120 THE ANALYST. INORGANIC ANALYSIS. The Gravimetric Determination of Mercury, and its Separation from Arsenic, Antimony, and Copper. Charles J. Pretzfeld. (Journ. Amer. Chem. SOC., xxv., 198.)-The author shows that the method of weighing mercury as siilphide gives high results, whilst the various methods depending on the reduction to mercurous chloride and weighing of this precipitate give somewhat low results when the chloride is precipitated from mercuric chloride solutions, though they give good results when the metal is present as nitrate and the reducing agent added before the sodium chloride. Precipitation with sodium arsenate, and weighing the Hg,(AsO,), obtained after drying at 100" C., also gives good results when applied to solutions free from hydrochloric acid, but is inapplicable to these.Both the electrolytic method, due to E. F. Smith, and that in which a solution of the double sulphides of mercury and potassium is used, give good concordant results. For the separation of mercury from antimony, arsenic, and copper, the author recommends the following method, which is more satisfactory than the ammonium sulphide method, and yields good results: To the clear solution of the metals a considerable amount of tartaric acid solution is added ; after stirring for one or two minutes, potassium cyanide is added gradually till a considerable excess is present, after which hydrogen sulphide is led into the solution till saturated, the liquid beiug kept cold. The precipitated mercuric sulphide is filtered off, dissolved, and the solution electroly sed.A. G. L. Rapid Determination of Molybdenum in Steel. G. Anchy. (Jown. Amsr. Chem. SOC., xxv., 215.)-In using the method previously described (THE ANALYST, xxvii., 205) the author now recommends taking only 0.8 gramme of the steel drillings,THE ANALYST. 121 making the caustic soda precipitation in a volume of 200 c.c., and taking only 100 C.C. of the filtrate for reduction and titration, whereby much time is saved. An improve- ment in the method due to Brakes (Journ. SOC. Chem. Ind., xxi., 832) is to expel the nitric acid from the original solution by evaporating to fuming with 3 C.C. sulphuric acid, as it is essential that the nitric acid should be completely expelled. A. G. L. A Chemical Method for determining the Quality of Limestones.Alfred M. Peter. (Journ. Amer. Chem. SOC., xxv., 143.)--Certain limestones containing ‘‘ blue ” layers are less valuable for exposed construction than ( ( gray ” limestones, on account of their more rapid disintegration on exposure to air. According to t h e author’s observations, the blue material may be shaly and moderately phosphatic, or very phosphatic and not especially clayey ; but it invariably contains easily soluble ferrous compounds, and a small, but very notable, amount of organic matter, which makes its presence known when the material is dissolved in dilute acids by the odour of petroleum imparted to the escaping gas, and by the brown scum floating upon the liquid. The author attributes the colour largely to the presence of ferrous phosphate, and is of opinion that the rapid weathering of ‘‘ blue ” limestone is due to the presence of easily oxidizable substances, such as ferrous compounds and organic matter, the oxidation of which is favoured by the greater porosity of this as compared with “ gray ” limestone.The porosity increases with the quantity of phosphates, and is possibly dependent upon the presence of minute shells, which sometimes give the rock a structure somewhat resembling that of chalk. Such limestone also sometimes. contains finely granular pyrites, which on oxidation assists disintegration. The author suggests that a good index of the relative stability of different limestones may be obtained by measuring the relative amounts of easily oxidizable organic matters, phosphoric acid, and clay or insoluble matters. He dissolves the limestone in dilute sulphuric acid in an Erlenmeyer flask in the presence of a known quantity of potassium permanganate, and by subsequent titration determines the quantity of permanganate consumed in the process of oxidation during half an hour at 100” C.He operates upon 1 gramme of the finely-powdered dried sample with 25 C.C. of decinormal per- manganate, and 100 C.C. of a 10 per cent. solution of sulphuric acid. More perman- ganate is added, if necessary, to maintain an excess. A blank of permanganate and