3*^8 ^^ "^^^^ OXIDES OF IRON"; 



cecds -vrhat was obiained by every other person, we must 

 suppose a mistake. I have not tried this method, being 

 ^ deterred by its uncertainty. 



Difficult to find The second method is easier, and more satisfactory. The 

 iron perfectly greatest difficulty, to which it is liable, is that of procuring 

 iron in a state of absolute purity, to make experiments upon. 

 I have tried many varieties, and have applied to those artists, 

 who were likely to have iron in the greatest purity. But 

 hitherto I have not been lucky enough to find a single spe- 

 Spedmens cimen absolutely pure. I was obliged therefore to analyse 

 analysed. ^j,g specimens which I employed, and to make allowance 



for the impurities, which varied in different specimens from 

 P.n^ViPfiiron -j4gth to jo75^th part of the whole. Polished iron wire is 

 wire best. most convenient. Iron filings, unless made on purpose, 



are not sufficiently pure, and it is more difficult to dissolve 

 them completely than iron wire. 

 Di';solved in 100 grains of iron wire were dissolved in diluted nitric 



nitric aciU. acid. The solution goes on rapidly, and is at first opake, 

 and almost black, owing to the nitrous gas which it retains. 

 This gas gradually separates, and then the liquid is nearly 

 colourless. When concentrated it becomes of a brownish 

 Reduced to red y®^^^^ ^*^^®°''* It was evaporated to dryness, and exposed 

 oxide for a quarter of an hour to a red heat in a platinum crucible. 



The red oxide thus obtained weighed 142*6 grains. In 

 another experiment made in the same way 100 grains of 

 gained near 45 '^^^ were converted into 144*75 of red oxide. This last 

 per c;;nt. result t consider as the most correct, because it coincides 



nearly with the result obtained by Hassenfratz in a different 

 manner, and because in experiments of this nature, where 

 liquids are evaporated to dryness, there is always a risk of 

 some loss during the evaporation. On this account, in 

 making choice of various results, that which gives the 

 greatest weight has the most chance of being correct. Upon 

 the whole then we may conclude with considerable proba- 

 bility, that the red oxide of iron is composed of 100 iron 

 and 45 oxigen. 

 It ') lid not ^ tniiil to deprive the red oxide of iron of part of its 



be dcoxi'ia ed oxigen by various methods, but without success. No degree 

 byheataUne, ^^ \ica.t, which I could raise, was capable of disengaging 

 oxigen gas from it, though the oxide acquired a black 



colour 



