OF COBALT, &c. 43* 



fenic, be converted into a green glafs; neither 

 is it a lit ingredient for fympathctic ink, nor doc 1 ? 

 it allord red folution with acids, or a green calx 

 limilar to that of cobalt. Befides, pure nickel 

 \vill melt niuf run into a mafs with filvcr, but 

 not cobalt; and to precipitate an hundred weight 

 of iiivcr, twice as much of nickel as of cobalt i 

 required. I^cad and bifmuth are much liker to 

 each other, yet no body doubts their diverfity. 



Although Brandt flic wed by experiments, fifty 

 years fincc, that cobalt is a peculiar metal; yet: 

 iome perfons, chiefly in Saxony ,have fmcedenicd 

 that part of cobalt which ftainsglafs to be metal- 

 lic. They have referred, too,to a certain ore of co- 

 balt (cobalt-muhttt) which communicates a green 

 tinge to glafs, and yet affords no rcgulus of co- 

 balt. But, although I have not, asyet,citherfcen 

 or examined this ore of cobalt, I fufpecl its pu- 

 rity to be the caufc of its ailbrding no rcgulus. 

 For, from what I have faid above it appears, that, 

 pure cobalt, without any intermixture of arfe- 

 nic is extremely difficult to melt. In allaying 

 many j'.lallcs tinged with cobalt, with an addition 

 of black flux, I ulwuyx obtained a rcgulus of that 

 metal, although but a very fmali quantity is nc~ 

 ccflary to ftain a large piece of glafs. The preci- 

 pitate too, produced in folutions of cobalt by the 

 admixture of phlogiilicated alkali afforded, upon 

 reduction, a regulus fit for ilaining glafs, and was 



iu 



