its _full_ effect at once, the contraction caused thereby would bring a
stress of 300 ÷ 15 = 20 tons per square inch upon the bottom plates of
the shell. But fortunately it cannot exert its full effect at once, but
it _can_ act to such an extent that we have known it to rupture the
plates of a new boiler through the seams on the bottom _no less than
three times in less than six weeks_ after the boilers were started up.

The effect in such cases will always be the most marked, especially if
the plant is furnished with a heater, when the engine is not running, for
then, as no steam is being drawn from the boilers, there is comparatively
little circulation going on in the water in the boiler, and the water
pumped in, colder than usual from the fact that the heater is not in
operation, spreads out in a thin layer on the lowest point of the shell,
and _stays there_, and keeps the temperature of the shell down, owing to
the fires being banked or the draught shut, while the larger body of
water above, at a temperature of from 300 to 325 degrees, keeps the upper
portion of the shell at _its_ higher temperature. It will readily be seen
that the strain brought upon the seams along the bottom is something
enormous, and we can understand why it is that many boilers of this class
rupture their girth seams while being filled up for the night after the
engine has been shut down. To most persons who have but a slight
knowledge of the matter, we fancy it would be a surprise to see the
persistence with which cold water will "hug" the bottom of a boiler under
such circumstances. We have seen boilers when the fire has been drawn,
and cold water pumped in to cool them off, so cold on the bottom that
they felt cold to the touch, and must consequently have had a temperature
considerably below 100° Fahr., while the water on top, above the tubes,
was sufficiently hot to scald; and they will remain in such a condition
for hours.