OF A VARIETY OF THE CORUNDUM. 213 



The alumina was dried gradually, and heated in a crucible, Precip. of earth 

 fo that the bottom of it was only of a low red : it weighed ^J^ 1 ^' 

 1 1 1 j- gr. Expofed to a ftrong red heat for half an hour, it =87^. and heat. 

 Sufpecting that a fmall portion of alkali was retained by it, I 

 rubbed it to a fine powder, and digefted it with acetic acid, 

 for many hours, and faturated it with ammonia. 



The alumina fufficiently edulcorated, dried and expofed to 

 a ftrong red heat for half an hour, now weighed 81 J. 100 

 grains of the original pounded fpar, from which I had taken 

 the 100 grains on which I operated, were expofed to the fame 

 heat, and for the fame time, and loft 4-f of a grain, as I had 

 found before. 



The proportion which the feveral ingredients of this ftone Component 

 bear to each other, according to the foregoing analyfis, will parts " 

 be, 



Alumina- - - - [Cd] 81,75 



[Bb] 1,625 7 



Silex [Cb] 10 J- - - 12,125 



[Cc] 0,50 ) 



Oxide of titanium -[Be?]- 4 



Water - - - - [A a] 0,937 



Lofs - - * - - < „.,,,-.,. 1,188 



100 



I have repeated the analyfis of this mineral * feveral times, Analyfis re- 

 and have always found it to contain the fame ingredients ; P ea . ted « 

 though there has been a flight variation as to the relative quan- 

 tities of each. This we might expe6t to be the cafe, not only 

 from the lofs infeparable from the operations to which the ftone 

 is fubje&ed, but alfo from the nature of the ftone itfelf, as it 

 is not perfectly homogeneous. In one inftance, I found the 

 alumina amount to 84 per cent, and the filex proportionably 

 diminifhed. — Until we have better means than we have at 

 prefent, of afcertaining the real quantities of water and oxygen 

 -which the earths and the oxides of metals feverally retain in 



* I am indebted to the kindnefs of Mr. Rafhleigh, for another 

 ipecimen of corundum from the fame place, which differs in fome 

 refpe&s in its external appearance from that which I examined. Its 

 colour is not of fo bright a lilac, and it feems of a more compact 

 and uniform grain. Its furface is covered with a coat of yellow 

 mica.— This I have not fubje&ed to experiment. 



the 



