thv Laws of Affi\iih\ 149 



filtered and evaporated, the alkaliiie part, which is fupcT-' 

 abundant to the conftitutioii of the carbonat of pota{h, that 

 i-6 to fay, all the portion not defended by a fufficiently large 

 mafs of carbonic acid, may be taken away by a weak affinity. 

 Alcohol has this property : by its means a feparation may 

 be efTejfted; the carhonat of potajh remains in folution in a 

 little water, while the alcohol of the fotajh is fupernatant. The 

 carbonat of potafli which is feparated might be treated atfo 

 with lime, arid by this fecond operation be reduced to a quan- 

 tity which might be neglected." 



C. Berthollet quotes alfo other experlmerfts refpe6ling che- 

 mical a6lion behig in the ratio of the mafs. 



" If a carbonat with excefs of potafh be treated with al- 

 cohol, a part only of its excefs of potafli is taken from it. 



^^ The other neutral falts have alfo the property of retain- 

 ing a part of the potafh when the latter is in excefs. 



*^ It is known alfo that the phofphat of lime cannot be 

 entirely decompofed by the fulphuric acid, though the latter 

 is ranked as having a ilronger affinity for lime than the phof- 

 phoric acid ha«. 



^^ The cafe is the fame when fulphat of alumine is decom- 

 pofed by ammonia : the precipitate always contains fulphuric 

 acid. 



*' If magnefia be precipitated from its fulphat by potaffi, 

 the magnefia retains alfo fulphuric acid ; for, when the mag- 

 nefia is urged by heat, it has afterwards a pretty flrong favour 

 of fulphat. 



** All thefe experiments prove that, in chemical analyfis, 

 chemills fall into an error when they take for the real weight, 

 cither of the alumine or of the magnefia which may be found 

 in the compound fubftance, that of the precipitate formed by 

 an cle61ive affinity. 



** It refults from the preceding obfervations, and many 

 others which might be quoted, that in elective affinity the 

 fubjeiSt of the combination divides itfelf between two fub- 

 flances, which a<^ on it in the ratio of the forces which they 

 may oppofe to each other. 



** One circumftance which merits attention, and which 

 particularly proves that chemical action depends as much ou 



the 



