ANALYTICAL CHEMISTRY A n a l y t i c a l Chemistry. 61 Improvements in Hemple’s Gas-Analysis Apparatus. GEORGE DE VOLDERE (Chem. Zentr. 1907 ii 11 13-1 114 ; from Zeitsch. chrn. Apparaternkunde 1907,2 344-346).-The levelling tube is replaced by a levelling reservoir fitted with a side tube having the same diameter as the burette. The burette is constructed as recommended by Pfeiff er62 ABSTRACTS OF CHEMICAL PAPERS. (Abstr. 1907 ii 194) but the capillary tube between the first enlargement and the stopcock has been omitted and a four-way cock has been put instead whilst a second tube serves for the purpose of applgifg suction or pressure. The chief advantage of this arrange- ment is that the burette can be fixed firmly by means of a clamp. The U-tube in the pipette has been also replaced by a four-way stop- cock.For further details the illustrations in the original article should be consulted. A convenient apparatus is also described for the combustion of gases by means of palladium asbestos. L. DE K. Use of Chloral Hydrate Chloral Alcoholate and Bromal Hydrate Solutions in Chemical Microscopic and Micro- chemical Investigations. EDUAXD SCHAER (Be?*. deuut. piharm. Oes. 1907 17 407-41 3).-Aqueous and alcoholic solutions of chloral hydrate and bromal hydrate dissolve a large number of inorganic and organic substances which have no other common solvent. Where complete solution of a mixture is necessary or advantageous as a preliminary step to the detection or separation of its components the employment of 60% to 80% aqueous or alcoholic chloral hydrate solutions as the solvent is recommended.The value of chloral hydrate as R clearing agent for microscopic [preparations depends on its penetrating power as also on its action as a solvent for the cell contents which render the preparation opaque. I n many cases the best results are obtained by successive treatment with concentrated aqueous and alcoholic solutions. I n microchemical analysis it has been found possible with the aid of chloral hydrate solutions of the alkaloid reagents to determine the location of alkaloids in plant structures. G. Y. Isolation of Traces of Mineral Substances from Saline Mixtures. Application to Medical Chemistry Toxicology and Hydrology Q. M E I L I ~ E (J. I’l~ccrm. Chim. 1907 {vi] 26 443-460).-A summary of the methods used in analysis for concentrating small quantities of substances which might otherwise escape detection.For instance traces of lead in water are carried down completely by adding a little copper sulphate before passing hydrogen sulphide and phosphoric and arsenic acids are precipitated completely in presence of ferric hydroxide. I n some cases extraction by means of an immiscible solvent in a suitable apperntus gives good results. Strong hydrochloric acid extracts lithia from alkali chlorides &c. L. DE Addition of Indigo in Titrations with Methyl- or Ethyl- orange. ROBERT LUTHER (Chem. Zeit. 1907 31 1172. Compare Kirschnick Abstr. 1907 ii 910).-Solutions of indigo-carmin and methyl-orange are mixed in such proportions that when added to a saturated solution of carbon dioxide in water a neutral grey tint is obtained (compare Kiister Abstr. 1897 ii 74).When such a mixture of methyl-orange and indigo-carmin is used the change of colour during the titration of an alkali wit.11 acid is as follows yellow,ANALYTICAL CBEMISTRY. 63 green colourless (or grey if a large quantity of indicator is taken) violet. The change from green or violet to the neutral grey or colour- less stage is said to be very decided and sharp. The addition of indigo t o solutions which are tinted yellow is also advantageous. Behaviour of Chlorates and Perchlorates during Reduction. DOMENICO VENDITORI (Gaxxetta 1907 37 ii 383-386).-1n presence of sulphuric acid finely-divided aluminium reduces chloric acid completely t o hydrochloric acid but has no reducing action on perchloric acid.Chloric acid may hence be estimated by converting it into hydrochloric acid and then estimating the latter by Trolhard’s method. T. H. P. W H. G. Detection of Fluorine Compounds in Wines. L. VANDAM (Ann. China. anal. 1907 12 466-468; Rev. irztem. Fahi$ 1907 20 147-14S).-Two hundred C.C. of the sample are mixed with a few drops of a 20% solution of sodium sulphate and then with 10 C.C. of a 10% solution of barium acetate. After twelve hours the bulk of the liquid is poured off and the precipitate collected on a filter washed dried and ignited in a platinum crucible. After moistening with a drop of water a little sulphnric acid is added and the crucible is at once fitted with an india-rubber ring supporting a paraffin-coated match-glass the convex part of which carries a n inscription made with a soft point.The watch-glass is kept cold by a current of cold water. After heating the crucible for half an hour on a boiling water-bath the watch-glass is gently heated to melt the paraffin coat- ing and wiped with a soft linen cloth. If the sample is pure no inscription will be visible or at most but very faintly on breathing on the glass but a visible etching is noticed even when there is as Estimation of Sulphur Dioxide in Wines. CARLO MENSIO (Gaxxefh 1907 37 ii 344-355).-The maximum amount of total sulphur dioxide per litre of mine permitted in ItaIy is 0.2 gram of which not more than 0.02 gram should be in the free state. According t o the official method of estimat’ion sulphur dioxide which is evolved from mine on boiling is regarded as existing in the free state i n the wine.The author shows that t h i s view is inaccurate since compounds such as that formed by sodium hydrogen sulphite with acetaldehyde give up from 30% to 75% of their suiphur dioxide when boiled in aqueous solution and from 90% to 95% when boiled with wine. The most accurate method of estimating the total sulphur dioxide in wine consists in acidifying with phosphoric acid and distilling in a current of carbon dioxide the distillate being collected in iodine solution in which the sulphur dioxide may be estimated either by titration of the excess of iodine with sodium thiosulphato or by Haas’s method (Abstr. 