1954} 



PLANTS COLLECTED IN NY AS ALAND 



433 



similar to that of C. aequatorialis on the leaves of all gatherings of C, nitidula 

 examined, though the "gland" is there often unpigmented, and in the Angola spec- 

 imens apparently always so. It should be noted that this pigment is not restricted 

 to the "gland" but may cover the whole leaf-tip. I have been unable to detect 

 any significant differences between C. nitidula, C. aequatorialis, and C. nitidula 

 subsp. angolensis. The width of the sepals in the :f flowers, the degree of their 

 connation, and the number of stamens seem to vary considerably. Kenya material 

 has four stamens and rather deeply divided lobes, but this may readily be matched 

 in other parts of the range. The narrower sepals of C. nitidula subsp. angolensis , 

 mentioned by Weimarck, are narrower only than in some specimens of typical C. ni- 

 tidula. Thus in Wehvitscb 1277 (Angola), one of the specimens on which the sub- 

 species was based, they are 1.6-1.75 mm. -wide; while, among specimens of typi- 

 cal C. nitidula, Brass 16457 and Wild 1442 (S. Rhodesia) have them 1.4 mm. wide 

 and in Fries, Norlindh & Weimarck 3685 (S. Rhodesia) they are 1.5 mm. wide. 



The three plants under discussion appear then to be geographical variations, 

 separable only by slight differences in foliage, not of specific, or in my opinion, 

 of subspecific significance. I therefore propose the two following varieties:- 



Cliffortia nitidula var. angolensis (Teim.) Brenan, stat. nov. 



Cliffortia nitidula subsp. angolensis Weim. Monogr. Gen. Cliffortia 47. 1934. 



Leaflets normally linear, parallel-sided, and green at the apex, not oblanceo- 

 late as in typical C. nitidula. 



Cliffortia nitidula var. aequatorialis (R. E. & T. C. E. Fr.) Brenan, comb. nov. 

 Cliffortia aequatorialis R. E. & T. C. E. Fr. Notizbl. Bot. Gart. Berlin 8: 649. 1923; 

 Weim. Monogr. Gen. Cliffortia 50. 1934. 



Leaflets as in the preceding variety but purple-tipped, especially when young. 



A single gathering, Welwitscb 1277c (on Morro de Lopollo, Apr. I860), has 

 been made in Angola of a plant in leaflet-shape very close to typical C. nitidula, 

 but differing in having most of the internodes glabrous. More material must show 

 whether this is a constant variation. 



Cliffortia nitidula subsp. pilosa Weim. Monogr. Gen. Cliffortia 49 (1934) ap- 

 pears to be quite distinct. 



SAXIFRAGACEAE 



Choristylis rhamnoides Harv. Lond. Jour. Bot. 1: 19. 1842. 



Choristylis shirensis Bak. f. Trans. Linn. Soc. II. Bot. 4: 13. 1894. 



Mlanje District: Mlanje Mountain, southwest ridge, in stunted forest on banks 

 of a stream, tree 3-4 m. high, 2120 m., June 28, 1946, 16507. Uganda to the Cape. 



There is a tendency for the tropical plants to have broader and more sharply 

 pointed leaves than those at the Cape, but the difference is at best very slight; 

 it is also inconstant and I consider decidedly not a specific one. Dissection of 

 the flowers shows no other difference worth worrying about. 



CRASSULACEAE 



Crassula pentandra (Royle ex Edgew.) Schonl. in Engl. & Prantl, Nat. Pflanzenf. 

 3 2a : 37. 1890. 



Tillaea pentandra Royle ex Edgew. Trans. Linn. Soc. 20: 50. 1846. 



Zomba District: Zomba Plateau, occasional in tufts on exposed mossy rocks, 

 herb 10 cm. high, fleshy, flowers green, 1400 m., May 30, 1946, 16092; ibid., con- 

 fined to dry sunny rocks and moist or wet open seepage slopes, herb 5-15 cm. 

 high, fleshy, flowers greenish, 1500 m., June 5, 1946, 16261. Anglo-Egyptian 

 Sudan to Nyasaland and Angola, also on the Cameroon Mountain in West Africa, 

 and in Socotra, Arabia, and India. 



