1954] 



PLANTS COLLECTED IN NYASALAND 



17 



yellowish, 1560 m., March 8, 1950, Wild 3240 (TYPUS in Herb. Kew.). Chindamore 

 Reserve, Ngomakuriram, in damp ground by rill over granite, flower pale mauve, 

 palate yellowish, 1680 m., March 25, 1952, Wild 3778. 



NORTHERN RHODESIA: Abercorn District: Kali Dambo near Kawembi Mis- 

 sion, in wet muddy marsh, sparingly, flowers mauve, 1550 m., May 6, 1952, 

 Richards 1635 A. Dhulu'miti, in marsh, flowers mauve, 1550 m., April 19, 1952, 

 Richards 1463A. 



GESNERIACEAE 45 



Streptocarpus goetzei Engl. Bot. Jahrb. 30: 406. 1901. 

 Streptocarpus rungwensis Engl. Bot. Jahrb. 57: 217. 1921. 



Zomba District: Zomba Plateau, herb, leaves solitary, pendent, apex pale 

 green, base dark green, frequent on mossy rocks in riverine rain-forest, 1400 m., 

 May 28, 1946, 16058; ibid., flower violet, very attractive, common on moist mossy 

 rocks in rain-forest, 1400 m., May 30, 1946, 16081*; ibid., herb, common, but 

 seldom seen in flower, flowers violet, on moist mossy rock faces in rain-forest, 

 1450 m., June 4, 1946, 16198*; ibid., herb seldom seen in flower, flower purplish- 

 violet, common on moist mossy rocks in riverine rain-forest, 1500 m., June 7, 

 1946, 16294*. Mlanje District: Mlanje Mountain; Luchenya Plateau, herb, an ex- 

 ceptionally large leaf measured 50 cm. long and 27 cm. wide, common on rocks 

 and tree trunks in primary forest, 1890 m., July 12, 1946, 16813. Cholo District: 

 Cholo Mountain, common on rocks in rain-forest, herb, flowers violet, leaves 

 solitary, pendent, 1400 m., Sept. 20, 1946, 27670*. Nyasaland and southern 

 Tanganyika. 



In some recent notes on Streptocarpus (Kew Bull. 1939: 70. 1939) I indicated 

 that, from the somewhat scanty material then available, I was unable to separate 

 S. goetzei and S. rungwensis. Brass' gatherings have greatly increased the ma- 

 terial available for examination, but they give no indication of belonging to more 

 than one species and I therefore maintain my previous treatment. 



When writing of S. comptonii Mansf. (I.e.), I mentioned that it showed a tend- 

 ency towards cleistogamy, a phenomenon that was then known among the culti- 

 vated unifoliate species of the genus (that is, subgen. Streptocarpus) and amongst 

 wild caulescent species (that is, subgen. Streptocarpella), but had not been re- 

 corded among wild species of subgen. Streptocarpus. This gap may now be filled, 

 for an examination of 16081 and 16198 shows clearly the short undeveloped 

 cleistogamous corollas adhering to the fruits. In 16081 chasmogamous flowers 

 are present on the same plant. 



The morphology of cleistogamous flowers in Streptocarpus is now well-known 

 (cf. Engler, Bot. Jahrb. 57: 202. 1921; Hill, Proc. Linn. Soc. 154: 1. 1942), and 

 in the caulescent S. nohilis C. B. Clarke, Lawrence (Gard. Chron. 113: 156. 

 1943) has demonstrated the controlling effect of the length of photo-period in the 

 early stages of development, for, if the daily light period falls below about 11 

 hours in the seedling phase, the whole plant is dwarfed and only cleistogamous 

 flowers are produced. It is nevertheless true that under more favourable condi- 

 tions some cleistogamous flowers may be found in a normal inflorescence (as 

 they are in this wild material of 5*. goetzei), and the precise conditions which 

 determine whether a particular bud shall become cleistogamous or chasmogamous 

 have yet to be investigated. 



Streptocarpus hirtinervis C. B. Clarke in Thiselton-Dyer, Fl. Cap. 4 2 : 446. 1904, 

 quoad descr.; in Thiselton-Dyer, Fl. Trop. Afr. 4 2 : 50 7. 1906. 



45 By B. L. Burtt, Royal Botanic Gardens, Kew (now at Royal Botanic Garden, 

 Edinburgh). 