sulphuric acid should be run under the same conditions. The solution obtained is filtered from the residue, which is ignited and weighed as total insoluble residue (clay and sand), phosphorus being estimated volumetrically in the filtrate.In the insoluble residue obtained from a duplicate determination the sand is deter- mined by digestion of the (unignited) residue first with sulphuric acid, and then with sodium carbonate solution, followed by ignition of the residue, the difference between this and the total residue being taken as clay. The conclusion arrived at is that limestone which consumes from permanganate much more than 0.3 per cent. of oxygen, and at the same time contains as much as 1 per cent. of phosphoric acid (more especially if it contains much “clay”), will disintegrate rapidly on exposure to air and moisture, and will be undesirable for road- building or other exposed construction.122 THE ANALYST. Average.. . ... ... 1.129 1 3.70 Highest . . .... ... 2.120 ’ 9-10 Lowest ... ... ... 0.666 ~ 0.96 Results are given of the examination of nine good and nine bad limestonee, of which results the following is a summary : 10-92 17.43 4.20 1.91 I P,Op I --- I I per Cent. Per Cent. 1 Per Cent. Tough, hard, compact stone, I resisting weathering, mostly I I gray ” : Average . . . ... ... 0.165 1 0.48 1 1.75 Highest.. . ... ... 0.341 1 1.01 2.26 Lowest ... ... ... 0.048 I 0.16 1.34 I -- I 1 Soft, porous, or shaly stone, weathering rapidly, mostly (‘ blue ”: I “ Clay. ” Per Cent. 0.52 0.66 0.13 7.05 10.23 _. - Total Insoluble. Per Ceut. -- 2-04 4.15 0-88 12.73 25.66 2.06 I A. G. L. __ Estimation of Free Lime in Thomas Meals. M. Bischoff. (Chem. Zeit., 1903, xxvii., 33.)-TT7hen a particular sample of Thomas meal was treated by the Scheibler process, agitating it with a 10 per cent.solution of sugar, the amount of lime recovered from the liquid was 6.12 per cent. after two hours’ shaking and 7.50 per cent. after twenty-four hours’ shaking. When, however, plain water was used, employing 1 litre per 1 or 2.5 grammes of meal, the yields were 2.46 and 3.10 per cent. respectively. On boiling the sugar solution, a precipitate was obtained, which proved to be calcium carbonate, and was equivalent to 3-20 per cent. of CaO. I t is therefore evident that a sugar compound with calcium bicarbonate passes into solution along with the ‘‘ saccharate ” of calcium oxide. From this it follows that when Thomas meals have to be examined for the lime content, water only should be used as the extracting liquid.F. H. L. Estimation of Citrate-soluble Phosphoric Acid. M. Passon. (Chenb. Zeit., 1903, xxvii., 33.) -The additions which Kellner and Bottcher have recently stated to be necessary (this vol., p. SO), when phosphoric acid is being determined by the direct magnesia process in Thomas meal (basic slag) liable to contain silica, seem to make the method so complicated that the oxidation process described by Mach and Passon in 1896 would appear to be simpler. This process has been somewhat modified, and iiow atands as follows : 100 C.C. of the Wagner extract are brought into a Kjeldahl flask and treated with 20 C.C. of strong nitric acid, 10 to 15 C.C. ol strong sulphuric acid, and a drop of mercury. The whole is boiled, and before it is cold a pinch of sodium chloride is added, which throws down most of the mercury. The liquid is diluted to 200 c.c., and 100 C.C.of the filtrate (=0*5 gramme of meal) are mixed with 50 C.C. of &he usual ammonium citrate solution and 20 C.C. of strong ammonia. When cold theTHE ANALYST. 123 phosphoric acid is precipitated with 25 C.C. of magnesia mixture, the whole is agitated for 5 minutes, and then filtered through a Gooch crucible. The concentrated acids separate out all the silica, whatever its proportion may be; and the method has an additional advantage in requiring only five minutes’ agitation instead of the thirty minutes’ stipulated for by Bottcher. F. H. L. Estimation of Citric Acid-Soluble Phosphoric Acid in Presence of Silica. Wm. Naumann. (Chenz. Zeit., 1903, xxvii., 121.)