130 M. Berzelius on the [Aug. 



To determine how far the sprudelstone represents the sub- 

 stance deposited by the water, when deprived of its carbonic 

 acid, I mixed a quantity of the water with caustic ammonia. 

 It instantly became turbid, and at the end of twenty-four hours 

 there had subsided a granular and slightly yellowish coloured 

 precipitate. The filtered liquid when concentrated deposited a 

 white earthy matter. The first of these precipitates contained 

 carbonate of lime and oxide of iron, but no magnesia. The 

 second dissolved without effervescence in acids, and left a 

 gelatinous silica. The solution contained magnesia ; oxalate of 

 ammonia produced no alteration in it, but phosphate of am- 

 monia precipitated from it the well known double salt of mag- 

 nesia. This experiment demonstrates that the substances held 

 in solution by the carbonic acid are precipitated in proportion 

 as the acid is dissipated, independently of the concentration 

 of the liquid ; but that the magnesia and silica do not make 

 their appearance until a portion of the water has been evapo- 

 rated. That the magnesia in this experiment was precipitated 

 in the state of silicate, proceeded obviously from the presence 

 of ammonia. 



The constituents of the sprudelstone represent therefore the ' 

 carbonate of lime and oxide of iron obtained in the analysis ; 

 and, consequently, by analyzing a larger quantity of that incrus- 

 tation, it might be possible to discover the proportion in which 

 the fluoric acid, the phosphoric acid, the oxide of iron, and the 

 strontian, exist in the water, when compared with the carbonate 

 of lime. 



a. I reduced a quantity of the above incrustation (pannsten) 

 to an impalpable powder, and boiled it repeatedly in distilled 

 water, in order to separate any soluble saline matter which the 

 water might have deposited among its particles. This was 

 afterwards thoroughly dried. 10 grammes of the powder thus 

 purified were dissolved in dilute nitric acid. Some oxide of 

 iron remain undissolved, but was speedily taken up on the ap- 

 plication of heat. After the carbonic acid gas had been com- 

 pletely expelled, the solution, which had a slight tinge of co- 

 lour, was filtered. A grayish powder, weighing 0*001 gramme, 

 was by this means separated : before the blowpipe with car- 

 bonate of soda in platinum foil it gave traces of manganese, 

 and on charcoal it left a globule of tin. 



b. The filtered liquid was decomposed in a close vessel with 

 caustic ammonia. A light yellowish coloured matter precipi- 

 tated, which, after ignition, became brown, and weighed 0T57 

 gramme. It was analyzed in the following manner. Sulphuric 

 acid mixed with it in a platinum crucible occasioned after a 

 few moments the disengagement of fluoric acid, and a glass 

 prepared in the usual way, when placed over the crucible, 

 became deeply etched. As the gas was not expelled instanta- 



