548 MR. G. BENTHAM ON COMPOSITA. 
universally endemic than in any other except insular ones. Above 
100 out of 149 genera (that is, above two thirds) are strictly 
endemic, or at most have a single species penetrating into tropical 
Africa; and out of 1400 species, if we deduct a very few, like 
Denekia capensis, which cross the frontier to the north, there are 
not, as far as I can estimate, above a dozen common to other 
regions ; and these are chiefly widely spread weeds, such as Bidens 
leucantha, Eclipta alba, Siegesbeckia orientalis, Sonchus arvensis, 
fic. The number of species to a genus is often also very large in 
endemic genera as well as in others which have a wide range in the 
Old World. Pteronia, Felicia, Athanasia, Othonna, Ursinia, Berk- 
heya have each from forty to eighty species; many others vary 
from ten to near forty ; Helichrysum has 187, and Senecio 190; so 
that, notwithstanding the number of well-marked monotypes or 
oligotypes, the average of species to a genus is very nearly ten, an 
average which exceeds that of the well-developed, but compara- 
tively little diversified, Composite flora of the Europzo- Asiatic, 
and is only surpassed by that of the Mediterranean region. 
One of the most striking features in the South-African Compo- 
site is the perfect isolation of many of the above-mentioned 
monotypic or small genera. Corymbicum, Brachylena cum Tar- 
chonantho, Denekia, (Edera, Cadiscus, Eriocephalus, Lasiospermum, 
Eumorphia, Alciope, Lopholena, Platycarpha, Anisocheta, Olden- 
burgia, Sic. have but a distant affinity with their cotribuals, or, 
rather, as I should say of many of them, with the genera with 
which they are technically associated. They may all be considered 
as the expiring remnants of long-lost races; but whether of races 
which have been differentiated, have grown flourished and gradu- 
ally been worn out in the region itself, or of races once ranging 
widely over adjoining regions which have there been generally 
destroyed by physical or other changes, but have left these few 
outlying survivors in situations where they have met with 
protection from such general causes of destruction, is a question 
for the solution of which we may as yet have no data. 
The prevailing tribes are, in the first place, those marked by the 
two large above-mentioned genera Senecio and Helichrysum, the 
Senecionide: and the Gnaphalioid Inuloideex ; in the next place, 
the almost endemic Arctotidex, then the Anthemidex, Asteroide:, 
radiate Inuloide*, and Calendulacez, all the above being either 
essentially Old-World or cosmopolitan tribes. There are also 
more Mutisiaces, a southern tribe, than in any other Old-World 
