OF COB ALT, Sec. 



of them is by infenfiblc degrees changed, till it 

 l>e transformed into fomc other. A variety of co- 

 lours too, red, yellow, green, mure, brown, and 

 others are produced in iron by different modes of 

 treatment, although not precifcly the fame as arc 

 obierved in the above three metals. It is, how- 

 ever, exceedingly diilicult to fcparate iron from 

 nickel, cobalt, and nianganefc ; as 1 have elle- 

 \vhere fhewn particularly, in the inftance of nic- 

 kel. Of regulus of nickel I have obierved, that 

 the more painfully it is purified, the more docs 

 it come to refemble iron in attraclability by the 

 magnet, and even the very fragments of the re- 

 guli then attract one another. I know that fomc 

 reguh of nickel are not fufceptible of magne- 

 tic at trad ion. But fuch need only to be puri- 

 fied in order to acquire this quality, efpecially 

 nickel precipitated with liver of fulphur, mull 

 be freed of all extraneous mixture before it can 

 become fubjccl to the power of the magnet. 



There has not, as yet, been in much pains ta- 

 ken to purify cobalt and mangancfe as to puri- 

 fy nickel. But the experiments which have been 

 madelhew clearly that iron adheres to thclV me- 

 f:ils with great obilinacy, and is often attracted 

 by them from among other matters. 



There is alfo another argument which mews 

 how great the affinity of nickel and cobalt with 

 iron is. The pure reguli of thefo metals, though e- 

 vuporatcd to drynefs with acids yield n ) ochre; 



which, 