-Five grammes of Thomas meal are agitated in a 4-litre flask with 2 per cent.citric acid solution for thirty minutes, and 100 C.C. of the filtered liquid ( = 1 gramme of substance) are collected in a 250 C.C. flask. About 8 C.C. (one-third of the quantity formerly used in the Naumann process) of nitric acid are added to prevent deposition of calcium citrate and to promote smooth boiling, and the mixture is concentrated over a small flame till the volume is reduced to about 25 C.C. and the liquid begins to boil badly. The flask is taken from the flame, allowed to cool slightly, and either 25 C.C. of strong sulphuric acid alone, or 25 C.C. of sulphuric acid and 5 C.C. of strong nitric acid are introduced. The flask is replaced over the flame and boiled for about ten minutes till the appearance of white vapours shows that the temperature has reached the boiling-point of sulphuric acid, and that the silica is thrown out.The liquid is cooled, cautiously diluted to 250 c.c., and 125 C.C. of the filtrate (= 0.5 gramme of meal) are treated with about 35 C.C. of strong ammonia, using rosolic acid as indicator, cooled, and agitated for half an hour with 50 C.C. of 24 per cent. ammonium citrate and 25 C.C. of magnesia mixture, filtering either immediately or when convenient. By adding the nitric acid in two portions as above described no oxidation of the citric acid occurs before the concentration of the solution. The process yields pure precipitates, which filter well. F. H. L. Separation of Quartz from Amorphous Silica. B.Sjollema. ( J . L U d W . , 1902, l., 371 ; through Chem. Zeit. Rep,, 1903, 2l.)-Quartz cannot be quantitatively separated from amorphous silica either by means of sodium hydroxide or by sodium carbonate solution. The former dissolves a little quartz as well as non-crystalline silica, and with the latter it is difficult to bring all the amorphous silica into solution. I t follows, therefore, that the work of Von Piedzicki, who determined the proportion of free silica in soils by boiling the samples’ for an hour and a half with 10 per cent. sodium hydroxide, is valueless. The separation, however, may be effected by the use of solutions of methylamine or, preferably, diethylamine. A 33 per cent. solution of diethylamine dissolves 0.6 gramme of amorphous silica to within 0.0002 gramme, if it is boiled therewith for about eight hours, and a 16.5 per cent.solution of msthylamine acts similarly. (cf. ANALYST, 1898, xxiii., 84.) F. H. L. Estimation of Bromine by means of Bromate Solution. W. Vaubel. (Chem. Zeit., 1902, xxvi., 1220.)-Since the hydroxyl and amido groups which exist in the ortho or para positions are rapidly displaced by bromine in acid solution from124 TEE ANALYST. bodies like phenol and the amines, while chlorine only acts in a similar fashion slowly, it might be thought that a mixture of potassium bromide and chloride could be analysed by adding standard potassium bromate solution in presence of a phenol and an acid until a reaction is obtained with potassium iodide starch paper- However, although potassium bromide, or bromine water, can be andysed in the nmnner indicated if sulphuric acid is used, the process does not work in the presence of hydrochloric acid, presumably because the bromine substitution product already formed increases the power of the phenol or aniline to take up chlorine.The Same method is nevertheless available for the titration of phenol and aniline, because an excess of bromide is always present. When it is employed, in the absence of hydrochloric acid, for the estimation of a bromide, only one-half the volume of bromate required must be calculated from, because the other half is used up in the formation of the tribromophenol derivative. F. H. L. The Detection of Hydrocyanic Acid in the Presence of Sulphocyanic, Wdroferrocyanic, and Hydroferricyanic Acids and their Salts.Louis E. Prsiss. (Amer. Chenz. Journ., xxviii., 240.)-All metals, except those belonging to the alkali group, are first removed by boiling with sodium carbonate sohtion. Ch-”tic potash and about 0.5 gramme aluminium filings are then added to the liquid As soon as a test with ferrous sulphate shows r e d u h ~ n to be compIete, which is usually the case after fifteen minutes, the solution is acidified with hydrochloric mid, and mercuric chloride added to precipitate the hydroferrocyanic acid. After shaking for a few minutes in the cold, the precipitate may be filtered off and washed with a little mercuric chloride solution. ?he solution is next made alkaline with caustic potash, the precipitated mercuric oxide filtered off, and the filtrate heated to boiling with a little ferrous sulphate, after which it is again filtered, acidified with hydrochloric acid, and ferric chloride solution added.After decolorizing any sulphocyanide formed by means of mercuric chloride, the blue Colour or precipitate due to the Prussian blue becomes apparent if any cyanide was present in the original substance. reduce any ferricyanide. A. G. L. Analysis of Cyanide Fume. L. Schneider. (Oesterr. ,%?its. Berg. U . Hiittenw., 1902, I., 498; through Chem. ,&it. Rep., 1902, 281.)