16024 shouldii. 94 ABSTRACTS OF CHEMICAL PAPERS. be employed in the conversion of the sodium oxalahe hgure into that of manganese; this corrects the tendency of the method to give low results although the author considers that gravimetrically standardised manganous sulphate is the most accurate standard that can be employed. The decomposition of ores by the hydrochloric and sulphuric acids method is suggested as being fully as accurate more rapid and possibly more convenient than the employment of hydrofluoric and sulphuric acids.Fusing the ore with sodium peroxide ie also recommended in the case of refractory samples. F. 34. G. 31. A New Rapid and Accurate Volumetric Method for the Estimation of Manganese and its Application to the Analysis of Iron and Steel. FLOYD J. METZGEB and L. E. MARRS (J. Ind. Eng. Chcm. 1911 3 333-335).-It has been observed that the presence of manganese disturbs the estimation of ferrous iron in rocks by the sulphuric-hydrofluoric acid method the more hydrogen fluoride employed the ;higher the percentage of iron found a fact which pointed to the possibility that manganese and permanganic acid react together quantitatively in the presence of a sufficient quantity of this acid.Solutions of carefully stsndardised munganous sulphate were titrated wit,h N/30-potassium permanganate in waxed beakers in the presence of 50% sulphuric acid (1 vol.) hydrofluoric acid (2.5 vole.) and varying quantities of ammonium fluoride (which increased the speed of reaction) and when the experiments were repeated in the presence of ferric iron identical results were obtained. The analysis of steel was conducted as follows About one gram of the sample dissolved in nitric acid was treated with ammonium per- sulphate (1 gram) and boiled during a few seconds evaporated to dryness and dissolved in a mixture of 50% sulphuric acid (20 c.c.) and 30 C.C. of water; the solution was boiled until clear transferred to a waxed beaker and treated with 5 grams of ammonium fluoride and 25 C.C.of hydrofluoric acid the volume made up to 100-150 c.c. and finally titrated with potassium permanganate. The results of numerous experiments are tabulated and the method is claimed to be simple accurate and rapid aiid a t the same time applicable to the cases-of spiegels and other manganese ores. F. M. G. M. Estimation of Ferrous Ions with Standard Iodine. GY~BEBT ROMYN (Chem. Zeit. 19 1 l,35,1300).-The process is more particularly intended for pharmaceutical preparations (Blnud’s pills syrup of ferrous iodide etc.). The preparation is dissolved in so much dilute sulphuric or hydrochloric acid that the free acidity amounts to about 20 C.C. of N-acid.The solution is then added to a mixture of 20 C.C. of iV/lO-iodine solution and 5 grams of powdered sodium pyro- phosphate and the mixture is shaken until the latter has nearly dissolved. After five minutes the excess of iodine is titrated with N/10-thiosulphate as usual; I C.C. of the latter=27*8 mg. of crgstallised ferrous sulphate. L. DE K.ANALYTICAL CHEMISTRY. ii. 95 Detection of Nickel and Gobalt in Mixtures. ANGEL DEL CAMPO and JAIME FERRER (Anal. Pis. Quini. 1911 9 201-213 272-276. Compare Abstr. 19 11 ii 825).-The value of potassium methyl xanthate as a delicate reagent for nickel and cobalt (compare loc. cit.) has been examined in respect to delicacy and in comparison with Tschugaeff’s reagent for nickel (dimethylglyoxime) and Vogel’s reagent for cobalt (ammonium thiocymate and amyl alcohol).In the case of nickel alone both reagents display about equaldelicacy detecting nickel in 1 C.C. of a 1 in 200,000 solution or in 4 C.C. of a 1 in 500,000 solution which is the limit. With mixtures of nickel and cobalt in the proportion of 1 to 100 Tschugaeff’s reagent fai!s but potassium methyl xanthate is trust- worthy arid with care may be employed to recognise nickel in the presence of even higher proportions of cobalt,. I n regard to cobalt the xanthate has about five tiuies the delicacyof Vogel’s reagent being capable of recognising the metal clearly in 1 C.C. of a solution of 1 in 250,000. With mixtures of cobalt with excess of nickel (10 to 2000 timesj the reagents display about equal value when working at the limits of sensibility oE Vogel’s test.G. D. L. Separation of Nickel and Zinc in German Silver and Other AlloyB. LA VERNE W. SPBING (J. Ind. Eng. Chem. 1911 3 255-256).-About 0.5 gram of the drillings is freed from tin with nitric acid the lead removed as sulphate copper by electrolysis and the iron by means of ammonium hydroxide. Five grams of ammonium chloride are then added and the solution just neutrslised with hydro- chloric acid aud treated with 0.4 gram of dimethylglyoxime (dissolved in alcohol) for every 0.1 gram of nickel supposed to be present; ammonium hydroxide is added drop by drop until just ammoniacal and the solution allowed to remain at just below boiling point during about half an hour.The precigitate is collected on a tarred filter washed with hot water driid at i05O and weighed C,H1,Ol,N,Ni X. 0.2031 = Ni. The filtrate is acidified with an excess of 10 C.C. of concentrated hydrochloric acid boiled to decompose the excess of dimethylglyoxime treated with 10 grams of microcosmic salt (in saturated solution) rendered exactJy neutral with ammonia and allowed to remain just below its boiling point until the precipitate has granulated; this is collected washed with hot; water ignited and weighed as zinc pyrophosphate. F. M. G. M. Estimation of Tin in Canned Foods. HERMAN SCHKEIBER and W. C. TABER (J. Id. Eng. CILenh. 1911 3 257-261).