Iron . 
45 
to an opaque brown glass streaked with white, 294 grains: 
loss in volatile mixtures 320 grains, or 36.5 per cent. 
The metallic button produced in this operation had a 
smooth crystallised surface ; and, when in fusion, com- 
bined less readily with oxygen offered by the contact of 
atmospheric air, than did the former two. The fracture 
was light grey, with a regular distinct grain ; the mass of 
fused earths were disposed after the manner of a crystalli- 
sation, in radii, shooting from the circumference of a mi* 
nute circle to the extremity of one larger. 
From these experiments it is obvious that reguli of 
crude iron may be obtained from the different classes of 
iron-stones, of a respectable produce, without any addi- 
tion whatever ; and that the tendency which these have 
to part with their iron is in the exact ratio of the quantity 
of lime present. The following statement exhibits tha 
difference betwixt the real and partial assay of the fore- 
going iron-stones. 
The siliceous iron-stone properly assayed, yielded 54.5 
per cent, but by fusion per se* only 25 ; leaving in the lat- 
ter method 9.5 mixed with the scoria. Argillaceous stone, 
properly assayed, gave 35.6, by fusion per se 28.5 ; leav- 
ing 7.1 mixed with the scoria. Calcareous iron-stone, 
properly assayed, yielded 33.7, by fusion per se 30 ; 
which left 3.7 mixed with the scoria. 
One reason may be advanced why the iron produced 
from the calcareous stone was so much more in pro- 
portion than the other two classes : both the siliceous and 
argillaceous buttons, when deprived of their scoria, decom- 
posed atmospheric air so rapidly as to suffer a little, in 
point of real metal, from the combination of oxygen. It 
was otherwise, however, when the surface of the calcare- 
ous button was exposed : the affinity of the oxygen to the 
metal being less than to the carbon united with it, carbonic 
acid was formed. A review of these experiments leads to 
* By itself : without addition. 
