ORGANIC ANALYSIS 51 INORGANIC ANALYSIS. Estimation of Amrrionia by the Boric Acid Method. L. Adler. (Zeitsch,. ges. Brauw., 1916,29, 161.-164, 169-172; through J . Inst. Brewing, 1916,22, 500.- In the volumetric estimation of ammonia-e.g., in the Kjeldahl method-it is usual to employ standard sulphiiric acid as absorbing liquid. Winkler, in 1913, suggested the use of boric acid, which is practically neutral to methyl orange; and the author, as the result of experiments described, recommends the followiug modification of Winkler’s method: The ammonia is absorbed in 50 C.C.of a 4 per cent. solution of pure crystallised boric acid contained in a 300 C.C. flask; these quantities, of course, need not be determined very accurately. The condenser tube should dip well below the surface of the absorbing liquid during the first fifteen minutes of distillation, but afterwards it can be raised so that the subsequent distillate serves to rinse the end of the tube.The condenser should be cooled by water, sufficiently to insure that the distillate reaches the absorption flssk at room temperature. T w d y52 ABSTRACTS OF CHEMICAL PAPERS minutes of vigorous ebullition suffices to distil all the ammonia.To insure the mas- mum of accuracy in the titration of the distillate it is advisable to employ a colour standard; this is prepared by treating 250 C.C. of distilled water with 0.15 C.C. of $G sulphuric acid and a few drops of a 0-05 per cent. solution of methyl orange. An oraage-yellow tint is thus produced, which remains unchanged for a long time.The disfiUa,te, treated with the same amount of methyl orange,is titrated with & sulphuric mid to the same tint, and the ammonia content is calculated from the volume of acid required, no correction being necessary for the boric acid present. The maxi- mum amount of ammonia which can be ahorbed without loss by the quantitg of boric acid mentioned above (50 C.C.of 4 per cent. solution) was not determined, but i t exceeds 70 mgrms. The results are as accurate as when sulphuric acid is used for absorbing t,he ammonia, Effteiency of the Aeration Method for Distilling Ammonia. P. A. Kober. ( J . Amw. Chem. SOC., 1916,38,2568-2572.)-The paper is a reply to various adverse criticisms on the efficiency of the aeration procedure in the ordinary Kjeldahl method for nitrogen determination a s a means of quantitatively separating the ammonia.To insure accurate results in distilling ammonia by aeration it is necessary t o use: (1) A sufficient volume of air; (2) a.6: high a column but as low a volume of liquid as is convenient; (3) a saturated solution of pure sodium hydroxide in adequate excess (ANALYST, 1913, 38, 581).An impure alkali containing sulphite i s liable to cause error because of the suIphur dioxide evolved and carried over into the standard acid before the acid of the Kjeldahl mixture is complet,ely neutralised. (4) Potassium hydroxide must not be used, hecauee the difficultly soluble pot,assium sulphate which separates may carry down ammonia, by occlusion or as a double salt.Errors of 10 per cent. may be caused in this way, but potassium sulphate a s ordinarily used in the preliminary acid treatment is not sufficient to produce appreciable error (ANALYST, 1908,33,442; 1909, %, 333; 1910,35,316; 1916,41,186,386). H. F. E. H. Estimation of Rare Gases (Argon). A. Sieverts and R. Brandt. (Zeittxh. angew. Chem., 1916,29, 402-406.)-The process described depends on the fact that metallic calcium at a dull red heat absorbs nitrogen and okher gases with the excep- tion of those of the argon group.When air only has to be removed from the argon the absorption may be made directly, but if the gas under examinat.ion contains large quantities of carbon monoxide, carbon dioxide, met<hene, etc., i t is advisable to remove the greater part of these previously.The apparatus employed consists simply of a tube made of combustion glass for containing tAhe calcium and a mano- meter attached to the tube by a movable joint; three-way taps are provided a t the top of the two arms of the manometer and connect with t.he calcium tube, an inlet tube for the gm, an exhaust tube, and a tube for admitting air to one