-A discussion of the difficulties experienced in obtn ining accurate results by the wet combustion methods usually employed for estimating tin in canned foods.The author gives a detailed description of an alkali fusion method which he finds gives excellent results the main procedure of which consists in mixing about 100 grams of the material weighed into an iron crucible with 50 C.C. of an aqueous solution coutaining 150 gramsii. 96 ABSTRACTS OF CHEMICAL PAPERS of sodium hydroxide and 100 grams of sodium carbonate per litre adding 75 c.c of alcoho’l (95%) evaporating carefully t o dryness with continual stirring slowly raising the temperature to about 1 60° and finally igniting in a muffle until all volatile matter is removed. The fused mass is transferred to a beaker and evaporated with 50 C.C. of concentrated sulphuric acid and 30 C.C. of nitric acid (D 1.42) until fumes of sulphur trioxide are apparent ; the mixture is cooled diluted almost neutralised with ammonium hydroxide and treated with hydrogen sulphide ; the collected precipitate is washed with a mixture of ammonium acetate and dilute acetic acid the tin re- dissolved by boiling with sodium hydroxide and after separation from other substances which may be present i t is finally re-precipitated collected ignited and weighed as stannic oxide.F. 31. G. M. Separation of Zirconium from Iron and Aluminium and the A n a l y s i s of Ferrozirconium. M. WUNDER and B. JEANNERET (Zeitsch. anal. Chern. 191 1 50 733-’735).-Se;uaration of Zirconiu’IIz h n and Aluminium.-About 1 gram of the mixed oxides is fused in a platinum crucible with 6 grams of sodium carbonate the mass is taken up with water and after addition of 1 gram of sodium carbonate boiled for a few minutes. The filtrate contains the aluminium (also any chromium) from which the alumina may be precipitated by adding excess of ammonium nitrate.If much is present it is advisable t o submit the insoluble mass t o a second fusion with sodium carbonate. The iron is dissolved by hot dilute hydrochloric acid (1 :l) and estimated as usual. The undissolved zirconium oxide is ignited and weighed as such. Analysis of ~erroaii.conium.-The alloy is heated on the water-bath with hydrochloric acid and broffiine when the zirconium and the iron pass into solution whilst the silicon is converted into oxide. After adding a little nitric acid the whoIe is evaporated to dryness and finally heated in an air-bath a t 110”.The mass is then dissolved in dilute hydrochloric acid and the silica is collected and ignited. From its weight should be deducted the small amount of zirconium oxide which is left after the usual treatment with hydrofluoric and sulphuric acid. The filtrate from the silica is precipitated with ammonia and the oxides are then submitted to the fusion with sodium carbonate as described. The process may be also applied to zirconium carbides. L. DE E( A Volumetric Method for E s t i m a t i n g Antimony in Alloys. GEORGE S. JAMIESON (J. Ind. Eng. Chem. 1911 3 250-251).-An application of Andrew’ (Abstr. 1903 ii 686) iodate method to the estimation of antimony in alloys especially ‘‘ hard leads ” and solders ; it is claimed to be rapid accurate and admissible in the presence of copper aud iron.Two other methods have been studied comparatively by the author the first which is based on getting the antimony into the quinquevalent form in dilute hydrochloric acid solution adding potassium iodide and titrating with sodium thiosulphate gave good resiilts only ind S ALY Ti C IL C HE Jl ISTLt Y ii. 97 the absence of copper or iron; the second method depends on getting the antimony into the tervalent condition in sodium hydrogen carbonate solution and titrating with iodine gave unsatisfactory results in the presence of lead owing to the co-precipitation of antimony which was found to take place even in the presence of tartaric acid. About 1 gram of an alloy containing less than 2% of antimony is weighed into a 200 C.C.Erlenmeyer flask and heated with 10 C.C. of concentrated sulphuric acid until decomposition is complete boiled gently for about two minutes after the lead sulphate has become whire cooled and diluted with 15 C.C. of water; 15 C.C. of dilute hydrochloric acid (1 1) are then added the mixture vigorously shaken and the lead sulphate collected on a Gooch crucible and washed with hydrochloric acid of the same concentration. The filtrate is transferred to a glass-stoppered bottle of about 250 C.C. capacity 5 C.C. of chloro- form 15 C.C. of concentrated hydrochloric acid and 5 C.C. of a solution of iodine monochloride added ; the bottle is shaken and after about five minutes the liberated iodine titrsted with standard potassium iodate until the chloroform in just decolorised after a vigorous shaking (which should be repeated in about a minute to ensure obtaining the correct end-point); the liquid can be poured off and the chloroform employed for further titrations.The concentration of the hydrochloric acid in the solution is of importance and the above quantities should be adhered to carefully. The iodine monochloride solution is prepared by dissolving 10 grams of potassium iodide and 6.44 grams of potassium iodate in 75 C.C. of water adding 75 C.C. of concentrated hydrochloric acid and a globule of chloroform and adjusting exactly to a faint iodine colour by shak- ing and adding minute quantities of potassium iodide or iodate until the required condition is obtained. Tne Geo-chemical Interpretation of Water Analyses.CEASE PALMER (Bull. U.S. Geol Survey 479; a Reprint 31 pp.).