364 On the different Kinds of Cadm'u, and. 



cobalt ore; and by a little fulphur, with al;undance of iron 

 without any arfciiic, in fulphurated cobalt ore. 



When fuch ore is treated in thefire^ the arfenic evaporate?, 

 and the metallic and earthy particles remain behind, which, 

 in order to be prepared for commerce, mull be mixed with 

 pounded quartz or fand, afterwards moiftened, and then 

 packed up in cafks under the name o^ fujlor or z.affera. If 

 very fufible glafs be melted with a little cobalt ore, and if 

 the mafs be finely pulverifed, it produces fniait, which, by 

 wafhing, may be fitted for an article of conmicrce of diffe- 

 rent qualities and forted. This fmalt, however, contains in 

 general an impure regulus of cobalt, which often confifts of 

 a mixture of cobalt, iron, arfenic, and nickel, and is called 

 cobalt fpe'sfe. If you wifli to obtain a pure regulus of cobalt, 

 \ou mull choofc for that purpofe a cobalt ore as free as pof- 

 lible from foreign metallic fubflances. You muft then roaft 

 it carefully according to the proportions of fulphur and arfe- 

 nic which vou find in it, and reduce it by adding carbona- 

 ceous matter to take up the oxygen, and a good alcaline flux 

 at a ftrong fufing heat. ' If the cobalt ore employed contains 

 bifmuth, and is free from nickel, the regulus of bifmulh will 

 be found at the bottom of the reduced cobalt in a fcparated 

 Ibtc. This regulus is not pure, and fufhciently free from 

 iron ; but it may be brought to puritv, though with trouble 

 *nd a confiderable lofs of the regulus of cobalt. To effeft 

 this, it muft be calcined with berax, on a teft, in the fame 

 manner as copper, until it is no longer attrafted by the mag- 

 net. Some, however, are of opinion, that iron is not tbe 

 onlv metal attracted by the loadltone, and that it exercifes 

 an attraftive power over the finell blue colouring regulus of 

 cobalt: but there are well-grounded objections aaainft this 

 idea, for that effeft takes place only when the latter is not 

 lufficiently freed from iron ; and according to the experi-. 

 mcnts of Berfimann, a compofition of one part of iron and 

 t^iree parts of cobalt is attracted by the magnet : but if the 



proportion 



