higher air speed that a less efficient atomisation can be used, and the speeding up of combustion has led to increased use of this type on high speed motors requiring a clear exhaust.

In work done on the size of the oil globules when the jet is broken up it is suggested (Sass) that to obtain good combustion 30% of the oil globules in the jet should not exceed 0 0002 inches, 30% o '0006 inches, and none should be larger than o 002 inches.

In the length and capacity of the fuel main from the injection pump to the sprayer, the important factor is found to be the length, as might be expected, and secondary Vibrational dis- charges are possible when the conduit exceeds a certain length

and bears relation to engine speed.

Fig. 14.

Gas Oil (Riedler 1916). K = Compression. D = Vapourisation. Z = Decomposition. V = Ignition and Combustion.

The lag period between the ending of compression and the beginning of ignition has occasioned much discussion. An earlier diagram (1916) is shewn in Fig. 14 as illustrating what probably takes place.

Though it will be gathered that a great amount of work has been done, there is still much to do in further ascertaining the action of forced turbulence on the jet, and in the determination of after-burning.

In high speed engines combustion goes on down the expansion stroke, lowering the efficiency so much that it may be in certain