Abstract
It is shown that recent models of the charge ordering in magnetite which can explain a- and b-axis doubling are likely to be unstable, even below the Verwey temperature, to the formation of a partially ordered structure which accords with single-crystal neutron results.