of the arms.About 5 grms. of metallic calcium are placed in the tube; the latter is placed horizon- tally in a small oven heated at 450" to 550" C., and the air is exhausted from the appa- ratue. The calcium tube if! then removed from the oven, allowed to cool, and in- serted vertically in a bath of cold water. The gas to be analysed is now admitted,INORGANIC AKALYSTS 53 the pressure noted on the manometer, the calcium tube is heated as before for about one hour or until the pressure no longer decreases, then cooled, and the pressure observed a t the same temperature as the previous reading.When the original pres- sure is denoted by P and t:he final pressure by p , the percentage by volume of rare gasea present is lOOp/P. w. P. s. Boiling Method for Estimation of Water-Soluble Arsenic in Lead Arsenate.G. P. Gray and A. W. Christie. .(J. Ind. andEng. Chew., 1916,8, 1109-1113.)- Arsenical insecticides for spraying trees must be substantially free from water- soluble arsenic compoundB, as these are violent plant poisow. For this reason the U.S. Legislature and those of the separate States of the Union have set limits to the permissible amount of soluble arsenic in such preparations, and, since methods of determining soh bility are necessarily more or less empirical, the Association of Official Agricultural Chemists has recommended two methods with a view t o insuring agreement between manufacturers' analysts and those of the States.One of t#hese methods occupies more than ten days, a very serious objection from the point of view of the works laboratory, whilst it is shown in the present paper that the second method fails to extract more than 60 to 80 per cent.of the soluble arsenic, when the amount of this is considerable-that is to say, of the order of 1 per cent., 0.75 per cent. being the permissible limit. The method now proposed coiisiste in drying the preparation, mixing 0.5 grm.of the dried material with 2:OO C.C. ammonia-free water and boiling the mixture for ten minutes, after which j.t is digested on the steam bath until the undissolved matter settles. The solution is then filtered off, mixed with 4 C.C. sulphurie acid and 1 grm. potassium iodide, and concentrated to about 40 c.c., diluted to about' 200 G.C. and any free iodine destroyed by the addition, drop by drop, of approxi- mately & thiosulphate solution.Methyl orange is added, and the solution nearly neutralised with sodium hydroxide. Sodium bicarbonate is then added in excess, and the arsenious oxide titrated with The method differs fro:m the earlier methods mainly in the use of boiling water for extraction of soluble matter, and in t,he great reduction of time required.It is shown that even prolonged boiling or digestion on the steam bath does not appre- ciably dissolve or hydrolyse lead arsenate. Yet the new method gives higher results, which lead to the conclueion that the earlier methods, which wed water a t room temperature or at 32" C., failed to extract the soluble areenic completely. G. C. J. iodine solution. Qualitative Separation and Detection of Tellurium and Arsenic.P. E. Browning, G. S. Simpson, and L. E. Porter. (Cbem. News, 1916, 114, 254.)- Tellurium is precipitated in the elementary form by means of sodium sulphite in dilute hydrochloric acid in presence of potassium iodide. The filtrate from the tellurium is boiled to remove excesa of sulphur dioxide, treated with hydrogen dioxide, boiled to remove the greater part of the iodine, made alkaline with sodium hydroxide, and treated with more hydrogen.dioxide to oxidise the arsenic to arsenate. The solution is then acidified, made alkaline again with ammonia, and treated with magnesia mixture to precipitate the arsenic. G. C. J.54 ABSTRACTS OF CHEMICAL PAPERS Estimation of Boron in Boron-Steel. C. Asehman, Jun. (Chim.Zeit., 1916,40, 960-961.)