"ISO EUFEOiunACKJE (Browu). [EHi^liorlia, 



characters tliat are of specific value. The four forms ahove characteiioed are 

 distinct in appearance, but grade into each other in such a manner that anysj>ecific 

 difToi-ence they may po.^sess vanishes in the dried material, and if specifically 

 different they must be characterised from living plants. The different f««rni.^ 

 occur in the same geographical area. 



I ani quite unable to distinguish E, rhomhifoJia and E. racemoso. Tlie former 

 was founded upon Drtge, 8217; and a young immature gnnvth figured for it, but 

 the distiibuted specimens of Di'ege, S217, that 1 have seen, consist of weak 

 branches corresponding to those of Boissier's figure and tiuwering branches 

 coiTesponding to the type of E. raccmosrt in E. Meyer's Herbarium, and GerrnriV& 

 1170 collected in Natal are partlj' exactly as in the figure of E, rliomlifull'i, partly 

 as in E. racemosa. As no desciiption of E. racemusa was published until 1S62, 

 when Boissier first described it in Be CandolWn PnxJrow^if^^ I take the earlier 

 published name of E, rhomlifoJIa for this plant, although certainly not so 

 applicable. Boissier has described the peduncles of £. raremosa as bearing a cyme 

 of 5-7 involucres, but this is quite inaccurate as to the type of E. ravemom in 

 E, Meyer's Herbarium, in which the pedinicles bear only 1-3 involucres^ and in 

 none of the spechuens that I have seen, which are conspecifie with that type, are 

 more than 3 involucres borne upon any i>edunc]e» usually there is only one. 

 In the original description (not in DC, Pvodr,) Boissier quotes *' EchJon (t* Zr/fher 

 DOS. 23, 83" as belonging to this species, but this specimen is F. arcctfthobioUles, 

 Boiss., and the quotation really refers to one specimen only, of which the number 

 is 23, the added number 83 is merely the locality number and not the ninnber of 

 a separate specimen (see Linvna, xix. 583 and xx. 258). Arthrotluunintii rf/rriosuft, 

 Klotzsch & Garclve (EcJdon t£- Zefjlier, 24 partly), is wrungly leferred to E, decv^sata^ 

 E. Meyer, by Boissier in DC. Pvodr. xv. ii. 75. E, rhombiftdkt is readily distinguished 

 from its nearest allies by its quickly deciduous bracts. 



59. E. aspericaulis (Pax in Engl. Jahi-b. xxviii. 26); a small 

 leafless and spineless succulent bush, api^arently 1-li ft. high, 

 dioecious ; branches alternate or perhaps sometimes opposite, 

 diverging from the stem at nn angle of 25°-40°, 1-l.V lin. thick, 

 6-angled, 'with slight furrows between the angles, rough from 

 minute hard papillrp, glabrous ; leaves opposite, rudinientar\-, about 

 1 lin. long, oblong, obtuse, with the apical half deflexed, dark brown 

 or purple-brown, glabrous ; cymes opposite, or from one axil, race- 

 jely scattered along the branches, up to about 4 lin. in diam., at 

 first with 1, finally producing 3-5 in\olucres; bracts about frMn. 

 long and nearly or quite as broad, oblong, very obtuse and apiculate 

 or subtruncate at the apex, rather flexible ; involucres unisexual, 

 about 1| lin. in diara., cup-shaped, glabrous outside, with 4-5 glands 

 and very small oblong or subquadrate slightly toothed lobes ; glands 

 lin. in their greater diam., transverse, oblong or elliptic-oblong, 



1-3 



entire, apparently yellowish or gieenish ; ovary and capsule not 



seen. 



Central Er giox : Calvinia Div. ; Hantani Mountains, J/^yrr ! 



Allied to E, murlcata, Thunb., by its rough branches, but the hubit h different, 

 the branches more slender and the papiilte upon them much more minute and 

 even, not compressed into irregular crenulation.s as in that species. 



60. E. catervifiora (N. E. Br.) ; a leafless, spineless, succulent 

 shrublet, apparently about 9-12 in. high, mostly trichotoraouslr 

 branched, dioecious, only a male specimen seen; main stems or 



