Compilation
Brachystegia goetzei

Recent authors have included B. glaberrima among glabrous forms of B. longi-folia, usually following my earlier opinion. I have now given more careful thought to the question of how to deal with the mass of forms in this group to which further collection continues to add new combinations of the characters of B. glaberrima and B. boehmii, sometimes with added complications apparently due to introgression from spp. 8–10, 14 and 15. The admitted scrap-heap grouped as putative hybrids under B. × longifolia seems to have originated through repeated crossing and back-crossing among the local forms of the two species mainly concerned, sometimes evidently to the point of swamping B. glaberrima sensu stricto in much of the area of overlap. Swarms of diverse forms occupy intermediate habitats and also seem to have invaded areas made available by destruction of forest, notably on the southern highlands of Tanganyika. Here B. D. Burtt (Iringa and Iringa-Mbeya road, Oct. 1936) collected a series of flowering specimens; Nos. 5762, 5759, 5811 and 5741, all with hairy panicles, range (in that order) from glabrous leaves with 7–10 pairs of leaflets to hairy leaves with 9–12 pairs.Boaler’s valuable series (Chunya District, Lupa Forest Reserve, transect I, fl. 22 Sept. 1962) includes Nos. 6, 8, 1, 4, and 5, ranging (in that order) from glabrous throughout with 6–7 pairs of leaflets to densely pubescent throughout with 8–10 pairs; No. 8 has a clearly defined tomentellous zone on the upper and outer part of each bracteole, but an otherwise glabrous panicle; this peculiar distribution of pubescence, so far unique in the genus, is correlated with tepals more like those of B. boehmii than in any other specimen in this series, while the leaves are intermediate between No. 6 and the associated B. spiciformis (Nos. 3 and 7) ! The possibility of rare crossing between the latter species and B. × longifolia is also suggested by Procter 1319 (Iringa District, Image Mt., July 1959) a remarkable, hairy specimen with 7–9 pairs of ovate leaflets and obscure, scarcely axillary stipule-bases, occurring where the putative parents occupy adjacent zones.The total range of forms included is largely covered by fig. 41, but Lindeman 712 (6 km. S. of Tabora, fl. 14 Sept. 1938), chosen as illustrative specimen of B. longifolia in T.T.C.L.: 93 (1949), is a good example of a borderline case, approaching B. glaberrima more closely than any of the forms illustrated. B. gossweileri Burtt Davy & Hutch, was believed to occur in Tanganyika on the basis of a small-leaved sterile specimen (Chunya District, 13 km. E. of Chunya, 22 Dec. 1961, Boaler 374) and the species was accordingly included when fig. 35 was prepared. More recent gatherings from the same locality are now available (20 May 1962, Boaler, Chunya transect I, Nos. 1–16, again sterile but comprising normal and coppice material in variety), as well as a specimen in young fruit (4 Dec. 1962, Boaler 759), stated to be from the same site but this time described as “10 miles [16 km.] east of Chunya”. These specimens show conclusively that the population forms part of the B. × longifolia complex as represented in the area generally and covers the leaf-variation of “ B. goetzei ”. Whether B. gossweileri, as understood elsewhere, can continue to be maintained as a species independent of the complex, is a problem still awaiting a satisfactory solution.