410 



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



young, usually leafless shoots flowering after grass is burned, flowers pink, 480 

 m., Sept. 3, 1946, 27525. Belgian Congo, Tanganyika Territory, Portuguese East 

 Africa, Nyasaland, N. Rhodesia, and Angola. 



Vigna nilotica (Del.) Hook. f. Niger Fl. 311. 1849; Bak. f. Leg. Trop. Afr. 404. 

 1929. 



Dolichos niloticus Del. Fl. figypte 253. pl> 38. 1812. 



Kota-kota District: Benga, Lake Nyasa, plentiful on sandy lake-shores, vine 

 1-2 m., flowers yellow, 470 m., Sept. 2, 1946, 17484. Egypt and Syria, southwards 

 through east Africa to Portuguese East Africa, Nyasaland, and possibly N. 

 Rhodesia. 



I am very indebted to my colleague Mr. R. B. Drummond for giving me the cor- 

 rect name for this plant. 



Vigna dekindtiana Harms, Bot. Jahrb. 30: 93. 1901; Bak. f. Leg. Trop. Afr. 407. 

 1929. 



Zomba District: Zomba, in Brachystegia woodlands on mountain slopes, not 

 common, small trailing vine, more or less scabrid, flowers purple, rather showy, 

 1100 m., May 26, 1946, 16028*. Zomba Plateau, twining in tall grass, vine about 

 1 m. high, flowers bluish-purple, conspicuous, 1400 m., June 11, 1946, 16331. 

 Cholo District: Cholo Mountain, occasional in rain-forest regrowths, vine 3 m. 

 high, flowers purple, standard brown below, 1200 m., Sept. 22, 1946, 27735. Proba- 

 bly widely distributed in tropical Africa. 



This is what has been called the wild form in Africa of the cow pea, Vigna 

 unguiculata (L.) Walp. [V. sinensis (L.) Savi ex Hassk., V. catjang (L.) Walp.]. 

 This is the view of C. V. Piper (U. S. Dep. Agr. Bur. PI. Ind. Circ. 124: 32. 1913; 

 U. S. Dep. Agr. Bur. PI. Ind. Bull. 229- 1912) and Harms [in Engl. Pflanzenw. 

 Afr. 3(1): 687. 1915]. This may well be true, but on account of the narrow de- 

 hiscent pods of V. dekindtiana, I feel that it is better kept as a distinct species. 



I thought that the correct name for V. dekindtiana might prove to be V. alba 

 (Don) Planch, ex Bak. f., described from S. Tome. The type, at the British Mu- 

 seum (Natural History), lacks flowers, but there are other later specimens — G. 

 Watt 7096, Exell 43, 52 — of what is no doubt the same species, also from S. 

 Tome. These have small flowers 17-19 mm. long and very small calyces 4-5 mm. 

 long, including the 1.5-2 mm. -long teeth. Although V. alba is probably not en- 

 demic to S. Tome, I feel it better at present to keep it separate from the large- 

 flowered V. dekindtiana, although later research may show that the differences 

 are only varietal. 



Another source of trouble is the South African Vigna triloba Walp. (Dolichos 

 trilobus Thunb. non L.) This also is I think at present to be kept apart from V. 

 dekindtiana, though I am anything but confident about it. Exell (Cat. Vase. Pi. S. 

 Tome, 163. 1944) was uncertain whether V. triloba and V. alba were or were not 

 conspecific. 



This complex of species urgently requires critical revision. 



Vigna ? gazensis Bak. f. Leg. Trop. Afr. 409. 1929. 



Mlanje District: Mlanje Mountain, southwest ridge, in rocky grassland, vine 

 trailing in grass, flowers purple, 2120 m., June 28, 1946, 16509; Luchenya Pla- 

 teau, twining on brushy growths in a forest-clearing, vine, flowers purple, 1890 m., 

 July 12, 1946, 16803. S. Rhodesia and ? Nyasaland. 



More material, including fruits, is wanted from both countries before the iden- 

 tity of the Nyasaland plant can be really certain. A specimen in Herb. Kew. (Purves 

 60, Mlanje Mountain, Tuchila Plateau, 1830 m., Aug. 1901, a blue-flowered 

 climber) is clearly the same species as the plants collected by Mr. Brass. 



