and Remarks on, the Mineral Species Tyrite. 93 



rated was now precipitated by hydrosulphate of ammonia and 

 filtered. The filtrate showed a trace of tin, which was not esti- 

 mated; the precipitate contained alumina, iron, uranium (glu- 

 cina?), and an earth which was probably zirconia, although its 

 oxalate was soluble. The alumina was separated by boiling the 

 whole in caustic potash and determined as usual. Iron was then 

 separated by carbonate of ammonia, and the solution, containing 

 then only zirconia (glucina ?) and uranium, precipitated together 

 and determined after ignition ; the uranium was then dissolved 

 out by hydrochloric acid, and determined as difference of weight 

 between the residue insoluble in acid and the former combined 

 amount. 



The oxide of iron was determined as usual, but on being re- 

 dissolved contained some insoluble earth, which was deducted 

 from its weight and added to the amount of zirconia. The quan- 

 tities thus found were as follow : — 



Water mean of four determinations as\^.gQ ^^^^^ 



loss on ignition J 



Mineral employed in analysis . . . 25*04 grains. '= 



Columbic acid obtained II* 12 ... 



Ignited oxalates obtained .... 8*20 ... 

 Carbonate of lime obtained . . . . 0*75 ... 

 Sesquioxide of cerium obtained . . 1'41 ... 

 Protoxide of lanthanium obtained . . 0-37 ... 



Yttria obtained 6-97 ... 



Alumina obtained 0'89 ... 



Zirconia (and glucina ?) obtained . . 0-70 ... ■"' 



Oxide of uranium (Ui'S 0"*) obtained . 1-55 ... 

 Sesquioxide of iron obtained . . . 0-59 ... 

 which, when tabulated, will afi'ord the following per-centage 

 composition : — 



Columbic acid 44-48 



Yttria 27-83 



Lime 1'68 



Protoxide of cerium . . . . 5-63 

 Protoxide of lanthanium . . 1'47 

 Protoxide of iron . . . • 2*1 1 

 Protoxide of uranium . . . 5*99 



Binoxide of tin trace 



Alumina 3 jo 



Zirconia (and glucina ?) . . 2-78 



Water 4-66 



100-18 

 The atomic equivalents of several of these bodies being so 