-This dust contains the follow- ing cyanogen compounds : Potassium ferrocyanide, thiocyanate, cyanide, and cyanate. These may be recovered by passing it through water, when the solution also contains potassium hydroxide, chloride, and carbonate.To estimate the cyanogen compounds the liquid is made alkaline with ammonia, and silver nitrate is added; the ferrocyanide is filtered off, washed, dried, and the silver in it deter- mined by the dry process. The amount of ferrocyanogen can also be checked by an iron estimation. Part of the filtrate is acidified with nitric acid, when silver cyanide, chloride, thiocyanate, and cyanate fall. This precipitate is warmed with aqua regia, and in the filtrate the sulphuric acid coming from the thiocyanogen is determined. Another portion of the ammoniacal liquid is acidified and the silver salts collected ; these are then digested for one hour at 100” C. with 10 C.C. of nitric acid and 200 C.C.THE ANALYST. 125 of water ; silver thiocyanate, cyanide, and chloride remain undissolved ; potassium cyanate appears in the solution.To the filtrate hydrochloric acid is added, and the amount of silver chloride, which corresponds with the original cyanate, is ascertained. If the above-mentioned precipitate is heated for one hour at 100" C. with strong nitric acid (1 : I), the silver from all the cyanogen compounds passes into solution, except that of the chloride and the yellow silver perthiocyanate ; so that the silver chloride thrown out of the hot filtrate corresponds with the original thiocyanate, cyanide, and cyanate. The silver cyanide can thus be estimated by difference. It may, however, be determined by mixing the aqueous extract of the fume with ammonia, and ammonium carbonate, precipitating the ferrocyanogen with silver, treating the filtrate with potassium iodide, shaking, and filtering.Only silver cyanide remains in solution, and its amount can be deduced from the silver chloride obtained on acidification with hydrochloric acid. F. H. L. The Volumetric Determination of Nitric Acid. J. I(. Phelps. (Zeits. Anoyg. Chern., xxxiii., 357.)-The author's method, which is a modification of Holland's (Chem. News, xvii., 219), depends on the oxidation, in an atmosphere of steam, of a ferrous salt by the nitric acid in the presence of hydrochloric acid. The apparatus used consists of a 250 C.C. flask closed by a double-bored rubber stopper, carrying a tap-funnel with constricted stem and a short wide tube bent over downwards outside the flask, to serve as exit tube. By dipping the end of this tube into mercury contained in a test-tube, it is possible to increase the pressure in the flask.To carry out a determination, the solution containing the nitrate is placed in the flask, the stem of the tttp-funnel filled with water, and the tap closed, after which the solution is boiled down to a volume of about 10 c.c., the exit tube being kept 3 centimetres below the surface of the mercury. After the air has thus been expelled from the flask, a known volume of standard ferrous sulphate solution in fairly large excess is drawn into the flask by allowing it to cool slightly, and raising the exit tube in the mercury. A quantity of concentrated hydrochloric acid, about equal in volume to the contents of the flask, is then added in the same way, and the whole boiled down to a volume of 10 or 15 c.c., to insure thorough reduction of the nitric acid, the exit tube being kept only just below the surface of the mercury during this boiling.The excess of acid is then neutralized by means of sodium carbonate, the flask allowed to cool, and the excess of ferrous sulphate determined with potassium permanganate or iodine and arsenious oxide solutions. The method gives very satisfactory results. For exact work, ammonium salts must be absent, and hence ammonium ferrous sulphate should not be used in place of ferrous sulphate. A. G. L. Notes on the Iceland Spar Method for Standa.rdizing of Hydrochloric Acid. W. Heber Green. (Chem. News, lxxxvii., &)--The author has used this