A possible new interpretation of the results of Δα/α from QSO absorption lines