OF COBALT, Sec. 



in all other refpcds very like pure cobalt. No\v f 

 fin cc experiments have fliuwn, that none but me- 

 tullic matters are precipitable from folutions by 

 phlogifticated alkali, when the fhturation is com- 

 plete ; it follows neceflarily, that the part of co- 

 "bait with which glafles arc flawed, mull be 



SOME 



