34 
one part of a furnace, to raise to a degree considerably 
higher, or indeed higher in the least, the temperature of 
another part of the same furnace, is perfectly incompatible 
with every theory and definition of heat that we have 
hitherto met with. Leaving, however, these statements to 
reconsideration, we return to the subject of our inquiry — the 
effects of the oxygen and hydrogen gases in the furnace, set 
at liberty by the decomposition of vapour, the subsequent 
actions of which, though different, are, we imagine, both bane- 
ful to the production of iron. 
To elucidate this, we must beg your attention to the well 
known experiment of passing steam or vapour through a 
tube heated to redness, and containing iron filings or wire. 
Here we find the component gases of the vapour separated, 
the oxygen combines with the iron, and therewith forms an 
oxide, and the hydrogen is liberated in the state of gas. 
There does not, however, appear to be any caloric generated 
or set free, for to continue this decomposition, it is absolutely 
necessary to furnish a constant strong external heat to the 
tube : now granting that the formation of hydrogen gas takes 
up a large quantity of caloric, yet the fixation of the oxygen 
ought to afford that quantity of caloric, and a surplus of 
nearly l-4th ; the comparative quantities of heat contained 
by oxygen and hydrogen, in proper proportions for composing 
water, being as 4 to 3.2 nearly. How happens it, then, that 
a continuation of external heat is requisite after decomposi- 
tion has once begun ? Again, iron is decomposed in water, 
and the same kind of oxide is formed as that produced from 
the preceding decomposition of vapour, and we have reason 
to suppose with no greater evolution of caloric ; but on the 
contrary, it is very probable that oxidation would not take 
place were the water perfectly deprived of heat ; for this 
oxidation is proportional to the heat of the water, though it 
is known that by heating water, the uncombined oxygen it 
contains is driven off in its gaseous state ; the oxidation does 
