anvil, and strike it slightly with a hammer, it explodes; but
instead of breaking either the hammer or the anvil, it is found
that that part of the face of each in contact with the
fulminating silver is damaged. In this case the velocity
communicated by the elastic matter disengaged may be greater than
the velocity of a wave traversing steel; so that the particles at
the surface are driven by the explosion so near to those next
adjacent, that when the compelling force is removed, the
repulsion of the particles within the mass drives back those
nearer to the surface, with such force, that they pass beyond the
limits of attraction, and are separated in the shape of powder.

25. i) The success of the experiment of firing a tallow candle
through a deal board, would be explained in the same manner, by
supposing the velocity of a wave propagated through deal to be
greater than that of a wave passing through tallow.

25. ii) The boiler of a steam-engine sometimes bursts even
during the escape of steam through the safety-valve. If the water
in the boiler is thrown upon any part which happens to be red
hot, the steam formed in the immediate neighbourhood of that part
expands with greater velocity than that with which a wave can be
transmitted through the less heated steam; consequently one
particle is urged against the next, and an almost invincible
obstacle is formed, in the same manner as described in the case
of the discharge of a gun. If the safety-valve is closed, it may
retain the pressure thus created for a short time, and even when
it is open the escape may not be sufficiently rapid to remove all
impediment; there may therefore exist momentarily within the
boiler pressures of various force, varying from that which can