1882 773). None of the methods given for estimsting free sulphurous acid in a mine yields accurate results.As the experiments of Franz and Eost (Abstr. 1904 ii 632) indicate that the toxic action of combined sulphur dioxide is not greatly dissimilar from that of the uncombined and as the presence little as 10 mgs. of ammonium fluoride per litre. L. DE K.64 ABSTRACTS OF CHEMICAL PAPERS. of the latter in a wine is detectable by the taste the fixing of a limit to the proportion of free sulphur dioxide should be abolished. T. H. Y. Digestion of Urine in the Estimation of Nitrogen by the Kjeldahl Method. PHILIP B. HAWK (J. Arner. Chenz. Xoc. 1907 29 1634-1 637).-The nitrogen in urine may be estimated equally well by digestion with sulphuric acid and copper sulphate sulphuric acid and metallic mercury or sulphuric acid metallic mercury and potassium sulphate added towards the end.In any case the boiling Alkalimetric Estimation of Phosphoric Acid by Neumann’s Method. J. P. GREGERSEN (Zeitsch. physiol. Clbern. 1907 53 453-463. Compare Abstr. 1903 ii 243; 1905 ii 68).-As the result of numerous series of experiments the conclusion is drawn that Neumann’s method gives excellent results when the following precautions are taken. I n the destruction of the organic material and the formation of ash 20 C.C. of Neumann’s acid mixture should be used and during the process concentrated nitric acid and not the mixed acids should be dropped in. I n the precipitation 250 C.C. of liquid should be used and this should contain 15% of ammonium nitrate only a slight excess of ammonium molybdate being employed; thus for 10-15 mgs.phosphorus 4 grams of molybdate suffice. I n the final titration it is advisable to add a slight excess of standard acid to boil in order to remove carbon dioxide and then to titrate back with 0*5N sodium hydroxide. Estimation of Phosphoric Acid in Basic Slags by Grete’s Method. C. H. KETNER (Chem. TKeekbZud. 1907 4 757-764).- Grete’s method (titration with gelatin-molybdenum solution until no further precipitate is formed) has been applied.successfully in the Dutch agricultural stations but for soluble phosphoric acid only. The author has tried to use it for total phosphoric acid also and after some trials has succeeded by working as follows. Ten grams of the basic slag are powdered moistened with a little water and then heated in a 500 C.C.Jena flask_with 50 C.C. of sulphuric acid for about half an hour and until white fumes are formed. When cold water is added and when again cooled water is added up to the mark and the whole is thoroughly shaken and filtered. Twenty-five C.C. of the filtrate are neutralised with ammonia 20 C.C. of ammonium nitrate solution (1 a) a teaspoonful of powdered potassium nitrate and 15 C.C. of nitric acid D 1.2 are added and the solution is then heated to boiling and titrated. In standardising the gelatin-molybdenum solution as usual with dihydrogen potassium phosphate in presence of the above substances there should also be added 16 C.C. of 10% ammonia previously neutralised with dilute sulphuric acid. should be continued for a t least thirty minutes.L. DE K. J. J. 8. L. DE K. Estimation of Arsenic by the Gutzeit Method. CHARLES R. SANGER and OTIS F. BLACK (J. Xoc. Chem. Incl. 1907 26 1115-1123).-The modification suggested by the authors in order t oANALYTICAL CHEMISTRY. 65 make the Gutzeit method quantitative consists in allowing the arsine to pass over the surface of a strip of paper impregnated with mercuric chloride. The band of colour thus obtained is compared with a series of bands prepared from known amounts of arsenic. The sensitised paper is inserted in the form of a strip in a horizontal delivery tube leading from the evolution flask The exact method of carrying out the estimation is given a t length in the original paper together with remarks on the precautions to be observed and on tho influence of certain interfering substances.w. P. s. Estimation of Arsenic in Urine. CHARLES R. SANGER and OTIS F. BLACK (J. SOC. Chenz. Ind. 1907 26 1123-1127 ; Zeitsch. anorg. Chew. 1907 56 153-157).-The evaporated urine is dis- tilled with hydrochloric acid the distillate is oxidised with nitric acid the latter is removed by heating with sulphuric acid and the remain- ing solution is used for the estimation of the arsenic. About 200 C.C. of the urine are evaporated to a volume of 35 C.C. and then introducedintoa distillation flask. One hundred C.C. of pure hydrochloric acid are added and the mixture is distilled the distillate being collected in a receiver containiug 25 C.C. of concentrated nitric acid. When about 100 C.C. have distilled over the distillate is treated with a further 25 C.C.of nitric acid and evaporated to a small bulk ; 5 C.C. of sulphuric acid are next added and the heating continued until the nitric acid has been expelled. The residue is diluted with water to a definite volume and EL part of it employed for the estimation of the arsenic the Marsh- Berzelius method being used for the latter purpose w. P. s. Estimation of Hydrofluosilicic Acid. SANUEL H i i ~ ~ a (Chenz. Zeit. 1907 31 1207-120S).