Given any rule, however �fundamental� or �necessary� for science, there are always circumstances when it is advisable not only to ignore the rule, but to adopt its opposite.