114 THE ANALYST. INORGANIC ANALYSIS. Volumetric Method for the Estimation of Barium. A. E. Hill and W. A. H. Zink. (Journ. Anzer. Chcm. SOC., 1909, 31, 43-49.)-To a cold, neutral, or filightly arnmoniacal solution containing about 0.1 gram of barium in 60 to 70 C.C. of liquid, a solution of potassium iodate is added in a fine stream, with continuous stirring, an excess of about 25 C.C. being used. The whole is stirred for one minute, allowed to stand, with frequent stirring, for five minutes, and filtered. The filter should be used dry or after moistening with potassium iodate solution, as barium iodate is appreciably soluble in water. The precipitate is transferred completely to the filter, using portions of the filtrate, and is then washed, first three times with concentrated ammonia solution, and then three or four times with 95 per cent.alcohol. I t is then washed with water into a flask, the filter-paper is added, and the whole treated with 50 C.C. of a 10 per cent. potassium iodide solution and 10 C.C. of hydrochloric acid. After standing for five minutes, with frequent shaking, the liberated iodine is titrated as usual. The method is applicable to all soluble barium salts, and is not affected by the presence of sodium salts, or of salts of magnesium, ammonium, or potassium, if their quantity does not exceed 0.5 gram. If larger quantities of these salts are present, precipitation is carried out in a hot solution, which is well stirred until cold. By proceeding in this way, occlusion of other iodates is avoided, but the barium iodate obtained does not dissolve so easily in hydrochloric acid as that precipitated in the cold.Calcium, strontium, and heavy metals must be absent. The limits of error of the method are given as 0.05 to 047 per cent., reckoned on the barium. A. G. L. Precipitation of Cobalt by Means of Potassium Nitrite. L. L. d e Koninck. (BUZZ. SOC. Chinz. Belg., 1909, 23, 11-13.)-1n precipitating cobalt a s potassium cobalt nitrite, the solution should be acidified before precipitation with nitric acid, according to the author, not with acetic acid, as is usual. The addition of nitric acid insures complete oxidation of the cobalt to the trivalent state. Suffi- cient potassium nitrite must be added to neutralise the free acid. A. G. L. Simplification of the Gravimetric Method of Estimating Halogens by Means of Silver Nitrate.E. Alefeld. (geeits. anal. Chem., 1909,48,79, 80.)- If 5 C.C. of ether are added to the solution of the halogen to be estimated, filtration of the liquid may be carried out immediately an excess of silver nitrate has been added, without heating or waiting for the precipitate to settle completely. The Gooch crucible used as filter should also be moistened with ether before filtration. The filtrate obtained is absolutely clear. A. G. L. Gravimetric Estimation of Nitric Acid. A. Hes. (Zeits. anal. Clze?n., 1909, 48, 81-98.)-The author’s experiments show that Busch’s nitron method forTHE ANALYST, 115 nitric acid yields accurate results (maximum error, 0.2 per cent.), if the solution contains about 0.1 per cent.of nitric acid. Ordinary weighed filter-papers may be used instead of Gooch or Neubauer crucibles. Solutions much more dilute than the above are not, or only incompletely, precipitated by nitron, but such solutions may be concentrated after the addition of the reagent. Dextrine and gelatine prevent the precipitation of nitron nitrate. The presence of aluminium sulphate, magnesium sulphate, ammonium sulphate, potassium phosphate, magnesium chloride, tartaric acid, citric acid, saccharose, and glucose does not interfere with the precipitation, but, if large quantities of these substances are present, the 10 C.C. of ice-cold water usually used may not suffice to wash the precipitate completely. I n presence of oxalic acid, high and very irregular results are obtained.Chloric acid is quanti- tatively precipitated by nitron from solutions containing about 0.25 per cent. of chloric acid, and the method can be used for the estimation of chloric acid, or of chloric and nitric acids together. Hydrobromic acid, if present, must be removed before precipitation by means of chlorine. Hydrogen iodide is destroyed by boiling with potassium iodate and acetic acid. Nitrous acid is removed by adding powdered hydrazine sulphate to the cooled liquid, or may be oxidised with hydrogen peroxide and sulphuric acid at 70" C. to nitric acid. Chromic acid is reduced with hydrazine sulphate. A. G. L. Action of Nitrogen on Commercial Barium Carbide. 