246 THE ANALYST. INORGANIC ANALYSIS. The Electrolytic Estimation of Bismuth. F. J. Metzger and H. T. Beans. { Jozcm. Anzer. Chem. Soc., 1908, 30, 589-593.)-Bismuth can be coiupletely pre- cipitated on a rotating cathode in an adherent form from a solution of the nitrate if fhe latter is first made alkaline to phenolphthalein with sodium hydroxide, and then acid again with sufficient acetic acid to give a practically clear solution. The further addition of boric acid yields a still better deposit. The conditions used were : Acetic acid, 20 C.C. of 50 per cent. acid; boric acid, 2 grams; dilution, 250 c.c.; current, N.D.,, = 0.2 to 0.16 amp6re a t 1.8 to 2-8 volts ; temperature, 63' to 78" C. ; bismuth, 0.1 grain ; time, one and a quarter hours. For 0.4 gram bismuth, 30 C.C.of acetic acid were used, and the time required was two and a quarter hours with the same cathode, which was shortened to one and a half hours when a cathode with a surface of 100 sy. cm. was employed. The cathode made 700 revolutions per minute, A. G. L.THE ANALYST. 247 The Analysis of Nearly Pure Gases. H. Franzen. (Zeits. hzorg. Clzem., 1908, 57, 395-397.) -Considerable quantities of impurities present in nearly pure and difficultly absorbable gases (c.g., com- pressed oxygen) are conveniently collected by means of the apparatus shown in the figure. I t consists of a 150 C.C. gas-sample tube provided with two taps, and a mark at 50 c.c., reckoned from the lower tap, and a Drechsel wash-bottle containing water. The sample tube is first filled with water.Gas is next passed through the wash-bottle until all air has been expelled, whan the sample tube is connected with the wash- bottle by means of a rubber tube, as shown. Both taps being opened, gas is passed into the sample tube to the mark. Both taps are then shut, a reservoir containing water is connected to the lower tap, which is then opened, and the residual 50 C.C. of water are sucked out. A small quantity of an absorbing reagent (e.g., sodium hyposulphite in the case of oxygen) is then introduced into the sample tube from below as usual, after which the lower tap is shut, the upper one opened, and gas allowed to flow into the sample tube, which is meanwhile vigorously shaken. When no more gas is absorbed, a fresh quantity of reagent is introduced, and this procedure repeated until the sample tube is completely filled with the unabsorbable impurities. A.G. L. A Separation of Iron from Manganese. R. B. Moore and I. Miller. (Jouriz. ilmer. Chem. SOC., 1908, 30, 593-594.)-A good separation of iron from manganese can be effected by adding to a fiolution of the chlorides containing a little free hydrochloric acid a sinall excess (0.5 c.c.) of pyridine. The solution may be gently warmed; the precipitate should be filtered off at once, and washed with pyridine water (1 : 500). The manganese in the filtrate cannot be precipitated with bromine and ammonia; it may be recovered by means of potassium carbonate or hydroxide. I n separating about 0.1 gram iron from its own or twice its weight of manganese, the error amounted to 0.5 mgm.at the most after a single separation. A. G. L. Methods for the Detection of Lead Dust and Fumes i n Works. Heim and Herbert. (Zeits. Gezoerbe-Hyg., 190 8, 15, 82-85 ; through Chem. .&it. Rep., 1908, 32, 193.)-Of the usual methods €or the detection of lead, the sulphide reaction is the most delicate, a reaction being obtained with 0.1 to 0.2 mgm. of lead, provided a layer of liquid 10 cin. deep be viewed from above. This method is, of course, inapplicable in the presence of other heavy metals. Trillat therefore propose3 t o248 THE ANALYST. utilise the colour developed by the action of peroxides on tetramethyldiamidodiphenyl- methane, the corresponding carbinol [C,H,N( CH,),],CH-OH being produced. The reagent is prepared by warming together for one hour 30 grams of dimethylaniline, 10 grams of formaldehyde, and 200 C.C.of water, acidified with 10 grams of sulphuric acid. An excess of soda is then added, and the unchanged dimethylaniline romoved by steam distillation. On cooling, the reagent crystallises out ; it is recrystallised from alcohol and dissolved in nitric acid (5 grams per 100 c.c.). The solution should be kept in the dark. In testing for lead, the substance is treated with sodium hypochlorite, the chlorine removed, and the reagent added. The air is led through cotton-wool and then through d p h u r i c acid. The cotton-wool is ashed after moistening with sulphuric acid, the sulphuric acid is evaporated to dryness, and both residue$ are tested as above. The method is stated to be exceedingly delicate.A, G. L. Purification of Mercury. W. Bettel. (Chenz. News, 1908, 97, 158.)