45 No. 'VI.-OY~ (x iV&od for Detecting Xmall Quantities of B73is~rufh. By M. M. PATTISON MUER,F.R.X.E. Assistant Lecturer on Chemistry The Owens College. 1.INvol. lxxxviii of Pogg e n d orf f '9 Animken page 45,there is a paper entitled "Researches upon Bismuth," by R. Schneider. The paper is mainly concerned with a detailed account of the author's investiga- tions upon an oxide of bismuth containing less oxygen than bismuthous oxide (Bi203). In describing one of the methods adopted for prepar-ing this oxide Xchneider mentions the fact that a solution of bismuth in nitric acid is precipitated by addition of tartaric acid but that the precipitate (tartrate of bismuth) so produced is soluble in caustic potash. He also shows that stannous chloride when mixed with excess of tartaric acid and then with such a quantity of potash as is sufficient to neutralise the acid yields a clear liquidwhich is unaltered on boiling or on dilution with water.He further shows that by mixing the prepared bismuth with the prepared tin solution, and adding an excess of potash a very dark brown liquid is obtained. Schneider proves that this liquid owes its colour to the presence of dissolved hypobismuthous oxide (Bi,O j contained in it. 2. As I have been lately and am still engaged with a study of the properties of this hypobismuthous oxide (I hope to be able to lay the results before the Society at an early date) it occurred to me that a colorimetric process for estimating small quantities of bismuth might be based upon the reaction described by Schneider.On trial how-ever I found that the reaction was not delicate enough to serve as the basis of a quantitative colorimetric method but I think we have in Schneider's reaction the means of detecting very small quantities of bismnth. 3. The fest liquid I prepare as follows:" 4 grams of stannous chloride are dissolved along with 12 grams of crjstallised tartaric acid in such a quantity of tolerably strong caustic potash solution as serves to produce a clear liquid having a distinctly alkaline reaction. This liquid must remain perfectly clear when it is warmed for some time to a temperature of 60" or 70". In order to prepare the solution which is supposed to contain bis- muth for the addition of the test liquid I add a considerable quantity of tartaric acid warm and then make alkaline by addition of potash.To t>his liquid I add a few C.C. of the test solution warm the mixture to 60" or 70" for a few minutes and place the beaker upon a piece of f Schneider's directions somewhat modified. 46 MCIR ON DETECTING SMALL QUANTITIES OF BISNUTH. white paper. The production of a brownish-black colour shows tlie presence of bismuth. 4. One part of bismuth in 210,000 parts of liquid may be thus detected. 5. I have examined the action of the stannous chloride solution upon solutions of metals other than bismuth prepared by adtling. tartaric acid to an acid or neutral solution of each metal followeJ by addition of excess of potash. The following are the results :-Cudmiurn yields a white precipitate without.any shade of brown or black. Mercury yields a grevish-black precipitate. Lead yields no precipitate nor is the liquid darkened. Copper yields a yelIowish-red precipitate. Arsenic yields 110 precipitate the liquid is not darkened. A?itimon!y yields no precipitate the liquid is not darkened. Irorc yields no precipitate the liquid is not darkaed. Cob& yields no precipitate the liquid is not darkened. Nickel yields no precipitate the liquid is not darkened. Chromium yields no precipitate the liquid is not cl'nrkeneci. Mangaaese yields a yellowish-red solution which gradually darkens. Mercury copper and manganesc can therefore alone interfere with the reaction the first-named metal if present must certainly be removed before applying the test although the others yield precipi- tates or coloured solutions yet this is only the case when they are present in comparabively large amount. 6. I had hoped to have perfected a volumetric method for deter-mining bismuth by precipitation as hypobismuthous oxide by means of the prepared tin solution employing mercuric:chloride as an indicator but as yet I have not succeeded. I would propose to call the prepared tin solution " Schneider's Reagent."