t 



224 EUPHORBTACE^ (Browii). [EnphorUcL j 



but as I have stated in the Flora of Tropical Afrira, vi, § i. 471, they are always 

 developed under and sometimes at a consideraLle distance below the leaf or leaf- 

 scar and cannot be stipules in the ordinary sense of the term, I do not understand 

 what their real relationship to the leaf is. Besides these pairs of spines, true 

 stipules are, however, sometimes developed on the succulent species ; sometimes 

 they take the form of small hard persistent auricles, at others are lepresented by 

 niinute points or small spines (well seen in E. Schinzii) and are always seated one 

 on each side of the leaf-scar or base of the leaf, but in most South African species 

 they are badly developed or entirely absent. 



The formation of a key to the South African species I have found to be 

 extremely difficultj as the distinctive characters that can be utilised are often 

 exceedingly few, not always present at all periods, and so need snpi>lementiDg by 

 others, and mostly cann»-)t ha stated in few words, as there are many cases where 

 it is perfectly obvious to the eye that two or more jdants placed side by side are 

 quite distinct species, yet the characters available for a key are so few, that it 

 becomes exceedingly diiricult to express what the eye instantly j)erceives in words 

 that will enable one to discriminate the species when seen separatel}''. This 

 particularly applies to those succulent species 1 have described as leafless, a term 

 that is not strictly currect, and must be understood to ajtply to the general 

 appearance of the plant, for durin<^ the growing season leaves are present, but 

 they are often so rudimentary and inconspicuous as not to be nf>ticeal)le, and are 

 usually deciduous. In the following key only characters that are apparently 

 absolute have been made use of. In a few cases where a plant varies or dried 

 specimens do not always give an ade<iuate idea of the plant and it might appear to 

 belong to either of two gi'oups iu the key, it has been inserted under both 

 headings to facilitate identification. 



A. Plant herbaceous or sometimes w^oody below, never succulent 

 nor spiny. 



I. Stuns and hrancJics evident ^ erect or pi^ostrate. Leaves alivays present, 

 cotispirifous and well developed^ except in 8, E, Pfeilii, 12, E. multifida, and 

 13; graveolens, whejr they are nnall and sometimes deciduous. 



Flowering leaves or bracts with a conspicuous white area 



at their base ... ... ... ... ,,. ♦,, (15) phylloclada. 



None of the leaves with a conspicuous white area at their 



base : 



♦Annual or perennial herbs, erect or prostrate ; rootstock 



not tuberous : 



fLeaves all opposite (except in 14, E, gluucella, and 



8, E, Pfeiliif where the few below the lowest 



branch are usually alternate) ; hnolucres not in 



umbels : 



JLeaf-blade from as broad as long to 4 times as 

 long as broad, ramble but not linear nor 

 linear-lanceolate ; 

 Involucres in axillary pedunculate cymes or 

 subglobose heads : 

 Stems erect or ascending, with conspicuous 

 spreading yellow hairs on their uj>per 

 part ; involucres in dense subglobose 

 heads (C) Mi'ta, 



Stems glabrous or finely adpressed-puberulous ; 



involucres in small leafy cymes, more 



■ rarely in dense heads : 



Stems erect or ascending, often puberulous ; 



leaves herbaceous, puberulous beneath 



or tm both sides (5) hyper icifolia. 



