On the choice of ingredients for a theory of the Ice Ages