A note on the geometric phase in adiabatic approximation [rapid communication]