The validity of Musher's model for the magnetic properties of aromatic molecules