0. Kuhling and 0. Berkold. (Zeits. Angew. Chem., 1909, 32, 193-196.)-Carbon present as carbide in barium and calcium cyanamide mixtures is estimated by evolution as acetylene, alcohol saturated with acetylene being used to decompose the substance, as, if water is used, explosions may occur.Twenty grams of the sample are treated first with 50 C.C. of 96 per cent. alcohol, and then with 100 C.C. of 80 per cent. alcohol. Cyanamide is estimated as previously described (Bericlzte, 1907, 40, 314). To estimate cyanide, 0.5 to 0.6 gram of the sample are distilled with 30 C.C. of a saturated solution of lead nitrate and 80 to 100 C.C. of water, the distillate being received in 25 C.C. of sodium hydroxide solution. An excess of ammoniacal silver nitrate solution is added to the distillate, and the silver acetylide filtered off after ten minutes' standing. The filtrate is gradually mixed in the cold with an excess of nitric acid, and the precipitated silver cyanide collected on a Gooch crucible, washed with cold water, and dried at, 100" to 110" C.This method was tested on solutions con- taining known amounts of cyanide and chloride, and proved accurate, the maximum error being 0.19 per cent. on the cyanide. A. G. L. Quantitative Estimation of Silver. F. Mawrow and G. Mollow. (Zeds. Anorg. Chem., 1909, 61 , 96-99.) - Hypophosphorous acid reduces solutions of silver nitrate instantly and completely on boiling, metallic silver, not silver hydride, being obtained. The precipitate is so finely divided that it passes through the filter on washing with water ; but it may be washed with 96 per cent. alcohol without loss. The results are accurate even in the presence of acetic and tartaric acids.Lead salts, treated in the same way, yield a precipitate of lead hypophosphite, difficultly soluble in cold water, more readily in hot water, and insoluble in alcohol. A. G. L.116 THE ANALYST, Colorimetric Estimation of Dissolved Oxygen in Water. G. B. Frank- forter, G. W. Walker, and A. D. Wilhoit. (Joz~rn. Anzcr. Chewz. SOC., 1909, 31, 35-43.)-The method depends on adding 2 C.C. of a solution of cuprous ammonium f- I colorinieter K, of 102 C.C. Capacity, is connected to G chloride to 100 C.C. of the sample of water, and com- paring the colour obtained with thatgiven by a known amount of cupric ammo- nium chloride suitably diluted. Air must be ex- cluded during the opera- tion. The process i s carried out in the ap- paratus shown : The bulb C contains copper wire introduced through the wide stop-cock B ; the bulb C is completely filled with a solution of cuprous chloride in hydrochloric acid. Burette A contains hydrochloric acid, and burette E ammonia.F is a bulb of 10 C.C. capacity, in which the reagent is prepared; it is initially filled with oxygen - free water from the inverted bottle H through the three-way tap G. The by a rubber tube leading to the tap I. The cork of the colorimeter carries a bent open tube M, and a short wide glass tube L, the lower end of which is closed by cementing on a glass disc. The colour of the liquid is observed through this tube L. K represents the com- parison colorimetier. In making a test, the bulb F is first filled with the reagent, by running in 2 C.C.of the liquid in C, after opening B, and then adding 8 C.C. of ammonia from E. The colorimeter K and the inverted bottle H, now containing the sample of water to be tested, are then connected to the apparatus, and the tube N, which reaches to the bottom of the bottle H, is joined to a hydrogen generator. The colorimeter is completely filled with the water, and 2 C.C. of the mixed reagent are added. For this purpose, the taps G, D, and B are manipulated so that 0-5 C.C. of liquid flows from C into F, and 1.5 C.C. of ammonia.from E into F. If the operation is carried out quickly, only water runs out of the bent tube M, and the colorimeter contains 1CO C.C. of water and 2 C.C. of reagent. Its colour is compared with that produced inTHE ANALYST.117 the second colorimeter by a known quantityof a solution of cupric chloride (contain- ing 1.1364 grams of copper per 1 litre of water), 1 C.C. of which corresponds to 0.1 C.C. of oxygen, and which is made up to 100 C.C. with ammonia and water. The standard is made up somewhat darker than the sample, and liquid drawn off through the stop-cock I’ until the tints are matched. The method is rapid; for waters con- taining up to 7 e . ~ . of dissolved oxygen per litre, the results obtained are generally about 0.1 C.C. higher than those given by Thresh’s method.