by Weight and by Volume. 



213 



Quantitative Ratio of Bichromate of Potash and Metallic Bismuth. 



In adopting centigrade methods of analysis, it becomes a 

 matter of vital importance to determine with the greatest accu- 

 racy the atomic weight of the body to be estimated, and of the 

 agent by which its determination is to be effected. The metal 

 to be estimated by the method now described is bismuth. If we 

 adopt with Gmelin the formula BiO^ for the oxide of bismuth, 

 we find that the following atomic weights have been assigned to 

 the metal :—Berzelius, 213; Kane, 213-3 ; Graham, 213-21: 

 Gmelin gives 210, or even less, as the atomic weight of bismuth. 

 To the metal chromium, which of course materially affects that 

 of bichromate of potash, the following atomic weights have been 

 assigned :— 26, 26-24, 27, and 28. The following Table gives 

 the quantities of bichromate of potash that would be required to 

 combine with 100 parts of bismuth, according to the various 

 atomic weights that have been assigned to the metals chromium 

 and bismuth. Potassium is taken at 39. 



Upon examining the above Table, it will be found that the 

 maximum quantity of bichromate of potash required for 100 

 parts of bismuth is 71-90, and the minimum quantity 68-10, a 

 difference too great to be attributed to experimental error when 

 pure materials are employed. No correct data being deducible 

 from the atomic weights given, it became necessary to ascertain 

 by experiment the exact quantity of bichromate of potash required 

 to combine with a known amount of bismuth. These data I ob- 

 tained by dissolving a certain quantity of pure bismuth in nitric 

 acid, and after removing excess of acid, I ascertained the quan- 

 tities required to combine in two or three ways ; first by addition 

 of bichromate of potash to the solution of bismuth until an excess 

 of bichromate gave rise to a yellow colour ; after which I applied 

 the method inversely, adding a solution of nitrate of bismuth of 

 known strength to a given volume of bichromate of potash until 

 the loss of colour indicated complete saturation of the solution. 

 From the weight of the precipitated chromate of bismuth I also 

 calculated the bichromate of potash that would be required. 



The mean of all my experiments gave 71-35 parts of bichro- 

 mate of potash as the quantity required to combine with 100 



