ANALYSIS OF MINERAL WATERS. 269 



had been complete. To establish its accuracy with more cer- 

 tainty, the following experiments were also made. 



Twenty grains of muriate of soda (pure rock-salt), which had 

 been exposed to a red heat, and ten grains of crystallised mu- 

 riate of magnesia, were dissolved in an ounce of water, at the 

 temperature of 100°. The phosphate of soda and carbonate 

 of ammonia were then employed to precipitate the magnesia in 

 the mode proposed by Dr Wollaston, that is, a solution of 

 the ammoniacal carbonate was first added, and afterwards a so- 

 lution of phosphate of soda, as long as any precipitation was 

 produced, taking care to preserve in the liquor a slight excess 

 of the ammonia. The precipitate being washed and dried, af- 

 forded, after exposure to a red heat for an hour, 5.4 grains of 

 phosphate of magnesia, equivalent to 2.15 of magnesia. The 

 clear liquor being evaporated, muriate of soda was obtained, 

 which, after exposure to a red heat, weighed 25/7 grains. 

 Phosphate of magnesia being composed of 39.7 of magnesia, 

 with 60.3 of phosphoric acid, 5.4 grains of it are equivalent to 

 6.4 grains of muriate of soda, and this deducted from the quan- 

 tity obtained 25.7, leaves 19.3 as the quantity originally dissol- 

 ved. 



A solution perfectly the same was prepared, and a solution 

 of carbonate of ammonia was added to it as before. A strong- 

 solution of phosphoric acid was then dropped in, as long as 

 any precipitation was produced, observing the precaution of 

 having always an excess of ammoniacal carbonate in the li- 

 quor. The precipitate being washed and dried, afforded, after 

 exposure to a red heat, 5.5 grains of phosphate of magnesia 

 equivalent to 2.19 of magnesia. The clear liquor being evapo- 

 rated, and the dry matter being exposed to a heat gradually 

 raised to redness, weighed, when cold, exactly 20 grains. 



In 



