SUMMARY
For each month of the year rainfall totals are regressed on sea-level pressure which has been redefined, using principal component analysis, as a set of uncorrelated pressure fields. This new set has been used, in the multiple linear regression procedure, as independent variables. It is shown that rainfall and sea-level pressure are related, linearly, in a way that is spatially dependent. The pattern of this association also varies from month to month but does so in a gentle way across the country. The strength of the relationship does not necessarily follow the seasonal march of rainfall across the country.