4 



2d8 



compressed, with flattened sides, rounded ou the back, taperin<r 

 upwards, rather abruptly ending in a stout hard spine-like point 

 4-5 m. long, with a triangular channel as broad as the leaf all 

 down the face and 5-12 shallow grooves or impressed lines down 

 tiie sides and back, dark green, with the grooves or longitudinal 

 lines blackish-green, without transverse markings; mam-ins of 



reddish-brown, with white mem 



usually spreading to a slightly wider breadth than the rest of the 

 leat at the basal part. Flower-stem up to 6i ft. hiffli, 

 paniculately branched at the upper three-fourths of its length; 

 branches ^-If ft. long, curved, ascending, the lower branched, 

 the upper simple. Flowers 4^7 in a cluster; pedicels U-2 lin. 

 ong, jointed above the middle; tube 21-3 lin. long; lobes 31 lin. 

 long^ linear, obtuse.-Baker in Kew Bulletin 1892, p. 129; Oliver 

 m Hooker, ^Ic. Plant, t. 2269; Schweinfurth in Bull. Herb. 



7 Ost-Afr. C. 



e Wilde- 



ih 7^p''4,^PP'.^^- 2' P- "^8; Engler, Pllanzenw Os 



17'J T l^' ?,*^^'V\^ ^'^"g^' P- 631, but not of pp. 624 

 tIJV -r''* a'.^ ^'^■- ^'}' Eritrea, 253; Fiori Boschi e Piante 

 Legn. Eritrea, 104; not of Gerome and Labroy. 



AW It ''' Za^}% '^*f'^°"' '^ Schweinfurth. Eritrea. 



SoMALiLAND ^\ ithout precise locality, St^ce ! 

 a ]wLl T f ascription is made partly from a photograph and 

 from ??Jn T^^'-'',^'^ ^* Kew, bott of which wer! received 

 from Somahland, and partly from Schweinfurth's specimen 31. 



hi. pditFn ^ w '^ h ft'^"!^^ h'^s figured and described in 

 his edit^n of Hemprich & Ehrenberg, s/mbolae Physicae, p. 27, 



necls thK^' ^- ^/'r ?-^^"^.i«> I believe, a tot^ally different 

 species, wnicli caiinnf. T^p rlofa^rv.^.^.! „,.i.n i- • i y « -. 



species, wliicli ca 

 obtained. 



Arabi.ltrr''^'''*^ t\'\''l'-' ^^^* ^^'^' V^^nt is a native of 

 Arabia and German East Africa, but he informed n,P iT^nf h^ 



hfld nnf or.rv,r.o>, 1 T • 1 , „ ' '"' ^^ miormea me that ne 



other no '3:^1 living plants from these two regions with each 



doubt as fn 1 ^^'''^ Pl"^*' ^ therefore htve very much 



evidenrent n! Trf""'"' °* t^^^ identification, as from the 



^':^ztj::::ri^:'^-^^^ -^^^ ^^ ^^^ -^^ous species, 



ml'ing'ro'er'lc.^'"'^'^'*^''^" '^ "^^^ ^^ ^^^^^ -^ Egypt for 



rootsto^L'^^t!; J;, is^r: .S:?h^^^^^ r\^ ^ ^*^^* r-p^ 



Wi spine Ho ;{f^t"^' ■7'''''^ from^ase to apex into a 

 band J ft t J^'*^'^ Pr"i r^'.^^^ ^^-^^^^ a chestnut-brown 

 the apex on t lie f..' l^°-.*i ^x.^^' ^^"^' channelled up to 

 he sides and Wk +]/"! '"'^^i '^^"^^ ^"^^ shallow grooves on 

 mar^fs at he T.i;.l '^^'^ f ^''''^°^ ^"^ acute with erect 

 Sorf obtul V X /"'^^ ^^d becoming more and 



^^arl/rtjafaf th; 7eaf tTCds leT"'^'^-^ ^"^"^^^'^^ ^^'^ 



icaj. lowaras the base; margins acute, 