method, which was proposed by Masson (Chem.News, lxxxi., 73), for the last three years, and finds it superior to any other. The use of a pipette instead of a burette is recommended for the measurement of the acid, the average error of a 20 C.C.126 THE ANALYST. pipette carefully freed from grease by means of an alcoholic soap solution being only 0.003’ C.C. The beaker used should be of Jena glass, and should be boiled for three or four hours in acidulated water before being used, since after this treatment its weight remains practically constant even when expmed to the action of the hydro- chloric acid and calcium chloride solutions. Finally, the molecular weight of calcium carbonate may be safely assumed as 100.0 (0 = 16) for this determination, since the error in the measurement of the acid is greater than the uncertainty in the atomic weight of calcium.A. G. L. Estimation of Perchlorates. M. Honig. (Chem. Zeit., 1903, xxvii. , 32.)-If the material to be analysed is a salipetre, and does not contain more than 5 per cent. of perchlorate, 5 to 10 grammes of it are taken; but in other cases the sample is diluted with sodium or potassium nitrate free from chlorine till the proportion of perchlorate does not exceed that limit, and 5 or 10 grammes of the prepared substance are melted in a small nickel crucible over a gas-burner, which brings the mass to quiet fusion without causing the ’ vessel to be visibly red. Next 2 to 3 grammes of iron filings are added and stirred i n ; the crucible is covered with a watch-glass and kept over the flame for half an hour, stirring at intervals.Finally, the melt is cooled, taken up in warm water, filtered, and the chlorine determined gravimetrically. Powdered zinc, aluminium, and lead (already recom- mended by Selckmann) are less convenient reducing agents, for the mixtures obtained with them are less fluid unless the temperature is pushed to a point where volatili- zation of chlorine is probable. Tin acts too energetically unless the perchlorate is diluted with a carbonate, and its salts are troublesome in after-analysis. Magnesium reduces the perchlorate with almost explosive violence, even when the diluent is a carbonate. The examples quoted show maximum errors of -1-0.8 and - 1.1 milli- gramme when estimating amounts of potassium perchlorate up to 0.497 gramme in weight.F. H. L. On the Solubility of Boric Acid i n Hydrochloric Acid. W. Herz. (Zeits- Anorg. Chem., xxxiii., 355.)-The author has made some experiments to see whether the statement in Dammer’s ‘‘ Handbuch der Anorganischen Chemie ” to the effect that boric acid is more soluble in hydrochloric acid than in water is correct, and finds, on the contrary, that the opposite is true, the solubility steadily diminishing with increasing concentration of the hydrochloric acid. The solubility of boric acid in pure water was found to be 0.907 gramme-equivalents per litre; while in 9.51 normal hydrochloric acid the solubility is only 0.338 gramme-equivalents. A. G. L. __ - _ _ _ _ - On the Simultaneous Volumetric Determination of Boric Acid and Strong Acids. W. Herz. (Zeits.Anorg. Chem., xxxiii., 353.) - In solutions containing boric acid together with some strong acid both may be determined volumetrically as follows : The quantity of standard sodium hydrate solution required to give a neutral reaction to the solution, using nitro-phenol as indicator, is first ascertained. Nitro-THE ANALYST. 127 phenol, which is yellow in alkaline and colourless in acid solution, is not affected by boric acid, and hence the quantity of sodium hydrate found above corresponds to the strong acid present. Mannitol and a few drops of phenolphthalein solution are then added to the liquid, and the titration continued to the appearance of the pink colour, the quantity of sodium hydrate now used corresponding to the b6ric acid in the solution. A. G. L.~ _____ A Rapid Method for determining Sulphur in Coal and Coke. Carl Sund- strom. (Jozcm. Amer. Chem. Xoc., xxv., 184.)