432 



MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 5 



form from Tanganyika Territory, Nyasaland, Portuguese East Africa, and S. 

 Rhodesia. 



Rubus ellipticus Sm. in Rees, Cyclop. 30: no. 16. 1819; Focke, Bibl. Bot. 17 ra : 

 198. 1911. 



Zomba District: Zomba Plateau, frequent in grassy clearings, shrub to 3 m. 

 high, very robust, canes stout, red-hairy, leaves grey-green beneath, flowers 

 white, 1400 m., May 28, 1946, 16065. Mlanje District: Mlanje Mountain; Luchenya 

 Plateau, plentiful in forest regrowth, shrub up to 6 m. high, canes red-hairy, very 

 stout, arched or scrambling, flowers white, fruit not seen, 1890 m., June 30, 

 1946, 16543. 



This species is not a native of Africa, but has a wide distribution in Asia 

 (India, Ceylon, Burma, China, Philippines), and is said by Focke to be natural- 

 ised in Jamaica. It has evidently been on Mlanje for some years. I am indebted 

 to my friend and former colleague Mr. A. C. Hoyle, Curator of the Herbarium of 

 the Imperial Forestry Institute, Oxford, for informing me of the following sheet in 

 that herbarium: Nyasaland: On Mlanje at Woodmen's camp, fruit yellow, small, 

 edible, agreeably acid, Sept. 24, 1929, Burtt Davy 22060. 



In the same herbarium there is a sheet of this species taken from a plant cul- 

 tivated at Amani, Tanganyika Territory (Greenway 1742). 



As Rubus ellipticus is unmentioned in Gustafsson's recent revision of the 

 African Rubi (Ark. Bot. 26A r : 1-68. 1934) and its identification may be thus a 

 matter of difficulty, it may be helpful to mention here the salient features of this 

 very striking species. The most obviously unusual character is the long red- 

 purple setae that more or less densely clothe the stems, petioles, and petiolules; 

 they are present on the inflorescence, but less dense. The leaflets are three per 

 leaf, usually grey- or white-tomentose beneath and obtuse or scarcely pointed at 

 the apex. The inflorescence is composed of short axillary branches and a dense, 

 many-flowered extra-axillary portion. The flowers are medium-sized. This com° 

 bination of characters (and especially the dense setae) will separate R. ellipticus 

 from all the species described from tropical Africa. 



Hagenia abyssinica (Bruce) J. F. Gmel. Syst. Nat. 2: 613. 1791. 



Banksia ["Bankesia"] abyssinica Bruce, Trav. Egypt, Arabia, Abyss. & Nubia 5: 

 73. 1790. 



North Nyasa District: Nyika Plateau, plentiful in upper montane forest of es- 

 carpment, also in second-growth forest, tree to 12 m. tall and to 30 cm. in diam- 

 eter, bark brown, flaky, leaves viscid, inflorescences on lateral branches which 

 dry and drop off after fruiting, 2350 m , Aug. 17, 1946, 17298. Mountains of east 

 and central Africa from the Anglo-Egyptian Sudan and Abyssinia to Nyasaland. 



Cliffortia nitidula (Engl.) R. E. & T. C. E. Fr. Notizbl. Bot. Gart. Berlin 8: 649. 

 1923; Weim. Monogr. Gen. Cliffortia 47. 1934. 



Cliffortia lineari folia Eckl. & Zeyh. var. nitidula Engl. Bot. Jahrb. 26: 376. 1899. 



Mlanje District: Mlanje Mountain; Luchenya plateau, plentiful in forest re- 

 growth and in forest-border shrubberies, shrub up to 4 m. high, leaves convex, 

 glossy, flowers greenish, 1870 m., June 27, 1946, 16457. Kenya to S. Rhodesia 

 and Angola; a distinct subspecies in South Africa, according to Weimarck. 



Weimarck (op. cit.) separates a species from Kenya, C. aequatorialis R. E. & 

 T. Co E« Fr., from C. nitidula by the usually more revolute .leaflet-margins and 

 by the leaflets having a small red gland at the apex. The leaflet-margins of C. 

 aequatorialis are more constantly and strongly revolute than in typical C. ni- 

 tidula, though exactly comparable with those plants from Angola which Weimarck 

 has called C. nitidula subsp. angolensis; and occasional leaves equally strongly 

 revolute may be seen on plants of typical C. nitidula. I have observed a "gland" 



