X-ray diffraction evidence for a critical end point for cerium I and cerium II☆