The physicist can never subject an isolated hypothesis to experimental test, but only a whole group of hypotheses.