1954] 



PLANTS COLLECTED IN NYASALAND 



21 



stylus filiformis, circiter 2-3 cm. Jongus, pubescens; stigmatis lobus anticus vix 

 2 mm. longus, posticus ad dentem minutum redactus. Fructus ignotus. 



Kota-kota District: Nchisi Mountain, in rain-forest undergrowth, herb 50 cm. 

 high, flowers with upper lip brown, lower lip purple, 1500 m., July 29, 1946, 

 17021*; ibid., locally common in rain-forest undergrowth, herb 40-60 cm. high, 

 young parts and inflorescence viscid, flowers with upper lip purple, lower lip 

 brown, 1500 m., July 30, 1946, 17042 (TYPUS in Herb. Kew.); ibid., common in 

 openings in rain-forest, Sept. 5, 1929, Burtt Davy 21281 (Herb. Kew.). 



I can find no close relatives of this species in eastern Africa. It appears to 

 come nearest to B. debilis Burkill from the Cameroons, but is abundantly distinct 

 from it. A poor hitherto un-named specimen collected by Burtt Davy is the only 

 other gathering of the species that I can find, and it also comes from Nchisi. It 

 seems that B. oligantha is an endemic species in the forests of that mountain. 



Mimulopsis solmsii Schweinf. Verh. Zool.-JBot. Ges. Wien 18: 677. 1868. 

 Mimulopsis violacea Lindau, Bot. Jahrb. 17: 185. 1893. 



Mimulopsis thomsoni C. B. Clarke in This el ton- Dyer, Fl. Trop. Afr. 5: 55. 1899. 

 Mimulopsis spathulata C. B. Clarke in Thiselton-Dyer, Fl. Trop. Afr. 5: 55. 1899. 

 Mimulopsis glandulosa (Lindau) Bullock, Kew Bull. 1931: 275. 1931, quoad synon. 



M. thomsoni, haud Epiclastopelma glandulosum Lindau, non M. glandulosa Bak. 



Jour. Linn. Soc. Bot. 25: 338. 1890, comb, illegit. 

 Mimulopsis velutinella Mildbr. Bull. Jard. Bot. Brux. 17: 86. 1943. 



Zomba District: Zomba Plateau, in grassy edges of rain-forest, herb 2 m. high 

 of weak scrambling habit, flowers lavender with a yellow blotch on the lower lip, 

 1450 m., June 3, 1946, 16181. British Cameroons and highlands from Abyssinia, 

 eastern Belgian Congo to S. Rhodesia. 



I have been unable - to consider the species whose names appear above to rep- 

 resent more than a single variable species. There has been much confusion in 

 the past caused by a number of unrelated circumstances. Schweinfurth described 

 M. solmsii as having yellow flowers. Whilst it cannot be proved that this was an 

 incorrect statement, circumstantial evidence strongly suggests that the colour of 

 the corollas was not yellow when the plant was growing. They dry, however, just 

 as if they had been yellow when fresh, which fact may have helped to suggest 

 his choice of name for the genus. All the specimens that have been examined of 

 the above-mentioned species, to which collector's colour notes are attached, have 

 white or white-tinged lavender, violet, or mauve corollas. There is, however, a 

 yellow spot or honey-guide inside the inferior lip. 



Epiclastopelma glandulosum Lindau, which is said to have yellow flowers, 

 was in my opinion wrongly placed by Clarke under Mimulopsis thomsoni. It seems 

 from description and from observations by Mildbraed (Notizbl. Bot. Gart. Berl. 

 11: 1081. 1934) that the genus Epiclastopelma is distinct from Mimulopsis and 

 should be maintained. Material collected recently by Mr. A. A. Bullock confirms 

 my opinion that Epiclastopelma is a distinct genus of which at least one species 

 has not yellow flowers. The Thomson specimens cited by Clarke under M. thomsoni 

 have no colour notes, so that the key character "corolla yellow" in the Flora of 

 tropical Africa relating to M. thomsoni was not founded on fact. Furthermore, 

 there is no evidence on the sheet of Scott Elliot 6960 at Kew that it had yellow 

 flowers, whilst it seems almost certain that a duplicate of this gathering at Berlin, 

 with the collector's number wrongly transcribed as t! 6060" was the holotype of 

 M. spathulata C. B. Clarke. 



Having disposed of the mythical character "corolla yellow," we now have to 

 try to distinguish the first five species of Mimulopsis enumerated in the Flora 

 of tropical ' Africa. Geography counts for nothing, as it is possible to match the 

 Abyssinian M. solmsii with specimens from the Cameroons and from Nyasaland. 



