Some Remarks on a Heuristic Point of View about the Role of Experiment in the Physical Sciences