Typical A. brevispica is widespread in tropical Africa, but occurs nowhere closer to our area than Angola and central Tanzania. In tropical Africa typical A. brevispica is relatively uniform and no difficulty is experienced in distinguishing it from A. schweinfurthii. A. brevispica subsp. dregeana however, is characteristically very variable and in our area it bridges many of the discontinuities which exist between typical A. brevispica and A. schwein­furthii in the tropics. Consequently difficulty is some­times experienced in distinguishing specimens of A. brevispica subsp. dregeana from A. schweinfurthii. A. schweinfurthii var. schweinfurthii differs from A. brevispica subsp. dregeana in having: 1. typically longer petioles (1,5)2,6-5,5 cm long; 2. a distinctly humped petiolar gland situated immediately above the pulvinus, sometimes two glands are present or, on occasions, the gland is absent; 3. leaflets usually larger, 2,5-8 x 0,8-2 mm; 4. leaflets glabrous on the lower surface apart from appressed marginal cilia. In A. brevispica subsp. dregeana when the lower leaflet surface is entirely glabrous the marginal cilia are usually lacking too so that the leaflets are completely devoid of hairs; 5. pods glabrous or almost so except for the glands; lacking the dense puberulence of A. brevispica subsp. dregeana. There is an overlap in petiole length and in leaflet width between A. brevispica subsp. dregeana and A. schweinfurthii var. schweinfurthii, but in combination the above characters usually enable A. schweinfurthii var. schweinfurthii to be disting­uished from subsp. dregeana. Certainly, most speci­mens can be readily sorted. Some specimens do, however, undoubtedly create difficulties and it is sometimes debatable whether they are robust speci­mens of A. brevispica subsp. dregeana or depauperate specimens of A. schweinfurthii. Ross 874 (NU) from 3 km S. of Mandini in the Lower Tugela valley is difficult to place as it has long petioles and the general facies of A. schweinfurthii, but the petiolar glands are not humped, the leaflets are appressed pubescent beneath and the pods are ± densely puberu­lous. The specimen seems best regarded as a robust specimen of A. brevispica subsp. dregeana. Ross 764 (NU) from the Weenen district of Natal is unusual in that some of the pinnae are themselves bipinnate. True A. pennata (L.) Willd. does not occur in Africa, but until fairly recently the name was used in a wide sense to cover A. brevispica, A. schweinfurthii and several other species which occur in tropical and subtropical Africa.