- A record of a large number of analyses of waters from American rivers. The results however are not given in terms of definite salts but as follows First the properties of the solution in % proportions [primary salinity secondary salinity tertiary salinity (acidity) primary alkalinity secondary alkalinity] ; second the percentage of reacting values of co-ordinate radicles irom which the properties of reaction may be directly derived accompanied by a statement of concentration values in parts per million ; third the character formula (the percentage of reacting values of the individual radicles determined) together with a statement of the concentration value ; fourth the base analyses. (1) Primary salinity (alkali salinity) is salinity not to exceed twice the sum of the reacting values of the radicles of the alkalis; (2) secondary salinity (permanent hardness) is the excess (if any) of salinity over primary salinity not to exceed twice the sum of the reacting values of the radicles of the alkaline earths group ; (3) tertiary salinity (acidity) is the excess (if any) of salinity over primary and.secondary salinity ; (4) primary alkalinity (permanent alkalinity) is F. 19. G. M. VOL. CII. ii 7ii. 98 ABSTRACTS OF CHEMICAL PAPEKS. the excess (if any) of twice the sum of the reacting values of the alkalis over salinity ; (5) secondary alkalinity (temporary alkalinity) is the excess (if any) of twice the sum of the reacting values of the radicles of the alkaline earth group over secondary salinity In distinguishing the special properties the values of racliclev of the same sign are doubled ; the positive and negative radicles which together induce their special properties thus receive their full value.I n the author’s opinion this form of stating results is more comprehensive to the geologist than the expression in definite salts which is always more or less hypothetical. L. DE K. Colorimetric Estimation of Phenols in Waste Liquors. HERMANN BACH (Zeitsch. anal. Chem. 1911 50 $36-740).-The process is based on the fact that 10 C.C. of a very dilute solution of phenol (a few rng. onlyj when heated to boiling with 0.2. C.C. of Millon’s reagent and 0.1 C.C. of nitric acid give when allowed t o cool a liquid of a permanent rose or reddish colour which may then be matched with a solution of phenol of known strength similarly treated.In order to apply the process to waste liquors 1 litre of the sample is mixed with 5-10 C.C. of strong aqueous potassium hydroxide and evaporated in a dish to 50 C.C. The contents are then transferred to a 500 C.C. Erlenmeyer flask having a mark at 150 c.c. a slight excess of sulphuric acid is added and when cold the liquid is diluted up to the 150 C.C. mark; one hundred C.C. are then distilled off. To this distillate which contains the phenol from one litre is then applied the above colorimetric test ; if the colour should be a dark red it must be suitably diluted. L. DE I(. The Estimation of Phenol and p-Cresol in Mixtures of the Two. HUGO DITZ and FRIEDRICH BARDACH (Biochem.Zeitsch. 191 1 37 272-312).-The authors maintain in opposition to Siegfried and Zimmermann and others that the original method proposed by Ditz and Cedivoda in 1899 for the estimation of phenol and p-cresol in mixtures of these substances gives accurate results provided that the details given in the description of the process are strictly adhered to. The method depends on the principal of estimating the amount of bromine taken up by the mixture and the success of the process depends on using the right excess of bromate and bromide mixture and of hydrochloric acid and on other details. The paper is largely controversial and the various points insisted on by the authors are illustrated by a large number of analyses of mixtures containing known quantities of the two phenols.S. B. S. Two Modifications of Fehling’s Solution. EDWARD FRANK HARRISON (Pharm. J. 191 1 [iv] 33 746-747).-Benedict has suggested recently that in the use of Fehling’s solution for the detec- tion and estimation of dextrose the actual reduction is effected by a substance resultirlg from the action of the alkali hydroxide on the sugar and that this change may proceed further especially in presenceANXLPTlChL CIIEXIlSTEtY. ii. 99 of substances which retard the normal action and give rise to non- reducing products. The reducing substance is produced more slowly by alkali carbonates but these have little or no further action and consequently Benedict has proposed two modified forms of Fehling:’~ solution for qualitative and quantitative use respectively in which among other changes sodium hydroxide is replaced by sodium carbonate (Journ.Amer. Med Ass. 191 1 Oct. 7). The preparation and application of these solutions to the detection and estimation of dextrose in urine are described and critically examined in this paper. The author considers that Benedict’s claim that his first solution is a far more delicate reagent for dextrose in urine than Fehling’s solution is justified. The solution for quantitative use is shown to give good results with solutions of dextrose in water but is less satisfactory when used for urine containing dextrose. WALTER F. SUTHERST (J. Ifid. Eng. Chem. 191 I 3 256).-The following is stated to be a convenient method of indicating the end-point when estimating sugar with Fehling solution.The Fehling solution is heated t o boiling and the sugar added until a faint blue colour is apparent at the edge of the porcelain vessel employed. As further sugar is added a drop of the mixture is placed on the top side of a filter paper folded in half the filtrate passes through and the spot is treated with a drop of a dilute acetic acid solution of 1% potassium ferrocyanide; on holding up to the light the faintest trace of copper ferrocyanide is plainly seen and the end of the reaction readily indicated. The Physiology and Pathology of Carbohydrate Metabolism [Colorimetric Estimation of Small Amounts of Sugar]. KARL REICHER and E. H. STEIN (Biochem. Zeitsch. 1911 37 321-344).