170 Bibliography. 
The relation between the fixed and volatile combustible matters of 
coals, is liable to considerable variation, according to the rate of distil- 
lation to which they are subjected. The more slowly this process is 
conducted, the higher (within certain limits) will be the proportion of 
fixed carbon.* The estimation of heating powers, therefore, from the 
quantity of fixed carbon which coals contain, if not wholly erroneous 
in principle, must be liable to considerable uncertainty in practice. — 
Many highly bituminous coals contain more than five per cent. of 
materials convertible into ammoniacal liquor by simple distillation with- 
out contact of air. This is proved on the largest scale in the manufac- 
ture of illuminating gas. That proportion, therefore, is not only una- 
vailable for heating purposes, but it also abstracts from the really com- 
bustible materials of the fuel, all the heat, sensible and latent, which 
the vaporized ammoniacal products receive during combustion. 
The proper water of combustion, namely, that derived from the hy- 
drogen in excess, and oxygen of the atmosphere, must in every instance 
where heat is applied to evaporate water above the boiling point, as in 
all ordinary steam boilers, be likewise incapable of giving up its latent, 
as well as much of its sensible heat. 
The average specific gravity of the six varieties of bituminous coals 
assayed is 1-31,—that of water at 60° being unity. Admitting the hy- 
drogen in its solid state to have a density of only 1:25, it must in pass- 
ing into the state, first of gaseous hydrogen, and then into that of wate- 
ry vapor, (still having the same bulk as the hydrogen,) undergo an en- 
largement to 2117 times its original bulk. This volume is further in- 
creased according to the usual law of gaseous expansion, by whatever 
heat above the boiling point is left in the vapor, when it passes away from 
the surface to be heated. In a well constructed evaporative apparatus 
producing steam of six pounds pressure, in which the circuit traversed 
by the gases after passing the grate, and before reaching the chimney, 
was one hundred and twenty one feet, the temperature was generally 
about 100° above the boiling point; and the watery vapor, being of 
course surcharged with heat, possessed 2431 times the bulk which it 
had in the solid state and at 60° of temperature. 
By the experiments of Dulong, (Comptes Rendus, tom. 7,)+ one 
gramme of pure carbon developes, in burning, heat enough to raise 
the temperature of 7170 grammes of water, 1° centigrade, or 12906 
grammes 1° Fahr. This latter number, is, therefore, used as a co-efli- 
cient, by which to multiply the numbers in the 12th column of the pre- 
ceding table to obtain those of the 15th. By the same authority, 1 
* See Proceedings of the Acad. of Nat. Sciences, Vol. II, pages 9, 10. 
t See also Peclet, Traite de la Chaleur, Tom. I, p. 50. . 
