"27. Mimosa spinis oppositis longitudine foliorum, foliis bipinnatis: partialibus bijugatis. Acacia maderaspatana, foliis parvis, aculeis e regione binis praegrandibus horrida, cortice cinereo. Pluk. alm. 3. t. 121. f. 4. Habitat in India I!." This appears to be one of those unfortunately rare cases where only one element composed Linnaeus's conception of a species. Afterwards confusion started, not surprisingly with so inadequate an original description. Vahl, Symb. Bot. 1, 81 (1790), wrongly identified with Mimosa horrida a distinct Arabian species, Mimosa orfota Forsk. (i.e. Acacia orfota (Forsk.) Schweinf.). Then, in Sp. P1. 4, I082 (I8o6), Willdenow transferred M. horrida to Acacia and enlarged its description, citing in its synonymy M. orfota and M. leucacantha Jacq., Hort. Schoenbr. 3, p. 75, t. 393 (I798), the latter being a South African species. Willdenow wrote of A. horrida "habitat in Africa", its occurrence in Arabia or India being neglected. In this way, Acacia horrida, originally based on an Indian plant, was gradually accepted as the correct name for a common South African acacia. In this sense the name was used by Harvey in the Flora Capensis, 2, 281 (1862) and by Bentham in his great revision of Mimoseae in Trans. Linn. Soc. 30, 507 (1875). Subsequently it was realised by certain authors, e.g. Burtt Davy in Kew Bull. 1922, 328 (1923) and in Fl. Transvaal 346 (1932), and E. G. Baker, Leg. Trop. Afr. 843 (1930), that the South African plant, for which they correctly used the name A. karroo Hayne, could not be conspecific with A. horrida. The confusion, as far as South Africa was concerned, was finally cleared away by Miss I. C. Verdoorn in Bothalia 6, 411-412 (1954). The true identity of A. horrida, however, remained unknown. There is no specimen named Mimosa horrida in the Linnaean Herbarium. M. horrida is in fact based