The mathematical physicist does not dictate to the world what it must be like. But he is guided by mathematical form to make suggestions to the experimenter. His peculiar role then ends. The experimenter decides.
The Aims of Mathematical Physics (p. 9), Clarendon Press. 1929