After reading the first two reports on the surveys of marine traffic in the Dover Strait conducted by the National Physical Laboratory, it was decided to investigate the possibility of applying the results to two points raised in recent issues of the Journal.

The first question studied was one suggested by Beattie, in his article on the first of the above traffic surveys. It was noted that there was quite a variation in the number of ships per hour passing through a given area. This could have been caused by a systematic tendency for ships to come through in clusters or could be a perfectly random variation. If the variation was completely random then the frequency distribution of ships per hour should follow a standard statistical distribution known as the Poisson distribution. For each of six separate distributions obtained from the results of the surveys on 26 February to 1 March 1971 and 27 April to 29 April 1971, the appropriate theoretical Poisson distribution was fitted and tests made to assess how close the theoretical and observed distributions were. The results of the calculations are summarized below in the table.