High-presicion Ar- Ar and U-Pb dating of magmatic events in the British Tertiary igneous province is difficult to reconcile with the magnetostratigraphy. The normally maganetised rocks of Mull Centres 2 and 3 are too old for C26N, much too young for C25N, and of too long a duration to fit into normal cryptpchrons in C26R. The most likely reason for this discrepancy is that the accepted time range of C26N (58-57.6 Ma) is too young by at least 0.5m.y. We propose to resolve the problem by re-determining the age of the Palaeocene-Ecocene boundary by Ar-Ar dating of sanidine from ash layers in northern Jutland. Our Ar-Ar dates will be checked against U-Pb ages of zircon and sphen from the Danish ashes and from the Hebrides.