ANALYTICAL CHEMISTRY. Analytical Chemistry. ii. 47 The Process of Preparation of Charcoal Sticks for Reduction. NESTOR C. ALEXANDRESCU (BUZZ. SOC. chim. Romdnia 1919 1 11-12).-The following process for the preparation of charcoal sticks for the reduction test in qualitative analysis is advocated. Thin wood chips such as are used in match making are boiled for ~ W Q minutes in a 2.5% solution of ammonium phos- phate and then dried a t a temperature not exceeding 60°. When it is required to carry out thr test half the stick is burnt and then the test is carried out in the usual manner on the substance previously mixed with sodium carbonate. W. G.ii. 48 ABSTRACTS OF CHEMICAL PAPERS. Titrations with Surface-active Substances as Indicators. Estimations of Acidity with the Homologues of the Fatty Acids Series.WILHELM WINDISCH and WALTHER DIETRICH (Biochem. Zeitsch. 1919 97 135-156).-Adopting Traube and Somogyi’s method (A.. 1915 ii 101) of estimating the reaction by means of the alteration in the surface tension brought about by the displacement of the acid or the base of the indicator by the acid o r base of the medium the authors tried the homologues of the salts of the fatty acids series up t o undecylic acid as indicators. The acids with C,-C have been found to be as sensitive as litmus and neutral-red. Free acids can be estimated in the Dresence of a primary phosphate by using the above acids as indicators. s. s. z. A Considerable Source of Error in Titrations in the Presence of Phenolphtbalein. HAEUSSERXANN (Sfiddentech.Apoth. Zeit. 1919 59. 361-362; froin Chem. Zenfr. 1919 iv 396).-The author directs attention t o the errors caused by the presence of free carbon dioxide in distilled water which occur in titrations in which phenolphthalein is used as indicator H. W. Preparation of Phenolphthalein Solution and Paper without the Use of Alcohol. H. CLAASSEN (Zentr. Zuckerilzd. 26 209 ; Arch. Suikerind. 1919 26 1588-1590).-When alcohol is unavailable an indicator solution may be prepared by dissolving 1 gram of phenolpht3halein in 12.5 C.C. of N-sodium hydroxide and diluting to 500 c:c. with water. The alkali content of the solution is so small t h a t it can be disregarded in factory work. Where greater accuracy is required the result can be corrected for the alkalinity of the indicator.Test paper can be prepared by immersing paper first in a solution made by diluting 200 C.C. of the above indicator solution to 1 litre drying immersing i t in a solution of sulphuric acid of the proper strength to give the desired acidity and drying again. The results obtained with the paper should be verified from time t o time by direct titration. CHEMICAL ABSTRACTS. Estimation of Bromine in Mineral Waters and Brines. W. F. BAUGHMAN and W. W. SKINNER ( J . Znd. Eny. Chem. 1919 11 954-959).-Chromic acid liberates bromine from bromides quantitatively a t the ordinary temperature and the bromine may be removed by aspiration. Chlorides under the same conditions yield only a trace of chlorine which probably forms chromyl chloride and remains in solution. When a mixture of chlorides and bromides is treated with the reagent some chlorobromide is formed and is removed together with the bromine on aspiration.I n the method of analysis described the residue from the mineral water is oxidised with chromic acid. with the addition of hydrogen peroxide and the liberated heloqens absorbed in a solution of socliLim sulphite and carbonate. This is evaporated and the residueANALYTICAL CHEMISTRY. ii. 49 agaiii oxidised with chromic acid and a current of air aspirated through the apparatus the absorption vessels of which contain potassium iodide solutioii In this second aspiration the quantity of chlorine present is so small t h a t only pure bromine is evolved and its amount is found by titration of the liberated iodine. [See also J .Poc. C'heni. Ind.. 1920 1 9 ~ 1 Test Paper for the Detection of Iodine. ANONYNOUS (J. Phnrm. Brlg. 1919 from ;Inn. Chini. niinl. 1919 [ii] 1 357). -One gram of starch is mixed with 10 C.C. of water 40 c c . of 1)oiling water are added the mixture is boiled f o r two minutes and 0.6 gram of sodium nitrite is then added. The paste obtained is painted on strips of paper by means of a brush. and the s t r i p are dried. To make the test a few drops of the solution under examination are poured on the paper followed by one drop of dilute sulphuric acid. A blue coIoration is obtained if the solution C. A . M. contains iodide. The test paper keeps well. m. P. s. The Spectrocomparator an Apparatus for the Esti- mation of the Percentage Saturation of Blood with Oxygen or Carbon Monoxide.AUGUST XROGH (J. Physiol. 1919 52 281-287; from Ch~n2. Zentr. 1919 iv 210-211).-A modification of Hartridge's method (A. 1912 ii 488) is proposed which gives satisfactory results and only requires small amounts of blood. The apparatus is fully described and figured in thc original which must be consulted for details. H. W. Method for bringing Elementary Sulphur into Solution for Analysis. A. P. BJERREGAARD (J. Id. Eng. Chrm 1919 11 ]OX).-The finely divided sulphur is dicsolved in a small quantity of dry bromine nitric acid is added and the mixture heated The oxidation requires a few minutes only after the excess of bromine has been expelled the mixture is diluted wlth water. boiled with the addition of hydrochloric acid t o expel nitric acid and the sulphuric acid IC then precipitated in the usual way as barium sulphate.w. P. s. Non-protein Sulphur of the Blood. M. KAHN (Proc. SOC. Eap. Biol. Med. 1919 16 139; from Physiol. A h s t r . 1919 4 374). -A method of fractionating and estimating the non-protein sulphur of the blood is described. The Titration of Sodium Thiosulphate Solutions. I. M . KOLTHOFF (Phnrm. Tl'eeX*hlad 1919 56 644-637) -The prepara- tion teqting and application of various substances for the standard- isation of thiosulphate solutions are described in detail and their relative merits for the purpose discussed. The purest forms of potassium dichroniate obtainable commercially contain free chromic acid or potassinin chromate. A method for the detection and esti- mation of these is given which depends on the location of a discon- tinuity in the conductivity curve on the addition of alkali or acid J. C.TI.ii. 50 ABSTRACTS OF CHEMICAL PAPERS. If chromic acid is present the addition of standard alkali causes no increase in the conductivity of a dichromate solution until the free acid is neutralised. Similarly no increase in conductivity is observed on the addition of acid to a solution containing chromate until all the latter has been converted into dichromate. For analy- tical purposes it is recommended to melt the pure dichromate in an electric furnwe before use. The other substances examined are iodine oxalic acid cyanagen iodide potassium iodate and potassiuni brcmate. All these are easily purified. and give results in the titra tion of thiosulphate with an error of less than 0.1%. The greatest emor 0'07"/ waa observed in titrating with dichromate.W. S. M. Influence of the Position of Substituente on the Behaviour of Aromatic Nitro-compounds in the Kjeldahl Estimation of Nitrogen. B. M. MARGOSCHES and ERWIN VOCEL (Bcr. 1919 52 [R] 1992-1998).