-Sodium peroxide is used as the oxidizing agent. In analysing coke, 0.7 gramme of the finely-powdered sample and 13 grammes dry sodium peroxide are intimately mixed in a 30 C.C. nickel crucible. A 3-inch fuse is inserted in the mixture, the crucible covered, supported on a triangle, and placed in about 9 inch of water; the fuse is then lit, and after three or four minutes, crucible and cover are placed in a small beaker with 30 C.C. water. As soon as solution has taken place, crucible and cover are removed, the solution is made acid with hydrochloric acid, and filtered ; in the filtrate, sulphuric acid is precipitated as usual. The fuses used are made by treating cotton wick with a mixture of 1 part fuming nitric acid and 2 parts sulphuric acid at 15' C. for twelve hours, washing in running water for twelve hours, drying at the ordinary temperature, then soaking in cold, nearly saturated potassium nitrate solution for an hour, pressing out the excess of solution, and again drying at the ordinary temperature.In analysing coal, the oxidation is effected in a small cylindrical steel bomb, the dimensions of which are : Internal diameter, la inches ; internal depth, I& inches; sides, bottom, and flange, Q inch thick; cover-plate, qK inch thick and 1% inches diameter. The cover does not touch the clamp, and is insulated from the bomb by a thin mica gasket, and from the screw by a piece of ordinary red fibre.For the deter- mination, 0-7 gramme of the coal and 13 grammes of sodium peroxide are mixed in the bomb, and compressed by means of a small vice. A thin iron wire, 4 inches long, is then inserted, one end being under the mica, touching the bomb, the other above the mica, in contact with the cover, the wire being looped so as to touch the mixture inside. A current of 4 ampAres is used to start the reaction, one terminal being applied to the cover, the other to the bomb. The product of the reaction is treated as for coke. The above methods give results practically identical with those given by Eschka's method in much less time and with cheaper apparatus, while there is also less chance of the in troductioa of extraneous sulphur. A. G. L. Determination of Selenium in Organic Compounds.H. Frerichs. (Arch. Phctrm., 1902, ccxl., 656 ; through Chem. Zed. Rep., 1903, 22.)-The organic matter in about 0.2 or 0.3 gramme of the sample is destroyed by the Carius process, by treating it with 1.4 nitric acid and about 0.5 gramme of silver nitrate. The mass is then brought into a porcelain basin by the aid of water, and evaporated to dryness. The residue is rubbed down with a few drops of water, rinsed on to a filter with128 THE ANALYST. alcohol, and washed with the latter as long as silver exists in the filtrate. The paper and precipitate are next transferred to a beaker, and boiled with 20 C.C. of nitric acid and 80 C.C. of water for about five minutes until everything has passed into solution. The liquid is diluted with 100 C.C.of water, mixed with 1 C.C. of a strong solution of iron-ainmonium'alum, and titrated with decinormal potassium thiocyanate. One C.C. of the latter corresponds with 0.00395 gramme of Se. The results are accurate. (Cf. this volume, p. 44.) F. H. L. Calculation of the Calorific Power of Coal. Goutal. (Ann. de Chim. anal., 1903, viii., 1-4. )-The formula originally proposed by the author: (Ann. de Chirn. anal., 1896, 169) was modified by De Paepe (ANALYST, xxiv., 107), who applied it to a wider range of coa,ls. Since then the author has examined more than 600 samples, and has found that the calorific power can be calculated with suficient accuracy for industrial purposes by means of the formula P = 82C + aV, where P represents the calorific power, C the percentage of fixed carbon, V that of the volatile substances, and n a factor depending upon the proportion of volatile substances. The author has plotted a curve of experimental results to determine the value of this factor, and from this has obtained the following values : Coal containing volatile Corresponding value of substances per cent. 5 10 15 20 25 30 35 38 40 a in calories ... ' ... 145 130 117 109 103 98 94 85 80 In the case of anthracite coals a is represented by a constant equal to 100 calories, and the formula becomes P = 82C + 1OOV. The difference between the values thus calculated and those actually determined is stated to rarely exceed 1 per cent., though in the case of certain anthracite or lignite coals it may exceed 2 per cent. The moisture is determined on 2 grammes at 110" C. ; the volatile substances by drying 2 grammes of coal and heating it in a covered platinum crucible over a Buasen burner, until flames no longer appear between the crucible and its cover, and then rapidly cooling the sample in a desiccator; and the ash by slow combustion in a muffle (cf. ANALYST, xxi., 21). C. A. M.