only is the specific heat of water much greater than that of any other
material present, but it possesses in a high degree the faculty of
absorbing heat throughout its mass, by virtue of the action known as
convection, provided that heat is communicated to it at or near the
bottom, and not too near its upper surface. Moreover, water is a much
more valuable substance for dissipating heat than appears from the
foregoing explanation; for reference to the experiment with the gas-
burner will show that six and a quarter times as much heat can be
absorbed by a given weight of water if it is permitted to change into
steam, as if it is merely raised to the boiling-point; and since by no
urging of the gas-burner can the temperature be raised above 100 deg. C. as
long as any liquid water remains unevaporated, if an excess of water is
employed in an acetylene generator, the temperature inside can never--
except quite locally--exceed 100 deg. C., however fast the carbide be
decomposed. An indefinitely large consumption of water by evaporation in
a generator matters nothing, for the liquid may be considered of no
pecuniary value, and it can all be recovered by condensation in a
subsequent portion of the plant.

It has been said that the quantity of heat liberated when a certain
amount of carbide suffers decomposition is fixed; it remains now to
consider what that quantity is. Quantities of heat are always measured in
terms of the amount needed to raise a certain weight of water a certain
number of degrees on the thermometric scale. There are several units in
use, but the one which will be employed throughout this book is the
"Large Calorie"; a large calorie being the amount of heat absorbed in
raising 1 kilogramme of water 1 deg. C. Referring for a moment to what has
been said about specific heats, it will be apparent that if 1 large
calorie is sufficient to heat 1 kilo, of water through 1 deg. C. the same
quantity will heat 1 kilo. of steel, whose specific heat is roughly 0.11,