Since all the localized collections definitely come from Lindi District, there may be some doubt as to the accuracy of Kirk’s locality “ opposite Zanzibar Island ” (see Kirkia 2: 197 (1961)).  This species differs from T. angustifolium Boj., of Madagascar, in its 1-foliolate, not 3-foliolate, leaves with a very long petiole and short leaflets. The flowers of T. paradoxum seem to be slightly larger and are apparently borne in terminal or pseudo-terminal inflorescences, never on leafless lateral ones.  It is known from only five gatherings, three of them made before 1915, but it was collected sterile in 1939 and again in 1944. It would be very interesting to rediscover this plant and investigate, by breeding experiments, its relationship with T. africanum, of which it seems to be an extreme narrow-leaved variant.