Good archaeology relies on ever more precise dates – obtainable, notably, from ice-cores and dendrochronology. These each provide year-by-year sequences, but they must be anchored at some point to real historical time, by a documented volcanic eruption, for example. But what if the dating methods don't agree? Here the author throws down the gauntlet to the ice-core researchers – their assigned dates are several years too old, probably due to the spurious addition of ‘uncertain’ layers. Leave these out and the two methods correlate exactly…