J^aphorhia.'] cxxii. euphorbiaciI/E (brown), 471 



following descriptions all the measurements are takou from flowers soaked in 

 boiling water or preserved in fluid, the diameter of the involucre usually includes 

 the glands and their appendages or processes, but not their two horns. Usually the 

 involucre-glands are free, but in a few of the terete-stemmed and s])iuy succulent 

 gpecics they are united into one continuous rim-like gland or with only a faint 

 indication of lobing; in these there is nothing but habit to separate them technically 

 from Sifnadenimn. The formation of a key to the species of this genus presents 

 exceptional difticulties, as the characters available are few in proi)ortion to the 

 number of species. It is often easy to see at a glance that two or more species are 

 ])erfectly distinct, but very difficult to find absolute characters whereby to distinguish 

 them in a key. Leaves are often so similar in a number of species or else so 

 dissimilar on the normal stems and those that spring up after the annual fires, that 

 distinct species have been founded upon them (as in E, zamhesiaiia, E. depauperata, 

 &c.), so that they can seldom be utilised. Pubescence is absent from the majority. 

 The involucre and its glands are very much alike in a large number of species, 

 whilst in the succulent group the number of angles and spines are very much the 

 same in many kinds and the flowers are often unknown, rendering distinctive 

 characters exclusive of measurements very hard to find. Among the succulent 

 species Pax has proposed groups characterised by tlie number of spines upon the 

 si)ine-shields, which 1 find to be quite invalid. For the difference in the number of 

 spines mentioned by him, depends upon the presence or absence of a pair of 

 prickles above the spines. The spines in books have been called " stipular spines," 

 but as they are always developed below, and sometimes at a distance below the leaf 

 or leaf -scar, they cannot be stipules in the ordinary sense of that term ; what their 

 veal relation to the leaf is I do not quite understand. But the prickles to which I 

 above allude are developed one on each side of the rudimentary leaf and are true 

 stipules. In some species they seem to be constantly well developed either as small 

 auricles or prickles, but with one exception are never so large as the true spines ; in 

 other species they are very frequently well developed on some branches and rudi- 

 mentary or absent on others or even on the same branch, so that it is often quite 

 impossible to use them as specific characters as has been done by Pax, with the 

 consequence that the same species is found placed under different names in different 

 groups. The term " flowering-eye " made use of for the succulent species, refers to 

 a more or less distinctly marked and usually depressed area or point, like the " eye " 

 of a potato, on the angles of the stem above the spine-pairs, from which the flowers 

 ultimately develop. In the following key only characters that are apparently 

 absolute have been made use of and where the plant is variable or cannot be dis- 

 tinguished by one unvarying character it is inserted under more than one heading. 



A* Plants without spine-shields, prickles or spines, 



with the exception of a few species with rigid 



woody spine-tipped branches. 

 Involucre wltb only 2-3 perfect erect fflands, 



(4fc-6 or more In all otber species) ; kerbs. 

 Leaves all petiolate; involucre-glands ^-f lin.long, 



2-lipped, appendagedj herb with tuberous roots 72. E. tnhenfera. 

 Leaves sessile at the flowering nodes ; involucre- 

 glands If lin. long, tubular, open down the 



inner side, without appendages ; main stem 



thick, fleshy, tuberculate .... l^^. J?, longetttberculosa. 



ZnTolucre with 4-5 grlanas, divided on tlieir 



outer margin into 3-1 5 simple or forked, 



filiform, linear or fing-er-llke processes, 



and including tl&e glands ^-1^ in. in 



dian&. 



