ces are reached and swept clean. There are numerous soot
blowing devices on the market which are designed to be permanently fixed
within the boiler setting. Where such devices are installed, there are
certain features that must be watched to avoid trouble. If there is any
leakage of water of condensation within the setting coming into contact
with the boiler tubes, it will tend toward corrosion, or if in contact
with the heated brickwork will cause rapid disintegration of the
setting. If the steam jets are so placed that they impinge directly
against the tubes, erosion may take place. Where such permanent soot
blowers are installed, too much care cannot be taken to guard against
these possibilities.

Internally, the tubes must be kept free from scale, the ingredients of
which a study of the chapter on the impurities of water indicates are
present in varying quantities in all feed waters. Not only has the
presence of scale a direct bearing on the efficiency and capacity to be
obtained from a boiler but its absence is an assurance against the
burning out of tubes.

In the absence of a blow-pipe action of the flames, it is impossible to
burn a metal surface where water is in intimate contact with that
surface.

In stoker-fired plants where a blast is used, and the furnace is not
properly designed, there is a danger of a blow-pipe action if the fires
are allowed to get too thin. The rapid formation of steam at such points
of localized heat may lead to the burning of the metal of the tubes.

Any formation of scale on the interior surface of a boiler keeps the
water from such a surface and increases its tendency to burn. Particles
of loose scale that may become detached will lodge at certain points in
the tubes and localize this tendency at such points. It is because of
the danger of detaching scale and causing loose flakes to be present
that the use of a boiler compound is not recommended for the removal of
scale that has already formed in a boiler. This question is covered in
the treatment of feed waters. If oil is allowed to enter a boiler, its
action is the same as that of scale in keeping the water away from the
metal surfaces.

[Illustration: Fig. 41]

It has been proven beyond a doubt that a very large percentage of tube
losses is due directly to the presence of scale which, in many
instances, has been so thin as to be considered of no moment, and the
importance of maintaining the boiler heating surfaces in a cl