1954] 



PLANTS COLLECTED IN NYASALAND 



35 



The sinking of species here may appear drastic, but, while W. paniculata is 

 unquestionably most variable, I have tried in vain to discover any satisfactory 

 characters to split it up, even varietally. Except perhaps for the dorsal calyx- 

 lobe, which may be present or absent, the floral characters are within limits very 

 uniform, but size, habit, and foliage are anything but. 



The amount of material of this species from tropical Africa is not large, but 

 it may be approximately sorted into three groups: 



(a) cauline leaves mostly 0.5-2 cm. long, flat, linear-lanceolate; heads or 

 spikes narrow, about 2-3 mm. across; corolla-tube about 1 mm. long; fruits small, 

 about 0.5 mm. long. This is S. lacunosa Klotzsch and as interpreted by Rolfe in 

 the Flora of tropical Africa, 



(b) cauline leaves mostly 0.5-1 cm. long, linear; heads or spikes often broader, 

 about 4 mm. across; corolla-tube 1-1.5 mm. long; fruits mostly larger, about 0.8- 

 1 mm. long. This includes Walafrida cecilae and W. fleckii, and, I consider, 

 typical W. paniculata. 



(c) cauline leaves very short and narrow, up to 3 mm. long; corolla as in (a); 

 fruits about 0.7 mm. long. This is represented by /. Gordon Read 22 (Herb. Kew.) 

 from N. Rhodesia; S. amboensis probably comes here also, and possibly also 

 S. dinteri. 



When specimens from other areas are examined these groups break down. Thus 

 the difference in fruit-size between (a) and (b) is altogether bridged by speci- 

 mens of 5. muralis from Madagascar, and the leaves vary enormously in S. African 

 specimens. In this connection I very much doubt whether the South African 

 Walafrida densiflora (Rolfe) Rolfe (see Fl. Cap. 5 1 : 127. 1901) can be maintained 

 as a separate species from paniculata; in tropical Africa certainly I have no doubt 

 that the various leaf-forms from filiform to linear-lanceolate belong to one 

 species.] 



[9. Walafrida alopecuroides (Rolfe) Rolfe in Thiselton-Dyer, Fl. Cap. 5 1 : 117. 

 1901. 



Selago alopecuroides Rolfe, Jour. Bot. 24: 175. 1886; in Thiselton-Dyer, Fl. Trop. 

 Afr. 5: 272. 1900. 



Angola and Southwest Africa. An exceedingly distinct species, immediately 

 recognisable by its long, dense, tail-like inflorescences.] 



VERBEN ACE AE 48 



Lantana trifolia L. Sp. PI. 626. 1753. 



Lantana salvifolia sensu Bak. in Thiselton-Dyer, Fl. Trop. Afr. 5: 276. 1900, pro 



parte; non L. salvifolia J acq. 

 Lantana meamsii Moldenke, Phytologia 1: 421. 1940. 



Zomba District: Zomba, Brachystegia woodlands of mountain sides, one plant 

 seen, shrub 1.5 m. high, flower white or pinkish, fruit purple, somewhat fleshy, 

 1100 m., May 26, 1946, 16030. Zomba Plateau, frequent in open woodlands, shrub 

 1.1-5 m. high, flower corolla white or pinkish, with yellow centre, fruit purple, 

 1400 m., May 28, 1946, 16069. Cholo District: Cholo Mountain, occasional in rain- 

 forest regrowths, shrub 1-2 m. high, flower pinkish white, fruit purple, 1200 m., 

 Sept. 21, 1946, 11119. Widespread throughout tropical Africa, America, and (? in- 

 troduced into) Asia. 



After some hesitation, I have decided to refer all the three gatherings listed 

 above to Lantana trifolia L. They are untypical specimens, with opposite leaves 

 and closer indumentum than is usual in the species. Typical L. trifolia has dis- 



48 Lantana and Lippia by R. D. Meikle, Royal Botanic Gardens, Kew. 



