In the Working Group Report (19, 411) a mistake in the figure quoted (page 415 , paragraph 5) for the number of collisions analysed in Appendix I (2648 not 2748) results in minor corrections to a number of the figures quoted in the first two pages of Appendix I. These in no way affect the arguments put forward. On page 415 of the Journal (page 5 of the separately published version) paragraph 5 , lines 1 and 2, should now read:

‘An analysis of some 2648 collisions selected randomly (Appendix I) showed that 70 per cent occurred in port areas and 30 per cent at sea.’