-The process is based on the fact that 1 mol. of hydrofluosilicic acid is neutralised by 2 mols. of sodium hydroxide when using methyl-orange as indicator whereas 6 mols. aro required if phenolphthalein is the indicator. Twenty-five C.C. of the sample are titrated with N-sodium hydroxide in presence of methyl-orange The result equals hydrofluosilicic acid and any other mineral acid present.The liquid is now heated to boiling and titrated while hot with phenolphthalein as indicator. Four mols. of sodium hydroxide now represent 1 mol. of hydrofluosilicic acid. The method is not strictly accurate owing to the slight dissocia- tion of the sodium silicofluoride formed but this does not interfere with its technical application. L. DE I<. The Carrasco-Planoher Method of Estimating Carbon and Hydrogen in Organic Suhtances. WILHELM LENZ (Zeitsch. ccnal. Chem. 1907 46 557-565).-The results of a number of estimations of carbon and hydrogen in salicylic acid by this method (Abstr. 1906 ii 201) are given. Whilst in some cases the quantities of carbon and hydrogen found agreed with those required by theory in others the carbon was as much as 3.5% too low.The loss was due to the forma- tion of carbon monoxide ; the quantity of the latter produced appeared t o depend on the speed at which the combustion was carried out the VOL. xciv. ii. 566 ABSTRACTS OF CHEMICAL FAPERS. longer the time taken for the estimation the greater the amount of carbon monoxide formed. Estimation of Carbon Monoxide in Atmospheric Air. J. LIVINGSTON K. TVIORGAN and JOHN E. MCWIIORTEH (J. Amer. Chem. Soc. 1907 29 15S9-1592).-The air is passed first through tubes containing sulphuric acid and potassium hydroxide. If it is desired to estimate also the carbon dioxide another tube containing standard barium hydroxide is placed in front of the sulphuric acid tube.The carbon monoxide is then absorbed in a U-tube cor~taining iodine pent- oxide and heated a t 150' in a glycerol-bath The products of the reaction free iodine and carbon dioxide are passed through a tube containing potassium iodide solution and the iodine is af terwards titrated with iV/lOOO thiosulphnte by way of a check. The carbon di- oxide is absorbed in standard barium hydroxide the excess of which is then titrated as usual with standard oxalic acid In estimating carbon monoxide by passing it over iodine pentoxide the ends of the lJ-tube should be sealed off after filling so as t o avoid traces of greasy matter from the lubricant used with glass stoppers. w. P. s. L. DE K Volumetric Estimation of Potassium as the Cobaltinitrite. %V. A. DRUSEIEL (Anzer.J. Sci. 1907 [iv] 433-438 ; Ze&tsc?$. anorg. Chem. 1907,56 223-329).-A moditication of the method proposed by Adie and Wood (Trans 1900 37 1076) The solution is mixed with excess of sodium aobaltinitrite and a little acetic acid and evaporated to a pasty condition on a water-bath. The residue is treated with cold water and the precipitate collected and washed on a n asbestos filter. The filter and contents are then heated with excess of N/10 permanganate for some five minutes when excess of dilute sulphuric acid ( 1 7) is added and the whole heated short of boiling until the manganese precipitate has completely dissolved. The excess of permanganate is estimated as usual by first bleach- ing with N/10 oxalic acid and then titrating with permanganate. Working in this manner 1 c,c.of N/10 permanganate=0*000856 gram of K,O The process is suitrible for the estimation o€ potash in manures not being interfered with by the presence of phosphoric acid. New Volumetric Method of Estimating Sodium Sulphide. E. PODRESCHETNIKOFF (Zeitsclh. 8 w b . Ind. 1907 6 3SS).-The new method permits of the estimation of both sodium hydroxide and sodium sulphide owing to the latter undergoing complete hydrolysis in dilute solution according to the equation Na,S e H,O = NaHS + NaOH. On titrating a very dilute solution of the sulphide with N/10 sulphuric acid in the presence of phenolphthalein as indicator the quantity o€ sodium sulphide plus the free hydroxide is obtained. The amount of sulphide alone is then determined by adding a slight excess of formalde- hyde and titrating by means of the acid the sodium hydroxide liberated according to the equation NaHS + CH,O + H,O = NaOH + OH*CH,-SH.The results obtained by this method are quite accu- rate. W. A. D. L. DE K.ANALYTICAL CHEMISTRY. 67 Biological Method for Estimating Alkali Carbonates in Soils. HARALD R. CHRISTENSEN (Cents.. Bakt. Par. 1907 ii 19 735-736).-1t was shown previously that azotobacter is unable to utilise calcium in the form of sulphate chloride and tribasic phosphate. It is now shown that scme soils deficient in cdcium in conjunction with mannitol and potassium phosphate will develop a growth of azotobacter when supplied with calcium sulphate similar to that obtained when calcium carbonate is added. Such soils show almost invariably an alkaline reaction and the conclusion is drawn that an alkali carbonate is present which interacts with tho calcium salt to form calcium carbonate.It is proposed to utilise the different behaviour of calcium sulphate in different soils for ascertaining the amounts of alkali carbonates present. N. H. J. &I. Volumetric Estimation of Magnesium. LEOPOLD ROSENTHALER (Zeitsch. and. C h m 1907 46 714-716).