-A method for estimating colorirnetrically small quantitiesof sugar by means of the Molisch reaction is described.Ten C.C. of concentrated sulphuric acid are placed in a graduated cylinder with glass stopper and t o this a tablet containing 0.05 gram of a-naphthol is added and then 2 C.C. of the solution to be investigated. Under these conditions the correct temperature for the development of the colour is attained. After gently mixing so that the colour is uniformly distributed the mixture is allowed to cool and is then diluted to 20 C.C. The colour is compared by means of a Plesch colorimeter or a Stein chromoscope with a solution prepared in a similar way with 0.02% dextrose solution. This is the most convenient strength when the sugar in blood is to be estimated. For the estimation of sugar in this fluid 2 C.C.of serum are titken and diluted to 10 C.C. with water and then 1-2 C.C. of concentrated potassium sulphate 6 t o 7% of dialysed ferric hydroxide solution (the method of preparing which is given in detail) and water up t o the total volume of 20 C.C. are added. Two C.C. of this filtered solution are employed for each test. S. B. S. The Eatimation of Sugars in Natural Products. CARL NEUBERG and MIGAKU ISHIDA (Biochem. Zeztsch. 191 1 37 142-169). -By precipitation with 50% mercuric acetate solution followed by T. A. H. Rapid Volumetric Estimation of Sugar. F. M. G. M. 7-2ii. 100 ABSTRACTS OF CHEMICAL PAPERS precipitation with 25% phosphotungstic acid solution all optically active derivatives of proteins as well as the proteins themselves nucleoproteins lecithin and other substances can be precipitated.The sugars remain in solution and can be accurately estimated polari- metrically in the filtrate. Glycosamine as me11 as the other carbohydrates can also be estimated in this way. The met1:od is illustrated by a large number of examples of analyses of mixtures of known sugar content with different varieties of sugar.:. A. method is also given for preparing the phosphotungstic acid reagent directly from sodium tungstate phosphoric and sulphuric acids. S. B. S. Estimation of Sugar in the Blood. DEXGO TAKAHASIII (BiochenL. Zeitsch. 1911 37 30-33).-The sugar in dog’s blood was estimated polarimetrically and after yeast fermentatiou both polarimetrically and by the methods of fhng and Bertrand. I n all cases no sugar was found after fermentation.The total sugar in dog’s blood was also estimated polarimetricitlly and by the methods of Kumagawa-Suto Bertrand and of Bang. The results by the first two methods agreed well with another and were always somewhat Iower than the results obtained by Bang’s method. S. B. S. The Estimation of Dextrose in Urine and Blood. BERTHOLD OPPLER (Zeitsch. yhysiol. Chem. 1911 75 71-134).-In order t o estimate dextrose in the urine polarimetrically before. and after fermentation the urine was treated with ghosphotungstic acid lead acetate and hydrogen sulphide ; this excludes pigment other reducing sub6tance8 and kevorotatory materials. Estimation by reduction gives too high a result. The existence of physiological glycosuria is very doubtful ; any sugar found in normal urine probably arises from other sources especially glycuronic acid.Similar conclusions are arrived at; in respect to the best method for estimating dextrose in blood. W. D. H. [Detection and] Estimation of Dextrose in Urine. G. -4. STUTTERHEIM (Yl/,arm. TPeekbZad 19 1 I 48 120 1- 1204).- The author uses an alkaline copper solution made as follows 17.32 grams of copper aulphate 100 grams of glycerol 20 grams of sodium hydroxide and water up to 250 C.C. This solution bhould be kept in brown bottles. When testing urice for dextrose 5 C.C. of the reagent are diluted with 5 C.C. of water heated to boiling in a test-tube and set aside. Five C.C. of the urine mixed with 5 C.C. of waterare also heated to boiling and then at once added to the copper solution.If dextrose is present the red copper. reduction will be noticed at the top of the liquid within one-half to one minute ; 0.5% of dextrose may thus be readily detected even when as frequently occurs the urine contains substances which interfere with the ordinary Fehling test. When estimating dext,rose in urine by the iodometric method the author recommends making an allowance of - 0°15:4 on account of dextrose orcurring naturally. L. DE I<.A S A LI’T LCA L CH E HI S1‘ it Y. ii. 101 The Detection and Estimation of Sugar in the Urine. A. C. ANDERSEK (Biochem. Zeitsch. 1911 37 262-265) -It was found t h a t the clarification of urine by charcoal in the presence of 5% hydro- chloric acid according to the method of Bang and Bohmannson leads to errors in the estimation of sugar owing t o adsorption by the charcoal especially when blood-charcoal is used.If however the clarification is carried out by charcoal in the presence of 10% of acetic acid (40c.c. urine,cliluted to 50 C.C. with50% acetic acid,and the mixture shaken with 4 grams of blood-charcoal for ten minutes) a perfectly clear colourless solution is obbained which on polarisation gives the Colorimetric Estimation of Sugar Creatine and Creaticine in Urine WJLHELN AUTENRIETH and GERHARD MULLER (Miinch. med. Wocl~. 1911 ; Reprint 12 pp.).-Whilst the process described by Autenrieth and Tesdorpf (Abstr. 1911 ii 159) yields trustworthy results when applied to colourless urines i t is recommended t h a t dark- coloured urine should be treated with blood-charcoal free from iron after the diluted urine has been boiled with Bang’s copper solution.The blood-charcoal when free from iron does not retain any of the Colorimetric Estimation of Lactose in Urine and Milk. WILHELN AUTENRIETH and ALBERT FUNK (Munch. med. Voch. 1911 Reprint Spp.).-The presence of lactose in urine may be ascertained by means of the test described by Wohlk (Abstr. 