Bothalia 23,2(1993) 235

posed by erosion of the soil above it and was probably not quite fully developed. It contained eight relatively large seeds (some dissected ovaries contained up to 10 ovules) and was extremely smelly when opened. In fact, when damaged all parts of the give off the unpleas­ ant foetid odour characteristic of Kedrostis. Meeuse men­ tions that C. humifructus is dispersed by antbears. I have, though, noticed no particular concentrations of K. psammophila around the groundsquirrel warrens which are common in the sandier parts of Namaqualand and its seeds may be dispersed by moles.

REFERENCES DE WET, B.C., ARCHER, R., FISH, L., GERMISHUIZEN, G., HER­ MAN, P.P., JORDAAN, M., PEROLD, S.M., REID, C., VAN ROOY, J., WTiLMAN. W.G. & GLEN, H.F. 1991. New taxa, new records and name changes for southern African . Bothalia 21: 211. GIBBS RUSSELL, G.E.. WELMAN, W.G.. RETIEF. E., IMMELMAN, K.L., GERMISHUIZEN, G.. PIENAAR, B.J., VAN WYK, M. & FIGURE 6.—Distribution of Kedrostispsammophila. NICHOLAS, A. 1987. List of species of southern African plants. Memoirs of the Botanical Sur\'e\' of South Africa No. 56:201 -203. gneissic sand (much more gravelly than the coastal sands) MEEUSE, A.D.J. 1962. The Cucurbitaceae of southern Africa. Bothalia 8: 1-111. higher up on the southern side of the Khamiesberge as far east as Kliprand. P. BRUYNS*

Finding fruit of this species proved difficult and I have * Bolus Herbarium, Univ. of Cape Town, Private Bag, Rondebosch 7700. only seen one (Bruyns 5174). This had been partially ex­ MS. received: 1992-09-07.

SAPOTACEAE

A BEQUAERTIODENDRON' BY ANY OTHER NAME?