-The magnesium salt is dissolved and treated with a measured volume (an excess) of potassium hydrogen arsenate KH AsO solution (containing about 9 grams per litre). The mixture is then diluted to a known volume with 10% ammonia and mixed. After three hours the solution is poured through a filter and a portion of the filtrate say 100 c.c is evaporated to dryness on the water-bath.The residue is dissolved in a very little water the solution is transferred t o a stoppered flask the basin is rinsed out with a little concentrated hydrochloric acid which is also added to the flask and the acid solution is then treated with con- centrated potassium iodide solution. If a precipitate forms a little more water is added ; if not hydrochloric acid is added before adding the water. The liberated iodine is then titrated back with thiosulphate solution. The difference between the amount of potassium hydrogen arsenate added and that found in excess represents the quantity precipitated with the magnesium ; 1 C.C. of N/10 thiosulphate solution corresponds with 0*002018 gram of magnesia (MgO). 'cv. P. s. Detection of Small Quantities of Zinc by Electrochemical Means.WALTEH. NEUMANN (Zeitsch. Elektrochem. 1907 13 751-752). -The solution to be tested containing sufficient potassium hydroxide t o render it roughly N / l O is electrolysed using a small platinum bead as anode and a brightly polished thin copper wire as cathode. Should zinc be present the copper becomes coated with a bright white deposit of zinc which may be removed by dipping the wire in 2iV potassium hydroxide solution. It is possible by this method to detect zinc in 0.1 C.C. of a N/40,000 solution of the metal. The solution to be tested must not contain chloride or nitrate. W. H. G. A Method of Estimating very Small Quantities of Zinc. GABRIEL BERTRAND and MAUKICE J AYILLIER (Conzpt. rend. 1907 146 924-926).-The method described previously (Abstr.1907,ii 53) of precipitating small quantities of zinc in the form of hydrated calcium zinc oxide can be employed for the estimation of the metal in solutions containing 0.1 mg of zinc in 3 litres in the presence of tin antimony 5-268 ABSTRACTS OF CHEMICAL PAPERS. arsenic copper cadmium cobalt lead silver iron manganese alumin- ium barium calcium magnesium sodium and potassium as carbonates chlorides sulphates nitrates phosphates silicates and sulphides. For details of the separation the original must be consulted. The final separation of the zinc from the calcium in the hydrated calcium zinc oxide is best effected by dissolving in hydrochloric acid evaporating t o dryness precipitating the zinc as sulphide in the pre- sence of sodium acetate and weighing as the sulphate.Rapid Estimation of Zinc by Electrolysis. FRANCIS C. FRARY (J. Amer. Chem. Xoc. 1907 29 1596-1603).-A solution contain- ing about 8 grams of sodium hydroxide and 0.1 gram of zinc deposits all the metal on a nickel gauze in thirty minutes with electromagnetic rotation (Abstr. 1907 ii 649) and a current of 4.5 amperes. The temperature should be as low as possible. For quantities of zinc up to 0.2 gram the strength of the current should be 4.5 amperes for fifteen minutes and then 1.5 amperes for twenty minutes. It is advis- able always t o test the electrolysed liquid to see whether the precipita- A. R. THORNEWELL (Chem. Zentr. 1907 ii 1269 ; from Chemist and Druggist 1907,71 413).-The solution should be perfectly neutral to methyl-orange and contain about 0.3 mg.of zinc per C.C. Two hundred C.C. are mixed with 25 C.C. (or more) of N-sodium hydroxide previously saturated with hydrogen sulphide and the whole is diluted to 250 C.C. and well shaken. Twenty-five C.C. of the clear supernatant liquid are then mixed with 25 C.C. of N/10 sulphuric acid the hydrogen sulphide is expelled by boiling and when cold the excess of acid is titrated mibh NIlO sodium hydroxide using methyl-orange as indicator. The zinc is then calculated from the amount of alkali Volumetric Estimation of Lead. €1. BOLLENBACH (Zeitsch. ccnab. Chem. 1907 46 582-588).-1n the process described ths lead solution is titrated with permanganate in alkaline solution. Sodium hydroxide is added to the lead solution until the precipitate which at first forms is redissolved and the whole is then diluted to a known volume.Fifty C.C. of this solution are run from a burette into a measured excess of potassium permanganate solution placed previously in a 500 C.C. flask together with 300 C.C. of hot water and 5 C.C. of N/2 sodium hydroxide solution. The addition of the lead solution is accom- panied with constant agitation of the permanganate mixture. After the lead solution has been introduced the solution is cooled diluted to 500 c.c. and the excess of the permanganate estimated either iodo- metrically or by the use of a lead solution of known strength. The per- manganate solution should contain 3 9 grams of the salt per litre and be titrated against the standard lead solution. The reaction takes place according t o the formula 2KMn0 + 6Pb(ONa) = 2Mn0 c 3Pb,0 + K,O + 6Na20.The presence of chlorides bromides and sul- phates has no influence on the results obtained by the method but iodides and substances which reduce permanganate should not be M. A. W. tion has been complete. L. DE K. Estimation of Zinc. absorbed. L. DE K. present. w. P. s.A N A LY T I C A L C H E R l I STRY . 69 Red Leads and their Examination. ALFRED PARTHEIL (Arch. P?tumn. 1907 245 519-528).-According to the Deutsche Arznei- buch IV the purity of red lead may be tested by adding 2.5 grams of the material mixed with 0.5 gram of oxalic acid to 10 C.C. of hot nitric acid and then adding gradually 25 C.C. of hot water ; not more than 0.035 gram should remain undissolved. It is shown that this test is quite useless owing to the formation of insoluble basic lead oxalates lead oxalonitrate C ~ C .The following method is said to give better results 10 C.C. of water 5 C.C. of lactic acid and 10 C.C. of 25% nitric acid are added to 2-5 grams of the red lead ; R violent reaction takes place after which the insoluble residue is weighed. JAROSLAV MILBAUER and VLADIM~R STAN~K (Zeitsc?