The problem of the improvement of signal to noise ratio in the method of sampling display is analysed on a more general basis. The noise within the duration of a sampling pulse is considered as a random process and the general expression for the factor of improvement of signal to noise ratio after the sampling process is obtained

Theoretical calculations and experimental results prove that the factor of improvement is higher than that is estimated by simply considering the effect of the increase in the number of sampling points. If the bandwidth of the incoming noise is greater than that of the signal

the pass-band of the input network of the sampling stage and the width of the sampling pulse may be adjusted for a desirable overall bandwidth of the sampling unit so that the output signal to noise ratio is optimum. The authors also suggest a method of sampling which over-jumps the empty space in the signal period and takes the samples of the signal only. By this method

the period of the low frequency signal may be greatly shortened and the time saved may be used for further integration to improve the signal to noise ratio. Experimental results agree well with the theory and the noise suppression effect of the sampling method is shown to be rather exciting.