The and nomenclature of the genus Be- The genus was described by Krause in quaertiodendron De Wild, have caused considerable con­ 1914 and typified by Englerophytum stelechantha Krause. troversy over the last thirty years. To complicate matters, It comprises five to ten poorly defined species in tropical two recent revisions, one of Bequaertiodendron (Liben Africa. Bequaertiodendron was reduced to synonymy 1989) and the other of the genera of the (Pen­ under Englerophytum by Aubreville in 1960. However, nington 1991), differ markedly from each other. Heine & Hemsley (1960, 1968) did not recognise this genus for the following reasons: the type material was Chrysophyllum magalismontanum was described by thought to have been destroyed, the diagnosis and Sonder (1850) and based on Zeyher 1849, a specimen illustrations did not fully agree, and the genus was de­ from the Magaliesberg. This species as well as C. scribed without fruits. natalense Sond. and C. glomeruliferum Hutch. & Dalz. have fruit characters which separate them from the rest Liben (1989), in a revision of the West and Central of the genus. The need to recognise this group as a sep­ African species, concluded that the material placed in arate entity prompted various researchers to describe new Bequaertiodendron magalismontanum by Heine & genera and sections in which members of the group could Hemsley (1960) could be divided into eight different spe­ be accommodated. Thus, Chrysophyllum L. sect. Zey­ cies. On the basis of macromorphological floral and seed herella Pierre ex Engl., Pachystela Pierre ex Engl. sect. characters, these species were grouped into three genera, Zeyherella (Engl.) Lecomte, Tisserantiodoxa Aubrev. & namely Englerophytum Krause. Wildemaniodoxa Aubrev. Pellegr., Zeyherella (Engl.) Aubrev. & Pellegr., Boivinella & Pellegr., and Zeyherella (Pierre ex Engl.) Aubrev. & Aubrev. & Pellegr. and Neohoivinella Aubrev. & Pellegr. Pellegr. Liben (1989) reduced Bequaertiodendron maga­ originated. lismontanum to synonymy under Zeyherella magalismon- tana (Sond.) Aubrev. & Pellegr. The genus Bequaertiodendron was established by De Wildeman (1919) and was based on Bequaert 2483 from Steyn (1990), in an attempt to retain Bequaer­ the Belgian Congo. Chrysophyllum magalismontanum tiodendron magalismontanum. used macromorphological Sond. was considered to be congeneric with the type spe­ characters of female and hermaphroditic flowers from cies of the genus, namely Bequaertiodendron congolense gynodioecious populations to evaluate Liben’s concept of De Wild. This meant that the concept of Bequaertioden­ B. magalismontanum sensu lato. According to Steyn dron had to be broadened, and this was done by Heine (1990), characters which distinguish the southern African & Hemsley in 1960, hence Bequaertiodendron De Wild, B. magalismontanum from the tropical African material emend. Heine & J.H. Hemsley. are the presence of a stylar compitum, an ovarial nectary. 236 Bothalia 23,2 (1993) introrse apiculate stamens, or staminodes derived from Meeuse: 339 (1960). Bequaertiodendron natalense (Sond.) Heine & such stamens, and a tendency in female plants to increase Hemsley: 308 (1960). to ten the number of corolla lobes, of epipetalous staminodes and of locules. It is also emphasised that the ACKNOWLEDGEMENTS staminal structure of southern African plants and En- Dr Donald J.B. Killick is thanked for constructive criti­ glerophytum (sensu Liben 1989) differ considerably. In cism. the latter genus short filaments are united in a staminal tube that adheres tightly to and completely covers the pis­ til and bears nonapiculate anthers. REFERENCES AUBREVILLE, A. 1960. Notes sur les Sapotacees de l'Afrique equatoriale. Notulae Systematicae 16: 253. Pennington (1991) reduced Bequaertiodendron to syn­ AUBREVILLE, A. 1971. Essais de geophyletique des Sapotacees 2. onymy under Englerophytum, because of the rediscovery Adansonia 11: 425. of authentic material of the type species of E. ste- AUBREVILLE, A. & PELLEGR1N, F. 1958. Rehabilitation de deux lechanthum (Aubreville 1971). He pointed out that closely genres de Sapotacees. Bulletin Societe botanique de France 105: related species groups in Sapotaceae are often highly vari­ 35-37. able in the number of their floral parts. An examination AUBREVILLE, A. & PELLEGR1N, F. 1959. Rectification au sujet de of the Englerophytum complex throughout its range Sapotacees africaines. Bulletin Soc iete botanique de France 106: showed that there was no clear distinction between free 22, 23. stamens and those fused into a tube, but rather a range of DE WILDEMAN, E. 1919. Sur quelques especes congolaises de la conditions from free, through slightly fused to completely famille des Sapotacees. Revue Zoologique Africaine 7, Sup­ plement Botanique: 21, 22. fused. As a whole, the complex is characterised by a ten­ dency towards fusion. HEINE, H. & HEMSLEY, J.H. 1960. Notes on African Sapotaceae II. The genus Bequaertiodendron De Wild. Kew Bulletin 14: 304. HEINE. H. & HEMSLEY. J.H. 1968. Sapotaceae. In E. Milne-Redhead Concerning the genus Bequaertiodendron in southern & R.M. Polhill, Flora of tropical East Africa: 18-24. Whitefriars Africa, it is clear that Englerophytum is the earlier name Press, London. and Pennington's treatment of the genus should be fol­ KRAUSE. K. 1914. Englerophytum. eine neue afrikanische Gattung der lowed. Nevertheless, on the basis of characters pointed Sapotaceen. Botanische Jahrbiicher fiir Systematik, Pflan- out by Steyn (1990), it is recommended that more research zengeschichte und Pflanzengeographie 50: 343. should be undertaken at species level within Engle­ LECOMTE, H. 1919. Quelques Sapotacees africaines. Bulletin du Mu­ rophytum. The NBI agrees with Pennington's views and seum d'histoire naturelle 25: 189-193. accepts the following name changes: LIBEN. L. 1989. La veritable identite des genres et especes confondus sous le nom de Bequaertiodendron magalismontanum (Sond.) Heine & Hemsley (Sapotaceae) en Afrique centrale et oc- Englerophytum magalismontanum (Sond.) Pen­ cidentale. Bulletin du Jardin botanique national de Belgique 59: nington, The genera of Sapotaceae: 252 (1991). 151-169. MEEUSE. A.D.J. 1960. Notes on the Sapotaceae of southern Africa. Chrysophyllum magulisnu>ntanum Sond.: 72 (1850). Pachystela Bothalia 7: 339. magalismontana (Sond.) Lecomte: 189 (1919). Zeyherella magalismont­ PENNINGTON, T.D. 1991. The genera o f Sapotaceae. Royal Botanic ana (Sond.) Aubrev. & Pellegr.: 37 (1958). Pouteria magalismontana Gardens, Kew, Surrey. (Sond.) Meeuse: 335 (1960). Bequaertiodendron magalismontanum SONDER, W. 1850. Beitriige /ur Flora von Siidafrika. Linnaea 23: 72. (Sond. I Heine & Hemsley: 307 (1960). STEYN. E M.A. 1990. Macromorphological floral characters in Bequaertiodendron magalismontanum: possible taxonomic sig­ (Sond.) Pennington, nificance. South African Journal of Botany 56: 440-442. The genera of Sapotaceae: 252 (1991). C. L. BREDENKAMP* andG.F. SMITH* Chrysophyllum natalense Sond.: 72 (1850). Boivinella natalensis (Sond.) Pierre ex Aubrev. & Pellegr.: 37 (1958). Neoboivinella natalensis ♦National Botanical Institute. Pri\ate Bag X101. Pretoria (XX) 1. (Sond.) Aubrev. & Pellegr.: 23 (1959). Pouteria natalensis (Sond.) MS. received: 1993-10-05.

ASCLEPIADACEAE

VALIDATION OF THE COMBINATION ASPIDONEPSIS REENENSIS (ASCLEPIADACEAE): THE TYPE SPECIES OF THE SUBGENUS UNGUILOBIUM

In our paper describing the new African genus As- Folia ascendentia, margine manifeste revoluta. In- pidonepsis (Nicholas & Goyder 1992), the basionym ot florescentia 4—11 -flora. Corolla reflexa: pagina abaxialis A. reenensis was inadvertently omitted, so invalidating the pubescentia. Coronae lohi ad columnam staminalem circa proposed combination. As this invalid combination is the 1 mm super insertionem corollae conjuncti. cucullati; ap­ pendix proximalis ad apicem deltato-falcata et apicem type of the proposed subgenus Unguilohium this too be­ styli aequans vel superans impendensque; extremum dis­ comes invalid. The present note is intended to correct tale coronae appendice parva omatum (A. reenensis) vel these omissions. appendice carente 04. shehae); sinus profundus rimi- formis.

Lnguilobium A. Nicholas & D.J. Goyder ex A. Nicholas & D.J. Goyder in Bothalia 22: 31, 32 (1992), TYPUS.—Aspidonepsis reenensis (N.H. Br.) A. Nich­ subgen. nov. olas & D.J. Goyder vide infra.