of Oxides of Iron. 283 
Alcohol has the same effect. But if the alcohol be diluted with 
water, though it may flame in the fire, it will be ineffectual, as 
it is driven off before the oxide becomes sufficiently heated to 
receive its action. 
Such combustible substances, as do not very readily part 
with their carbonic element, require rather longer continuance 
of heat than others : for example, charcoal and cinders, well 
burnt, must be longer in the fire to have their full effect on 
the oxide, than dry wood, coal, or sulphur. 
But such substances as may be sublimed with facility, will 
gradually quit the oxide, by a continued application even of a 
low heat, leaving it unmagnetic, as at first. 
How very small a portion of inflammable matter is requisite 
to render a considerable quantity of oxide magnetic, is evident, 
since one grain of camphor dissolved in an adequate portion of 
alcohol, and mixed with a hundred grains of the oxide in a 
glass mortar, will, by a red heat, render all the particles of 
the oxide magnetic. 
As oxides of iron therefore are rendered magnetic by heat, 
when mixed with inflammable matter, it may be understood 
why Prussian blue, sulphurets, and ores of iron containing 
inflammable matter, become magnetic by the agency of fire ; 
while at the same time it is observable, that these same ores 
revert to their unmagnetic state, when the heat has been con- 
tinued sufficiently long to drive off the whole of the inflammable 
matter: thus we find among the cinders of a common fire cal- 
cined sulphurets of iron, distinguishable by their red colour, 
unmagnetic when all the sulphur is sublimed. 
My intention in this communication is to prove generally 
that mere oxides of iron are not magnetic ; that any inflam- 
O o 2 
