taken the value 0.103 in certain calculations.] both measured at
temperatures where water is a liquid. Putting the foregoing facts in
another shape, for a given rise in temperature that substance will absorb
the most heat which has the highest specific heat, and therefore, in this
respect, 1 part by weight of water will do the work of roughly 9 parts by
weight of iron, and of about 4 parts by weight of calcium carbide.

From the practical aspect what has been said amounts to this: During the
operation of an acetylene generator a large amount of heat is produced,
the quantity of which is beyond human control. It is desirable, for
various reasons, that the temperature shall be kept as low as possible.
There are three substances present to which the heat may be compelled to
transfer itself until it has opportunity to pass into the surrounding
atmosphere: the material of which the apparatus is constructed, the gas
which is in process of evolution, and whichever of the two bodies--
calcium carbide or water--is in excess in the generator. Of these, the
specific heat at constant pressure of acetylene has unfortunately not yet
been determined, but its relative capacity for absorbing heat is
undoubtedly small; moreover the gas could not be permitted to become
sufficiently hot to carry off the heat without grave disadvantages. The
specific heat of calcium carbide is also comparatively small, and there
are similar disadvantages in allowing it to become hot; moreover it is
deficient in heat-conducting power, so that heat communicated to one
portion of the mass does not extend rapidly throughout, but remains
concentrated in one spot, causing the temperature to rise objectionably.
Steel has a sufficient amount of heat-conducting power to prevent undue
concentration in one place; but, as has been stated, its specific heat is
only one-ninth that of water. Water is clearly, therefore, the proper
substance to employ for the dissipation of the heat generated, although
it is strictly speaking almost devoid of heat-conducting power; for not