1906 ii 122); dextrose when present to the extent of not more than I% does not yield a coloration wit,h the test neither does it interfere with the lactose reaction. For the estimation of lactose 10 C.C. of the diluted urine (containing not more than 0-04 gram of the sugar) are boiled fur three minutes with 50 C.C. of Bang’s copper solution cooled diluted with potassium thiocyanate solution to a volume of 50 c.c.0.5 gram of blood-charcod free from iron is added the mixture is shaken for five minutes and filtered. The quantity of unreduced copper in the filtrate is then estimated by comparison with standard solution in a colorimeter (compare Abstr. 1911 ii 159). I n the case of milk 25 C.C. of the sample are diluted with 400 C.C. of water acetic acid is added the mixture is boiled and diluted when cold to a volume of 500 C.C. After filtration 10 C.C. of the filtrate are boiled Estimation of Sucrose and the fDetection of Dextrin in Foods. ANDRE F. VOLLANT (Ann. J’aZsifi 1911 4 504-509).-A method proposed by Lemeland (Abstr. 1910 ii lOOS) which is based on the destruction of the reducing and rotatory powers of arabinose dextrose l~vulose galactose and lactose when these sugars are heated with manganese dioxide and a1 kaline hydrogen peroxide solution.whilst sucrose is not affected may be employed for the estimation of sucrose in commercial sugars syrups confectionery dried milks and honey. It is pointed out that acetic acid should be used for neutralising the sugar solution after correct sugar values. s. B. s. copper. w. P. s. with 50 C.C. of Bang’s solution as described above. w. P. s. The results obtained are trustworthy.ii. 102 ABSTRACTS OF CHEMICAL PAPERS. treatment and not nitric acid as prescribed in the original paper. After clarification the solution must be acidified by acetic acid before being examined in the polmimeter. Dextrin may be detected when mixed with other sugars by inverting the sucrose present and then subjecting the solution to the above-mentioned treatment ; any optical activity exhibited by the resulting solution is due solely to the presence of dextrin.w. P. s. Approximate Estimation of Starch by Iodine. LESTER REED (Chem. News 1911 104 271).-A quantity of the substance con- taining about 0.1 gram of starch is heated with 5 C.C. of glycerol to a temperature of 190’ for five minutes; the mixture is then diluted to a volume of about 50 c.c. and filtered. To the filtrate when cold is added concentrated iodine solution (in potassium iodide) the precipi- tated starch-iodine compound is collected on a filter washed with boiling 90% alcohol and then rinsed into a platinum basin with a jet of boiling water. The contents of the basin are boiled to expel iodine then evaporated dried and weighed.Allowance is made for mineral matter present. The process cannot be applied to bread or cocoa or t o substances in which the starch has been heated previously. In the case of cocoa theobromine or some other substance is precipitated together with the starch. w. P. s. Estimation of the Acidity of Wine. FERNAND REPITON (Ann. PaZsif. 1911 4 578-580).-The following method is recommended for the estimation of the tota,l acidity of red wines Five C.C. of the wine are placed in a large beaker 5 drops of a cold saturated alcoholic phenolphthalein solution and 3 drops of a 1.2% alcoholic fluorescein solution are added and the mixture is heated just to boiling; 5 C.C. of cold water are then added and tbe solution is titrated with N/20- alkali solution.The red colour of the wine changes gradually to green during the titration and the end-point is taken when the red colour of the indicator appears in the green sdution. This point is readily seen if a white tile is placed under the beaker. w. P. s. Estimation of the Acidity of Milk. 0. RAMAfsTEDT (Chem. Zeit. 1911 35 121s-1219).-Soxhlet’s method (titration of 50 C.C. of milk with Nl4-sodium hydroxide solution using phenolphthalein as indicator) is the most trustworthy of the many processes which have been proposed for the estimation of the acidity of milk. I n the case of Thorner’s method (titrating a mixtnre of 10 C.C. of the milk anci 20 C.C. of water with N/lO-sodium hydroxide) the results obtained are lower than those yielded by Soxhlet’s method when both sets of results are calculated to the same basis; this lower result is due to the presence of the added water and the loss increases with the quantity of water.w. P. s. Estimation of Benzoic Acid. EDUARD REMY (Chem. Z e m t r . 191 1 ii 1379-1380 ; from Apoth. Zeit. 1911 26 835-836).-An iodo- riietric method is described. The benzoic acid is dissolved in 50% alcohol and the solution is treated with 5% potassium iodide andANALYTICAL CHEMISTRY. ii. 103 potassium iodate solutions ; the benzoic acid is converted into mono- iodobenzoic acid and an equivalent quantity of iodine is liberated. This free iodine is then titrated with thiosulphate solution; 1.22 grams of benzoic acid liberate 1.269 grams of iodine. w. P. s. Estimation of Cinnamic Acid in Aqueous Solutions by m e a n s of Bromine.ANNE W. K. DE JONG (Rec. tyau. chim. 1911 30 223-224).-The cinriamic acid is obtained in a finely divided state by solution in sodium hydroxide and precipitation by hydrochloric acid. Bromine water (N/lO) is then added until the solution does riot decolorise after five minutes. Potassium iodide is added and the liberated iodine estimated by titration. From this is calculated the amount of bromine used and thus the amount; of cinnamic acid in the solution. W. G. Estimation of Uric Acid by means of Iodine and the Action of Iodine on Creatinine. C. J. REICHARDT (Phwm. Zeit. 1911,56 922).-h the method described by Pizzorno (Abstr. 