-1t is found that when boric acid ie heated at 1,000" C. with an exce8s of ammonium phosphate, the excess of the latter is volatilised completely and boron phosphate remains. A method based on this fact is described for the estimation of boron in boron-steel. A quantity of about 3 grms. of the steel is dis- solved in dilute sulphuric acid in a flask attached to a reflux condenser ; care should be taken to use the minimum quantity of acid for dissolving the metal.The ferrous eulphate is then oxidised by the addition of hydrogen peroxide and the mixture is distilled as far as possible; the distillate is collected in a receiver containing 0.5 grm. of ammonium carbonate dissolved in 20 C.C. of water, with the addition of a few drops of ammonia, and the outlet tube of the receiver is connected with a second receiver containing water.After the contents of the flask have cooled, 10 C.C. of abso1ut.e methyl alcohol, free from acetone, are added, and the distillation is continued; this addition of methyl alcohol followed by distillation is repeated five times. A current of air is drawn through the apparatus for some minutes between each distillaticn, and for thirty minutes after the final distillation. The contents of the receivers are then transferred to a weighed platinum basin containing 1 grm.of ammoniuni phosphate, the mixture is evaporated, and the residue heated at 1,000" C. in an electric oven until constant in weight. Results of experiments with &eel and added quantities of boric acid ahow that the method is very accurate.w. P. s. Qualitative Separation and Detection of Iron, Thallium, Zirconium, and Titanium. P. E. Browning, G. S. Simpson, and L. E. Porter. (Chem. News, 1916,114, 254-255.)-The hydroxides are dissolved in the least possible amount of sulphuric acid and the eolution treated with hydrogen peroxide, a red coloration indicating titanium.The solution is made faintly ahline with sodium hydroxide, and sodium phosphate is added. Sulphuric acid containing hydrogen peroxide is added until the solution is acid, the latter reagent serving to keep the titanium in the higher state of oxidation. The zirconium phosphate remains as a precipitate, and is filtered off. The iron and thallium in the filtrate are reprecipitated as hydroxides or phosphates by addition of sodium hydroxide, and the filtrate from the hydroxides is acidified with sulphuric acid and treated with sodium sulphite and a little.more sodium phosphate to precipitate titanium as phosphate. The hydroxides or phos- phates of iron and thallium are dissolved in sulphuric acid, the solution treated with sodium sulphite, and thallous iodide precipitated by addition of potassium iodide.The filtrate from the thallous iodide is boiled to remove sulphur dioxide, treated with hydrogen dioxide, again boiled to remove iodine and oxidise the iron, which is finally identified by the thiocyanate reaction. G. C . J. Separation of Lithium from the Other Alkali Metals. S. Palkin. ( J . Amer. Chem. Soc., 1916,38,2326-2332 .)-Sodium and potaesium chlorides are progreseively precipitated from an aqueous solution by addition of alcohol and ether, the filtrate containing the whole of the lithium chloride and only traces of the other chlorides is evaporated to dryness, the residue extracted with absolute alcohol containing a drop of hydrochloric acid, and the last trace of sodium and potassium chloride in the resulting extract is precipitated by addition of ether.INORGANIC ANALYSIS 55 The mixed chlorides (:U.5 grm.) are dissolved in 1.5 C.C.water, one drop of con- centrated hydrochloric acid is added, and then absolute alcohol (20 c.c.) drop by drop into the centre of the beaker, whilst its contents are rotated. Ether (60 c.c.) is added in a similar manner, and the mixture allowed to stand for five minutes.The precipitate is filtered off and washed with ether alcohol (4 : 1). The filtrate and washings are evaporated to dryness, and the residue taken up in 10 C.C. of absolute alcohol, warming if necessary, so that practically everything passes into solution. Ether (50 c.c.) is added in the manner above described, and one drop of hydrochloric acid. After allowing to &and for an hour, with frequent rotation, the solution is filtered from the small precipitate and the latter washed with et2her-alcohol.The filtrate and washings are evaporated to dryness, and the residue is converted into lithium sulphate and weighed. G. C . 5. Separation of C%siu.m and Rubidium by Fractional Crystallisation of their Alums. (Amer. J. Science., 1916,42, 279; through Chem.News, 19143,114, 285-286.)