-The behaviour of the isomeric nitropheriols nitrobenzoic acids and nitrobenzaldehydes when treated with sul- phuric acid and potassium sulphate has been investigated ; it is found that the nitrogen content of the ortho-derivatives can be cor- rectly estimated i n this manner but t h a t the results for the meta- and para-derivatives are much too low. The favourable action of the hydroxy- or alkyloxy-group in the ortho-position t o the nitro-group is further shown by the fact t h a t the process gives exact results with 2-nitroresorcinol 3-nitro-p-cresol and o-nitrophenetole but low results with p-nitrophenetole.A further series of experiments shows t h a t the presence of o-nitrophenol or of o-nitrobenzoic acid has a favourable influence on the behaviour of the corresponding meta- and para-derivatives and t h a t this action cannot be ascribed solely to their content of phenol or benzoic acid. The possibility t h a t condsnsation products of the o-nitro-compounds may be the active agents has led t o an investigation of the behaviour of phthalic acid phthalic anhydride and phenolphthalein under like conditions biit the expected favourable action was not observed. The use of sali- cylic acid appears particularly advantageous in the case of meta compounds.It is somewhat remarkable t h a t the three isomeric iiitrocinnaniic acids yield accurate results. A further series of determinations is recorded with 2 4-dinitro- phenol 2 4-dinitro-a-naphthol 2 4-dinitrobenzoic acid 2 4-dinitro- toluene 3 5-dinitrobenzoic acid 2 6-dinitrotoluene and 2 4 G-tri- nitrophenol; the results are considerably too low in every case although coilcordant among themselves. The facts however that tht 2 4-dinitro-compounds give higher values than the correspond- ing 3 ' 5 - or 2 ' 6-dinitro-derivatives and t h a t 2 4-dinitrophenol gives results almost identical with those obtained with 2 4-dinitrotoluene arc in accordauce with the observations made with the mononitro- compounds. H.W. Grete's Volumetric Method (for Estimating Phosphoric Acid). R. W. TUINZING (Imzdir. VeCielauchs.-stat. 1919 94 191-195. Compare A 1916 ii. 490).-This method was foundANALYTICAL CfIEXISTRY. ii. 51 to be trustworthy. Directions are given for the preparation of the ammonium molybdate-gelatin solution used and for the recovery of ammonium molybdate from t,he residual solutions. W. P. 5. Behaviour of certain Organic Arsenic Compounds in Marsh’s Test. DOWENICO GANASSINI (Boll. Chint Farm. 1919 58 385-390).-The addition of platinum chloride to promote the evolution of hydrogen is inadvisable even in the case of mineral arsenic since it may fix part of the arsenic as platinum arsenide. I n the case of cacodylic acid platinum chloride in large excess com- bines to form a double compound cacodylplatinochloride which may remain undecomposed. An analogous compound is formed between platinum chloride and methylarsinic acid. Cacodylic acid in Marsh’s test produces an orange-yellow deposit in addition t o the ordinary black deposit.This probably consists of erythrarsine. By heating the hydrogen flask to about 335O i n a bath of melted lead only the orange-yellow deposit is obtained. Methylarsinic acid and neosalvarsan under the same conditions also yield yellow deposits but these may be distinguished from the cacodylic acid deposit by the fact that during the formation of the former the gas does not form dense white fumes. The gas from the methylarsinic acid test gives a yellow precipitate with Bettendorf’s reagent (stan- nous chloride solution saturated with hydrogen chloride) whilst the lemon-yellow ring from neosalvarsan is only formed at a high tem- perature does not change to black and is readily soluble in ammonia solution.Atoxyl and salvarsan yield only the ordinary black rings in the test. [See also J. Soc. C h m . Z n d . 1920 4 2 ~ . ] C. A. 31. Separations in the Arsenic Group. WILIIELX S m E c K m and ADOLF RIEDEMBNN (Rep. 1919 52 [U] 1935-1947).-The authors recommend the following modification of the distillation method f o r the estimation of arsenic. The arsenic solution is placed in a flask provided with a dropping funnel and connected through ail efficient worm condenser with a vessel containing water; the flask is half filled with concentrated hydrochloric acid potassium bromide (1.5 grams) and a few pieces of porous earthenware are added.The contents of the flask are heated to gentle boiling and thionyl chloride (10 c.c.) is gradually run in a t such a rate that the time of addition is half an hour. The distillate is diluted with water to 700-800 c.c. boiled under reflux in a brisk current of carbon dioxide until sulphur dioxide is completely expelled and the sulphur has collected (this should be the case in half an hour a t the most) and filtered. Arsenic is estimated in the filtrate as ths trisiilphide. Phosphorus trichloride (25 c.c.) may replace the thionyl chloride in which case the subsequent boiling of the distil- late is unnecessary; on the other hand the presence of phosphorous acid i n the residue often causes complications in subsequent estima- tions.The inethocl gives accurate results and can be applied t o the separation of arsenic from antimony tin copper lead mercury,ii. 52 ABSTRACTS OF C'HENICAL PAPERS. and iron aud for the estimation of arsenic and iron in arsenide of iron. The separation of antimony from tin can be effected by taking advantage of the fact that antimony chloride is volatile a t 155-165O from solutions in which the tin can be rendered non- volatile by the presence of phosphoric acid; the tin can be volati- lised subsequently a t a somewhat higher temperature by taking advantage of the fact that the addition of hydrobromic acid destroys the restraining action of the phosphoric acid. The authors have attempted to shorten the time required for this separation by utilis- ing the catalpsing action of hydrobroniic acid from the commence- ment but the results are unsatisfactory since tin distils over with the antimony the catalpsing influence of hydrogen bromide out- balancing the restraining effect of phosphoric acid.On the other hand the use of hydrobromic acid is very advantageous in the separation of antimony or tin alone from other elements and does not require the absence of nitric acid. Thus i n the separation of antimony aud copper the hydrochloric acid solution of the metals is added to concentrated sulphuric acid (6 c.c.) and phosphoric acid (D 1.78. 7 c.c.) cmtained in a distillation flask connected with a receiver containing hydrochloric acid. The contents of the flask are heated a t l G O o and a mixture of concentrated hydrochloric acid (10 volumes) and hydrobroniic acid (I3 1.78 1 volume) is added a t such a rate that the temperature remains constant.The antimony is completely removed in thirty minutes. The separation of antimony from lead is accomplished similarly. Tin is quantitatively separated from copper and lead in a n analo- gous manner; the solution of the metals in concentrated hydro- chloric acid is treated with sulphuric acid (12 c.c.) and distilled a t 1 GOO with the regulated addition of the hydrochloric-hydrobroniic acid mixture (20 c.c.). The process is complete in half an hour; the tin is precipitated from the distillate as the sulphide and weighed as the oxide. The method call be applied to the estimation of tin in brass. The separation of antimony and tin from mercury cannot be effected in this manner since the latter is also partly volatilised. A novel type of filter tube is described which is constructed by drawing out the end of a piece of wide glass tubing so as t o form a stem; a filter disc covered v i t h asbestos rests on the shoulder of the wider tube which is further provided with a ground-in cap in which a capillary opening is made.Early Developments in Organic Macro- and Micro-analysis. J. V. DUBSK~ (Chem. Tl'eeX blad 1919 16 1482-1493).-The author gives a retrospect of the work of various experimenters in connexion with the analysis of organic substances and indicates the lines along which research progressed. E. OTT (./. Grcsbelricclit 1919 62 89-90; from C'hriii. Z e n f ? 1919 iv 171-172).