POTASH IN SCHIST. ' SO£ 



The results of the different operations wero 



Silex i 68 



Alumiue* 15 



Oxide of iron • • • • $ 



Magnesia • 2 



Lime (at most) 1 



91 



consequently there was a loss of 9 per cent. I knew in- Great deficiea* 

 deed, that tiie whole of the silex was not collected; but ^ 

 what might be supposed to have been left was fur fiom an- 

 swering to this deficiency. 



Accordingly I took one hundred decig. of the substance Water expel| 

 reduced to powder, and calcined them Strongly in a platina !-*• 

 crucible. They lost 3 dec, and were slightly agglutinated. 

 Six still remained to be accounted for, and, suspecting the 

 presence of an alkali, I sought for it after Mr. Davy's me- 

 thod. 



5 grs. were fused in a silver crucible with 10 of boracic Second analy- 

 acid. The whole was diluted in water; muriatic acid was SIS * 

 added in excess ; it was evaporated to dryness ; an excess of 

 acid was added afresh ; and the silex was separated by filtra- 

 tion. The liquor, when sufficiently evaporated, deposited a 

 great deal of boracic acid, which was removed. The whole 

 was then precipitated by carbonate of ammonia, boiled, and 

 filtered. The liquor, rendered again acid, and evaporated 

 to a pellicle, deposited boracic acid, which was removed: 

 and, the evaporation being continued, the residuum was cal- 

 cined, to drive off the ammoniacal salts. What remained 

 still contained boracic acid; and, whatever precautions were 

 taken, it was impossible to separate it by evaporation. Hence Not -^ 

 this method, though very convenient for detecting the pre- adapted to as* 

 sence of an alkali, appears to me not well calculated for ^^L ^ 

 finding its proportion. Into the liquor, reduced to a fe*W alkali, 

 grammes, muriate of platina was poured, which occasioned 

 a considerable precipitate, that was found to be the triple 

 muriate of platina and potash, as will be related in the third 

 analysis, that was made of the same schist. 



Vol. XXIII.— Aug. 180Q. X ThU 



