d be frozen, it would be
condensed, and more atoms of oxygen could be crowded into the cubic
foot, and the fire would receive a corresponding advantage." Mr.
Williams proceeded upon this theory, and died without solving the
perplexing mystery of as frequent failure as success which attended
his experiments with steamship boilers. The only successes which he
obtained were misleading, because they were made with boilers so badly
proportioned for their work that almost any change would produce
benefit.

Successful combustion requires something more than the necessary
chemical elements of carbon, hydrogen, and oxygen, for it requires
something to cook the elements, so to speak, and that is heat, and for
this reason: When the coal is volatilized in the furnace, what would
be a cubic foot of gas, if cold, is itself heated and its volume
increased to double its normal proportion. It is thin and attenuated.
The cold air which is introduced to the furnace is denser than the
gas. With dampers wide open in the chimney, and the gases and air
passing into the flues with a velocity of 40 feet per second, they
strike the colder surface of the tubes, and are cooled below the point
of combustion before they have had time to become assimilated; and
although an opponent in a debate upon steam boiler tests once stated
that his thermometer in the chimney showed only 250 degrees, and
indicated that all the value that was practical had been obtained from
the coal, I took the liberty to maintain that a chemist might have
analyzed the gases and shown there were dollars in them; and that if
the thermometer had been removed from the chimney and placed in the
pile of coal outside the boiler, it would have gone still lower; but
it would not have proved the value to have been extracted from the
coal, for it was not the complete test to apply.

The condition of things in the furnace may be illustrated thus: If we
should mingle a quart of molasses and a gallon of water, it would
require considerable manipulation and some time to cause them to
unite. Why? Because one element is so much denser than the other; but
if we should mix a quart of the gallon of water with the quart of
molasses, and render their densities somewhere near the density of the
remaining water, and then pour the masses together, there would be a
more speedy commingling of the two. And so with the furnace. I have
always maintained that every furnace should be lined with fire-brick,