~ anal. Clmt. 1907 46 644-656).-The well-known blue coloration produced when copper sulphate solution is treated with ammonia has not always the same intensity for equal quantities of the metal. Increasing quantities of ammonia or ammonium chloride diminish the depth of colour whilst ammonium carbonate has the opposite effect the maximum coloration obtained with ammonia and copper sulphate being increased by one-third on the addition of ammonium carbonate.Potassium hydroxide in the presence of amines gives with copper sulphate a coloration varying from 56% to 66% of that obtained with ammonia. Many other instances are given but in each the coloration is less than that yielded by ammonia. An ammoniacal copper solution has a coloration one hun- dred and fifty times more intense than a copper sulphate solution Estimation of Mercury by Reduction with Hydrogen .Per- oxide. A. KOLB and A. FELDHOFEN (Zeitsch. angem. Chenz. 1907 20 1977-1 gSO).-Twenty-five C.C. of mercuric chloride solution con- taining about 0.1 gram of mercury are mixed with 25 C.C. of 2iY hydro- chloric acid and 25 C.C. of 10% tartaric acid solution. The liquid is neutralised with ammonia and again faintly acidified with tartaric acid.The whole is heated on the water-bath and 10 C.C. of hydrogen peroxide are added and then from time t o time another5 c.c. so that 25-30 C.C. are used in the space of forty-five minutes. The precipitated calomel may be collected and weighed or better titrated iodometrically . The process is also useful for the separation of mercury from arsenic Detection of very Small Quantities of Mercuric Chloride. K. KOF and HUGO HAEIIN (Arch. Pharm. 1907 245 529-533).- The vapour arising from a mercuric chloride solution can not only be caught on a photographic plate (compare Abstr. 1907 ii 7321 but also on moist filter paper. A moistened filter paper resting on a thin glass slide and placed for one hundred and sixty-five hours at a distance of about 1 cm.above the surface of a 2% aqueous solution of mercuric chloride turns dark when treated with hydrogen sulphide except where it has been protected by the strip of glass. The photographic effect may be employed W. H. G. Colorimetric Estimation of Copper. containing the same quantity of copper. w. P. s. antimony tin cadmium and bismuth. L. DE K.70 ABSTRACTS OF CHEMICAL PAPERS. for the detection of small quantities of mercuric chloride. A dist,inct white patch is obtained on developing a photographic plate which has been kept at a distance of 2 to 3 mm. from one drop of a 0.01% solution of mercuric chloride for twenty-four hours. The same effect on the plate is produced if instead of the drop of mercuric chloride solution one drop of the distillate obtained by steam-distilling 100 C.C.of a 0.01% mercuric chloride solution until 100 C.C. of distillate are OF- tained is employed. Acidimetric Assay of White Precipitate. ERWIN RUPP and F. LEHMANN (Phccmt. Zeit. 1907 52 1014).-0+2-03 Gram of the powdered sample is dissolved in 50 C.C. of water and 2-3 grams of potassium iodide and the ammonia and potassium hydroxide liberated in the reaction are titrated with N/10 hydrochloric acid with 1-2 drops of 0.2% alcoholic methyl-orange solution as indicator. 0.2 Gram of the commercially pure article requires 1506-16 C.C. of N/IO acid. W. H. G. L. DE K. Complete Analysis of Mercury Cyanides. ERWIN RUPP and F. LEHMANN (Pharrn. Zeit. 1907 52 1020).-One gram of the com- pound is dissolved and the solution made up t o 100 C.C.In order to estimate the mercury 20 C.C. are placed in a flask and after adding 1-2 grams of potassium iodide and 3-6 C.C. of officinal potassium hydroxide the mercury is reduced to the metdlic stato by adding 2-3 C.C. of officinal formaldehyde previously diluted with 20 C.C. of water. After shaking thoroughly for two minutes a slight excess of acetic acid is added and then 25 C.C. of N/lO iodine. When after repeated shaking the mercury has redissolved 10 C.C. of dilute sulphuric acid are added and the excess of iodine is titrated with N/10 thiosulphate. I n order to estimate the cyanogen 10 C.C. of the solution are mixed with a little water and 5-10 C.C. of officinal sodium hydroxide and 25 C.C. of N/10 iodine are added slowly with con- stant agitation.The whole is allowed to remain for two t o three hours or else heated for twenty to thirty minutes in the water-bath. The liquid is diluted to 100 c.c. acidified with about 25 C.C. of hydrochloric acid and titrated with thiosulphate ; 0.1 gram of mercury Borax Bead Test for Praseo- and Neo-dymium Salts. JAROSLAV MILBAUER (Zeitsch. anal. Chem. 1907 46 657-658).-The following colorations are obtained when these s a h are heated in the usual way on a borax bead. Neodymium in oxidising flame no colour; in reducing flame violet if sufficient of the salt is used. Praseodymium oxidising flame yellowish green ; reducing flame green. The colorations are seen only when the bead is cold. Beads consisting of phosphates may be employed in place of borax.cyanide = 15.87 C.C. of N/10 iodine. L. DE K. w. P. s. Detection and Estimation of Small Quantities of Man- ganese. MAURJCE DUYE (Ann. Chim. anal. 1907 12 465-466).- A pink or red coloration is obtained when a trace of manganese salt is heated with moderately alkaline potassium hypochlorite solution (10%)ANALYTICAL CHEMISTRY. 71 to which has been added 1 drop of copper sulphate solution (10%). A precipitate of cupric oxide is produced and the coloration is observed in the Fupernatant liquid. The reaction may be made quantihative by comparing the coloration obtained with that produced by known Modiflcation of Volhard's Method of Estimating Man- ganese. ERWIN W. MAYER (Zeitscl~ angew. Chem. 1907 20 1980-1 981).