This paper reports an experimental study of the distribution of time intervals between bursts of 43.5‐Mc/s signals received over a 690‐km path from Stanford, California, to San Diego, California, with the object of determining whether or not the bursts occurred completely at random.The distribution of time intervals observed is consistent with the distribution expected if the bursts are produced by scattered radiation from the ionized trails of meteors which enter the upper atmosphere at random. The number of time intervals observed less than one second is appreciably less than the number calculated. This would be the case if some bursts were not observed when the interval between bursts is comparable to the duration of the bursts. The difference between observed and calculated values for large intervals for which the actual number of observations is small is not statistically significant. The observed number of triplets of equally spaced bursts substantially agrees with the number expected if the bursts occurred at random.The data used in this analysis were the intervals preceding 778 bursts observed during 0400–0600 hours on 3 August 1956 and those preceding 446 bursts observed during 0500–0700 hours on 6 Novemb