A short exposition is given of the foundation of the causal description in classical physics and the failure of the principle of causality in coping with atomic phenomena. It is emphasized that the individuality of the quantum processes excludes a separation between a behaviour of the atomic objects and their interaction with the measuring instruments denning the conditions under which the phenomena appear. This circumstance forces us to recognize a novel relationship, conveniently termed complementarity, between empirical evidence obtained under different experimental conditions. An appropriate tool for a complementary mode of description is provided by the quantum-mechanical formalism which allows us to account for regularities of definite or statistical character beyond the grasp of classical physical explanation. – N. B.