-The pronounced difference in t,he solu- bility of potassium aluminium sulphate and of the alums of czesium and rubidium enables a satisfactory separation of potassium to be made, but the difference in the solubilities of cmium and. rubidium alums (0.62 and 2.3 parts in 100 parts of water at 15" to 17" C.) is too small for a rapid separation.About 22 fractional crystallixa- tions were required before the czesium had been completely eliminated from the rubidium. There is a, much greater difference in the solubilities of the corresponding iron alums. Thus, 100 paxts of water a t 25" C. dissolve 2.7 parts of the czsiuia alum and about 17 parts of the :rubidium alum. After eight recrystallisations of t,he mixed alums the final fraction consisted of pure iron rubidium alum free from cssium.On treatment with a strong solution of ammonium aluminium su1phat.e the more insoluble alums are readily thrown out of solution, and this affords a convenient method of separating caesium from pollucite. It was found that 5 C.C. of a saturated solution of ammonium duminium sulphat,e would give a percept<ible precipit,at,e with 1 C.C.of a solution of rubidium chloride conhining 0.0002 grm. of rubidium, and with 1 C.C. of czesiiurn chloride solution cont,aining 0-00005 grm. of ca?sium. P. E. Browning and S. R. Spencer. ' C. A. M. Estimation of Thorium and Mesothorium in Monazite. K. L. Kithil. {Tech%. Paper, 110, U.S. Dept. Interior, Bureau of Mines; Chem. News, 1916, 114, 283-285.)-l?or the estirnetion of thorium in monazite the mefhod of Metzger (J.Amer.Chem. Xoc., 1902,24, 901) is recommended. The quantitat>ive estimat,ion of mexo- thorium is effected in the 13ame way as that of radium (Soddy , " The Chemistry of the Radio-Elements," p. 46 ; Wutherford, '' Radio-active Substances and their Radiations," p. 550), the results being expressed in terms of the y-ray activity of radium in equilibrium.If both radium and mesothorium are to be estimated, the y-ray method is used for both substances together, and afterwards the radium is estimated alone by the emanation method. After the solution has been boiled, the y-ray method gives only the amount of mesolthorium, since the radium emanation is expelled by boiling, and radium C,, which emits the y-rays, is practically disintegrated after three hours'56 ABSTRACTH OF CHEMICAL PAPERS distance between poles, about.3 mm. time. The mesothorium obtainable from a ton of monazite sand is about 5-4 nigrms., but in practice only about 50 per cent. of that yield is obtained. It is sold on the basis of its activity compared with that of the purest radium bromide as estimated by the y-ray method.For the separation of the minerals in monazite sands the conglomerate is first freed from larger gravel and clatys, and is graded, before concen- tration, through sieves of 20, 50, 80, and 100 mesh, while clay and mica particles are removed by a sliming process. The properly graded material, when treated with a Wetherill electro-magnetic separator having two magnets and an 18-inch belt gives the following results: Current.Specific Points at wrhicli Qravi t J . Mineral Sqwatecl. St ineral. Magnetite . . . . .>a16 to 5.13 - Very weak. First pole, first magnet; dis- - tance between poles, about I ~ m m . Ilmeni t e . . . . 1 4.5to5 Haematite . . .. j - Garnet . . .. 3.8 to 6.3 E pido t e .. . . 3.2 to395 3.2 to3.25 3-3 to 3-6 Apatite .. .. Olivine .. Tourmaline .. Monazite (92 t.0 95 Platinum (if any) :: I - cellf.) and traces of zircon and rutile. Platinum, etc., if any. , Residues off the belt are quartz (2.651, felsper (2.5 to 2-7), gold, zircon (4.5 to 4-7), rutiIe (4.2 to 4*25), etc. C. A. M. Oxalate-Iodide Method for Paris-Green Analysis. C. A. Peters and L. E, Fielding. ( J . I d .and Eng. Chem., 1916, 8, 1114-1115.)