1911 ii 667) care must be taken in heating the urine with sodium carbonate as uric acid is converted into glyoxylic acid when boiled with alkalis.When the urine contains creatinine this is changed into creatine by the treatment prescribed. The author finds that creatinine is converted into creatine by the action of iodine; in this reaction the iodine at first combines with the creatinine and only after the lapse of some length of time is hydrogen iodide liberated with the formation of methylguanidineacetyl iodide. w. P. s. Estimation of Formaldehyde. ELIAS ELVOVE (Amer. J. Pharm. 1911 83 455-471).-The cyanide process was found to be trustworthy provided that an excess of cyanide is employed ; when this excess is not less than one-half equivalent the reaction is not influenced by moderate variations in temperature neither does the time during which the mixture is left appreciably affect the results obtained.This process is considered to be preferable to the hydrogen peroxide method both for dilute and concentrated formaldehyde solutions (compare Abstr. 1904 ii 98). w. P. s. Presence of Furfuraldehyde as an Indication of Adulter- ation in Some Fermented Alcoholic Beverages. V. PASQUERO and A. CAPPA (Gaxxetta 1911 41 ii 349-358).-Beer or wine gives a distillate which usually contains traces of furfuraldehyde as indicated by the coloration with aniline acetate. If however the beverage has previously been neutralised with magnesium carbonate the distillate no longer shows a furfuraldehyde reaction. The distillate from a neutralised wine or beer to which caramel or commercial alcohol has been added does give the furfuraldehyde reaction so that an indication of the presence of these adulterants can be obtained by neutralising the beverage with magnesium carbonate and testing the distillate with aniline acetate.R. v. s.ii. 104 ABSTRACTS OF CHEMICAL PAPERS. Microchemical Analysis of Plants. 111. D e t e c t i o n of Asculin by Micro-sublimation in the Examination of Rhizoma Gelsemii. 0. TUNMANN (Chenz. Zenti.. 191 1 ii 1384 ; from Apoth. Zeit. 1911 26 812-814. Compare Abstr. 1911 ii 669 1022 1023).-Asculin may be detected in the bark of Asculus Hippocu8tanurn by means of the sublimation test described by the author (loc. c i t . ) ; tbe spring is the only time of the year when the bark contains the substance. The latter is more rezdily obtained from Gelsemiurn sempervirens.The temperature employed for the sublima- tion should be from 50” t o 60° when the plucoside is obtained i n the form of prisms or rods with square ends. The crystals are soluble i n warm water alcohol and ether and the aqueous solution exhibits a bluish-green fluorescence which disappears on the addition of hydro- chloric acid. Chlorine colours the aqueous solution light red but the glucoside is best identitied by disholving the crystals in 2 drops of ammonia and adding 1 drop of nitric acid when a bluish-red coloration appears a t the point of contact. Bsculetin could not be detected in Radix Belladonnae but the seeds of Euphorbiu Lutlqris yielded a granular sublimate which became crystalline after about ten days. The subterranean portlions of Gelsemiurn elegnns which are sometimes sold as a substitute for Rhixomu Gelsemii yielded neitker aesculin nor Detection and I d e n t i f i c a t i o n of ‘‘ Saccharin ” and “ Dulcin ” in Beverages Foods Drugs Cosmetics etc.STEFANO CAMILLA and C. PERTUSI (Chew. Zents.. 1911 ii 1269-1270; from Giorn. Farm. Chim. 19 11,60 285-‘293).-“ saccharin ” (0-benzoicsulphinide) and ‘‘ dulcin ” (p-phenetolecarbamide) may be detected in solid sub- stances by heating about 30 grams of the sample with 1 gram of barium hydroxide and 50 C.C. of waher filtering the mixture and extracting the acidified filtrate with a mixture of ether and benzene. On evaporating the solvent the two sweetenirig materials a r e obtained as a crystalline residue. Liquids must be evaporated before being tested.“ Saccharin ” and ‘’ dulcin ” may be identified as follows 100 C.C. of the liquid or 10 grams of the solid substance under examination are mixed with 5 grams of magnesium oxide (in the case of solid substances 10 C.C. of water are also added) and the mixture is evaporated t o dryness. The residue is extracted with acetone contain- ing 10% of water the extract is evaporated to remove the acetone and the residual aqueous solution is shaken with ether; “ dulcin ” goes into solution in the ether. The 6‘ saccharin ” is obtained by shaking the acidified aqueous solution with a mixture of ether aiid light petroleum after the “ dulcin ” has been removed. Should salicylic acid also be present in the sample the residue of “mcchar-iri ” obtained on evaporating the ether-light petroleum solution is dis- solved in dilute sulphnric acid treated with a n excess of potassium permanganate in order to destroy the salicylic acid and the “ saccharin” then separated by again shaking the solution with a mixture of ether Estimation of Carbamide.ALEXANDRE DESGREZ and FEUILLIE (Compt. rend. 191 1 153,1007-1010).-A modification of Bouchard’s gelsemic acid. w. P. s. and light petroleum. w. P. s.A N A L Y T I C A L C HE MI STRY. ii. 105 method is described. The ureometer consists of a wide-stemmed pipette having the bulb near the point and a tap at the other end. The pipette is plunged into a narrow jar coctainiug chlorofoim when the lower part becomes filled and then 1 C.C. of urine 5 C.C. of water and 8-10 C.C. of Millon’s reagent are introduced through the tap.The chloroform is warmed to 30-35’ by heating a metal collar passing round the jar. After twenty to twenty-five minutes reaction is complete the pipette is transferred t o another vessel and the liquid displaced by water then by sodium hydroxide solution and finally by water after which the volume of the gas is measured. Under these ,conditions the ammonium salts uric acid amino-acids purine deriv- atives and undialysable substances in the urine are not decomposed. Allantoin is slowly decomposed but the amount of nitrogen from this source is iusufficient to affect the accuracy of a urine analysis. w. 0. w. Estimation of Amino-nitrogen in Polypeptides by van Slyke’s Method. ENIL ABDERHALDEN and DONALD D. VAN SLYKE (Zeitsch.ph,ysioZ. Chem. 19 11 74 505-508).