on certain assumptions is sometimes sufficient to
indicate unusual losses.

The largest loss is ordinarily due to the chimney gases, which depends
directly upon the weight of the gas and its temperature leaving the
boiler. As pointed out in the chapter on flue gas analysis, the lower
limit of the weight of gas is fixed by the minimum air supplied with
which complete combustion may be obtained. As shown, where this supply
is unduly small, the loss caused by burning the carbon to CO instead of
to CO_{2} more than offsets the gain in decreasing the weight of gas.

The lower limit of the stack temperature, as has been shown in the
chapter on draft, is more or less fixed by the temperature necessary to
create sufficient draft suction for good combustion. With natural draft,
this lower limit is probably between 400 and 450 degrees.

Capacity--Before the capacity of a boiler is considered, it is necessary
to define the basis to which such a term may be referred. Such a basis
is the so-called boiler horse power.

The unit of motive power in general use among steam engineers is the
"horse power" which is equivalent to 33,000 foot pounds per minute.
Stationary boilers are at the present time rated in horse power, though
such a basis of rating may lead and has often led to a misunderstanding.
_Work_, as the term is used in mechanics, is the overcoming of
resistance through space, while _power_ is the _rate_ of work or the
amount done per unit of time. As the operation of a boiler in service
implies no motion, it can produce no power in the sense of the term as
understood in mechanics. Its operation is the generation of steam, which
acts as a medium to convey the energy of the fuel which is in the form
of heat to a prime mover in which that heat energy is converted into
energy of motion or work, and power is developed.

If all engines developed the same amount of power from an equal amount
of heat, a boiler might be designated as one having a definite horse
power, dependent upon the amount of engine horse power its steam would
develop. Such a statement of the rating of boilers, though it would
still be inaccurate, if the term is considered in its mechanical sense,
could, through custom, be interpreted to indicate that a boiler was of
the exact capacity required to generate the steam necessary to develop a
definite amount of horse power in an engine. Such a basis of rating,
however, is obviously impossible when the fact is cons