9 B CHEMICAL AN ALTS IS, ^c. 



may be united to carbonic acid, cryftallizes, and alfo any neu- 

 tral fait that may remain ; and the folution being decanted a 

 fecond time, is obtained perfectly pure, colourlefs, and tranfpa- 

 rent. 



In the lad boiling it is fomewhat difficult to obferve the ex- 

 act degree of concentration at which all the alkaline carbonate will 

 cryftallize ; and if the evaporation is carried too far, the cauftic 

 alkali, if potafh, cryftallizes itfelf: fo that feveral evaporations* 

 are fometimes requifite. 



When no more water remains in the folution than is juft 

 fufficient to hold the cauftic alkali difTolved, it contains nearly 

 half its weight of alkali ; but the exact quantity is eafily known, 

 by evaporating a portion to drynefs in a (ilver crucible. Before 

 ufing fuch a folution for analyfes, I afcertain its purity in the 

 following manner : Sortie of it being fuperfaturated with per- 

 fectly pure nitric acid, is examined with nitrate of barytes and 

 of filver ; with neither of which, if properly made, it will give 

 the fmalleft cloud ; confequently it can contain no fulphuric or 

 muriatic acid. 



Another portion being faturated exactly with a pure acid, 

 the whole is evaporated to drynefs, and the fait left is rediffol- 

 ved in a little water. If any earth were contained in the cauftic 

 alkali, it would remain thus undhTolved ; but when made as 

 above defcribed, I have never, in this examination, obferved the 

 fmalleft fediment. 



After the alkali is purified from neutral falts, and from the 

 part united to carbonic acid, it may itfelf be cryftallized by far- 

 ther evaporation, as Mr Lowitz has fhown. But this procefs 

 feems of no ufe in chemical analyfis, as the alkali is previoufly 

 obtained altogether pure. 



V. 