-The author finds that gold and the base metals present as impurities in metallic mercury may readily be removed by a weak cyanide solution, the purification being hastened by the addition of sodium peroxide to oxidise the base metals. I n one experiment 7 kilogms. of impure mercury were treated in a shallow, flat-bottomed receptacle with 1,500 C.C. of a 3 per cent. solution of potassium cyanide, 20 grams of sodium peroxide being added at intervals and the solution occasionally agitated. During the first four days the base metals were dissolved and the gold contents remained practically stationary, but on the fifth day the amount of gold dropped from 150 to 90 mgms. per kilogm.of mercury. On the sixth day, 12 mgms. ; seventh, 6 mgms. ; eighth day, 5 mgms. of gold per kilogm. were found. The cyanide solution was then removed, and replaced by a fresh 0.2 per cent. solution. On the fourteenth day of treatment the gold remaining in the mercury was less than mgm. per kilogm., and the metal thus purified was superior to the vacuum-distilled product. As there is practically no loss of mercury, the cost of the operation is very small. A. R. T. The Detection of Nickel in Ores and Nickel Steel. H. Grossmann. (Chenz. Zeit., 1908, 32, 315-316.)---The author criticises the method of Pozzi-Escot (ANALYST, 1907, 32, 432; 1908, 106), which he regards as unsuitable for the estimation of nickel in New Caledonian ores. For its detection in ores and nickel steel he gives the following modification of his dicyandiamidine method (ANALYST, 1907, 32, 273, 394).The filtrate from the metals of the hydrogen sulphide group is concentrated to a small volume, freed from too large an excess of mineral acids, and treated with Rochelle salt in the proportion of at least 4 grams to 1 gram of the mineral or alloy, The solution is then rendered strongly alkaline with ammonia, and treated with about 1 gram of dicyandiamidine sulphate and with potassium hydroxide. The nickel separates immediately, or, if present only in sinall quantity, after a short time. If R large amount of manganese is present, it is best to treat the liquid with hydrazine sulphate, which prevents the oxidation of the amrnoniacal tartrate solution and separation of higher manganese oxides.C. A. M.THE ANALYSTo 249 The Gravimetric Estimation of Tellurium. V. Lenher and A. W. Hom- berger. (Jozmn. Anzer. Chem. SOC., 1908, 30, 387-391.)-The authors find that tellurium can be quickly and accurately estimated by heating a fairly concentrated solution of either oxide, containing 10 per cent. of hydrochloric acid, to boiling, adding 15 C.C. of a saturated solution of sulphur dioxide, then 10 C.C. of a 15 per cent. solution of hydrazine hydrochloride, and then another 25 C.C. of the solution of sulphur dioxide, The solution is boiled until the precipitate settles in a form suitable for washing, which should take about five minutes. The precipitate is then filtered off on a Gooch crucible, washed first with hot water, then with alcohol, dried at 105" C., and weighed.A. G. L. Quantitative Separation of Thallium from Silver. J. F. Spencer and Margaret Le Pla. (Proc. Chem. Suc., 1908, 24, 75.)-A quantitative separation of the salts of silver and thallium is effected by treating the mixture with a stream of chlorine, whereby the thallium is converted into the very soluble thaliic chloride and the .silver is precipitated as silver chloride. After removing the latter, the solution is concentrated, and reduced by means of sulphur dioxide (either liquid or gaseous). The excess of sulphur dioxide is removed by boiling, and the thallium precipitated as thallous iodide, filtered, washed, and dried at 104". The results are quantitative, and are obtained very quickly. Thallous chloride was found to be more soluble in potassium carbonate than in water, for, whereas water dissolves 3-86 grams per litre at 25", a 5 N-solution of potassium carbonate dissolves 21.84 grams per litre at the same temperature.Volumetric Estimation of Titanium. H. D. Newton. (Zeits. Anoiy. CJzeiiz., 1908,57,278-280.)