-Uncertaiii'Lies in fractionaI combustion with copper H.W. W. J. W. Gas Analytical Combustion with Copper OxideANALYTICAL CHEMISTRY. ii. 53 oxide arise in consequence of the dissociation CuO = Cu + 0 which can be eliminated by subsequent repeated passage of the nitrogen over the material in the quartz tube which is maintained a t a low red heat. Preparation of chemically pure nitrogen by means of phosphorus and heated copper oxide is liable to error and recourse should be had to the pyrogallol method. Copper oxide does not appear to be suitable f o r gas analyses in which the carbon dioxide formed by combustion is measured. Estimation of Carbon Monoxide in Blood. D. I). VAN SLPKE and H. A. SALVESEN (Proc. Soc. Exp. Biol. X e d . 1919 16 140; from Physiol.d ? ) s t v . 1919 4 374).-The blood is treated as in the estimation of oxygen by Van Slyke's method. A mixture of carbon monoxide and oxygen with a slight amount of nitrogen is obtained. The oxygen is absorbed by alkaline pyrogallol solution. The residual gag after a correction has been made for nitrogen is carbon monoxide. J. C. D. A Micro-method for the Estimation of Calcium in Blood- serum and other Organic Substances. I). J . DE WAARD (Biochem. Zeitsch. 1919 97 176-186).-The substances are eva- porated and incinerated and the salts of the ash extracted with dilute hydrochloric acid. The calcium is precipitated from the solu- tion as the oxalate which is centrifuged washed dissolved in dilute acid and titrated with potassium permanganate.0.1 Mg. of calcium can be estimated by this method with an error of 4%. A Direct Micro-estimation of Calcium in the Serum. D. J. DE WAARD (Biochem. Zeitsch. 1919 97 lf36-189).-hll the calcium can be precipitated directly from the serum by means of ammonium oxalate and be estimated by the method described in the preceding abstract. The difference between an estimation carried out on tbe incinerated serum and on the original serum was only Ferrous Sulphide as an Indicator in Acidimetry and a New Volumetric Method for the Estimation of Zinc. J. HOUBEN (Ber. 1919 52 [B] 1613-1621).-Since the produc- tion of ferrous sulphide is prevented by the slightest traces of acids it is possible to titrate an acid by adding a crystal of pure ferrous ammonium sulphate passing in well-washed hydrogen sulphide (not enough to saturate the solution) and running in alkali until the black colour of ferrous sulphide is permanent. Experiments with hydrochloric and sulphuric acids on the one hand and alkali hydr- oxides and carbonates or borax on the other show that the end-point is quite as definite as it is in the case of methyl-orange or phenol- phthalein.The only indistinct end-point is found when titrating sulphuric acid with borax but here the difficulty may be overcome by running in more acid until the black colour just disappears again. The method is obviously useful ill the case of colourecl liquids or when working in a bad light. H. W. S . S. Z. 0.001 nig. of calcium per C.C. s. s. z.ii. 54 ABSTRACTS OF CHEMICAL PAPERS. I n the case of zinc salts hydrogen sulphide completely precipitates zinc as sulphide from the sulphate if the solution is less than 0.2AT and from the chloride if the concentration is below 0'067N.A titration of the free acid produced would therefore be a means of estimating ziiic in its neutral salts and here the above use of a ferrous salt proves its worth. The solution is saturated with hydro- gen sulphide some ferrous ammonium sulphate is added and alkali is run in preferably borax solution until the white turbidity becomes brown. A '' back titration '' with an acid is not possible in this case however without first filtering the zinc sulphide as a double zinc-ferrous sulphide is formed which is not quickly broken up. Examples of the application of the method to metallic zinc zinc salts zinc ash and blende show t h a t the results approach those obtained by gravimetric analysis very closely in fact much more so than those given by any other volumetric method.Blende is examined as follows. About 5 grams of the mineral are digested with 40 C.C. of concentrated hydrochloric acid and 40 C.C. of water the solution filtered and saturated with hydrogen sulphide filtered again and diluted to 1000 C.C. with dilute hydrochloric acid. Por- tions of 50 C.C. are now neutralised with sodium carbonate using methyl-orange treated with hydrogen sulphide f o r half an hour mixed with ferrous ammonium sulphate and titrated with borax or sodium carbonate. The alkalimetric titration of ferric salts may be carried out by employing the same principles.Electrometric Analysis with Potassium Ferrocyanide. ERICH MULLER (Zeztsch. nngezu. Ckem. 1919 32 351-352).- Electrometric titration of lead and zinc salts with potassium ferro- cyanide solution is trustworthy but the method fails in the case of other metals owing to the fact that the precipitates produced do w. P. s. Colorimetric Estimation of Lead Dioxide in Litharge. WALLER V. MORGAN (J. Incl. ICny. C'hem. 1919 11 1055).-The method depends on the fact that lead dioxide oxidises aniline to aniline-purple and t h a t the coloration obtained is proportional t o the amount of the dioxide present. Five grams of the litharge con- taining lead dioxide are boiled for one minute with 2 grams of aniline hydrochloride dissolved in 10 C.C. of water and 5 C.C. of concentrated hydrochloric acid ; the mixture is then filtered and the coloration exhibited by the filtrate compared with those given under the same conditions by litharge containing known amounts of lead Complex Internal Salts in Quantitative Analysis.I. I. BELLUCCT and A . CHIUC~NI (G'azirttn 1919 49 ii 187-216).- The authors summarise and criticise published work on the itpplica- ticjns of a-nitroso-P-naphthol and ' I ciipferron ' to the separation and estimation of metals. Of course cobalt and nickel salts must be absent. J. C. W. not have definite compositions. peroxide. w. P. s.ANALYTICAL CHEMISTRY. ii. 55 For the separation of copper neither of these reagents offeis appreciable advantages over ordinary analytical methods except perhaps for the separation of copper from arsenic and more parti- cularly from antimony.For the separation of iron cupferron is t o be preferred t o u-nitroso-P-naphthol chiefly because the ferric precipitate obtained with the former withstands a greater degree of free mineral acidity and is therefore more easily freed from extraneous metals and also because the precipitate is less volumin- ous. a-Nitroso-P-naphthol serves principally as a reagent for cobalt and also as a reagent for palladium whilst cupferron acts as a reagent for titanium zirconium and vanadium as well as for iron. The two compounds assume indeed the character of general reagents for groups of elements which contain however elements quite different from those usually classed together. T. H. P. Detection of Cerium.FR. FEIGL ( d s t e ~ r . Chem'Zezt. [iii 22 124-126; from Chem. Zemtr. 1919 iv 592).-Lecoq's reaction is much more distinctly obtained after the addition of a small quan- tity of an aluminium salt but is not applicable in the presence of iron. A vivid blue coloration is obtained when cerium dioxide is moistened with a solution of benzidine in acetic acid ; other cerous and ceric compounds with the exception of cerous fluoride cerous carbonate and double sulphates of cerium behave similarly. The reaction is most sensitive when the solution under investigation is made just alkaline with sodium or potassium hydroxide heated to boiling filtered and the filter paper treated with a drop of benzidine solution; 0.02 mg. of cerium per litre can be thus detected. The other metals which belong to the ammonium sulphide group of the rare earths with the exception of thallium do not show the reac tion The absence of other oxidisiiig agents and of manganese cobalt thallium and chromate is essential since the reaction depends on oxidation.Faintly acid solutions of iron also give the reaction ; after previous addition of sodium or potassium hydroxide a coloration is not observed. I n the presence of iron it is advisable to precipitate cerium as the fluoride and to convert the latter by alkali hydroxide into the hydroxide since precipitated ferric hydr- oxide carries down notable amounts of cerium which are thus removed from the sphere of action. H. W. Soil Analysis. F. 11 uxmn (Laitdzc. Versuclis.