A comment on the Born model treatment of the polymorphic transitions of the alkali halides