-The sample (0-25 grm.) is treated with 1 C.C. of dilute (1 : 10) sulphuric acid and 50 C.C. water, and the mixture boiled for two minutes. Whilst still hot, about 2 grms. of oxalic acid crystals are added, little by little, as the first addition is apt to give rise to violent frothing. The mixture is again heated to boiling and then allowed to stand overnight.The copper oxalate is filtered off on asbestos, washed with water, and the crucible, asbestos, and precipitate are then placed in the original beaker, water and 5 to 10 C.C. of dilute { 1 : 1) sulphuric acid are added, the temperature is raised nearly to boiling, and the oxalic acid titrated with permanganate. From the consumption of permanganate the copper is calculated.The filtrate from the copper oxalate is neutralised with sodium bicarbonate, of which an excess is then added, and the arsenic is titrated with iodine solution. Test numbers show that the method is as accurate as others commonly used for the analysis of Parip green. G. C. J. Estimation of Phosphorus Pentoxide after Citrate Digestion, 0. C. Smith. ( J . Id. and Eng. Chem., 1916,8,1127-1128.)-Difficulty is sometimes experienced in obtaining a clear solution of the fertiliser and filters left after the removal of revertedINORGANIC ANALYSIS 57 phosphat,e by the digestion 4ith neutral ammonium citrate. By the use of the follow- ing method a clear solution can be obtained in less than an hour.The precipitate and papers are treated with 10 C.C. sulphuric acid and 50 C.C.dilute (1 : 1) nitric acid, and t'he mixture is digested in a flask on a hot plate until fumes of sulphuric acid arise. One or two C.C. of nitric acid are added and digestion continued until the appearance of sulphuric acid fumes, a,fter which the treatment is repeated. A clear solution nearly always results a t this stage. If it does not, further treatment with nitric acid will readily clear it,.After neutralisation wit.h ammonia and addition of am- monium nitrate, the phosphorus is precipitated in the usual way. G. C. J. Quantitative Estimation of Small Quantities of Sulphate. H. J. Hamburger. (Proc. K . klkad. Wetensch. Amsterdam., 1916, 19, 115-125; Biochem. Zeitsclz., 1916, 77, 168-188; through J . Chem. Soc., 1916,210, ii., 641.)-The micro- volumetric method, already applied in the estimation of small quantities of potassium, has been found to yield satisfactory results in the estima\tjon of sulphate by precipitation as barium sulphate.The precipitate is forced into a cali- brated capillary tube by centrifuging, and, when constantr, the volume of the pre- cipitate is read off. The tube is calibrated by preliminary experiments with solutions containing known quantities of sulphate.Experiments showed that the volume of B given quantity of precipitated barium sulphate depends on the size and shape of the crystals, and it has been found necessary t o carry out tahe precipitation under certain definite conditions. In particular, it is essential that the precipitated barium sulphate should consist of very small crystals, and this result may be atta.ined by the addition of acetone.The procedure adopted is to add 2.5 C.C. of hydrochloric acid (1 : 1) to 5 C.C. of the sulphate solution; to this solution are added 5 C.C. of a barium chloride solution (2.4 per cent. BaC12,2H20) containing three to five drops of acetone. When the estimation is carried out. in this manner the result, obtained is not affected by the prepence of sodium, potassium, calcium, magnesium, chloride, or phosphate in the original solution.Sublimed Sulphur rind its Adulteration. Fonzes-Diaeon. (Ann. F~d@f'., 1916, 9, 333-339.)-Sublimed sulphur (flowers of sulphur), when well prepared, contains about 32 per cent. of sulphur which is insoluble in carbon disulphide ( c f .Bruno, ANALYST, 1910, 35, 279). The proportion of this insoluble modification, however, decreases slightly when the sulphur is kept, and is less in specimens collected in the condensing chamber at a point near the inlet for the hot vapour. The latter specimens contain much coarse matter, and the author finds that the finer portions of the powder, after sifting, still show a high percentage of insoluble sulphur.Ground sulphur, on the other hand, is almost entirely soluble in carbon disulphide. If sublimed sulphur is sifted through a 100-mesh sieve and then through a 140-mesh sieve, and the insoluble sulphur determined in the portions which do not pass through the two sieves, respectively, the ratio of the insoluble sulphur in the coarser