30 
920=8565° ; or as much as would elevate about three times 
that weight of iron to the point of fusion, taking the same 
at 20577° temperature.* 
From which it appears, that to decompose the vapour in- 
troduced into the furnace in weather of 13° and 80°, and 
when the atmosphere is fully saturated with moisture at those 
temperatures, as much heat is requisite as would fuse in the 
first, 12.2 oz., and in the second case 122 oz. of iron per 
minute. This caloric, which we conceive lost to any useful 
purpose, we confess seems an immense quantity ; but when 
it is considered that the air is very rarely if ever found to 
be perfectly saturated with vapour, that even this large 
quantity bears no very great proportion to the whole mass of 
caloric generated in the furnace ; and that the heat sufficient, 
either in quantity or degree, for fusing iron, is considerably 
less than that necessary to produce iron from ironstone, as 
is abundantly evident from a comparison of the blast and 
produce of a blow-hole and blast furnace, — when these con- 
siderations are attended to, we think the efi'ect of vapour, as 
stated, is by no means incredible. 
Admitting then their reality, we have now some prospect 
of tracing to their source the mischievous effects of a sum- 
mer's air in blast furnaces. 
We do not indeed attribute these effects to the mere ab- 
straction of heat we have just calculated ; for though that 
certainly is an evil to some extent, yet it is of inferior mag- 
* VARIABLE ESTIMATES OF THE DEGREE OF HEAT REQUIRED TO FUSE CAST 
IRON. — ALL FROM DR. THOMPSON, VIZ. : — 
20577° Encyclopaedia Britannica 7 article " Chemistr " 
17997° Ditto Ditto. 3 ^ ^ emib ry. 
3470° in a late work of Dr. Thompson. 
All by Fahrenheit. 
The last ought to be most coiTect ; the first numbers appear too high, probably 
obtained by guess only ; the last may possibly be below the truth, though ascer- 
tained by later and more correct means. The difference between the first and 
last being as much as 6 to 1, will make an enormous difference in our calculation, 
which we have not either time or inclination to ascertain at present. 
