ON THE PURITY OF THE FIXED ALKALIS. 27 



tralize them ; and that 100 parts of soda, prepared in the 

 saine manner, take 88. From the results of the experi- 

 ments above given however, we must infer, that 84*2 of 

 muiiatic acid are required, to saturate 100 parts of potash; 

 and 120*5 of the same acid, to neutralize 100 of pure soda: 

 whence it follows, that, the strengths of these alkalis being 

 represented by other, numbers, when they are compared 

 with those of bases, the nature of winch is fully ascertained, 

 they must give different proportions from those mentioned 

 in the work :Of Mr. Berthollet. 



The capacities of saturation of t,he ^carbonates, being as- but the car- 

 certained by analysis, are liable only to little variation; and ^°".^J^^ P""^" 

 then the degree of energy of the muriatic acid approaches 

 nearer to that of the carbonic acid, which has some influence 

 on the results deduced from a comparison of them. 



The same reasoning applies to experiments made on the Othsr neutral 

 .sulphates, nitrates^ and phosphates, with potash or soda ^ 

 for their base; but I shall confine myself to a few observa- 

 tions on the experiments, which Mr. Berthollet has pub- 

 •Jished in chap. 18 of the work I have mentioned. To find Berthollet's 

 ihe quantity of water muriatic acid gas can retain, Mr. Ber- of^j'r'latic'^ 

 ithoUet neutralized 100 parts of potash prepared with alco- gas for water - 

 Jbol, and \:ept so.yie time in fusion. The muriate obtained "^'*^'^^'^®* 

 was carefully dried, and weighed only 126*6, instead of 

 l6l*5, which ought to have been its weight. Ought not 

 this difference, which is in some measure owing to the 

 water contained in the muriatic acid gas, to be attributed in 

 part to the water, or foreign matter, which forms 0*27 of 

 .the {jotash employed ? And may we not thus account for the 

 great differences, that exist between the numbers represent- 

 ing the component parts of muriate of potash in the experi- 

 ments of Berthollet, Kirv/an, and Richter? At least we are 

 naturally led to presume so, from the facts I have given 

 above. 



1 ies>iet the not having been able to ascertain the nature The addition 



of the foreign matter, which is always found combined vvith *° '^^,^'^^^'* 



, , , probably 



soda and potash prepared by means of alcohol. I cannot water, but 



venture therefore to assert any thins: on the subject; but I ''^'^ not ascer- 

 , ,. , • 1 . , . • 1 , tained. 



believe, that water acts a considerable part in these phe- 

 nomena ; and I could have wished to have had time to ex- 



