40 



/ 



MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 9, No. 1 



Plectranthus cf. zombensis Bak. in Thiselton-Dyer, Fl. Trop. Afr. 5: 402. 1900. 



Zomba District: Zomba Plateau, one plant found in open grasslands, perennial 

 herb 80 cm. high, flowers blue, branches of the inflorescence and calyces dark 

 purple, 1800 m., May 31, 1946, 16115. 



P. zombensis is endemic to Nyasaland. The type (in bad condition) shows 

 the young stage with the flowers just coming out, and the inflorescences far more 

 contracted than in Brass 16115. More material is needed to prove that the two 

 stages link up, as I believe they will. 



Plectranthus albo-violaceus Gurke, Bot. Jahrb. 30: 397. 1901. 



Mlanje District: Upper Ruo River, herb, some leaves purple beneath, corolla 

 white with purple throat, about 850 m., July 4, 1946, Vernay 16659*. Mlanje Moun- 

 tain; Likubula-Tuchila Divide, plentiful in forest-openings, shrub 2-3 m. high, 

 strongly aromatic, flowers white with a purple spot on upper lip, 2000 m., July 9, 

 1946, 16755. North Nyasa District: Nyika Plateau, common in montane forest 

 undergrowth, shrub 2 m. high, flowers white, upper lip violet, 2400 m., Aug. 18, 

 1946, 17317. 



P. albo-violaceus has been previously recorded only from the Rungwe Moun- 

 tains in southwestern Tanganyika, and there is a sheet of the type-number — 

 Goetze 1140 — in the herbarium of the British Museum (Natural History), which 

 I have examined. A sheet collected in the Ocotea forests of the Eastern Aberdares 

 in Kenya Colony (H. M. Gardner 1395 in Herb. Kew.), previously and wrongly 

 named P. elegans Britten, and another sheet collected at 1200 m. on Mount 

 Chiradzulu, Nyasaland, by A. Whyte (s.n.) in Herb. Kew., which has until now 

 masqueraded as an unnamed species of Coleus, are obviously conspecific. 



I feel sure that our plant is rightly identified with P. albo-violaceus, but men- 

 tion must be made of P. mandalensis Bak. in Thiselton-Dyer, Fl. Trop. Afr. 5: 

 405 (1900). This was described from a cultivated specimen originally from 

 Mandala, Nyasaland, and no other specimens have since been identified with it. 

 The type of P. mandalensis has the leaves shortly pointed or subacute at apex 

 and subcordate or truncate at base, not rather attenuate at apex and base as in 

 P. albo-violaceus. The inflorescences in the two species seem the same. I had 

 at first, before seeing P. albo-violaceus , identified Mr. Brass* specimens with 

 P. mandalensis, though with reservations. The leaves of P. albo-violaceus vary a 

 good deal in depth of serrations (3-10 mm.), which in some are so deep that they 

 might be called lobes. The length of the lower calyx-teeth also varies. I still feel 

 it quite likely that P. albo-violaceus and P. mandalensis may be conspecific, 

 especially as the latter was grown under the alien conditions of Kew; but until 

 more evidence has come I think it better to use a later certain name than an un- 

 certain earlier one. 



P. albo-violaceus is clearly related to P. elegans Britten, q.v., but with 

 smaller and paler corollas, and bracts fringed with long flexuous multicellular 

 hairs ± violet-suffused on the cross-walls. 



Brass 17317 is more hairy than the other two numbers cited, but not signifi- 

 cantly so in my opinion. 



Plectranthus manganjensis §ak. in Thiselton-Dyer, Fl. Trop. Afr. 5: 406. 1900. 

 Plectranthus albocaeruleus N. E. Br. Kew Bull. 1901: 130. 1901. 



Cholo District: Cholo Mountain, occasional in rain-forest regrowths, herb 1 m. 

 high, plant fleshy, flowers white and purple, 1200 m., Sept. 19, 1946, 17651. 

 Endemic to Nyasaland. 



I cannot distinguish the above-named plants. P. manganjensis is very close 

 to P. violaceus Gurke, but with narrower inflorescences, different indumentum on 



