42 



to test paper ; boil, allow the precipitate settle, decant the clear fluid 

 on a filter, wash with 50 c. c. of hot water, by decantauon, then trans- 

 fer the precipitate to the filter and wash well with hot water, until all 

 the chlorides are removed : testing a few drops of the wash -water with 

 argentic nitrate. 



Evaporate the filtrate to about 75 c. c., and add 25 c. c. of a saturated 

 solution of ammonium carbonate, sufficient to precipitate the excess of 

 barium present; boil, allow the precipitate to settle, decant the clear 

 fluid, wash by decantation with 50 c. c. hot water, transfer the precipi- 

 tate to a filter, and wash with hot water until all the chlorides are re- 

 moved. 



Evaporate this filtrate to dryness in a platinum dish, and, when dry, 

 drive off the ammonium chloride at a low red heat. Cool, take up with 

 water, filter, and wash well to remove the carbonaceous matter, and 

 test the fluid with a few drops of the barium hydrate solution ; if this 

 produces a precipitate, add 10 c. c. more, or until the barium produces 

 no further precipitate ; filter off the precipitate, and repeat the treatment 

 with ammonium carbonate. 



Finally evaporate the solution to dryness in a weighed platinum dish 

 after adding a few drops of hydrochloric acid, expel any ammonium 

 chloride present at a low red heat, cool, and weigh the chlorides of 

 potassium and sodium. 



DETERMINATION OF THE POTASH. 



Dissolve the mixed chlorides in 25 c. c. of warm water, filter, if nec- 

 essary, and transfer to a small lipped porcelain dish, add 2 c. c. of dilute 

 hydrochloric acid and 8 c. c. of platinum tetrachloride solution, pre- 

 pared by dissolving one part by weight of the platinum tetrachloride in 

 ten parts of distilled water, and evaporate to a pasty consistency on 

 the water-bath. Then pour into the disli about 50 c. c. of 85 per cent, 

 alcohol, without removing the dish from the bath, and heat for two or 

 three minutes. Care must be taken that there are no ammoniacal fumes 

 floating about in the air of the laboratory, as they would form a precipi- 

 tate with the platinum tetrachloride, thus increasing and vitiating the 

 result. Allow the precipitate to settle, and the fluid shows by its yel- 

 low color that a sufficient amount of the platinum tetrachloride has been 

 used; decant the clear fluid through a weighed Gooch crucible, transfer 

 the precipitate to the crucible, and wash well with 85 per cent, alcohol. 

 (Crookes, p. 2.) 



Dry the crucible and contents in the air-bath at 100 C., cool, and weigh 

 the potassium platinum chloride. This weight multiplied by 0.30559 

 will give the weight of potassium chloride present. The weight of the 

 potassium chloride is deducted from the weight of the mixed chlorides, 

 leaving the weight t>f the sodium chloride present. (Fres., 152, la, 

 97, 98.) 



