In recent work involving micromaser 'which-path' detectors it was show
n that the loss of interference in a two slit experiment is based comp
letely on complementarity. The position-momentum uncertainty relation
plays no essential role. Storey et al, have argued that the loss of in
terference is due to atomic recoil originating in the emission and abs
orption of microwave photons. This was refuted by Englert er al., who
showed that the kicks due to these microwave photons were, to a good a
pproximation, negligible. But the argument continues in a recent Natur
e article, and that is the reason for this detailed calculation. The r
esult proves that the correlation between detector and measured system
, i.e., between micromaser cavities and atom, is the only reason for t
he loss of interference, Transverse momentum kicks play no role.