264 an inch high, ''."hsn et rest the flowering growths are covered be- 

 low v.'ith 2-4 corisceous, blackish sheaths, that are often oblique 

 and longitudinally rugoSe. Growths, v;hen at rest, half -moon- shaped, 

 2-2|- lines broad and 1-li line thick, length not stated, i'ree 

 pairs of leaves sub-clavate; the second pair united to their tips. 

 5'lowers sessile. Calyx 6-lobed, with the ovary-part globosely pear- 

 shaped. Petals white, only a^very old flower examined. Stigmas 

 6, t-isted together, nerrly 2-^- lines long. Oyary slightly det;res- 

 sed at the centre-on the top. Oapsule 3|-4j lines in diameter; 

 valves blackish within; expanding keels sub-triangular, adpressed 

 and flattened, v.'ith toothed edges and broad marginal vangs ; cells 

 roofed with cell-wings. Seeds i- line long, globosely obovate, 

 pallid. 



Conophyllum nanum, L, BqI,, Notes on ^-esemb. , VqI, II^ p. 51, 

 Clanv/illiam Division : -^etv^^een Twenty-four ^ivers and Porterville, 

 on clsyey soil. Leipoldt. 



I have not seen this plant, but it seems to be closely allied 

 to one that ^v, Fischer has described or is about todescribe, under 

 the name of ''. Schickianum, ^isch, , v/hich is e stemless species 

 about li inch high, v:ith the cone-leaves about 5 lines thick, tipped 

 with leaf-points 5-6 lines long, 



N. S. Brown 

 (To be continued.) 



MSSEI..:BRYANTHEI,im.I. 

 Gard. Chron. HI. 8?: 12. 1930. 

 (Continued from nage 264, "^ol. L^iX^I.) 



13 Since the publication in T^e Gardeners* Chronicle, 1925, Vol. 

 IX^VIII, p. 412 and 433, of the nrovisional key to the genera into 

 w'-'ich I had divided the old genus I'^esenbryanthemum, errors in it 

 have been discovered which I wish here to rectify, i'or, as might be 

 expected, in constructing for the first time a key to such a very 

 complicated group of plants as this is, it would be scarcely possible 

 to avoid errors, especially as the material available for the pur- 

 pose was often incomDlete. Iv:ost, however, are but trivial mistakes 

 that do not affect the general accuracy of the key, but there are 

 two that require explanation. They v/ere not discovered until long 

 after the key vras published, and -^ had not intended to correct them 

 until dealing v/ith the genera concerned, but as one has been much 

 exploited by other writers, it 



14 is deemed better to correct them novi and explain how they were made. 

 One error concerns 5 — 



I'-esembryanthemum, Linn, (emended). 



^'''hen I emended the genus ^"esembryanthemum, as detailed in The 

 Gardeners' Chronicle, Vol, LXXVIII^ jp, 232, I selected, on the usual 

 logical r>rincirle as there stated, i". umbellatum, ^inn., as the type 

 of the reformed genus, and explained why. Tet it most unfortunately 

 happens that the structural characters belonging to that species and 

 its allies are not included in the published key, but only those be- 

 longing to other species that I had associated with it, although its 

 own characters v.'ere als: set dov/n in my original key. This error 

 of omission was certainly made during the copying of one of the 

 earlier of the many keys constructed, because it appears in l?hillips* 

 Genera of 3, -^-frican ^lowering Plants, and the ^.SS. for that v;ork 

 I sent to South Africa either in 1924 or early in 1925, months be- 

 fore the key in The Gardeners' Chronicle was published. As the i^S. 



