[O]ne would get into hopeless difficulties if one tried to describe what happens between two consecutive observations.