Bond orders and valences in the SCF theory: a comment