e position of the two main valves--the air valve F and the
exhaust E. These valves, as may be seen from the drawing, are capable of
withdrawal after the cover of the combustion chamber has been removed.
The latter is an iron casting, shaped and faced up to make an absolutely
tight joint; no asbestos or any packing is used to make this joint--and
is held in place by four studs, as shown. Thus, all that is necessary is
to remove the four nuts, lift the cover off, then pull out the pins
which keep the spiral springs in position, and withdraw the valves. The
latter are seated direct on to the metal of the cylinder casting, the
gun-metal bushes A and B acting as guides. Further reference to A (the
mixer), which serves a twofold purpose, will be made later on.

The gas valve and cock are mounted in a separate casting, which is
carried by a couple of studs, the joint between this and cylinder being
made with a piece of rubber insertion. The gas enters at the gas-cock,
passes through the valve and port G, and round the annular space in the
bush or "mixer" A, previously mentioned, and thence through a number of
small holes in same, immediately below the seat of the air valve F. At
the same time, pure air is drawn in _via_ the air box (as explained
hereafter), through port L (fig. 11), and thence up the centre of bush A
and over the small holes through which the gas is flowing. The two then
thoroughly mix and enter the combustion chamber together as the air
valve F is opened. This device produces a perfectly homogeneous mixture,
which conduces in no small measure to perfect combustion when the
explosion takes place, and upon which, to a very great extent, depends
the efficiency of the engine. Besides possible loss in this direction,
however, there is another source of waste which cannot be eliminated,
and that is the heat taken away by the cooling water which surrounds the
cylinder. As this loss is inevitable, the best thing we can do is to
make it as small as possible. Theoretically, it would be no small
advantage if we could work at very much higher temperatures than we do
at the present time, and it is only certain mechanical difficulties
which bar the way and so effectually prevent the already high thermal
efficiency of the engine being greatly increased.

It is no easy matter to overcome these difficulties completely, but
improvements in this direction are continually being made, so that
troubles which attended the gas-engin