■ LACK SAND ON THE BANKS OF THE DON. 23 



Bla<-k oxide of iron 98'70 



_ - "^yhite oxide of titanium 1 2*65 



Arsenic •> I'OO 



Silica and alumina 1'50 



Total, 113.85 



HerQ ther? W »n excess of nearly 14 grains, owiqg, witboHt 

 doubt, to the combination of oxigeu with the iron and the 

 titanium during the analysis. 



Had the iron in the Q»*e been in the metaUic state, the state of the 

 excess of weight, instead of 14, could not have been less iron in the ore. 

 tiian 30. For the black oxide is known to be a compound 

 of 100 metal and 37 oxigen. Heqce, I think, it follows, 

 that the iron in our ore must have been in the state of an 

 oxide, and that it must have contained less oxigen thaa 

 black oxide of iron. A good many trials, both on iron- 

 sand, and on some of the other magnetic ores of iron, in- 

 <]uce me to conclude, that the iron in most of them is com* 

 bined with between 17 and 18 per cent of oxigen. This Protoxide of 

 compound, hitherto almost overlooked by chemists, I con- ""'^» 

 aiider as the real protoxide of iron. Thenard has lately de- 

 monstrated the existence of an oxide intermediate between 

 the black and the red ; so that we are now acquainted with 

 four oxides of this metal. But the protoxide, I presume, 

 does not combine with acids like the others. Analogy leads ^ ^^jj^ oxide 

 us to presume the existence of a fifth oxide, between the presumable. 

 green and the red. 



As to the titanium, it is impossible to know what increase st^teof the 

 of weight it bus sustained, because we are neither acquainted titanium in the 

 with it in the metallic state, nor know how much oxigen^ 

 its different oxides contain. It is highly improbable, that, 

 iiriron-saud, the titanium is in the metallic state, if it be 

 .made out that the iron is in that of an oxide. The experi^ 

 luents of Vauqneljn Slid IJecht, compared with those of 

 Klaproth, have taught us, that there are three oxides of 

 titanium, namely, the blue, the red, and the white. 

 From an experiment of Vauquelin and Hecht, and from 

 some of ray own, I am disposed to consider thCse oxides 

 as composed of the following proportions of metal and 

 #xigen ; 



h Blue 