-stut. 1919 94 181-1 €39).--For the estimation of iron aluminium calcium mag- nesium pot3assium and phosphoric acid in a soil i t is recommended that 300 grams of the sample be mixed with 900 C.C.of concentrated hydrochloric acid and shaken occasionally during forty-eight hours ; the solution is then decanted filtered and 300 C.C. of the filtrate are eva.porated with the addition of ammonium chloride. The residue obtained is treated with nitric and hydrochloric acids again eva- porated this operation is repeated and the final solution in hydro- chloric acid is used after separation of the silica for the estimation of the above-mentioned constituents. w. P. s.ii. 56 ABSTRACTS OF CHEMICAL PAPERS. Application of Hsematoxylin to the Detection of Iron in Tissues. ,I. MAWAS (Compt. rend. SOC. Bzol. 1919 82 155-158; from L'krm.Z e n t y . 1919 iv 245).-The author considers t h a t the reaction of hematoxylin with ferrous ions is to be ascribed t o salt formation and not as Macallum assumed to oxidation. The latter's statement t h a t the reaction only occurs with inorganic iron is incorrect since iron compounds of the albumins which are not acted on by hydrogen sulphide or ammonium sulphide but are affected by potassium ferro- or ferri-cyanide also show this change. The formation of Prussian-blue is a more delicate test for the pres- ence of iron in tissues; nevertheless the author has obtained useful results with hematoxylin particularly i n certain cases of disease of the eye. The affinity of the haematoxyliii f o r the chromatin of the n~icleus whereby a similar coloration is developed is disadvan- tageoiis.H. w. Brazilin and its Iron Lakes. J. MAWAS (C'ompt. rend. SOC. JIloZ. 1919 82 158-159; from C'hem. Zentr. 1919 iv 245).- Rrazilin is a more sensitive and more rapid reagent than hnniat- osylin (preceding abstract) ; its aqueous solution I'ale red with an orange fluorescence gives dark brown insoluble lakes with ferrous salts. I n aqueous or alcoholic solution it colours tissue containing iron dark brown within a few minutes whilst the nucleus becomes reddish-violet. Differentiation which may be necessary owing to over-colouring is effected with alcohol and chloroform or prefer- ably with alcohol containing 1% of hydrogen chloride. The nucleus becomes practically decolorised whilst the iron pigment is un- affected ; the former again beconies reddish-violet when washed with dilute aqueous alkali.H. W. Iodometric Estimation of Iron. I. M. KOLTSIOFF (Pharm. IVeelcblad 1919 56 1565-1568).-The estimation of ferrous iron by means of iodine with addition of a pyrophosphate gives results which are about 37k too low owing to oxidation of the iodine by dissolved oxygen in the solutions. If sulphuric acid and sodium hydrogen carbonate are added to eliminate the oxygen the error is reduced t o 0.8-1:L. A more suitable method which gives accurate results is the following 25 C.C. of 0.liV-potassium bromate and 10 c c . of 25% phosphoric acid are added to 10 C.C. of O.lll;-ferrous solution to which a few drops of acid have been added. After remaining for five minutes in a stoppered flask 5 C.C.of potassium iodide and 2 drops of molybdate solution are added the iodine being titrated after five minutes with thiosulphate. The purp0.e of the phosphoric acid is to combine with the ferric ions and prevent them interacting with the potassium iodide and the molybdate is added to accelerate the reaction of the bromate and iodide. w 3 w. Volumetric Estimation of Iron. J. HOUBEN (Ber. 1919 Compare this vol. ii 53).-The process 52 [R] 2072-2076.ANALYTICAL CHEMISTRY. ii. 57 perinils the estimation of ferrous and fewic iron and free mineral acid in a solution. A portion of the solution is completely reduced by hydrogen SUI- phide and titrated with -V-potassium or sodium hydroxide solution until the dark coloration due to the incipient precipitation of ferrous sulphide is observed which does not disappear after vigor- ous shaking; the free acid plus one-third of the acid originally com- bined with the ferric salt is thereby estimated A second portion of the solution is reduced with sulphur dioxide (air being completely escluded) most of the excess of the latter is removed by boiling and the cooled solution is titrated with S-alkali as before after being treated with hydrogen sulphide which completes the removal of sulphur dioxide.The free acid plus two-thirds of the acid origiii- ally united to the ferric iron is thus estimated. The difference between the two titrations multiplied by three gives the number of C.C. of .T-ferric chloride solution originally present. If this number is subtracted from the smaller of the two burette readings the dif- ference is the number of C.C.of S-acid which were uncombined iu the solution The ferrous salt is estimated in the usual manner Test analyses show the method to have a high degree of accuracy. H. W. Separation of Iron Aluminium Chromium Glucinum Titanium and Zirconium by the Sodium Carbonate Method. P. WENGER arid J. WURRMANN (Ann. ChLim. nnnl. 1919 [ii] 1. 337-339).-A method described previously by Wenger and Wunder (A. 1912 ii 687) may be applied when the mixture contains titanium or zirconium hut not both. After fusion with sodium carbonate the insoluble iron and zirconium oxides may be treated with hydrochloric acid (1 l ) in which the zirconium oxide is insoluble. I n the absence of zirconium iron and titanium may be separated by one of the known methods even in the presence of ghicinum.[See further .7. S'or.. P l i ~ w . Zn/7. 1920 4 6 ~ . ] w. P. s. A New Method for the Volumetric Estimation 01 Nickel. JOSEF HOLLWTA (Monntsli. 1919 40 281-291).-The process depends on the fact that when a nickel salt reacts with dimethyl- glyoxime acid is liberated which is estimated by titration with alkali in the presence of phenolphthalein or methyl-red. The standard iT / 50-solution is prepared by dissolving 4,6400 grams of pure dimethylglyoxime in the requisite quantity of 97:A alcohol (300-400 c.c.) ; 20 C.C. of AT-potassium hydroxide solution are added wihh constant shaking and the solution is diluted to 1000 C.C. with distilled water free from carbon dioxide and filtered after twenty-four hours from any slight precipitate of potassium carbonate.(The alcohol used must be free from aldehyde and neutral in reaction.) The alkali content of the solution is checked by titration with standard acid using phenolphthalein or methyl- red as indicator. The nickel solution under examinat.ion is diliitedii. 58 ABSTRACTS O F CHEMICL4L PAPERS. to the requiied extent a i d exactly iieutralised with N 10- or N / 5 0 - potassium hydroxide solution. 2-5 C.C. of the standard solution are added and after vigorous agitation the solution is gently warmed whereby the precipitate is caused to collect leaving a clear liquor ; the standard solution is then gradually added with frequent agitation until the solution acquires a permanent pale pink color- ation.I n consequence of the bulky nature of the nickel precipitate the amount of the metal in the solution should not exceed 0.03 gram. The method is particularly advantageous in dealing with very m a l l quantities of nickel and titration can readily be effected with X / 100-solutions if methyl-red is substituted for phenolphthalein as indicator. The results are not affected by the presence of the alkali salts of strong acids. Further details are promised in a subsequent communication. H. W. Iodometric Estimation of the Chromic Acid in Lead Chromate. MAX GROGER (Zeztsch. anorg. Chem. 1919 108 267-272) .