powder to that in the finer powder will always be less than unity.Por instance, the coarser portion of a sample of genuine sublimed sulphur contained 15.1 per cent. of insaluhle sulphur, and the finer portion 21.3 per cent. giving the ratio of 15.1/21*3. In other58 ABSTRACTS OF CHEMICAL PAPERS words, the finer portion of the powder contains the mbre insoluble sulphur. With ground sulphur, when i t contains any insoluble sulphur, the coarEer portion will contain more insoluble sulphur than does the finer portion, and the ratio will be greater than unity.If, therefore, a sample of sulphur be sifted through sieves of the dimen- sions given, and the insoluble sulphur estimated in the portions which remain on the sieves, the presence of not leFs than 25 per cent.of ground sulphur is indicated when the ratio between the two quantities of insoluble sulphur approaches near to or exceeds unity. With more than 30 per cent. of ground sulphur, the ratio is always above unity. w. P. s. Quantitative Estimation of Small Quantities of Sulphide Sulphur. W. A. Drushel and C. M. Elston. (Amer. J . Sci., 1916, 42, 155; through Chern. News, 1916, 114, 272-273.)-The method depends on the liberation of hydrogen sulphide and its quantitative absorption by lead acetate paper, the stain being then compared with standard stains produced by decomposing solutions of sodium sulphide con- taining from 0-0002 to 0-004 per cent.of sulphide sulphur. The apparatus consists of a round-bottomed flask through the cork of which is passed the smaller end of a Liebig condenser, slightly shortened and drawn downwards, while across the upper end is tied a filter paper moistened with dilute lead acetate solution. In preparing the standard stains the sodium sulphide solution, containing exactly 0.01 per cent.of sulphide sulphur, is diluted with nitrite-free distilled water to the required extent, and quantities of 1 to 5 C.C. are gently boiled with 25 C.C.of 0.5 per cent. hydrochloric acid in the flask, so that the steam passes not too rapidly through the lead acetate paper. In test experiments with sodium sulphide the method was found to be accurate within 04003 per cent. for solutions containing not more than 0.001 per cent. of sulphide sulphur. In using the method for the estimation of hydrogen sulphide in air, 25 litres of the air are slowly drawn through an absorption bulb containing dilute potassium hydroxide solution, and this is subsequently diluted, and an aliquot portion is boiled with hydrochloric acid in the apparatus.The method is capable of detecting less than 1 part of hydrogen sulphide in 5,000,000 of air, and gives con- cordant results. For the estimation of sulphur in coke from 5 to 10 grms.of the sample are boiled with dilute hydrochloric acid and the hydrogen sulphide absorbed by dilute potassium hydroxide solution, as in the ordinary method of Fresenius. The a.bsorbed sulphur is subsequently estimated colorirnetrically in aliquot, port.ions of the alkaline liquid. The results quoted are practically identical with those given by the gravimetric method.A further application of the method is in the analysis of paper, portions of 1 to 2 grms. being tested in the apparatus. Samples of tissue paper thus examined contained from 0@002 to 04W1 per cent. of sulphide sulphur, and it was found that when papers containing the larger amounts were used for wrap- ping articles of polished silver a pronounced tarnish appeared on the metal in the course of two to three weeks.c. 9. N. Estimation of Sulphur in Iron and Steel. H. B. PuMer. ( J . I d . and Eng. Chem., 1916, 8, 1115-1123.)