A revision of Lobostemon Lehm., and a discussion of the Species Problem

Abstract

SUMMARY The previous accounts of the genus Lobostemon are shown to be inadequate, and a complete revision has been made. A small genus Echiostachys is separated from Lobostemon on floral and vegetative features. A more precise definition of Lobostemon is given, and it is shown that the so-called Lobostemons of Tropical Africa have no claim to inclusion in the genus, which is purely South African. Lobostemon is subdivided into five well-marked sections, the characters of which are given. While the sections may be easily recognised, many of the species show ill-defined boundaries. A cytological investigation of half the species of Lobostemon and all the species of Echiostachys has been made. The basic number of chromosomes is seven in both genera. The species problem is discussed, and it is shown that Vavilov's Law of Homologous Series in Variation is applicable here. Detailed descriptions of the two genera and the species are given, together with their synonomy. Herbarium specimens are cited under each species. Keys to the species are given.