Comments on de Fontaine's “an approximate criterion for the loss of coherency in modulated structures”