-Four modifications of the process are described of which the following two are the most suitable for technical purposes. (1) One gram of ferromanganese 4 grams of pig-iron or 8 grams of steel are dissolved in nitric acid (D 1.2)'; thesolution is evaporated to R small bulk diluted with water introduced into a litre-flask and mixed with a sufficiency of zinc oxide emulsion to precipitate the iron.The whole is then diluted to 1 litre and in 250 C.C. of tho filtrdte the manganese is titrated with standard permanganate (2) The sample is dissolved as before in nitric acid but before boiling down cz few Clay Analysis. Residue Left After Volatilisation of the Silica with Hydrofluoric and Sulphuric Acids. W. R. BLOOR (J. Amer. Chem. Soc. 1907 29 1603-1 606).-Experiments showing that the residue obtained on heating the separated silica with hydrofluoric and sulphuric acids does not consist of iron and aluminium oxides only but contains also the other constituents of clay such as calcium and Precipitation of' Iron by Ammonia in Presence of Tartaric Acid.WILHELM STRECRER (Chem. Zeit. 1907 31 1217. Compare Grossmann and Schuck Abstr. 1907 ii 819).-Ferric hydroxide is precipitated completely on addition of ammonia to a soIution con- taining 0.1110 gram of iron in the ferric state and 0.1911 gram of tartaric acid in 53 C.C. If a larger proportion of tartaric acid is present the precipitation is incomplete and is finally inhibited Attempts to isolate a complex compound from the resulting red solu- tion containing excess of tartaric acid have been unsuccessful as the products obtained do not have a constant composition (compare MAhu Colorimetric Comparison of Copper and Nickel Solutions.JAROSLAV MILBAUER (Zeitsch. anal. Chem. 1907 40 656-657).- Nickel salts when treated with ammonia give a similar blue coloration to that obtained with copper salts. The author finds that one part of copper gives a coloration equal t o that yielded by eighteen parts of nickel and this fact renders the test of little use in testing very small quantities of the latter metal. I n fact a N/100 ammoniacal nickel solution is practically colourless whilst a copper solution of equivalent concentration has a marked blue colour. amounts of manganese. w. P. s. drops of hydrochloric acid are added. L. DE K. magnesium oxides also titanium. L. DE K.Jalmesber. Chem. 1873 569). a. Y. w. P. s. Brunck's Dimethylglyoxime Procees for the Estimation of Nickel. HERRMANN GROSSMANN and BERNHARD SCEI~CK (Xeitsch. angew. Chsna 1907 20 1981-1982).-~ reply to Grunck (Pbstr.,72 ABSTRACTS OF CHEMICAL PAPERS. 1907 ii 989) in which the authors uphold the accuracy of their own dicyanodiamidine process. L. DE K. Volumetric Estimation of Bismuth. RICHARD EHRENFELD (Zeitsch. unal. Chern. 1907 46 710-71 l).-The process consists in precipitating the bismuth with excess of sodium phosphate and afterwards titrating back this excess. The bismuth solution is acidified with nitric acid to prevent the formation of basic salt a known excess of sodium phosphate solution is added and the mixture is diluted to a known volume. After filtration or settling the excess of phosphate is titrated in a portion of the clear solution; uranium acetate solution is used for the titration and suflicient sodium acetate is added to the solution to be titrated in order to remove the free nitric acid {compare Abstr. 1906 ii 55 ; 1907 ii 403).w. P. s. Apparatus f o r the Prevention of Acid Fumes in Gold and Silver Assaying. DARD (Ann. Chim. ancd. 1907 12 425-426).- A portable apparatus which may be placed in front of any kind of chimney. The hydrochloric or nitrous vapours are made to traverse a layer of fragments of white marble and the carbon dioxide liberated is carried off by a central chimney-funnel which will act readily owing to the heat given off by the burners. Several assays can be made simultaneously without suffering from irritatiog fumes. L.DE K. Refractometric Analysis of Organic Mixtures. ADOLF BEYTHIEN and R. HENNICKE (Pharm. ZentT.-h. 1907,48,1005-1013). -The composition of a mixture containing any two of the following substances may be ascertained from its refractive index the constants taken being acetone ng 1.3620 nz 2.3590 ; carbon disulphide n:” 1.6315 ; oil of turpentine ng 1-4690 ; carbon tetrachloride nz 1.4630 ; xylene qag 1.49680 ; ethyl bromide n1,5 1.42540 ; chloro- form ng 1.4466. Mixtures containing three of these substances may be analysed in the same way if one of the constituents can be estimated by a separate method. w. P. s. Reaction of Phloroglucinol-Hydrochloric Acid with Essential Oils. KARL KOBERT (Zeitsch. anal. Chem. 1907 46 711-714).- The author finds that only those essential oils which contain an allyl- group give a red coloration when treated with a hydrochloric acid solution of phloroglucinol.A red coloration is obtained with the essential oils of mustard cloves pimento dill orange-blossom jaborandi leaves tarragon basil bay lavender Peru balsam geranium parsley and sassafras. w. P. s. Detection of Methyl Alcohol in Fermented Liquids. JULES WOLFF (APLPZ. Chin2. anccZ. 1907 12 470-471).-Any alcoholic liquid containing sucrose invert-sugar or caramel should be redistilled before testing it for methyl alcohol by means of chromic acid &c. (formation of formaldehyde polymerides). These substances yield formaldehydeANALYTICAL CHEMISTRY. 73 when thus tested and consequently methyl alcohol may be stated to be present in a spirit when such is not the case.w. P. s. Normal Tubes for the Polarimetric Estimation of Sucrose. H. ROUSSET (Ann. Chim. a n d 1907 12 468-470).