Iron, 
203 
Iron when treated in the coke furnace ; while on the other 
hand argillaceous iron-stone being much more refractory, 
does not melt till it comes into the very hottest part of the 
furnace, and therefore has had full time to absorb, the desi- 
rable quantity of carbon. Another thing to be attended 
to is the proper regulation of the blast,, and this depends 
upon its dryness, its temperature, its compression, and its 
direction. The dryness and temperature appear to be 
principally governed by the season of the year, and there- 
fore are but little capable of being modified by the manu- 
facturer. It is plain that the dryer and colder the air is, 
I the greater will be its effect on the combustion, and it is 
found by constant experience, that the produce of iron 
during the summer months is greatly inferior in quantity, 
; and materially so in quality to that which is manufactured 
in the winter : a clear, dry and severe frost is the most fa- 
vorable period in every respect for the working of the fur- 
nace, and a change to snow or rain is infallibly followed by 
a corresponding deterioration. The higher the tempera- 
ture of the blast is. when it is delivered into the furnace. 
« y 
the smaller (the degree of compression and other circum- 
stances being equal) will be the quantity of oxygen con- 
tained in every cubic foot, and of course the vigour of the 
combustion. Nor is the force of the blast and its direction 
a subject of less importance ; it is obvious that in propor- 
tion as the charge descends, the carbonaceous matter is com 
tinually diminishing, hence the proper situation for the 
focus of the blast is that part of the furnace, where, when 
the ore shall have arrived, it will be fully carbonized and 
surrounded with a sufficient quantity of fuel to excite an 
intense heat, and absorb nearly the whole of the oxygen of 
the air, and thus prevent it from either oxydating the iron, 
or carrying off the carbon with which it may be combin- 
ed. This precise situation, in a furnace properly com 
structed,, will be found to be just within the expansion of 