* A. G. L. Estimation of Sulphide in Alkali Cyanide. T. Ewan ( J o z ~ m . SOC. Chm., I d . , 1909, 28, 10-13.)-0f the three methods studied, that of titration with lead nitrate, with lead paper as an external indicator, is the most rapid and convenient, and, if suitably performed and corrected, it is sufficiently accurate for technical purposes with quantities of potassium sulphide below 0.1 per cent. G.W. Williams’s modification of the colorimetric process (with sodium hydroxide and lead acetate) is more accurate but less rapid, and it is necessary to use boiled water and to check the strength of the standard solution very frequently. Lastly, Feld’s method, depending on the expulsion of sulphuretted hydrogen by distilling with magnesium chloride in an atmosphere of carbon dioxide, is the most accurate, but too laborious for technical purposes. The author describes the following procedure for performing the titration method with lead nitrate : 10 grams of the powdered cyanide are stirred quickly with 15 C.C.of ordinary water, any lumps being broken up with a glass rod. Any small undissolved residue will generally go into solution before the titration is finished. The lead nitrate solution (1 C.C. equivalent to 1 mgm. of potassium sulphide) is then run in quickly so long as it produces an increase in colour; after this it is added gradually, the solution being tested by placing drops of it in contact with drops of lead nitrate solution on thick blotting-paper. The author has carefully investigated the errors involved in the lead titration method, due to oxygen in the water, oxidation whilst the substance is being dissolved, and oxidation during titration. These errors are very considerable ; but if the procedure described above be adopted, one and a half to two and a half minutes being required for dissolving, and three to four minutes for titrating, the errors are sufficiently constant at from 18 to 26 per cent.of the sulphide present. If then the results be multiplied by the factor 1.25, or if, as is more convenient, a solution containing 2.4 grams of lead nitrate per litre be assumed to correspond with 0.01 per cent. of potassium sulphide per 1 C.C. used, the results are accurate to 2 or 3 units in the third place of decimals, and compare favourably with those obtained by Williams’s method. J. F. B. On the Titration of Sulphites with Permanganate. J. Milbauer (Zeits. ai2n.l. CZienz., 1909, 48, 17-24.)-The results of the experiments recorded show that correct results can only be obtained when the sulphite solution is added to a large excess of the permanganate solution acidified previously with sulphuric acid, and the excess of the permanganate then titrated.Ten times the quantity of per- manganate solution actually required for the oxidation of the sulphite is acidified with concentrated sulphuric acid (the amount added being such that the mixture * Dae probably to oxygen dissolved in the solution of ammonia.118 THE ANALYST. contains 20 per cent. by volume of sulphuric acid), and a very dilute solution of the sulphite is then added. The excess of permanganate is next titrated back with standardised oxalic acid, ferrous sulphate, or hydrogen peroxide solution. The excess of permanganate cannot be titrated with thiosulphate solution after the addition of potassium iodide.The permanganate solution should be of such strength that each C.C. is equivalent to 0.001 gram of sulphur dioxide. The addition of a mere trace of a catalytic agent, such as bromine or iodine, quickens the rate of the oxidation and renders the reaction more complete at low concentrations. w. P. s. Estimation of Sulphuric Acid as Barium Sulphate. E. Ruppin (Chen~. Zeit., 1909, 33, 17-18.)