-The paper includes a bibliography with 285 entries. The methods available are classified and briefly discussed. Particular attention isINORGANIC ANALYSIS 59 directed to the relation bertween fineness of sample and probable error of sampbg, in view of the fact that sulphur in iron and steel tends t,o segregate in that portion of ail ingot which solidifies last.If it, be desired that the third figure to the right of the decimal point should have any significance in a statement of percentage of sulphur in steel, and if 5 grms. of metal be taken for analysis, t,hen t,be sample should pass an 80-mesh sieve.With coarser material, several analyses must be made to estimate the probable error due to irregular distribution of the sulphur. Experiments are described demonstrating the accuracy of the nitric acid method. Bamber'a method is shown to give results which are nearly always high, but not uniformly high, and this is attributed to the access of sulphur from dust in the air daring the prolonged digestions required by this method. I n the author's hands, the chlorate method gives excellent results, its chief advantages being the almost instant and gentle solution of the iron and the avoidance of the use of nitric acid.It is also rapid. On the other hand, the accompanying use of hydrofluoric acid raises difficulties. Not only is glass apparatus athcked, but the barium snlphate may become con- taminat'ed by sodium fluosilicate, which is difficult to wash out.G. C. J. Analysis of Certain Tungsten Derivatives. 0. R. Sweeney. (J. Amer. Chem. SOC., 1916,38,2377-2383.)-Arsenic can be separated from mixtures of sodium tungstate and arsenate as follows: The mixture is heated in a combustion-tube to 300" C. in a current of hydrogen chloride.I n this way the arsenic js fully removed to a suitable receiver in four hours. The contents of the combustion-tube are rinsed into a dish with dilute ammonia, the solution evaporated to dryness, and the residue digested with dilute (1 : I.) nitric acid, adding small quantities of hydrochloric acid from time to time. When all action has ceased the liquid is evaporated and the residue repeatedly moistened with nitric acid, which is evaporated to expel the last traces of chlorine.The residue is dissolved in sodium hydroxide, the solution diluted, filtered, neutralised with nitric acid, and the tungstic acid precipitat,ed by addition of mercurous nitrate solution. Complex salts, such as sodium arsenio-tungstate or antimonio-tungstate, should be weighed into small tubes open a t both ends.These tubes are placed in a com- bustion tube and heated to 200" C. for an hour in a current of dry hydrogen chloride. The apparatus is cooled, the residue moistened with water, and heated again for thirty minutes in a current of hydrogen chloride. This treatment is repeated three times. The most convenient way to secure at correct temperature of the furnace is to turn the gas up slowly until tungstic acid just begins to volatilise, and then to check it some- what.Sodium vanadio-tungstafe cannot be analysed as above described, but if chlorine be mixed with the hydrogen chloride, the vanadium can be completely expelled a t about 200" C. A suitable form of combustion tube and a convenient oven (formed of a tin box) are figured in the text, which includes detailed directions for insuring that the separations shall be quantitative. The tungsten in, the residue is then estimated as already described. G. C. J.60 ABSTRACTS OF CHEMLCAL PAPERS Separation of Vanadium from Phosphoric and Arsenic Acids and from Uranium. W. A. Turner. (Amer. J . Sci., 1916, 42, 109; through Chern. News, 1916, 114, 261-262.)-The method of estimating vanadium in a metavanadate solution by precipitation with " cupferron " (ANALYST, 1916,41,261) is also applicable to its separation from phosphoric and arsenic acids, the only difference being t,hat a more thorough washing of the precipitate is required. For the estimation of vanadium in presence of uranium (uranyl nitrate) i t is necessary to make the precipitation from a fairly strong (10 per cent.) sulphuric acid solution, to prevent the formation of insoluble umnyl vanadate, and to use an acid solution of about the same strength, containing 1.5 grms. of cupferron per litre, for washing the precipitate. To effect a more complete separation the washed precipitate is dissolved in ammonia solution, and the liquid rendered nearly acid, and cooled to 20" C., before completing the acidi- fication and again precipitating the vanadium. The uranium in the filtrate may be estimated by the ammonium hydroxide method. The results t'hus obtained were about 0*002 grm. too low on 0.110 grm. of vanadium, while the amounts of uranium were in close agreement witlh theory. C. A. M.