-Determinations of the amino-acid nitrogen in a number of polypeptides have been made by van Slyke’s method ; these are in agreement with the theoretical values except in those cases in which the amino-group is present as glycine. When the values for the amino-nitrogen in glycyl polypeptides are multified by the factor 0-8 values more in accordance with the calculated are obtained. HERMANN FUHNER (Arch. exp. Puth. Pliarm. 19 11 66 178-190).-ChemicaI methods for the detection of aconitine give results which are not altogether iatisfactory in forensic cases ; the taste is characteristic but veratrine produces a very similar although weaker effect on the tongue The alkaloid can however be identified with certainty by its action on the isolated frog’s heart and 0*0005 mg.gives the effect. The effect in question is a characteristic peristalsis of which some tracings are given. W. D. H Reaction of Quinine and Other Alkaloids with Calomel. E. BARONI and 0. BORLINETTO (Chem. Zentv 1911 ii 392-395; Erom Giorn. Fawn. Chim. 1311 60 241-244).-Various salts of quinine give a brown colour when rubbed together with moist mer,ciirous chloride. The free base does not give the reaction. Sydrastine and heroine hydrochlorides pilocsrpine cocaine in the presence of ether and heroine in the presence of dilute alcohol give the same reaction. S. B. S. A. W. VAN DER HAAR (Pharm. Weekblccd 1911,48,1302-1307).-After criticising the official methods the author proposes the following process 10 grams of the extract are mixed with 20 C.C.of water and &hen boiled in a flask until the weight of the solution amounts to about 10 or 11 grams. Four C.C. of dilute hydrochloric acid are added and when cold she liquid is diluted to 20 grams The liquid is then shaken with E. F. A. The Toxicological Detection of Aconitine. Estimation of Hydrastine in H y d r a a t i s Extract.ii. 106 ABSTRACTS OF CHEMICAL PAPERS. 1 gram of talc and filtered. Ten grams of the filtrate are placed in a 100 C.C. flask 7 C.C. of ammonia are added and also 25 C.C. of ether when the whole is shaken for two minutes. Twenty-five C.C. of light petroleum (b. p. SO-SO0) are now added and after shaking for half a minute two grams of powder of tragacanth are introduced and the whole is again thoroughly shaken.Forty C.C. ar0 now taken off and after adding another 5 C.C. of light petroleum the solution is submitted t o distillation until 35 C.C. have passed over. The f h k is then put for eighteen to twenty-four hours in a cool place and after pouring off the liquid and washing the crystals with 2 C.C. of light petroleum Alkaloid . Reactions (Hydrastinine). C. REICHARD (Pharm. 2entr.-h. 1911 52 1253-1260).-A large number of reactions for hydrastinine are communicated of which the two following ones are the most characteristic and establish the identity of the alkaloid. A few particles of the hydrochloride are triturated with a minute crystal of potassium ferro- or ferri-cyanide and a drop of water is then added. With ferricyanide a splendid reddish-green precipitate is obtained which is characterised by its dichroism.The crystals form elongated prisms and are beautifully developed. When held up to the light they appear bluish dark green with total reflexion of the light ; the coloiirs are permanent. With the ferrocyanide crystals are obtained resembling mother-of-pearl ; these are particularly Characteristic for hydrastinine. Sodium nitroprusside gives a compound not unlike the ferricyanide. The behaviour of sulphuric acid is also noteworthy as the yellow colour given on warming again disappears on cooling and the test may Estimation of Morphine in Opium. H. WIEBELITZ (Chem. Zentr. 1911 ii 1383-1384; from Bpoth. Zeit. 1911 26 S24).- The alkaloid separated by the process described by Frerichv and Mannheim may be estimated volurnetrically by dissolving the moist precipitate in 25 C.C.of iV/lO-hydrochloric acid washing the precipitate until the filtrate and washings amount to 100 c.c. and titrating Colorimetric and Physiological Estimation of the Active Principle of the Suprarenal Gland. WORTH HALE and ATHERTON SEIDELL (Anaer. J. Pharm. 191 1,83,551-558).-0€ the various colour reactions of adrenaline that with iodic acid which has been described by Krauss and independently by Frankel and Allers (Abstr. 1909 ii 628) is the most trustworthy for the colorimetric estimation of this substance. The addition of phosphoric acid as recommended by Frankel and Allers is however not advisable a8 it causes the colour to fade. With specimens of the gland containing from 0.2 to 0.8% of adrenaline a quantity of 0.01 gram is heated just to boiling with 5 C.C.of dilute hydrochloric acid (2.5 C.C. of N/l-acid per 100 c.c.) and 5 C.C. of 0.2% potassium iodate solution. After fifteen minutes the mixture is filtered and the colour produced compared with a series of standards containing from 0.01 to 0.1 mg. of adrenaline per C.C. The standard solutions may also be prepared by mixing 1 volume of a solution they are dried on the water-bath and weighed. L. DE K. be repeated over and over again. L. DE K. 50 C.C. of this solution in the usual way. w. P. s.ANALYTICAL CHEMISTRY. ii. 107 containing 20 grams of potassium platiriichloride and 100 C.C. of hydro- chloric acid per litre with 3 volumes of a solution containing 12 grams of crystallised cobaltous chloride and 100 C.C.of hydrochloric acid per litre; quantities of this mixture are diluted so as to yield colora- tions corresponding with those given by definite amounts of adrenaline. The results obtained by this colorimetric process are lower than those found by determining the relative rise in blood-pressure as compared with that produced by a given amount of the pure substance when the solutions were injected into dogs. w. P. s. Catalytic Action of Light in the Oxidation of Phenolphthalin to Phenolphthalein. GIUSEPPE ROSSI (Chem. Zentr. 