52 AJhanij Museum Rvvonh. 



could have taken ])lace, and yet capsules and seeds were produced. 



Self-fertiliaatUm is therefore possible, as in all other species of 

 AnaccDnpscros, and is, I think, th" rule in all of them, even in the 

 species with showy flowers. Accurate observations in their native 

 habitats are, however, very desirable. As far as my observations 

 go, the beautiful show-apparatus of the flower in this and other 

 species is solely used by the plant to press the anthers against the 

 stigma when the flower closes again. It takes over G weeks from 

 the time the flower has opened, until the capsule is ready to shed 

 the ripe seeds. Just before this takes place, the upper internodes 

 of the flowering branches elongate to quite twice their original size or 

 even more, the stipules and involucral leaves become patent, in- 

 stead of being closely appressed. 



A fifth species of Anacampseros belonging to the section 

 A von i a was sent to me by Mr, G. Alston some years 

 ago, but it has never flowered yet. Its leaves and 

 stipules are multifarious. Its caudex is branched after the manner 

 of A. ustulata. The leaves are imbricated, green, semicircular 

 in outline, convex on the back, flat on the face, about 2 mm. broad, 

 1mm. high, the stipules are lanceolate, about 3 mm. long, hot 

 imbricate, curved outwards in the upper portion, scarious with a 

 broad, median, pale yellowish-green line. 



As these characters distinguish it from all other known species 

 of Anacam2)seros, I venture to give it a specific name and ca 1 it 

 Anaccunpsercs (AvoniaJ recur vata, Schonl. n. sp. 



Anacainpseros ustulata, E. Mey. (Flora Capensis, II, p, 383). — 

 This species appeals to be fairly common on the m.ountains of the 

 Eastern Karroo^ and extends towards the East to the Stormberg 

 range, whence I received live specimens from Mr. T. R. Sim, F.L.S., 

 who also sent it from the neighbourhood of Naauwpoort. I find 

 that the flowers are somewhat lai-ger than the involucre, not equal- 

 ling it, as stated in the Flora Capensis ; the petals are white, 

 obovate, slightly rccu ved at the apex. There are eight stamens, 

 alternating with the petals, and 3 belonging to an epipetalous 

 incomplete who/l ; the filaments are subulate, the anthers oblong. 

 The stamens are all placed with the petals on a disc, which sur- 

 rounds a dei)ression in which the ovary is sunk to about half its 

 length. The ovary is nearly globose, the style cylindrical, the 

 stigma 3-branched. The })laceiitation, is free-central (without any 

 counection with the wall of the ovarv above. This observation 



