596 CXXUI. EUPHORBIACEX (BROWN). [ Haphorbia, 
long and 8-10 lin. in diam., deeply 3-lobed as seen from above (globose- 
triangular and reddish when alive, Schweinfurth), glabrous; cell-walls 
very thick and woody. Seeds slightly compressed-subglobose, 1 3-2 lin. 
in diam., with a slight furrow on one side, smooth, light grey or whitish, 
slightly marbled, with a dull surface.—Z. abyssinica, Schweinf. in Bull. 
Herb. Boiss. vii. App. ii. 317, partly; Pax in Engl. Jahrb. xxxiv. 
77, partly. 
Wile Land. Eritrea: vicinity of Acrur, Schweinfurth, 1351, partly! 
There appears to be some confusion with regard to the succulent  tree-like 
Euphorbias growing in Eritrea which Schweinfurth and others have referred to 
E. abyssinica. None of the specimens which I have seen belong to tht species. 
Schweinfurth and also Pax certainly include two or more species under that name. 
The former states that it varies in appearance, growing 10 to 30 ft. in height, 
sometimes with a short trunk and a large crown of longer branches, sometimes with 
the trunk as tall as the obconic flat-topped crown. Seedling plants are 3- scon 
becoming 4-angled. Young unbranched trunks are sometimes up to 9-angled. ‘The 
main branches are 4~6 in. in diam. and 6—9-angled and the secondary branches about 
5-angled. Leaves cf young shoots and seedlirgs are up to 2 in. long, oblanceolate, 
slightly cuspidate at the apex, create at the base, passing into petiole. The spines 
are rarely more than 4 lin, long, grey with black tips, in pairs about 7 lin. apart. I 
do not know if this description refers to the Acrur plant or not, as I have not seen 
leaves of the plant I describe, but only a seedling plant, with stem-sections and locse 
flowers and fruit. I therefore restrict my description of H. acrurensis to the Acrur 
plant with 6-7- or more angled stems, to which I suppose the pedunculite cymes 
distributed with the stem-sections may beloug, Schweinfurth & Riva, 1351, 
collected in 1892 (not 1891). Unfortunately fragments with sessile fruiting ¢ymes 
have also been distributed under the same number and date, but the structure of the 
spine-shields and flowering-eyes on the scraps of stem-angle with these fragments 
and that of the cymes and involucres is identical with those of B. disclusa, N. E. Br., 
to which species I believe they belong. Schweinfurth 1351 of the 1891 collection 1s 
probably a variety of EH. Erythree, N.E. Br. All these have been distr.buted as 
H. abyssinica, Raeuschel, from whieh they differ totally in having the branches 
constricted into parallel-sided or slightly conical (net elliptic or orbicular) segments. 
182. E. Erythreez, V. #. Br. Arborescent, succulent, leafless: 
spiny. Branches ascending like the arms of a candelabrum, slightly 
constricted into subparallel-sided joints of unequal length, 2-3 1. 
diam. and 5-7-angled in the specimens seen (but according to Seas 
up to 44 in. in diam., and at first 3-angled, becoming 4-angled), ie 
the solid central part 1-14 in. thick and as much or more nc 7 
the angles are broad, dark sap-green; angles wing-like, $-1} 1D- bie 
and 14-2 lin. thick at the edges, straight or wavy, with even oF igre 
tocthed margins and separated by triangular channels 2-1 in. be 
perhaps more) deep, growing out into flat faces with age. eee 
rudimentary, scale-like. Spines 14-44 lin. long, stout, in pairs 3 ly 
apart, diverging, on suborbicular shields 2-3 lin. long and bron 
placed, but not connected by a horny border, brown, becoming ee 
Flowering-eyes touching or merged into the spine-shields. ae 
and fruit not seen nor described.—F. Candelabrum, var. Erythret, 
Berger, Sukk. Euphorb. 73. 
Nile Land. Eritrea; Harassa Valley, Schweinfurth, 226! 227! and " 
specimen ! 
ivated 