1911 ii 1659; from Giorn. Farm. Ci~im. 1911 60 433-436). -Phenolphthalein is reduced to phenolphthalin by the action of nascent hydrogen whilst the reverse reaction takes place in the presence of hydrogen peroxide and may be employed for the detection of blood.In the absence of the latter the oxidation proceeds very slowly and may take days but when blood is present the action is much more rapid. The test is carried out as follows An alkaline solution of phenolphthalein is shaken with zinc dust until colourless filtered and the filtrate mixed with the solution to be tested to which a few C.C. of 12% hydrogen peroxide have been added previously. If blood is present a red coloration develops within three hours (compare Abstr. 1909 ii 195). Exposure t o bright light such as an arc-light also increases the rapidity of the oxidation of phenolphthalin. w. P. s. Estimation of Methzemoglobin. JOSEPH BARCROFT and FRANZ MULLER (Proc. physiol. Xoc. 1911 xx; J . Physiol. 43).-See this vol. i 58.Use of Aniline Dyes as Reagents for Bile Pigments in Urine. OTTO V. C. E. PETERSEN (Chem. Zemtr. 1911 ii 1490; from Deutsch. rned. Woch. 1911 37 1891-1892).-Methyl-violet methylene-blue Nile-blue as well as indigo-carmine and Prussian-blue yield a green coloration when added to urine containing bile pigments. The green coloration is also obtained in the case of urine free from bile pigments and the intensity of the colour increases with the quantity of urine used. The test is therefore not a chemical reaction but the result of mixing blue and yellow colouring matters; it is not very sensitive but as a rough test a dye solution containing 1 part in Detection of Albumoses in Urine. EMIL HUGO FITTIPALDI (Chem. Zentr. 1911 ii 3489-1490; from Deutsch. med. Woch.1911 37 1890-1891).-Twenty C.C. of the urine are mixed with 120 C.C. of absolute alcohol and after the lapse of twenty-four hours the precipitate foi.med is collected on a filter; the precipitate is then dissolved in the least possible quantity of 32% sodium hydroxide solution and treated with freshly-prepared ammoniacal nickel solution (equal volumes of ammonia and 5% nickel sulphate solution). If albumoses or peptooes are present in the urine a reddish-orange colora- tion develops immediately. Other proteins uric acid cyanuric acid 10,000 may be employed. w. P. s.ii. 108 ABSTRACTS OF CHEMICAL PAPERS. xanthine hypoxanthine sarcosine and urobilin do not give a reaction. The test may also be used for the detection of albumoses in blood after the latter has been heated with animal charcoal cooled and filtered.w. P. s. The Different Methods for Estimating Pepsin and Trypsin with a Description of a New Simple Method. WILHELM WALD- scHmDT (f"22yer'~ Archiv 191 1 143 189-229).-The very numerous methods devised for the purpose are described discussed and classified. The new method is ft modification of Griitzner's carmine-fibrin process ; if diphenylrosaniline (spirit-blue) is substituted for carmine the fibrin so stained can be employed for the estimation of both pepsin and try psin. W. I?. H. J. DUMONT (Compt. rend. 191 1 153 889-891).-'l'he following method is stated to be rapid and accurate. Two to five grams of the dried soil are treated with about an equal weight of oxalic acid; the volume of carbon dioxide liberated gives the amount of carbonate present. The mixture is diluted until the solution contains 2% of acid boiled for thirty minutes filtered the residue washed with dilute nit,ric acid and finally with water. Calcium etc. may be estimated in the filtrate. The residue is washed with 75-80 C.C. of dilute ammonia and boiled with 20 C.C. of ammonia to remove humic acid. After two hours the liquid is decanted the residue suspended in dilute ammonia and the coarse sand separated by centrifugation and weighed. The colloidal clays in the decanted liquid are again centrifugated at two different speeds i n order to separate two deposits of different degrees of coarseness. Colloids remaining in suspension are coagulated by ammonium carbonate and removed after a third centrifugation. The final liquid is employed for determination of soluble extract and ash. New Method for the Physical Analysis of Soil. w. 0. w. The Application of Bio-chemical Methods in the Analysis of the Bilberry (Whortleberry). (Mlle.) A. FICHTENHOLZ (J. Pharm. Chinz. 191 1 [vii] 4 441-446).-The glucosides arbutin and methyl- arbutin are estimated in the dried leaves of the bilberry (Arctostaphylos Uva ursi) by the changes in rotation and reducing power produced by the action of emulsin on an extract of the leaves allowance being made for the reducing power of the quinol and methylquinol also pro- duced by the hydrolysis. The dried and powdered material is repeatedly extracted with boiling 90% alcohol and the extract after evaporating off the alcohol and dissolving in water purified with basic lead acetate the excess of lead removed and the liquid made to a known volume. It is then treated with yeast invertase for several days to remove the sucrose the amount of this substance present being determined by the change in rotation and reducing power which the liquid undergoes during this time. The invertase is then destroyed by boiling a pre- pnxation of emulsin is added and the rotation and reducing power determined a t intervals of several days until hydrolysis is complete. W. J. Y.