-The iodometric method of estimating chromic acid has not hitherto been used for lead chromate on account of the impossi- bility of completely decomposing the chromate by means of dilute sulphuric acid.Dilute hydrochloric acid dissolves lead chromate completely but hydrochloric acid has a reducing action on chromic acid with liberation of free chlorine and has therefore not been used to dissolve lead chromate. It is now found however t h a t if the hydrochloric acid is sufficiently dilute its reducing action is negligible. Experiments showed t h a t when acid of 1.25N strength or less was used in the proportion of 25HC1 to 1 mol. of chromate no reduction took place after boiling the solution f o r three hours. With stronger hydrochloric acid the rate of reduction increased rapidly with the concentration of the acid. For carrying out an aualysis about 0.3 gram of lead chromate is dissolved in 50 C.C. of 1.25AV-hydrochloric acid by gently heating.The solution is cooled 1 gram of potassium iodide added allowed t o remain for ten minutes diluted with 100 C.C. of water and the free iodine titrated with sodium thiosulphate. There is no need to filter off the pre- cipitated lead iodide as the colour of this does not interfere with the end-point. After the end of the titration free iodine again appears after a short time. This is due to atmospheric oxidatiou of hydriodic acid a reaction which appears to be catalysed by the lead iodide present. The results obtained by this method are generally about 0.5% too high. Detection of Molybdenum by means of Xanthic Acid. J. KOPPEI (C’hem. Zeit. 1919 43 777-778).-Molybdenun1 com- pounds give a characteristic red to plum-coloured reaction with xanthic acid which was first described by Siewert (Zeitsch. ges.iYaturw. 1864 23 5). The test is preferably applied by acidify- ing the solution with a mineral acid after the addition of a trace of solid potassium xanthate. and is capable of detecting molybdenum in a solution containing 0~00000064 gram per c.c The reaction is E. H. R.ANALPTICAL CHEMISTRY. ii. 59 weaker when phosphoric formic or acetic acid is used for the acidification. and much weaker with salicylic acid whilst when oxalic citric or tartaric acid is used there is little if any color- ation. Analogous xanthic compounds are formed by copper iron cobalt nickel etc. but the colorations do not interfere with the molybdenum test. Vanadic acid forms a yellow compound and uranic acid a brown compound which like the molybdenum com- pound are soluble in ether and other organic solvents.Tungstic acid if present in very large preponderance may interfere with the test. The molybdenum compound may be obtained as a black oil by mixing a solution of 2 grams of ammonium paramolybdate with 1 gram of potassium xanthate. When vigorously stirred this oil solidifies forming black crystals with green surface reflections. The mean analyses of Siewert's solid preparation agree with the formula Mo0,,(OEt*C*S*S*H)2. [See also J . SOC. Cl~em. Zu(7. 1920 January.] C. A. M. Colour Reactions of Molybdenum and Tungsten. 11. G. A. BARBIERI ( A t t i R. Accnd. Lincei 1919 [vl 28 i 390-392. Compare A. 1919 i 549).-According t.0 Pkchard (A. 1894 ii 319). replacement of the hydrochloric or siilphuric acid used in Braun's reaction by acetic acid results in the formation of a yellow coloration and with concentrated solutions of a yellow. crystalline precipitate due to conversion of t.he molybdate into trimolybdate and union of this with the thiocyanate present to form an additive compound for instance KCNS,K20.3M00,,4H,0.On repetition o f PBchard's reaction with ammonium thiocyanate and ammonium permolybdate or tri- or tetra- or normal molybdate the author finds that in all cases the additive compound NH,SCN,(NH,),0,4Mo03,5H20 is formed and in no case an additive compound of the trimolybdate. From potassium trimolybdate and thiocyanate in presence of acetic acid the corresponding cornpozind KSCN,K,0,4Mo03,5H,0 is formed ; both potasium and ammonium compounds form yellow acicular crystals.Ammonium tetramolybdate. (NH4),0,4Mo0,:2~H,0 may be obtaiiied by the action of acetic acid on ammoilium permolybdato ill aqueous solution (compare Junius A 1905 ii. 826). T. H. P. Qualitative and Quantitative Test for Molybdenum in Steel and Iron. SIEGFRIED LAURENS MALOWAN (Zeitsch. anory. CILeni. 1919 108 73-80).-A specific and very sensitive test for molybdenum is given by xanthic acid. The test is best carried out with freshly prepared xanthate solution in the following manner. Absolute alcoholic potash is shaken with excess of carbon disulphide until no more of the latter is dissolved. To the solution so obtained 30% acetic acid is added until it shows a slight yellow turbidity and the reagent is added drop by drop to the solution t.0 be tested.If molybdenum is present an intense red colourii. 60 ABSTRACTS OF CHEMICAL PAPERS. develops which is quite stable in water and the intensity of which is proportional to the quantity of molybdenum present. The sensitiveness of the reaction is such t h a t 0.000005 gram of molybdenum in 0.00070/ solution can be detected with certainty in presence of ot$her elements. The coloured product is readily soluble in ethyl or amyl alcohol ether or chloroform less easily in petroleum and benzene. It is readily extracted from its reddish-violet ethereal solution by alkalis from which acids precipitate it. The ethereal solution decomposes on keeping with formation of molybdenum sulphide. The reaction is not given by vanadium tungsten titanium or uranium; chromates give a dark coloration with xaiithic acid and should be reduced if present before the test is made.The reaction can be used for the colorimetric estimation of molybdenum the coloured product being extracted with a mixture of 65Oh ether and 35% light petroleum and a mixture of 30% ether and 70% light petroleum used for dilution for comparison with the standard solution. If pure ether is used decomposition takes place too rapidly for accurate determinations to be made. E. H. R. Colorimetric Method for the Estimation of Small Quantities of Uranium. MULLER (Chem. Zezt. 1919 43 739-740) .-The method depends on the red coloration obtained when a uranyl salt is treated with sodium salicylate; the color- ation yielded by any uranyl salt solution is compared with t h a t produced by a known amount of uranium under the same coiidi- tions.Free mineral acid acetic acid iron alcohol or acetone must not be present but neutral alkali salts do not interfere. The method is suitable for the estimation of uranium in solutious con- taining as little as 0,0276 of the metal. w. P. s. Estimation of Zirconium by the Phosphate Method. G. E. F . LUNDELL and H. B. KNOWLES (1. Amer. Chen~. SOC. 1919 41 1801-1808).-Zirconium can be quantitatively pre- cipitated as secondary zirconium phosphate in cold or tepid soh- tions containing froin 2% t o 20% by weight of sulphuric acid pro- vided t h a t a ten- to one hundred-fold excess of the precipitant diaiumonium hydrogeij phosphate is used. Hydrolysis which occurs when the phosphate precipitate is washed with water can be almost entirely avoided by the use of a cold 5% ammonium nitrate solution for washing.Zirconium pyrophosphate for which the ZrO factor is 0.4632 is obtained on ignition of a secondary zirconium phosphate which has been washed with ammonium nitrate solution. No definite composition can be ascribed t o the com- pound resulting when secondary zirconium phosphate which has been mashed with water is ignited. Zirconium can be quantita- tively separated as phosphate in a 20% sulphuric acid solution from iron. aliiiniuinm chromium cerium. and thorium. The separatioll froin titanium Cali alio be effected provided hydrogen peroxide is p~eseiit. J . F. 8.ANALYTICAL CHEMISTRY. ii. 61 Gas-analytical Separation of Acetylene Ethylene and Benzene.W. D. TREADWELL and F. A. TAUBER (Hclv. Chim. A cta 1919 2 601-607).