70 IX.-Ort a Fourtk New Method for Estimating Bismuth Vohmetrically . By M. M. PATTISOK MUTR, F.R.S.E., Praelector in Chemidry, Caius College, Cambridge. 1. WHEN a saturated solution of oxalic acid is added to a solution of bismuth in nitric acid, a white crystalline precipitate slowly forms this precipitate, according to Souchay and Lenssen (Ann. Chern. Pharm., cv, 245) is normal bismuth oxalate, Bi,3C20a + ag. This reaction appeared to afford a mean whereby bismuth might be accurately and readily determined. I f the oxalate were produced under conditions which might be realised without too much difficulty, and if this oxalate were of constant composition, every requisite of an accurate volumetric process would be fulfilled. 2. I carried out several experiments, using a solution of bismuth in a small excess of nitric acid ; precipitating with a saturated solution of oxalic acid ; washing by decantation ; dissolving in warm dilute sulphuric or hydrochloric acid ; titrating with standard permanganate ; and calculating on the assumption that three molecules of oxalic acid found corresponded with two atoms of bismuth thrown down as oxa- late.The results were: not, however, very satisfactory, the amounts of bismuth found being generally too low, and the numbers exhibiting considerable discrepancies . 3. Sonchay and Lenssen (loc. cit.), have shown that the action of hot water upon normal bismuth oxalate results in the production of a basic oxalate having the composition Bi2O3.2C2O3 + aq. By taking advantage of the production of this basic oxalate I thought it would be possible to estimate bismuth accurately.To a measured volume of a solution of a weighed quantity of bis- mnthous oxide in nitric acid, an excess of a saturated solution of oxalic acid was added; the supernatant liquid was poured off from the, precipitate, which soon settled to the bott)om of t<he vessel ; the precipitate was boiled with successive small quantities of water, until the decanted liquid ceased to exhibit an acid reaction ; the residue was dissolved in dilute hydrochloric acid ; and standardised permanganate of potassium was run in-the liquid being maintained at a tempera- ture of 60" or so-until a permanent pale pink colour was produced. The results obtained were accurate, and showed no discrepancies. I quote a few :- This oxalate is stable ; it is only slightly soluble in nitric acid.MUIR ON ESTIMATING BISMUTH VOLUMETRIGALLY.71 (1.) 50 C.C. bismuth-solution = 0.28404 gram Bi : used 51 C.C. stan- dard permanganate = 0.17008 gra,m H2C,0a.2H20 = 0.28348 gram. Bi. difference = - .00056 gram. (2.) 50 C.C. bismuth-solution = 0.28404 gram. Bi.: used 50.7 C.C. permanganate = 0.28182 gram Bi. difference = - -00222 gram. (3.) 10 C.C. bismuth-solution = 0.056364 gram B i : used 10.2 C.C. permanganate = 0.056697 gram Bi. difEerence + *000333 gram. The calculation is based on the assumption of the accuracy of the formula assigned by Sonchny and Lenssen to the basic oxalate of bismuth. I have, by independent analysis, proved the accuracy of their formula. One molecule of oxalic acid found corresponds with one atom of bismuth.4. The solution containing bismuth must be free from hydro- chloric acid as the basic oxalate is readily soluble in that, acid. A large excess of nit.ric acid must also be avoided. Oxalic acid must be added in considerable excess. If the precipitate be thoroughly shaken up with the liquid, and the vessel be then set aside, the precipitate quickly settles, and the superna,tant liquid may be poured off through a filter in a very short time. I find that if the precipitate be boiled for 5 or 10 minutes with successive quantities of about 50 C.C. of water, it is quickly transformed into the basic salt. So soon as the supernatant liquid ceases t o show an acid reaction, the transformation is complete. It is well to employ a solution of permanganate so dilute that at least 50 C.C.are required for the titration. The basic oxalat2 may be dissolved in dilute sulphuric acid in place of hydrochloric ; it is more soluble, however, in the latter acid ; if the solution contains but little hydrochloric acid there is no danger of chlorine being evolved during the process of titration. 5. In applying this process to the estimation of bismuth in solution containing other metals, it is necessary if the solution contain sub- stnncescapable of acting upon, or of being acted on by permanganate, to separate the bismuth from the other metals present. This is easily done by precipitating in a partially neutralised solution with much warm water and a little ammonium chloride. The precipitate must be dissolved in nitric acid, and the liquid boiled down once or twice with addition of the same acid in order to expel all hydrochloric acid, before precipitating as oxalate.The liquid should contain just su&- cient nitric acid to prevent precipitation of the basic nitrate before oxalic acid is added. 6. The failure of those experiments in which I sought to determine bismuth after precipitation as normai oxalate, is evidently to be traced to the partial decomposition of the normal salt by the wash-water with the production of varying quantities of basic oxalate. My own 0 272 THOSlAS ON CUPROUS CHLORIDE AND THE experiments have convinced me that the normal oxalate very readily undergoes partial decomposition by contact even with cold water. 7. Of the various methods which I have devised for the volumetric estimation of bismuth, I regard the oxalate method as the most generally applicable. No special standard solution is required, as permanganate solution is in constant use where volumetric analysis is practised ; the process is simple and accurate, and may be quickly carried out : bis- muth is readily separated from other metals and brought into a con- dition suitable for estimation by this method.