It is the rule which says that the other rules of scientific procedure must be designed in such a way that they do not protect any statement in science against falsification.

A pocket Popper (ed. 1983)