Evidence for influenza-associated excess mortality in the three centuries before the 20th has been sought from parish burial registers in Cumbria, Devon, Dyfed, East Anglia, Gloucestershire and Northumbria, compared with inter-epidemic years. Most of the registers showed excess of burials concordant with eight historic influenza epidemics.
Comparison of the dates of these epidemics, deduced from the burials data in different areas, showed a rate of spread difficult to reconcile with direct personto-person spread of influenza from the sick. An alternative explanation based on development of latency of the virus in the sick person and subsequent seasonal reactivation is discussed.