232 
Mr Mushet on the Deoxidation 
nearly 40 per cent, of oxygen, nor have I been able to satisfy 
myself that ironstone properly roasted, unites in the slightest 
degree with oxygen, though there is no doubt but that long 
exposure and high temperature, will cause such a combination, 
though to a limited extent. 1 hope, therefore, you will indulge 
me wdth room for the insertion of some experiments and re- 
marks on the roasting, deoxidation, and reduction of ironstones, 
tending to elucidate the subject. 
Ironstones in general in roasting, lose from SO to 35 per cent. 
of their weight. Torrefaction commences with a change of 
colour upon the surface, wliich keeps penetrating as the process 
advances. When the ironstone is of one uniform colour through- 
out, tlie operation may then be considered as perfect ; if a mass 
of th.e ironstone be broken in the interim, the unroasted part 
w'ill be found occupying the centre, and of a blackish colour. 
Some hours are necessary to roast thoroughly, even a small 
piece of ironstone ; the same effects, however, may be produ- 
ced in tw'o minutes, by exposing ironstone finely pounded, upon 
a red-hot plate. The change of colour in this case is rapid ; it 
almost instantly becomes brown, then black, though in cooling 
it usually returns to a brownish red, or dirty purple colour. 
An ironstone that had lost 32 per cent, in roasting, was 
poimded very small, fused, and yielded as follows : 
400 gi-aiiis of ironstone, 
50 grains of finely pulverised coke 
300 grains of ironstone, 
60 ditto of coke 
300 grains of ironstone 
7 5 ditto of coke 
139 grains of iron, =: 33| cent. 
123 grains of iron, 41 1 per cent. 
132 grains of iron, m 44 per cent. 
The last reduction may be considered as nearly a perfect 
assay of the ore, which, in many cases, where the earthy matters 
are nearly in equal proportions, may be accurately performed 
without the addition of any dux. 
Some masses of the same substance not previously roasted, 
wxnghing 6024 grains, were put into a crucible along with coke- 
dust, and a pyrometer roll fitted at 1°. After an exposure of 
ten hours to a bright heat, the pyrometer was found to indicate 
28'’, and the ironstone had lost in weight 42 per cent, or 10 per 
cent, more, tlian when roasted in the common wav. This I 
attributed to the loss of oxygen, in consequence of the metallio 