-The author quotes experiments to show that Treadwell's method, in which the boiling acid solution of the sulphate is precipitated with a dilute solution of barium chloride, added drop by drop, leads t o low results, only 96 per cent. of the theoretical amount being obtained. This is due to the occlusion of sulphates other than barium sulphate, especially of calcium and potassium.Correct results are obtained by modifying Hintz and Weber's procedure, a 50 per cent. excess of a 10 per cent. barium chloride solution being at once added to the boiling acid solution of the sulphate ; the precipitate should be repeatedly boiled out with water containing hydrochloric acid. Correct results are also obtained by pre- cipitating in a boiling solution made alkaline with sodium bicarbonate, acidifying with hydrochloric acid, and boiling for fifteen minutes. A. G. L. Estimation of Tin in Tin-plate. K. Meyer. (Zeits. mzgcw. Clze~n., 1909, 22, 68-69.)-Twenty to fifty grams of the tin-plate are freed from fat and dirt, cut into pieces of not more than 50 square cni. area, placed in a large dish or flask, and covered with water.After heating to 80" C., 1 gram of sodium peroxide for every 20 grams of sample is stirred into the liquid and the heating continued. As soon as the violent evolution of gas has ceased, another 1 grain of sodium peroxide is added and the liquid boiled. If any tin is still visible, a little more sodium peroxide is added, and the boiling repeated. The residual pieces of iron are removed from the liquid, washed first with water, then with alcohol, dried and weighed. The loss in weight corresponds to the tin present, not more than 0.009 per cent. of iron being removed. If lead and zinc are present, they will also pass into solution; in this case the dissolved tin must be actually estimated. A. G. I;. A New Reaction Distinguishing between Salts of Uranium and of Cadmium.P. Lemaire. (Ann. de Chim. 9naZ. Appl., 1909, 14, 6, 7.)-The reagent consists of a 5 per cent. solution of thiosinamine rendered alkaline, just before use, by the addition of 5 C.C. of 60 per cent. sodium hydroxide per 100 C.C. On adding to 2 or 3 C.C. of this alkaline reagent a few drops of the liquid under examination, and boiling and shaking the mixture, a permanent yellow precipitate is obtained in the presence of salts of cadmium or uranium. No yellow precipitate is obtained under the same conditions with arsenites, arsenates, or salts of lead, bismuth, antimony, copper, silver, mercury, tin, chromium, zinc, aluminium, iron, manganese, nickel, cobalt, alkalies or alkaline earth metals. When a yellow precipitate has been obtained, its nature is readily determined by the fact thatTHE ANALYST.119 hydrogen sulphide precipitates cadmium from a hydrochloric acid solution, whilst ammonium hydrosulphide precipitates uranyl salts, but not cadmium. In the case of mixtures of cadmium or uranium salts with other compounds, a preliminary separa- tion of the bases is usually necessary, since t.he thiosinamine precipitate may then be brown or black. C. A. M. Estimation of Zinc by Means of Ferrocyanide. E. Rupp. (Chern. Zeit., 1909, 33, 3, 4.)-The author shows that oxidation of potassium ferrocyanide to ferri- cyanide is effected quantitatively by allowing an excess of iodine to act on the neutral solution, to which 5 grams of Rochelle salt are added for a voluine of 50 to 100 C.C. At the end of one hour the excess of iodine is titrated with standard thiosulphate solution. To apply the method to the estimation of zinc, 10 C.C. of a neutral solution of zinc-sulphate, containing about 0-1 gram zinc, are added to 20 C.C. of a standard 5 per cent. potassium ferrocyanide solution after diluting the latter with 50 C.C. of water and adding 5 grams of Rochelle salt. The mixed solutions are allowed to stand for twenty to thirty minutes, to insure complete conversion of the zinc ferro- cyanide first precipitated into potassium zinc ferrocyanide, lC,Zn,Fe,C,,N,, ; 25 C.C. of a solution of iodine are then added, and the excess of iodine is estimated after another hour’s standing. Since iodine has no action on the precipitate, 2 atoms iodine correspond to 3 atoms zinc. The results vary between 99.9 and 100-2 per cent. of the zinc (Cf. ANALYST, 1902, 27, 291). A. G. L.