-The quantitative absorption of acetylene from mixtures containing it together with ethylene and benzene vapour is conveniently effected with a solution of mercuric cyanide (20 grams) in 2N-sodium hydroxide solution (100 c.c.) ; the gas is shaken with the mixture (6 c.c.) for one to two minutes a t the end of which the acetylene is completely absorbed whilst ethylene and benzene vapour are unaffected. A solution of mercuric nitrate (20 grams) in 2iT-iiitric acid (100 c.c.) which has been saturated with sodium nitrate is recom- mended for the absorption of ethylene in the presence of benzene vapour; the gas is shaken with 5-10 C.C. of the solution for two to three minutes. Propylene appears to behave similarly to ethylene since it is indifferent to alkaline mercuric cyanide solu- tion whilst readily absorbed by mercuric nitrate solution.H. W. Dehydrogenation of Hydrocarbons by means of Palladium-black. JENO TAWZ and NIKOLAUS VON PtrmoKY (Ber. 1919 52 [B] 1573-1583).-Zelinski has shown that cycle- hexanes are converted into benzene hydrocarbons by exposing them to the action of palladium-black a t 300° whereas hexane aiicl c!/cZopentanes are unchanged (A. 1913 i 167). This seemed to offer a process whereby cyclohexanes could be estimated in the presence of paraffins and a method has now been developed which gives satisfactory results within certain limits. The first difficulty to be overcome was the preparation of a sufficiently active catalyst for Zelinski’s agent is spoilt by mixtures of paraffins and cyclohexane long before the latter is completely dehydrogenated. I n the reduction of palladium solutions it is usual to employ formic acid any exceps being destroyed by the palladium-black itself.It seemed possible therefore that the activity of the catalyst would suffer in such a treatment and as it was found that the activity towards formic acid is parallel to the activity towards cyclohexane it was decided to use no more than the theoretical quantity of the reducing agent. This has the desired effect the catalyst being much more active. For example two specimens of 1 grain each prepared by Zelinski’s method pro- duced 1-8 and 11-2 C.C. of gas from formic acid but ten samples made by the new process which is described in detail gave 011 an average 458 c.c.one portion developing as much as 840 C.C. With such an active catalyst the quantitative dehydrogenation of c?/c/ohexane in various mixtures with hexnne may be easilv realised ; but then another disturbing factor appears namely that a t 300° the saturated paraffins themselves are to a certain extent converted into unsaturated hydrocarbons by palladium black. The sinall amount of hydrogen set free by the paraffins however reachw a constant volume after an hour or two and although a little more gas is liberated when the catalyst is renewed ail equilibriuinii. 62 ABSTRACTS OF CHEMICAL PAPERS is soon established. A similar equilibrium is reached when the olefines are heated with an excess of hydrogen t h a t is the hydrogenation is never quite complete. The volume of hydrogen produced by the paraffins is relatively small compared with the yield from the cyclohexanes so within certain limits the propor- tion of the latter hydrocarbons in a mixture can be gauged from the amount of hydrogen liberated.The limit is generally about 1% of cyclic hydrocarbon but in some experiments it was as low as 0.576. For mixtures containing less than this. only the chemical proof of the presence of benzene will suffice. The reaction with trioxymethylene and sulphuric acid is recommended. The catalyst (12 grams) is contained in a tube heated in a siniple electric furnace and the vapour is continually circulated. Thc volume of oil employed varies from 10 C.C. for rich mixtures to 100 C.C.I n the experiments described the following volumes of hydrogen were produced by 1 C.C. of hydrocarbon a t 300° (or 270O in the case of the methyl derivatives) cyclohexane 614 c c . ; methylryclohexane 618 C.C. ; 1 3-dimethylcyclohexane 504 C.C. ; ivopentane 38 C.C. ; n-hexane 40.8 C.C. (after renewing the catalyst five times) ; n-heptane 32.5 c c. ; n-octane 36 C.C. As an important application of the method the analysis of a sample of hexane b. p. 69-7Io from American petroleum is recorded. Whereas 30 c c. of synthetic n-hexane gave only 338 c.r. of hydrogen 30 C.C. of the natural oil yielded 446 C.C. under exactly the same conditions. This corresponds with 2.4% of cf/cZo- hexane in the sample and actually 0.5592 gram of pure dinitro- benzene was obtained from the dehydrogenated 30 C.C.of oil. For quantitative purposes a special apparatus is described. J. C. W. Criteria of the Degree of Purity of Commercial Toluene. JOHS SCOTT LT-NSDEN (T. 1919. 115 1366-1372). Differentiation of Methyl and Ethyl Alcohols. PANWWITZ (Ph arm. Zrntr.-h. 1919 60 441-442) .-Although crystallised copper sulphnte is insoluble in ethyl alcohol and dissolves in methyl alcohol t o give a blue solution this test is useless for dis- tinguishing methyl alcohol from ethyl alcohol. owing to the fact t h a t a small quantity of water destroys the blue colour of the methyl alcohol solution. The quantity of added water must be increased to 35% by vol. before the blue colonr reappears and a t this dilution ethyl alcohol also begins t o dissolve copper sulphate.The same effects are noticed when ferrous sulphate is used in place of copper sulphate. The borax flame test for methyl alcohol alio fails when water is present. w. P. s. Influence of the Presence of Trimethylene Glycol on the Estimation of Glycerol by the Zeisel isoPropyl Iodide Method. C. A. ROJAHN ( B e y . 1919,52 [B] 1454-1460).-AXlthough trimethylene glycol produces no silver iodide precipitate whenANALYTICAL CHEMISTRY. ii. 63 digested by itself in the Zeisel apparatus at 130-126° it does so behave when added to glycerol. In fact the excess of silver iodide corresponds with the amount of s-di-iodopropane produced if the quantity of trimethylene glycol is small. As a consequence Zeisel's method is not applicable to the estimation of fermentation glycerol (the German '' protol "-glycerol) for this contains tri- rnethylene glycol.J. C. W. Chromatic Reactions of Formaldehyde with certain Aromatic Compounds. ARTURO ROSSI (Boll. Chiin. Farm. 19 19 50 265-270).-Colorations which are not necessarily distinctive of the aldehyde are given by formaldehyde in 0.1 or 0.0Ol:A aqueous solution when added to sulphuric acid containing a trace of certain aromatic compounds in solution. Applied as a zone reac- tion the following colorations are obtained Gallic acid a reddish- brown ring and on shaking the tube a green liquid changing to greyish-green. Tanmc acid a reddish-brown ring and a liquid bluish- green a t the bottom but subsequently brownish-red throughout. Pyrogallol a light rose-coloured ring and liquid. The test is cap- able of detecting 0.00001% of formaldehyde.Salicylic acid a light rose-coloured ring and solution. CateciLol and resorcinol various red rings and liquids. P-NaphthoZ a violet-red ring and brownish- black liquid with green fluorescence. Belzzonaphf hoZ similar colour reactions. Phenyl salicylate carminered ring and reddish-yellow liquid becoming pink on dilution. Phenolphthalein no ring but an orange liquid changing to carmine-red on shaking. Dilution with water causes a carmine-red precipitate to separate. C . A. M. Estimation of Formaldehyde in Blood. E. SALKOWSKI (Biochem. Zeitsch. 1919 97 129-135).