42 Mr. ConnePs Analysis qfLevyne. 



with muriatic acid. Speedy gelatin ization took place with 

 rise of temperature. Silica was separated in the usual man- 

 ner, and, after ignition, weighed 4*7 1 grains. From the resi- 

 dual liquid ammonia threw down alumina, which when ignited 

 weighed 2*4 grains. By dissolving it in muriatic acid, and by 

 the subsequent action of caustic potash, and other necessary 

 steps, there were separated from it *03 of silica, *03 of oxide of 

 iron, and *03 of red oxide of manganese, equivalent to *027 of 

 protoxide, leaving 2*31 for the amount of the alumina. To 

 the liquid which had been precipitated by ammonia, oxalate 

 of ammonia was added, its action being aided by heat. The 

 precipitate obtained yielded by calcination under the usual 

 precautions 1*84 of carbonate of lime, from which, however, 

 •05 of oxide of iron were separated by solution in diluted mu- 

 riatic acid, and the agency of hydrosulphuret of ammonia, 

 leaving 1*785 of carbonate of lime, equivalent to 1-004 of lime. 

 The remaining liquid was evaporated to dryness, and the am- 

 moniacal salt driven off" by heat, when a residue of *61 re- 

 mained, which, when dissolved in water, left *02 of silica, with 

 a trace of lime and magnesia. The solution gave cubical cry- 

 stals by spontaneous evaporation; and by the agency of chlo- 

 ride of platinum and of alcohol, these crystals were found to 

 consist of *20 chloride of potassium, equivalent to *13 potash, 

 and '39 chloride of sodium, equivalent to *16 soda. 



We have thus in 10*28 grains of the mineral, exclusive of 



water, Silica (4*71 -r-*03 + '02) 4*76 



Alumina 2*31 



Lime 1* 



Soda -16 



Potash -13 



Oxide of iron (-03 + '05) -08 



. — manganese '02 



8-46 

 And in 100 parts: 



Silica 46*30 contains oxygen ... 24 05 ... 7 



Alumina 22*47 ... 10*49 ... 3 



Lime 9*72 2*73 



Soda 1*55 -39 



Potash 1*26 *21 



3*33 ... 1 



Oxide of iron # 77 



manganese *19. 



Water 19*51 17*23 ... 5 



TovrT* 



* In operating upon small quantities, any accidental loss or excess of 

 course becomes magnified when reduced to 100 parts. The excess on 

 the analysis of 10*28, taking the proper proportion of water into account, 

 was only '18. 



