232 
James Small. 
As time went on this new genus spread to the arid, semi-desert 
regions which were increasing in South Africa in the evening of the 
Miocene day; the dryness and strong insolation had the same 
result as they had at an earlier date in the Mediterranean region. 
The type became spinescent (Berkheya) ; reduction and aggregation 
continuing, as they did again later to produce Echiuops among the 
thistles, the dense glomerules of the spinescent Gundelia and 
Platycarpha originated. 
In addition to the origin of Ursinia the upper Miocene in 
South Africa saw the origin of the Athrixia line. The Inula group 
was by this time well represented in South Africa and a reduction 
in the corolla of the ray florets occurred, similar to that which had 
previously taken place in the similarly arid part of the Mediterra¬ 
nean region. This with othervery slight changesgave the Athrixiinae. 
On account of the continued drying up of the region, which produced 
a distinctly xerophytic habitat, the flat or slightly recurved leaves of 
Athrixia became much reduced and practically ericoid. The semi- 
shrubby habit changed at the same time into the typical xeromorphic 
shrubby habit. These conditions also led to a reduced or depau¬ 
perate condition in the capitulum : very few flowers were developed 
in each head and this led to the usual aggregation. By these 
epharmonic adaptations Relhqnia and its congeners were originated. 
Helichrysuni by this time was well developed in South Africa 
and the xerophytic conditions had almost the same effects on this 
genus as they had on Athrixia. So much so that epharmosis, in 
this case as in many others, led to convergent evolution, and the 
present systematic group Relhaniinae includes both Relhania and 
Metalasia, which latter was the product of the second series of 
epharmonic variations. 
A neighbouring event of this time was the origin of the 
Grangeinae in tropical Africa. The Asters and Erigerons had 
followed Senecio along the lower levels of the mountains and on 
attaining the very favourable conditions yielded by the tropical 
African scrub underwent the usual reductions. The pappus first 
became reduced or caducous and then disappeared completely. At 
the same time the tendency to reduction of the rays, which was so well 
developed in the Aster plexus resulted in reduced ray florets. 
Meanwhile the Senecioneae in Mexico were again being affected 
by the mesophytic conditions which had previously led to the fusion 
of the pappus setae in the origin of the Galinsoginae, Madiinae and 
Liabinae near the beginning of the Miocene. In this case the 
