Its Origin , Migrations , and Evolutionary Tendencies. 215 
glands, which, however, grade into one another completely, and sometimes 
vary in the same species or on the same individual. The systematic works 
say little or nothing regarding the probable evolutionary history of the 
genus, but if we follow it step by step along the lines of its invasion into 
South Africa we can give the following outline of its evolution. The more 
mesophytic tropical species of shrubs with erect, leafy, woody 9 spineless 
stems probably come nearest to the ancestral form. Some of them 
have invaded Eastern South Africa, e. g. E . epicyparissias. The purely 
herbaceous type has been derived, but has not diverged very far. Of the 
herbaceous forms the small annuals are most recent, and of these E. inae- 
quilatera is one of the commonest and has spread all over Africa. Many 
are weeds of cultivation, e. g. E . peplus and E. helioscopia . A section of 
the perennial herbaceous types has developed tuberous root-stocks. The 
shrubby types have in another direction developed spines which, according 
to N. E. Brown ( 4 ), are of three types in the South African species : 
(1) where the apex of a branch becomes spiny, as in E. lignosa and 
E. spinea , two dwarf shrublets from Namaqualand (one or two transitional 
forms have tapering branches not acutely spine-tipped) ; (2) where the 
peduncle becomes transformed into a spine ; and (3) the so-called ‘ stipular 
spines ’, which are in pairs but are probably not stipular in origin. 
Spine development, as we have noted, is one of the general reactions to 
drier conditions. The main evolutionary tendency in the genus, however, 
has been towards succulence. It is seen in the large variety of low-growing 
forms that have probably developed from the perennial herbaceous forms 
and in the large succulent tree Euphorbias which come nearer to the primi- 
tive shrubby forms. The succulent, c leafless and often spiny Euphorbias 
have developed from the tropical types partly in the dry areas of the 
western side and partly in the dry river valleys of the eastern side where the 
species are usually distinct. The Karroo species are also distinct. 
In support of the view that the numerous succulent South African 
species are recent and derivative we have not only the general origin of the 
subtropical South African flora, but the further fact that when cultivated 
under moist conditions succulent species show a tendency to revert to 
a shrubby type and develop slender leafy branches with no trace of succu- 
lence. This has been noted particularly by N. E. Brown in connexion with 
E. gorgonis as cultivated at Kew (see 4 , and Gardeners’ Chronicle, 1914, 
lvi. 230, fig. 91, p. 312). 
Though succulence, thorn development, and other xerophytic characters 
are common features of derivative subtropical forms, it must not be assumed 
that evolution in the South African flora has always been in this direction. 
There are numerous extra-tropical mesophytic forest situations where recent 
endemic forms have been produced. Some genera show development 
in both directions. 
R 