-The protein of the blood is hydrolysed with gastric juice. The hydrolysed mixture is dis- tilled and N/lO-iodine added to the distillate.The residual iodine is then titrated with thiosulphate. The formaldehyde found amounted to only 85.73% of the quantity added to the blood. 5. s. z. Estimation of Ace tone and Acetoacetic Acid with Auten- rieth's Colorimeter. H. SCHALL (Miinch. m d . Woch. 1919 66 812-814; from CJzem. Zentr. 1919 iv 562).-Description of the application of Autenrieth's colorimeter to the colorimetric estima- tion of acetone according to Legal and to the estimation of aceto- acetic acid. Gerhardt's iron chloride reaction is suitable for the estimation of considerable concentrations of the latter (above 0.5%)) whilst at greater dilutions Rimini's ethylenediamine test is used. The former has the advantage of easy manipulation and con- stancy of colour whilst the latter is more sensitive and gives a pure tint which however rapidly loses in intensity.Empirical solutions of dyes are used as standards. When acebone and acetoacetic acid are present simultaneously the estimation of acetone requiresii. 64 ABSTRACTS OF CHEMICAL PAPERS. the correctioii of a €actor since acetoacetic acid gives a positive I,egal’s reaction which is 5.55 as inteilse as that of acetone. H. W. An Acetoneurometer. A. ADLER (JIiiizch. med. Woch. 1919 66 722-723; from Chem. Z ( ~ r i t r . 1919 iv 398).-A colorimetric method of estimating acetoiie is described which is based on Legal’s test. -4 mixture of urine (10 c.c.) sodium hydroxide solution (20%) 1 c.c.) and sodium nitroprusside solution (33% 0.5 c.c.) is well shaken and acetic acid (20 drops) is added. The colour is compared with t h a t given by mixing Neutral-red New-blue and Diamond- phosphine.Solutions of the dyes of the following concentrations are piepared Neutral-red 1 1000; Xew-blue 1 1000; Diamond-phos- phine D 1 :loo. Ten C.C. of the first 3 C.C. of the second and 0’6 C.C. of the third solution are mixed with water 5 C.C. (solu- tion I) and correspond with 0.1% of acetone. Ten C.C. of solution I diluted with an equal volume of water correspond with 0.08% of acetone; 5 C.C. of solution I with 10 C.C. of water are equivalent in colour to 0.05% of acetone; 5 C.C. of solution I with 15 C.C. of water correspond with 0.01% of acetone. Five C.C. of a solution of neutral- red (10 c.c.) new-blue (1 c.c.) and diamond-phosphine (0.1 c.c.) with water (50 c.c.) correspond with 0’005% of acetone.The solu- tions are placed in tubes similar to those used in the experiment which are subsequently sealed. The five standards are in practice sufficient for ascertaining the acetone content of a urine. H. W. Methylene-blue as a Reagent for the Detection of Sugar in Urine. F. UTZ (Siiddeut. Apoth. Zeit. 1919 59 280-281 ; from CI’Lem. Zentr. 1919 iv 212-213).-According t o the author’s observations urines which contain sugar always give a positive reac- tion with solutions of methylene-blue even when Lieber’s modifica- tion of the test is used. Aqueous solutions of dextrose behave simi- larly but sucrose carbamide uric acid and saliva do not give the reaction. Milk behaves similarly to urine containing sugar. Whilst aqueous solutions of hexamethylenetetramine do not react with methylene-blue the urines of patients to whom the drug has been administered behaire similarly to saccharine urines but do not give a positive result with Fehlirig’s solution or Nylander’s reagent.Normal urines free from sugar also give a positive reaction with methylene-blue when they have been preserved for some days and are partly decomposed and this is particularly noticeable if they have been rendered alkaline with potassium hydroxide. Artificial gastric juice and glycuronic acid are also able to decolorise methylene-blue. The substance therefore is not a suitable reagent for the detectior of sugar in urine. Rapid Method for the Estimation of Sugar in Urine. OTTO MAYER (Arch. Hyg. 1919 88 184-197; from Chem. Zemtr.1919 iv 211-212).-The detection of sugar in urine is best carried out by Trommer’s test for the performance of which as also for Fehling’s and Nylander’s reactions practical directions are given. H. W.ANALYTICAL CHEMISTRY. ii. 6 3 The estimation of sugar is effected in the following manner urine ( 1 0 c.c.) is placed in a 100 C.C. stoppered measuring cylinder sodium hydroxide (1576 10 c.c.) is added and the solution is diluted t o 50 C.C. with water; copper sulphate solution (25 grains per litre) is gradually added with gentle shaking until the precipitate which denotes the end of the reaction practically completely redissolves ox vigorous agitation leaving only a just perceptible but permanent turbidity which slightly increases in course of time. Under these conditions each C.C.of copper sulphate solution corresponds with 0.1% of dextrose. If the urine contains more than 4O/ of sugar it is preferable to take only 5 c.c. whilst if the content is 0.5-17L 20 C.C. may be measured out. Separation of alkaline earth phos- phates from urines which have been treated with alkali and diluted from 2.5 to 10 times occurs only when these are present in excep- tional quantity. In this case since the sensitiveness of the elid- point may be greatly reduced it is advisable to treat 20 C.C. of the urine with 20 C.C. of sodium hydroxide solution to filter the solution after dilution to 100 c.c. and to perform the titration with 50 C.C. of the clear filtrate. For the detection of acetone powdered sodium nitroprusside (0.05-0.1 gram) is shaken with about 10 C.C.of urine about 1 C.C. of sodium hydroxide (15%) is added to the solution which is then acidified with acetic acid (3ci% 2 c.c.). I n this form the test is so sensitive t h a t distillation is only necessary with such urines as are deeply coloured or contain only small amounts of acetone. The most suitable reagent for the detection of acetoacetic acid is a mixture of liquor ferri. sesquichl. (5 c.c.) and sodium chloride solution (1 3 95 c.c.) the reaction being performed as a ring test. Volumetric Estimation of Sugar in Milk. J. L. JIAYER ( J . Amer. I’harm. Assoc. 1919 8 551-553).-BenedictJs method for the estimation of dextrose i n urine as modified by the author was applied to the estimation of lactose i n milk. The results were J.D] VAN LEEUWEN (Chem. ll’eekhlad 1919 16 1425-1426) .-In the titration of invert sugar solution with Fehling’s solution dilution of the latter with four times its volume of magnesium sulphate solution (0’5 gram per litre) causes the precipitated cuprous oxide to settle rapidly and thus gives a clear solution f o r testing with potassium ferrocyanide and acetic acid ; the solution must be cooled for this test. W. J. W. Estimation of Carbamide by means of Sodium Hypo- bromite. E. DEKEUWER and L. LESCOEUR (Compt. rend. SOC. Baol. 1919 82 445-447; from Chem. Zentr. 1919 iv 350).-In the estimation of pure carbamide by bromine and alkali hydroxide a slight deficit is always observed which has about the same magni- tude when calculated from the nitrogen evolved or from the bromine used; the error is doubled when the calculation is based on the H.W. trustworthy. CHEMICAL ABSTRACTS. [Estimation of Invert-sugar by Fehling’s Solution.]ii. 66 ABSTRACTS OF CHEMICAL PAPERS. amount of carbon dioxide formed. The course of the reaction appears to depend on the relative quantities of alkali and carb- amide. A factor can be applied based on the results obtained. I n the estimation of carbamide in urine high results are obtained when the calculation is based on the bromine which is utilised (particularly in the case of urines containing sugar) since bromine is also used in the oxidation of substances other than carbamide. The results deduced from the amounts of nitrogen and carbon dioxide formed respectively generally agree well among themselves ; in the case of urines containing sugar the values calculated from carbon dioxide are lower than those reckoned from nitrogen.11. m. Estimation of Carbamide in Blood and Urine. C. BRAHM (Deut. med. Woch. 1919 45 803; from Chent. Zetitr. 