86 Analysis of Warwickite. 



3. A portion of the solution 1, was treated with tartaric acid, 

 neutrahzed with ammonia, and examined for hme and magnesia, 

 without the detection of either. 



4. A portion of sohition 1 , was precipitated by potassa, and the 

 precipitate treated with a sohition of the same agent. The clear 

 fluid, on being separated from the insoluble matter by the filter, 

 was boiled with hydrochlorate of ammonia, which produced only 

 a faint troubling of the fluid, proving that a trace of alumine was 

 present. 



5. The insoluble matter, after digestion in solution of potassa 

 4, was agitated repeatedly with a solution of carbonate of ammo- 

 nia in a close vial. The clear fluid, after many hours standing, 

 was withdrawn by a dropping tube and boiled. It threw down 

 -a yellowish white precipitate, which was separated by the filter 

 and ignited. No glow, like that evinced by zirconia, was observ- 

 able during the process. The powder was redissolved by hydro- 

 chloric acid, without difficulty, excepting a few particles of titanic 

 acid. Its solubility subsequent to ignition showed that it was not 

 thorina, while its white color proved it not to be the oxide of 

 cerium. It was therefore considered as yttria, although the tinc- 

 ture of nut-galls threw down from its hydrochloric solution an 

 orange red precipitate of titanium, evincing that traces of titanic 

 acid had also been, taken up by the carbonate of ammonia. 



6. A portion of the neutralized solution 3, was tested for man- 

 ganese by the chloride of lime, without aflbrding any indication 

 of that metal. Another portion was treated with hydrosulphate 

 of ammonia, which gave a black precipitate. It was separated 

 from the fluid, and fused with borax before the blowpipe, whereby 

 the reaction of iron was obtained. 



7. A portion of the mineral was fused with anhydrous carbo- 

 nate of soda, supersaturated with acetic acid, and tested with ace- 

 tate of lead for phosphoric acid, which acid was not detected. 

 Examination was also made for silicic acid, but with the like 

 result. 



Having thus ascertained that the ingredients of the mineral are 

 fluorine, titanium, iron and yttrium, with a trace of aluminium, I 

 proceeded as follows to determine their relative proportions : The 

 yttria in the sulphuro-tartaric solution was precipitated by tartrate 

 of potassa, although it was found impossible to obtain it in this 

 way, entirely unmixed with titanic acid. Another method of 