-The use of tubes 1’79.2 mm. and 358.4 mm. in length is recommended for the estimation of sucrose by means of Laurent and Duboscq’s or Vivien’s polarimeters. The weight of the sample taken is 20 grams in 100 C.C. of water and the percentage of sucrose is obtained without calculation. w. P. 8. Estimation of the Most Important Acids that Occur in Wine in Presence of Alcohol and Glycerol. A. HEIDUSCHKA and GEORGE QUINCKE (Arch. Phccrm. 1907 245 458-461).-The method described was tested with a mixture of acetic lactic tartaric malic and succinic acids with alcohol and glycerol (and water) and found t o give good results.Of such a mixture 50 C.C. are neutralised with aqueous barium hydroxide and evaporated to 15 c.c. care being taken to keep the liquid neutral. The residue is rinsed with a little hot water into a measuring cylinder and mixed with so much alcohol that the mixture contains 80% (by volume) of the latter. The precipitate of barium tartrate malate and succinate is collected on a filter and washed with 80% alcohol. The filtrate which contains barium acetate and lactate and glycerol is freed from alcohol by evaporation and diluted t o a known volume. An aliquot portion is distilled under diminished pressure the distillate is diluted to a known volume an aliquot part of it is oxidised with a large excess of potassium dichromate and dilute sulphuric acid (say 5 C.C.with 50 C.C. N / l O dichromate solution and 50 C.C. 20% sulphuric acid) by heating for one hour in a pressure-bottle in t.he water-bath and the excess of dichromate is estimated by adding potassium iodide and titrating with N/10 sodium thiosulphate solution; in this way the glycerol is determined. An aliquot portion of the filtrate is oxidised in the same way without being distilled first ; in this way the sum of t h e glycerol and the lactic acid is determined (C,H,O requires 20). The precipitated barium salts with the filter are brought into a measuring cylinder stirred with 20 C.C. of water containing some sulphuric acid diluted with alcohol to 100 c.c. and filtered. Of the filtrate 80 C.C. are mixed with 0.5 C.C. of 20% potassium acetate solution 15 grams of powdered potassium chloride and 2 C.C. of acetic acid and allowed t o remain for twelve hours ; the Precipitated potassium hydrogen tartrate is then filtered off and its amount determined by titration.The filtrate is freed from alcohol by evaporation mixed with 50 C.C. of dilute sulphuric acid and titrated with N / l O potassium permanganate solution; in this way the malic acid is determined (C4H605 requires 60). The oxidised liquid is evaporated to dryness with quartz sand the residue is extracted with ether the extract is evaporated and the succinic acid left is weighed. The alcohol is estimat,ed by neutralising another portion of the original mixture distilling off the alcohol and oxidising a portion74 ABSTRACTS OF CHEMICAL PAPERS.of the distillate with dichromate and sulphuric acid as described above (C2'EI,0 requires 20). The acetic acid is estimated by distilling another portion of the original mixture and titrating the distillate. C F. B. Estimation and Colour Reactions of Qlycuronic Acid. K. U. L E F ~ V ~ ~ E and BERNHARU TOLLENS (Ber. 1907 40 4513-4523 ; Zeitsch. Fer. deut. Xucke:e,.irnd. 1907 57 1097-11 10. Compare Neuberg and Neimann Abstr. 1905 ii 436; Tollens ibid. 559).- When boiled with hydrochloric acid glycuronic acid yields fur- furaldehyde and carbon dioxide (Giinther and Tollens Abstr. 1890 1352; Chalmot and Tollens Abstr. 1891 5G8; Mann and Tollens Abstr. 1895 ii 335). On this reaction the authors have based two methods of estimating glycuronic acid ; in the first the furfuraldehyde is converted into andmeighed as its phloroglucide; in the second the carbon dioxide evolved is absorbed and weighed.The weight of the phloroglucide is one-third and that of the carbon dioxide one-quarter of the glycuronic acid Details are given of estimations with gly- curonic acid euxanthic acid magnesium euxanthate and sodium urochloralate in which the error does not exceed 1-1*5%. By tlie first method piuri (Indian-yellow) is found to contain 18% to 19.5% of glycuronic acid equivalent to 43.1% to 46.7% of euxanthic acid ; Graebe Aders arid Heyer (Abstr. 1902 i 39) found piuri to contain 39% to 46% of euxanthic acid. To estimate pentoses and glycuronic acid in presence of each other the amount of glycuronic acid is determined by the carbon dioxide method and one-third of its weight substracted from the phloroglucide the remainder of this being equivalent to the pentosos.Piuri is found to contain 1.39% of nitrogen which is only partly accounted for by the presence of small amounts of hippuric acid and a substance which is probably m-toluic acid; another specimen con- tained benzoic acid. Contrary to Bial's (Abstr. 1902 ii 703) and Sachs's (Abstr. 1907 ii 135) statements glycuronic acid reacts with Bial's orcinol reagent but more slowly than the pentoses and only on continued heating of the reaction mixture. The green coloration can be detected with 0.0025 mg. of xylose 0.005 mg. of arabinose and 0.035 mg. of glycuronic acid. With phloroglucinol in presence of ferric chloride glycuronic acid forms a dirty red turbid solution which gives a spectrum with a band atl the 11-line less marked than the band lying to the right of the' U-line in the pentose-phloroglucinol spectrum. G.T. Detection and Estimation of Benzoic Acids in Ketchups Fruits and Ciders. HAHKY S. REED (J. Amzes-. Ciiem. Xoc. 1907 29 1626-1 629)-The sample is acidified with sulphuric acid the benzoic acid is extracted with chloroform and the latter is evaporated in a current of air in a special apparatus. The residue is then heated at 260" in a current of air and t h e rc)littilised henzoic acid absorbed in aqueous sodium hydroxide. Prom this it is recovered by acidifyingANALYTICAL CHEMISTRY. 