1824.] Mineral Waters of Carlsbad. 131 



neously and with effervescence, it is probable that the precipi 

 tate contained no silica. When the filiate of lime had been 

 fully decomposed, the residual saline mass was boiled in as 

 much water as was sufficient to take up the whole of the sul- 

 phate of lime. The solution, mixed with ammonia, gave a 

 yellow coloured precipitate resembling oxide of iron, and weigh- 

 ing after ignition O06 gramme. 



c. The solution, separated from the above precipate, was de- 

 composed by oxalate of ammonia. The oxalate of lime thus 

 formed, left, after calcination, 0-127 gramme of carbonate of 

 lime, equivalent to 0-099 gramme of fluate of lime. 



d. The oxide of iron from b, was dissolved in muriatic acid ; 

 a white matter, weighing O'OOl gramme, remained undissolved, 

 which, when heated with an alkali on charcoal before the blow- 

 pipe, was converted into a globule of tin. The solution was 

 combined almost to saturation with sal ammoniac,* and triple 

 prussiate of potash was added, until the whole of the oxide of 

 iron was precipitated. The whole was then filtered, and the 

 precipitate was washed with a solution of sal ammoniac. The 

 filtered liquid, mixed with ammonia, gave a white flocky pre- 

 cipitate, weighing after ignition 0015 gramme. This was dis- 

 solved in muriatic acid, and the solution was mixed with an 

 excess of caustic potash. 0*006 gramme phosphate of lime 

 precipitated. What remained dissolved in the alkali was sepa- 

 rated by saturation with muriatic acid, and by the subsequent 

 addition of ammonia. It fell as a white precipitate, which, 

 however, gradually became pale amber coloured on being dried. 

 Before the blowpipe, nitrate of cobalt developed in it a deep 

 but rather impure blue colour, with carbonate of soda on the 

 platinum foil it indicated traces of manganese, and with boracic 

 acid and iron it yielded a fused button of phosphuret of iron. 

 It consisted therefore of subphosphate of alumina, containing 

 traces of phosphate of manganese. The liquid from which the 

 subphosphate of alumina and phosphate of lime had been preci- 

 pitated, being mixed with lime water, gave 0"003 gramme of 

 phosphate of lime, whose acid (0*00135 gramme) must have 

 been combined with oxide of iron. Subtracting; this along- with 

 the weight of the other substances separated in d, from the 

 0*06 gramme in b, we obtain for the quantity of oxide of iron 

 0-0420 gramme. The sum of the weights of all these sub- 

 stances corresponds almost exactly with the quantity originally 

 submitted to analysis : an additional proof that the fluoric acid 

 existed in the precipitate uncombined with silica. Had the 

 contrary happened, a considerable loss would have been sus- 



" Sal ammoniac was added, because a liquid containing ;in excess of the triple prus- 

 siate has the property of dissolving a considerable quantity of the blue precipitate ; but 

 tlii^ is prevented by the presence of the dissolved salt. 



k2 



