40 
Iron , 
of operation more perfect than that, wherein a few 
lost or gained ( which error will more likely take place 
from the inequality of heat, than the reckoning of time,) 
make a variation from truth equal to I -5th or l-6th of the 
whole. The manufacturer would be in a lamentable pre- 
dicament were he thus circumstanced, and obliged, by 
some means or other, to take the crude iron from the blast- 
furnace immediately upon being separated from the ores, 
lest a considerable portion of it should totally disappear. 
If the degree of heat produced from a smith’s forge was 
at all times the same, though supplied with fuel of various 
natures, and under different changes of temperature, then 
it is most probable that, by reckoning Bergman’s time to 
a minute, a button of iron, accurate in its results, might 
be obtained from some ores by the flux he has directed to 
be used. But I cannot conceive that this should be an 
universal consequence : far less can I comprehend, after 
a \ list separation has been effected, that five minutes will 
destroy l-10th of the produce of metal, while the incum- 
bent fluid protects the surface of the metallic button from 
the action of the atmosphere. 
It is not, however, my province to enter into a minute 
examination of the products obtained from the use of such 
vitreous fluxes, which are always productive of the most 
oxygenated state of crude iron ; and the accuracy of whose 
results, under a state of such high oxygenation, are ah 
ways to be suspected. I have frequently proved that, in 
using them, the affinity of the metal was so great to oxy- 
gen, that a slight derangement of the crucible, which had 
thrown the vitrified fluid from any point of the surface, was 
immediately attended by a rapid deflagration, and a consi- 
derable portion of the iron oxydated. 
I shall chiefly confine itself to a communication of 
those facts, which I have repeatedly confirmed, with a 
view of extending my results and observations to a more. 
