22 



MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vo*l. 9, No. 1 



Shape, size, and indumentum of calyx, all very variable characters, seem quite 

 uncorrelated, as do the degree of toothing of the leaf margin and the presence 

 or absence of a cordate base to the blade. The largest variation seems to occur 

 in East Africa. It may be significant that more than one reliable collector has 

 obtained unwittingly a mixed gathering showing two distinct "forms." Humbert 

 in his gathering 7441 from the mountains west of Lake Kivu has both M. 

 velutinella Mildbr. and M. violacea var. kivuensis Mildbr. These species differ 

 chiefly in the former's not having the glandular and long jointed hairs on the 

 calyx lobes. The Kew specimen of this gathering is somewhat intermediate in 

 this respect. Maitland under his 447 has collected near Kampala a similar mixture 

 of forms, whilst Purseglove under his 723 has both the extreme forms and inter- 

 mediates. Similar mixtures of forms have been collected in Kenya. The fact that 

 these different forms occur within a population of similar-looking plants suggests 

 most strongly that they represent but a single species. It is not, in my opinion, 

 wise at the present state of our knowledge to give names to these variants, and 

 I accordingly adopt the earliest specific name for the aggregate species without 

 attempting to differentiate it further. 



Dyschoriste verticillaris (Oliv.) C. B. Clarke in Thiselton-Dyer, Fl. Trop. Afr. 



5: 75. 1899. 



Calophanes verticillaris T. Anders, ex Oliv. Trans. Linn. Soc. 29: 126. 1875. 

 Dyschoriste decora S. Moore, Jour. Bot. 51: 212. 1913. 



Zomba District: Zomba Plateau, in Brachystegia woodlands, herb 40-60 cm. 

 high, flower with reddish calyx and pale purple corolla, viscid, 1100 m., June 

 17, 1946, 16334. Kota-kota District: Nchisi Mountain, occasional in Brachystegia 

 woodland, perennial herb 70-100 cm. high, stems several, erect from a thick woody 

 stock, flowers dark purple, 1400 m., July 24, 1946, 16889. Kasungu District: 

 Kasungu, local in Brachystegia woodlands, perennial herb 60-80 cm. high, stems 

 several, erect from a thick woody stock, 1000 m., Aug. 25, 1946, 17416. Cholo 

 District: Cholo Mountain, common in Brachystegia woodland, shrub 1-1.5 m. high, 

 stems several, erect, flower pale purple, 1200 m., Sept. 26, 1946, 17826. Southern 

 Tanganyika, south to Portuguese East Africa, and S. Rhodesia, and west to 

 Katanga and N. Rhodesia. 



Phaulopsis imbricata (Forsk.) Sweet, Hort. Brit. .1: 327. 1827. 

 Ruellia imbricata Forsk. Fl. Aegypt.-Arab. 113. 1775. 



Phaulopsis parviflora Willd. Sp. PI. 3: 342. 1801; C. B. Clarke in Thiselton-Dyer, FL 

 Trope Afr. 5: 83. 1899. 



Zomba District: Zomba Plateau, common on moist roadsides in woodlands, 

 perennial, herb about 30-40 cm. -high, branches ascending, several from a more or 

 less fleshy stock, bracts green, viscid with glandular hairs, flowers white, 1450 

 m., June 5, 1946, 16253. Kota-kota District: Nchisi Mountain, gregarious among 

 rocks in Brachystegia woodland, perennial herb, 20-70 cm. high, bracts green, 

 viscid, flowers white, 1400 m., July 26, 1946, 16950. Cholo District: Cholo Moun- 

 tain, frequent in rain-forest second growths, herb more or less 1 m. high, calyx 

 and bracts viscid, flowers white, 1200 m., Sept. 19, 1946, 17644. Widely spread 

 through tropical Africa and Asia. 



As treated in the Flora of tropical Africa, the species of Phaulopsis are highly 

 unsatisfactory. The characters "strobilate, " "hardly strobilate, 99 and "not strob- 

 ilate" are difficult to assess, so that the key is of little use for determining the 

 first nine species. I agree with Clarke in considering his P. longifolia T. Thorns, 

 (not Sims, and therefore an illegitimate name) a variant of P. imbricata (which 

 Clarke calls P. parviflora). Time at present will not allow the dissection and ex- 

 amination of the large amount of material of these "species" now at Kew, so I 



