e shell is kept electro-positive. Care
must always be taken that the fragments which break off the zinc as it
wastes away cannot fall upon the heating surfaces of the boiler.

_Evaporators._--In marine boilers the waste of water which occurs from
leakages in the cycle of the evaporation in the boiler, use in the
engine, condensation in the condenser and return to the boiler as
feed-water, is made up by fresh water distilled from sea-water in
"evaporators." Of these there are many forms with different provisions
for cleaning the coils, but they are all identical in principle. They
are fed with sea-water, and means are provided for blowing out the brine
produced in them when some of the water is evaporated. The heat required
for the evaporation is obtained from live steam from the boilers, which
is admitted into coils of copper pipe. The water condensed in these
coils is returned direct to the feed-water, and the steam evaporated
from the sea-water is led either into the low-pressure receiver of the
steam-engine or into the condenser.

[Illustration: FIG. 18.--Handhole Fittings.]

_Efficiency of Boilers._--The useful work obtained from any boiler
depends upon many considerations. For a high efficiency, that is, a
large amount of steam produced in proportion to the amount of fuel
consumed, different conditions have to be fulfilled from those required
where a large output of steam from a given plant is of more importance
than economy of fuel. For a high efficiency, completeness of combustion
of fuel must be combined with sufficient heating surface to absorb so
much of the heat produced as will reduce the temperature of the funnel
gases to nearly that of steam. Completeness of combustion can only be
obtained by admitting considerably more air to the fire than is
theoretically necessary fully to oxidize the combustible portions of the
fuel, and by providing sufficient time and opportunity for a thorough
mixture of the air and furnace gases to take place before the
temperature is lowered to that critical point below which combustion
will not take place. It is generally considered that the amount of
excess air required is nearly equal to that theoretically necessary;
experience, however, tends to show that much less than this is really
required if proper means are provided for ensuring an early complete
mixture of the gases. Different means are needed to effect this with
different kinds of coal, those necessary for properly b