This Euclea tree in fruit was seen in the Biedouw Valley during September.
At first glance the leaves appeared too wide for it to be Euclea lancea, too glabrous for E. tomentosa and too acutely pointed at the tips and flat-bladed to be E. undulata. Other species that might occur in the area are E. racemosa that should not grow this far from the coast and its leaf-tips would be rounded or at least bluntly tipped. E. pseudebenus has narrow leaves and does not occur this far south. E. linearis leaves are even narrower, although its inland range may come close.
After seeing the quite hairless leaves of E. tomentosa in the same area, this identification became the verdict, in spite of no leaf undulations. The grey bark and the fruit also favour this choice over E. lancea (Coates Palgrave, 2002; Van Wyk and Van Wyk, 1997; iNaturalist; http://redlist.sanbi.org).