I have long been curious about the variation in Scadoxus multiflorus, but was never quite sure what constituted subsp. katherinae. Prompted by the recent correspondence, I looked up the species in Flora of Tropical East Africa: Amaryllidaceae, by Inger Nordal (of the University of Oslo: she knows the tropical African amaryllid flora better than anyone, I think). She adds a note to the formal description, which reads: "S. multiflorus subsp. multiflorus is widespread and heterogeneous, occupying a wide range of habitats. Lack of discontinuity prevents further delimitation. Only in the lowland rainforests of Sierra Leone to Ghana and in the coastal areas from Swaziland to the East[ern] Cape is differentiation sharp enough to justify the separation of subsp. longitubus (C.H. Wright) Friis & Nordal and subsp. katherinae (Bak.) Friis & Nordal respectively. In East Africa the typical savanna form is slender with flowers before the leaves, the mountain form more robust with leaves and flowers contemporary. Intervening habitats have intermediate forms, and the differences disappear in cultivation." To my mind, that is a model of communication of scientific information: brief and to the point. Seeing the great globes of S. multiflorus (not particularly 'slender') emerging from dusty dry soil in the Serengeti remains one of the oddest of my botanical memories; they seem much more at home in the lush forest understorey of the East African mountains. It was abundant in the forest around my former home on Kilimanjaro, and the bulbs I grow now accompanied me home from there. They readily adopted a summer-growing regime in the UK, and flower reliably. John Grimshaw Dr John M. Grimshaw Garden Manager, Colesbourne Gardens Sycamore Cottage Colesbourne Nr Cheltenham Gloucestershire GL53 9NP Website: http://www.colesbournegardens.org.uk/