Note on the interpretation of the rotatory dispersion of potassium dithionate