It has been shown by earlier workers that the observation of the statistical distribution of time lags to breakdown of small spark gaps may provide information concerning the emission of electrons from the cathode. This paper discusses the restrictions imposed by this method of measurement of cathode emission. Since the measuring technique causes the surface being investigated to undergo a change, efficient spark quenching should be provided to reduce this change to a minimum. A circuit which provides efficient spark quenching is described, together with a novel circuit to record a set of time lags automatically. Preliminary experiments show that conditioning of the cathode surface in air is now more rapid than that obtained by previous workers, and the reasons for this are discussed. Further, application of the method to the breakdown of small gaps in n-hexane indicates that the behaviour is similar to that of a gap in air in that conditioning of the cathode surface and a statistical variation of time lags occur.