24 BLACK SAND ON THE BANKS OF THE DON* 



METAL*. OXIGEN. 



Proportions of 1. 151 ue, 100 iG 



itsoxide». 2. Red, 100 33 



3. White, 100 4<) 



I find, that, when the white oxide of" titanium is reduced 

 to the state of red oxide, it loses one fourth of its weij^ht ; 

 and that red oxide, when raised to the slate of white oxide, 

 increases exactly one third of its weight. It was the ktjow- 

 letlge of these facts, that led me to the preceding numbers. 

 And I think they may be used, till some more direct expe- 

 riments lead us to precise conclusions. 



State of the U- Red oxide being the only state in which this metal has 



koivsuiitT '^* y^^ occurred separate, we may conclude, that it combines, 

 in this state, with metallic oxides, and that the titanium in 

 iron-sand, is most probably in this state. But white oxide, 

 diminished by one fourth, gives us the equivalent quantity 

 of red oxide. On this supposition, the titanium present, 

 before the analys-is, in the J 00 grains of ore, weighed 9*5 

 grains. 



The arsenic The appearance of the arsenic surprised me a good deal, 



from acciden- ^g \^ yf^s altogether unexpected. I am disposed to ascribe 

 tal arsenic py- . ° . , ' . . , . , . , , 



rites. It to some parlicies ot arsenic pyrites which might nave 



been accidentally present. This conjecture will appear the 

 more probable, when we reflect, that arsenic pyrites very 

 frequently accompanies iron-sand. Before the microscope, 

 the iron-sand appears to contain some white shining parti- 

 cles, which, probably, are arsenic pyrites. 



The small quantity of silica and alumind I ascribe, with^ 

 out hesitation, to grains of quartz and felspar, which had 

 adhered to the iron-sand, and been analysed along with it. 

 Some such grains were actually observed and separated. 

 But others, probably, escaped detection, 



12. If these suppositions be admitted as well founded, the 

 iron-sand was composed of 



Component Protoxide of iron 85-3 



parts of the Red oxide of titanium 9*5 



iron-sand. Arsenic ♦.... 1*0 



Silica and alumina.f i . f. . , j'5 



Loss* •••• r *> f •>••*• •f«« • 2*7 



100-0 

 ♦. Probably, from what follows, an errour of the press for oxide. C. 



The 



