REGIONS AND AREAS OF DISTRIBUTION. 547 
dc.) are little more than weeds which spring up rapidly in nume- 
rous individuals where a forest has been cleared or cultivated 
ground abandoned; and even these decrease eastward. If the 
known Composite of the Indian archipelago were reduced to our 
ordinary standard they would not probably extend to above 110 
to 120 species. Beccari's collection of 1849 Sarawak plants as 
received at Kew contained only six Composite, including such 
ubiquitous weeds as Ageratum, Spilanthes, and Crepis ( Youngia), 
and a very common Blumea and Vernonia. 
The principal Composite genera or races of a higher grade 
prevalent in tropical Asia are Vernonia, Blumea and allies, Conyza 
and allies, Grangeines, Gnaphalioid Inuloidez, and Senecionidez. 
No others can count ten species, the most remarkable among them 
being a few Mutisiacese mostly allied to South-African ones, but 
with some special types, such as .4insliea and Tricholepis (the 
only genus of Cynaroidex which is chiefly tropical). Three tribes, 
the American Helenioidez& and the South-African Calendulacee 
and Arctotidex, are wholly unrepresented in tropical Asia. 
lt is not probable that future investigations will add very 
materially to the Composite of the region. Even if the moun- 
tains of New Guinea should exhibit any Australian character in 
their vegetation, it would be more probably exemplified in Prote- 
acere, Myrtacex, and other woody orders now common to New 
Caledonia and Australia than in the herbaceous Composite. 
5. South-African Region. 
This may be generally described as extratropical South Africa, 
although we are not in a position to assign to it any precise limits 
to the northward. In the west it may be naturally bounded by 
the dry deserts assigned by Grisebach to the Kalahari region, 
which appear to be unfavourable to the development or extension 
of Composite. Eastward a few of the southern genera seem to 
penetrate further north, intermixing with the tropical genera, and 
represented by some species even in Abyssinia; but these are so 
few in comparison with their strictly South-African congeners, that 
the region is better defined, and the Composite statisties, owing 
to the recent elaboration of the order by Harvey in the ‘ Flora 
Capensis,’ better established, than in almost any other region. 
As the tropical-Asiatie is the poorest, so is the South-African, 
in proportion to its extent, at once the richest and the most 
diversified of all our regions, and the genera and species more 
