IN matching one fuel against another directly with a double-float carbureter, the knock sometimes either disappears or increases in intensity when the change from one fuel to the other is made. Results of a study of this subject, which was made recently, are presented in the paper.
From the tests the conclusion is drawn that means should be provided to dry out the mixture between the carbureter and the cylinder-block of the engine used for antiknock measurement. Before a satisfactory answer to the problem can be reached, however, several other points require further investigation.