1919 iv 442).-A criticism of Citron’s method of estimating carb- amide. The latter is not quantitatively decomposed by sodium hypobromite according to the scheme CO(NH,),+ 3NaBrO = l3NaBr + CO + 2H20 + N since a proportion of the nitrogen is retained as nitric acid or cyanuric acid. I n the author’s opinion all methods which are based on the reaction between carbamide and sodium hypobromite are to be rejected. The form of apparatus proposed is also open t o many objections. H. IN. Estimation of Carbamide in Urine by means of Urease. ARNOLD HAHN (Deut.med. Woch. 1919 45 911-912); from Chem. Zenfr. 1919 iv 562-563).-The author’s method has been modified by Horvith and Kadletz by the introduction of a receiver containing acid ; the alteration is regarded as adding an unnecessary coniplication to the process. H. W. Detection of Hydrocyanic Acid in a Case of Poisoning. Its Post-mortem Transformation into Sulphocyanic Acid. k. CHELLE (Contpt. rend. 1919 169 862-854. Compare A. 1919 11 529 530) .-A dog was poisoned by administering potassium cymide and an autopsy immediately performed the various organs being divided up and kept in stoppered sterile flasks for various lengths of time. Using the methods previously described (Zoc. cit.) hydrocyanic acid as such but no thiocyanic acid was detected after two hours in the brain lungs and blood but not in the liver.After eight days no hydrocyanic acid as such could be detected in any of the organs or the blood but thiocyanic acid was found in the blood lungs and brain but not in the liver. Further the amount of thio- cyanic acid found in the lungs and brain but not in the blood corresponded with far more than the amount of hydrocyanic acid found a t the end of two hours. This is explained on the grounds that a certain amount of hydrocyanic acid becomes enclosed in the cells in the first few hours after death and is not set free by the methods of estimation used but after eight days as a result ofANALYTICAT CHEMISTRY. ii. 67 putrefaction and consequent cytolysis it is all liberated by the reagents and thus estimated. Estimation of Ferricyanide and Ferrocyanide.I. M. KOLTHOFF (Pharm. Weekblad 1919 56 1618-1626).-The author has invest.igated the iodometric estimation of potassium ferricyanide. The acidity of the solution has an appreciable influence on the reaction ; if the concentration is sufficiently high accurate results are obtained after allowing the solutions to remain for only half a minute. Rupp's method for the estimation of ferrocyanide is inac- curate. An outline is given of the theoretical interpretation of the iodometric estimation of ferri- and ferro-cyanide. The reducing action of ferrocyanide is increased by rise of temperature. W. J. W. W. G. Proteinogenous Amines. 11. A Micro-chemical Colori- metric Method for Estimating Iminazole Derivatives.KARL K. KOESSLER and MILTON T. &NKE ( J . Biol. Chem. 1919 39 497-519).-The method depends on the colour readion between compounds containing the iminazole ring and sodium p-diazobenzenesulghonate. It can be applied directly to practic- ally any iminazole derivative gives equally good results whether applied to pure solutions or mixtures and is capable of estimating with a fair degree of accuracy as little as 0.00001 gram of any of the iminazoles. The iminazole derivative is added under certain conditions to a mixture of sodium carbonate and the pdiazo- benzenesulphonic acid in one cup of a Duboscq colorimeter and the height of liquid in this cup is set a t 20 mm. The other cup is filled with an appropriate indicator solution and the height varied until the colours match.From tables which have been pre- pared the amount of the iminazole derivative can be read off. Certain substances interfere with the method notably ketones alcohols ammonium salts and proteins. Animal charcoal must not be used to decolorise solutions which it is desired to test for iminazole derivatives as loss by adsorption occurs. J. C. D. Proteinogenous Amines . 111. A Quantitative Method for Separating$ Histamine from Histidine. KARL K. KOESSLER and MILTON T. HANKE ( J . Biol. Chem. 1919 39 521-538).- The colorimetric process described in the preceding abstract may be utilised for estimating certain components of mixtures of iminazole derivatives. A mixture oQ these derivatives such for example as that which is formed by the action of bacteria on histidine in the presence of salts and dextrose is treated with sufficient sodium hydroxide to make the whole a 20% solution of alkali.This solution is extracted repeatedly with amyl alcohol whereby the histamine ammonia methyliminazole and other amines are removed. This extract is termed the histamine frac- tion whilst the residual liquid containing histidine iminazole pro- pionic acetic and lactic acids is called the histidine fraction. The amount of histidine can be determined by an estimation of theii. 68 ABSTRACTS OF UHEMICAL PAPERS. amino-nitrogen in the second fraction. A colorimetric estimation on the same fraction will indicate whether there are other iminazole derivatives present other than histjdine. The histamine and methyliminazole are extracted from the amyl alcohol by 1 *ON-sulphuric acid and a colorimetric estimation will give on this extract an indication of the total iminazole sub- stances present.The methyliminazole may be estimated directly by the fact that it is volatile in steam. The colorimetric estimations on these two fractions will give by difference the amount of histamine present. J. C. D. Direct Estimation of Non-amino-nitrogen in the Products of Protein Hydrolysis. ALMA HILLER and DONALD D. VAN SLYRE ( J . Biol. Chem. 1919 39 479-488).-An extension of the method of protein analysis introduced by Van Slyke (A. 1915 ii 851). The filtrate from the phosphotungstates of the diamino-acids repre- senting 2 or 3 grams of protein is concentrated to 150 C.C. Por- tions of this solution are used for estimations of the total nitrogen and amino-nitrogen as described in the original paper. Of the residue 100 C.C. are measured into a separating funnel of 500 C.C. capacity and are rendered free from phosphotungstic acid by extraction with amyl alcohol-ether mixture in the presence of hydrochloric acid. The solution is concentrated in a vacuum to about 50 c.c. and washed into a 100 C.C. volumetric flask. Of this solution 20 C.C. portions are used for the non-amino-nitrogen nstimation. Each portion is placed in a small Kjeldahl flask with 1-2 C.C. of 30% sodium nitrite and 5 C.C. of concentrated hydro- chloric acid. When deamination is completed the solution is neutralised to alizarin with 10% sodium hydroxide and reduced with zinc-copper couple. After making alkaline with magnesium oxide the ammonia formed by the reduction is removed by boiling and residual nitrogen is estimated by the Kjeldahl process. The results by this direct method agreed closely in analyses of caseinogen and gelatin with the results obtained indirectly by the original procedure. J. C. D. Apparatus for Rapid Gastric Analysis together with a Method for the Preservation of Starch Solution. RAYMOND J. MILLER ( J . Znd. Eng. Chem. 1919 11 963-964).-A con- venient arrangement of burettes indicator reservoirs and delivery tubes an automatic pipette etc. attached to one stand and intended for the estimation of total acidity free acidity and amino-acid nitrogen (formaldehyde method) in gastric fluids. The starch solution is preserved under a laper of liquid “petrolatum,” with or without the addition of a small quantity of toluene. w. P. s.