in
tubular boilers with advantage?

_A._--By no means. The tubes of tubular boilers are almost always more than
4 feet 6 inches long, but then the calorimeter is almost always less than
18 square inches per horse power--generally about two thirds of this.
Indeed, tubular boilers with a large calorimeter are not found to be so
satisfactory as where the calorimeter is small, partly from the propensity
of the smoke in such cases to pass through a few of the tubes instead of
the whole of them, and partly from the deposit of soot which takes place
when the draught is sluggish. It is a very confusing practice, however, to
speak of nominal horse power in connection with boilers, since that is a
quantity quite indeterminate.

EVAPORATIVE POWER OF BOILERS.

270. _Q._--The main thing after all in boilers is their evaporative powers?

_A._--The proportions of tubular boilers, as of all boilers, should
obviously have reference to the evaporation required, whereas the demand
upon the boiler for steam is very often reckoned contingent upon the
nominal horse power of the engine; and as the nominal power of an engine is
a conventional quantity by no means in uniform proportion to the actual
quantity of steam consumed, perplexing complications as to the proper
proportions of boilers have in consequence sprung up, to which most of the
failures in that department of engineering may be imputed. It is highly
expedient, therefore, in planning boilers for any particular engine, to
consider exclusively the actual power required to be produced, and to
apportion the capabilities of the boiler accordingly.

271. _Q._--In other words you would recommend the inquiry to be restricted
to the mode of evaporating a given number of cubic feet of water in the
hour, instead of embracing the problem how an engine of a given nominal
power was to be supplied with steam?

_A._--I would first, as I have already stated, consider the actual power
required to be produced, and then fix the amount of expansion to be
adopted. If the engine had to work up to three times its nominal power, as
is now common in marine engines, I should either increase correspondingly
the quantity of evaporating surface in the boiler, or adopt such an amount
of expansion as would increase threefold the efficacy of the steam, or
combine in a modified manner both of these arrangements. Reckoning the
evaporation of a cubic foot of water in the hour as equivalent to an actual
hor