-The titanium solution, containing about 10 per cent. of sulphuric acid, is reduced by warming with zinc in a 100-c.c. flask, closed by a stopper carrying an inlet-tube and a small separating funnel. During the reduction and subsequent cooling a current of hydrogen is led through the flask. An excess of a solution of ferric sulphate is then added, the flask is completely filled with cold water, and its contents are transferred to a larger flask in which the ferrous iron produced is titrated with potassium permangmate solution.The zinc used must be free from iron, or else contain a known amount of this metal. The test results quoted are very good. A. G. L. Notes on Wijs' Solution. H. Ingle. ( J o z L ~ . SOC. Chenz. ITzd., 1908, 27, 314.)-In reply to Harvey's objections (ANALYST, 1903, 28, 41) to the method of procedure proposed by the author, the latter states that no advantage is t o be gained by adding water to the blank test before the potassium iodide! as there are no iodochloridea of unsaturated compounds present. Iodine chloride being soluble in both chloroform and water, it is not removed from the sphere of action of either water or potassium iodide. The loss of titre which takes place in this case is due to the decomposition of the hypoiodous acid formed by the action of the water on the iodine chloride.(1) IC1+ H,O = HI0 + HCI, and (2) 4HIO = 2H20 + 21, + 0,. The reactions which take place may be represented thus :250 THE ANALYST. The first is a reversible reaction. If the amount of hydrochloric acid present in the solution is increased less hypoiodous acid is formed, and consequently there is less tendency for the second reaction to take place. If Wijs’ solution and water be mixed in equal volumes, the solution gradually becomes redder in colour, owing to the separation of iodine, and bubbles of gas rise to the surface. Harvey’s contention that the reduction in the titre in the blank test may be due to substitution in the acetic acid by the action of the free iodine and iodic acid does not appear to the author to be sound, for it is impossible for iodic acid to exist in the presence of hydrochloric or hydriodic acid.w. P. s. The Preparation of Conduetivity Water. H. Hartley, N. P. Campbell, and R. H. Poole. (Trans. Chenz. Soc., 1908, 93, 428-431.)-With the still described below, the authors have found no difficulty in preparing 2.5 litres of water with a specific conductivity not greater than 0.75 gemmho (1 gemmho = 1.0 x 1 0 - 6 reciprocal ohms), and 3 more litres of fairly good water of 1.5 gemmhos, in a, single distilla- C - -71 -T A tion, without any special precautions for keeping the air of the laboratory free from impurities. The boiling vessel ( A ) is of copper or tinned iron, holding about 10 litres. This is filled with ordinary distilled water (of approximately 5 gem- mhos).From A the steam passes through the glass trap (B), where any dust is stopped by the glass wool (W), and any con- densed water escapes at P. The steam then passes into the condensing vessel (C), which is made of the best tinned iron. The con- denser tube ( T ) is made of block tin and is soldered through the lid. I t is cooled internally by a stream of tap water. P represents baffle plates. Water condenses on T, and the small tin cover prevents any water, except that con- densed on the block tin surface, from dropping into the tin funnel and delivery tube (E). The waste water runs away through R. A fine capillary tube is attached to both D and 12, to prevent steam blowing off. The water is collected in a steamed-out Jena flask,THE ANALYST.251 holding 3 litres. The tin delivery tube passes through a piece of Jena tubing in the rubber stopper of the flask, a small piece of rubber tubing making an air-tight joint. The water is drawn off through a siphon