45 the solution and shaking with chloroform and the residue left on evaporation is then tested for benzoic acid by converting it first into m-dinitrobenzoic acid and this into ammonium diaminobenzoic acid which has a characteristic appearance.The quantitative estimation is based on the fact that calcium bendoate is fairly soluble in cold water. The chloroform residue of the acid is treated with a definite amount of milk of lime and the benzoic acid is calculated from the amount of calcium found in the solution after allowing for dissolved calcium oxide as found by a blank experiment L. DE K. Estimation of Fat in Skimmed Milk. R. L E Z ~ (Cornpt. rend. 1907 145 817-S18),-Three litres of the skimmed milk are heated to a temperature of 40" and treated with 60 C.C. of a solution prepared by dissolving 80 grams of sodium hydroxide in 180 C.C. of ammonia and diluting to a volume of 250 C.C. The treated milk is then subjected to centrifugal action a small apparatus being used and the cream thus separated is used for the analysis the estimation of the fat being made by Gerber's process.w* P. s. Elstimation of Some Alkaloids by Means of Picrolonic Acid. HERMANN MATTHES and 0. RAMMSTEDT (Zeitsch urrzd. Clwn 1907 46 565-574). -Gotarnine codeine and morphine give insoluble pre- cipitates when treated with picrolonic acid and this reaction may be used for the estimation of these alkaloids in preparations such as tablets powders Stc. A tablet or a quantity of the powder equivalent to about 0.05 gram of alkaloid is dissolved in 10 C.C. of water the solution is filtered and after washing the filter and inaoluble residue the filtrate is treated with 2.0 C.C.of 2c'/lO alcoholic solution of picrolonic acid. The mixture is placed aside for fifteen hours a t the ordinary temperature after which t h e the precipitate is collected on a weighed filter washed with water dried a t l l O o ant1 weighed. The precipitates have the following formuls cotsrnine picrolonate C1,Hl,O,~,C,,H,O,N ; codeine picrolonate and morphine picrolonttte Cl,H,,O,N,CloH,O,N (comp:zre Abstr. ~18%?1O,N ,ClOH*O,N4 1907 ii 592). w. P. 8. Estimation of Cystine in Urine. J. F. GASKELL (J. P?qsioZ. 1907 36 142-149).-The addition of acetone renders the acetic acid method of estimating cystine a trustworthy one. I n five cystine calculi cystine crystallised in hexzgons but i t may crystnllise in other forms. The specific rotation of cystine dissolved in ammonia is - 97*5" and dissolved in hydrochloric acid - 223".Hair-cystine and calculus-cystine are identical. W. D. H. Estimation of Indigotin in Indigo-yielding Plants. CYRIL BERGTHEIL and RICHARD V. BRIGGS (J. SOC. ClLem. Ind. 1907 26 1172-1 174).-The modifications of the persiilphate strid isatin methods suggested by Orchardson Wood and b'loxani (hbstr. 1907 ii 415) are considered by the authors to be useless in ordinary practice both76 ABSTRACTS OF CHEMICAL PAPERS. leading to untrustworthy results unless the exact indican-content of the sample is known previously and the methods adjusted accordingly. w. P. s. Analysis of Indigo (111) and of the Dried Leaves of Indogofera Arrecta and Indigofera Surnatrana. R. GAUNT P.THOMAS and W. POPPLEWELL ELOXAM (J. h’oc. Chem. I d 1907 26 11 74-1 182). -As the result of continued use of the tetrasulphonate method the authors consider that the claims for accuracy made for this process (Abstr. 1906 ii 819) are fully substantiated. For the analysis of the leaf the isatin method is at present the only trustworthy one; the product obtained from it has been proved to be almost chemicallypure indirubia. A s the formation of indirubin can only originate from the indoxyl liberated from the indican on hydrolysis the quantity of indirubin obtained must be a true expression of the amount of this substance which is present. w. P. s. Titration of Tannin by Means of Iodine. FERDINAND JEAN (Ann. China. and. 1907 12 426-427).-9 reply to Cormimbeuf (Abstr. 1907 ii 994). The author states that the iodine process is trustworthy but the titration should take place in a solution contain- ing excess of sodium hydrogen carbonate. 1,. DE K. Vanillin-Hydrochloric Acid as a Test for Albumin and Tryptophan. LEOPOLD ROSENTHALER (Chem. Zentr. 1907 ii 946 ; from Apoth. Zeit. 1907 22 67S).-Winckel has shown (Apoth. Zeit. 1905 22 209) that several fruit juices and substances con- taining ferments give a violet coloration with vanillin and hydro- chloric acid and has concluded that this is a characteristic reaction of ferments. I n the present paper it is pointed out that albumin globulin and casein give a violet coloration with vanillin and hydrochloric acid ; likewise also tryptophan but not phenylalanine tyrosine histidine or pyrrolidine-2-carboxylic acid. W. H. G. Estimation of Pepsin by Mean8 of Edestin. EKNST FULD and LOUIS A. LEVISON (Biochema. Zeitsch. 1907 6 473-501).-The method is based on the principle that edestin (the crystallire protein from hemp seeds) is precipitated by sodium chloride from its solution in 0.3% hydrochloric acid but that the products of its digestion are not so precipitated. Accordingly the minimum quantity of gastric juice is found which when added to a definite volume of the edestin solution prevents the formation of a precipitate when solid sodium chloride is added after half an hour’s incubation. G. B.