1824.] Mineral Waters of Carlsbad. 133 



radiated crystals, which was not deliquescent ; these were 

 somewhat soluble in alcohol, and cotton, moistened with the 

 solution, burned with a red coloured flame ; and finally, which 

 indeed I consider the most decisive character of all, they dis- 

 solved in a saturated solution of sulphate of lime, but at the 

 same instant rendered it exceedingly turbid, in consequence of 

 the formation of a difficultly soluble sulphate of strontian. 

 Muriatic acid rendered the liquid again transparent: in proof 

 that the precipitate did not consist of sulphate of barytes. It 

 follows from this experiment that 1000 parts of the water con- 

 tain 0'00096 of carbonate of strontian. 



3. Quantity of the Oxides of Iron and Manganese. — The ana- 

 lysis of the sprudelstone already demonstrated that the 0'004 

 gramme regarded as oxide of iron in the original analysis of the 

 water, was not that substance in a state of purity, but contained 

 sensible quantities of silica, phosphate of alumina, phosphate 

 of manganese, and phosphate of lime. To determine the quan- 

 tity of oxide of iron with greater precision, 4*107 grammes of 

 the insoluble earthy matter from the Carlsbad water were dis- 

 solved in nitric acid, and the solution was mixed with an excess 

 of ammonia. From the unignited precipitate sulphuric acid 

 expelled some fluosilicic acid, which, with a view to ascertain 

 its quantity, was made to pass through a solution of carbonate 

 of soda. The sulphuric solution being precipitated by ammo- 

 nia, the filtered liquid contained no lime : consequently the 

 fluoric acid and silica had been combined with oxide of iron. 

 Potash boiled on the precipitate, left 0-02 gramme of oxide of 

 iron undissolved ; and, from the alkaline solution, there was 

 separated in the usual manner 0*004 gramme phosphate of alu- 

 mina, mixed with some manganese. Lime water added to the 

 remaining liquid, threw down a small quantity of phosphate of 

 lime. As the oxide of iron might have been contaminated with 

 phosphate of lime, it was dissolved in muriatic acid, and pre- 

 cipitated by the triple prussiate of potash ; but no traces of 

 phosphate of lime could be detected in the filtered liquid. 



The solution from which the oxide of iron and alumina had 

 been precipitated, was now mixed with oxalate of ammonia. It 

 yielded 2 - 514 grammes of carbonate of lime. If we estimate 

 the quantity of oxide of iron according to this quantity of lime, 

 we shall find its amount in 1000 parts of the water to be 

 0-00248. The quantity of phosphate of alumina here found is 

 rather greater than that obtained in the analysis of the sprudel- 

 stone ; but this is partly owing to its not having been so com- 

 pletely freed from silica. 



That in these analyses the oxide of iron was always mixed 

 with subphosphate of iron, is no proof that such a salt existed 

 in the water : it is merely in consequence of the property which 

 oxide of iron possesses, when precipitated from a solution con- 



