14 



REPORT — 1877. 



Precipitate. Factor = KOI per cent. 



Committee 3-2723 x -30560* 100-00 



Fresenius 3-2723 x -30560 100-00 



Frank and Berrand 3-2723 x -30507 99-83 



The Committee is informed that the factor -194 is adopted by some chemists 

 for calculating the chloroplatinate to anhydrous potash, a plan which would 

 cause the result of 100-52 of chloride of potassium to be obtained instead 

 of 100-00. 



With the view of testing the relative accuracy of the different modifications 

 of the platinum process when applied to the estimation of potassium in the 

 form of pure chloride, the following experiments were performed : — ■ 



The letters P and W refer to the mode of taking the required quantity of 

 chloride of potassium, P signifying pipette measurement and W the weighing of 

 the solution used. In the former case the percentage of chloride of potassium 

 was obtained by calculating the chloroplatinate precipitate to potassium 

 chloride, dividing the result by -6D7 and multiplying by 100. When a 

 weighed quantity of potassium chloride solution was employed, the following 

 equation was used for calculating the percentage of chloride of potassium 

 found (S is the weight of solution used, P that of the precipitate obtained) : — 



Px -3056x11x100 _ P X -3361 _percentage of 

 g " s = ~ KG found. 



Results bracketed together in the following tables were obtained from 

 experiments executed side by side. 



Table I. — Results of Experiments on pure Chloride of Potassium, using 

 considerable excess of Platinum Solution. 



These results, so far as they go, are decidedly in favour of Tatlock's method, 

 and conclusively prove that it is capable of great accuracy. 



At a later period of the investigations it was supposed that the exces- 

 sive results obtained by some of tho methods might be due to the fact that 

 a very considerable excess of platinum solution was employed — a condition 

 not in accordance with the directions of Fresenius and of Frank, but essen- 

 tial in Tatlock's method. The experiments made to elucidate this point did 

 not immediately succeed those already detailed, but it is convenient to record 

 the results hero rather than in another place. 



In the following experiments the quantity of platinum solution employed 



* This factor was adopted by Messrs, Chalmers and Tatlock as long ago as 1868. 