of Jena glass tubing, and connection between the flask and the air is made through the soda-lime tube (8). The small flask ( K ) contains nioist glass wool, to prevent any soda-lime dust from being carried over into the large flask. For the first half-hour after boiling begins the water collected is impure, and is rejected. The flask ( F ) is then attached to the delivery tube, and the water is collected for two hours. The two litres obtained serve to wash out the flask, and are then run off, and during the next three hours the best water collects.This usually amounts to about 2.5 litres, having K,, = 0.75 gemmho. The flask is then removed, the small tube is quickly closed with a rubber plug, and the water may be kept ready for use without its conductivity increasing to any extent. By putting another flask under the delivery tube at this stage, another litre may be obtained with Kls= 1 gemmho. w. P. s. Volumetric Estimation of Zinc. W. H. Keen. (JOZ~Y~Z. Amer. Clzem. Soc., 1908, 30, 225233.) -In titrating zinc by potassium ferrocysnide the author standardises the latter solution on varying quantities of zinc, since a slightly different factor is obtained according to the amount of zinc present. A " blank " experiment should also be made with the ferrocyanide solution, about 0.20 C.C.being generally consumed. All titrations are carried out on a, volume of about 150 C.C. of solution at a temperature of 85' C. Zinc in Ores.-After the removal of lead, silica, copper, and iron, the titration of the zinc may be carried out directly in the absence of manganese and nickel. If manganese be present, it may be got rid of (together with the iron if in small quantity only) by adding to the acid filtrate from the hydrogen sulphide separation, after evaporation, 5 grams of potassium or sodium bromide, rendering the liquid strongly ammoniacal, and stirring for one hour, or allowing to stand overnight. The precipitate should be washed with ammonia, and the manganese determined by dissolving the precipitate in sulphurous acid solution, adding nitric acid and expelling any nitrous fumes, and the estimation completed by the bismuthate process.If manganese be present in the ore in considerable quantity, or if nickel be present, the best plan is first to precipitate the zinc as sulphide. The precipitate is dissolved in hydrochloric acid, and hydrogen sulphide passed through the solution to ensure the removal of all the copper, Any occluded iron is removed from the filtrate, which is then neutralised, and 3 C.C. of hydrochloric acid added in excess, and titrated with ferrocyanide as usual. Analysis of Brasses. --After the removal of tin, copper, and lead, the evaporated filtrate is taken up in hydrochloric acid, the iron removed, and the zinc in the filtrate titrated with ferrocyanide. [In the case of manganese bronzes a similar method may be followed, but the manganese must be separated by the alkali-metal bromide process, vide supra.] Alternatively, after the removal of tin, copper, and lead, the filtrate is rendered ammoniacal, colourless ammonium sulphide added, and the liquid kept hot for an hour. The zinc sulphide is then redissolved in hydrochloric252 THE ANALYST. acid, any copper present removed by hydrogen sulphide, and, after precipitating the iron from the filtrate, the ferrocyanide titration is proceeded with. .Zinc in Alzhminium Alloys.-In the absence of iron the titration may be m d e directly, since aluminium does not interfere. Iron is, however, generally present in considerable quantities, in which case 1 gram of the sample is dissolved in nitric or hydrochloric acid, a little citric acid added, followed by ammonia and then ammonium sulphide to the boiling liquid. The precipitate is allowed to settle, the liquid filtered, the last traces of citric acid being carefully washed out with a wash- water containing a little ammonium sulphide. The precipitate, which contains the zinc and any copper and iron originally present, is treated as described under “Brasses.” Or the metal may be dissolved in hydrochloric acid, and the zinc precipitated from a slightly acid solution, as directed in the separation from nickel in ores. A. R. T.