Subsp. capensis occurs in South Africa and S Mozambique. Subsp. multiflora is very different from typical E. capensis from Cape Province which has small almost round leaves with revolute margins and invariably solitary flowers. Further up the coast and particularly on Inhaca Island, what is clearly the same species displays much greater variability and larger leaves and 1–3-flowered fasciculate inflorescences are common. E. mossambicensis Engl. based on material from Beira is presumably similar although it has been kept up in some recent works on South African trees. Specimens occur in Mozambique which are very close to Tanzanian material, a fact d by Greenway long ago (adnot. on Greenway 2610) so it is clear that the East African populations cannot be separated specifically from E. capensis; the problem that remains is how far subsp. multiflora extends southwards. The difficulty of placing intermediate collections does not in my opinion invalidate recognising subspecies for the extremes.