excelled, explosions, which may sometimes be fire-extinguishers. In such
cases detonation may be prevented by there being ample space to receive
the suddenly ignited vapor, lessening the tension of it, but carrying
the flames much more rapidly than otherwise to inflammable materials at
great distance.

If disastrous results have arisen from the vapor of turpentine as a fire
spreader in vaults without windows, it is possible that if a quantity of
hot water were suddenly converted into steam in closely confined spaces,
effects of pressure might be observed, less destructive perhaps, but
resembling those which other explosives might produce. If the immense
temperature attained in some conflagrations be considered--sufficient
to melt iron and vitrify brick--it is possible to conceive of water as
being instantly converted into steam. Even a very small quantity of
water thus expanded could produce most disastrous results. While such
formation of steam, if it happened, would certainly extinguish most
flames in direct contact, the general phenomena shown would be
explosive.

A curious circumstance occurred at the Broad street (N.Y.) fire in 1845,
previously mentioned. The fire extended through to Broadway, and almost
to Bowling Green. A shock like a dull explosion was heard, and by many
this was attributed to the effects of gunpowder and saltpeter. Several
firemen were, at the moment of the shock, on the roof of the burning
building, when the whole roof was suddenly raised and then let down
into the street, carrying the men with it uninjured. One of the firemen
described the sensation "as if the roof had been first _hoisted_ up
and then squashed down." _Query:_ Was this like the common lifting and
falling back of the loose lid of a tea-kettle containing boiling water?
Was it from steam--at a low pressure perhaps--seeking vent through the