568 MR. G. BENTHAM ON COMPOSITE. 
above Table, although not so absolute as in several ofthe preceding 
insular groups. Ofthe fifteen species common to other lands eight 
(belonging to Gnaphalium, Cotula, Bidens, Picris, Taraxacum, and 
Sonchus) are widely spread weeds which may have been earlier or . 
later received by means of transmission still in operation ; the re- 
maining seven are Australian, some of them extending also north- 
wards to the Indian archipelago. The common genera with ende- 
mic species are also mostly Australian ; several, however, are also 
in extratropical or Antarctic South America, the New-Zealand 
races forming, as it were, a connecting-link between the Austra- 
lian and the Chilian regions. The three endemic New-Zealand 
genera are not very highly differentiated. Brachyglottis is near 
some forms of the cosmopolitan Senecio: Pleurophyllum and Cel- 
misia (which last, the largest New-Zealand genus of Composite, is 
almost, but not quite, endemic) are nearly allied to some Andine 
forms of Erigeron. The Gnaphalioid group is the most exceptional 
in its relations; the endemic genus Haastia appears more nearly 
allied to those Mascarene species of Psiadia or Conyza upon which 
Cordemoy has founded his genus Frappiera, than to any other 
New-Zealand genus. The New-Zealand Cassinie have the cha- 
racters of the South-African section Rhynea, rather than of the 
Australian typical Cassini; and some of the Helichrysum forms 
have their nearest counterparts in some mountain-species of the 
distant Mediterranean region. Sonchus, another northern race, 
has one well-marked species endemic in New Zealand. 
Some of the genera present highly developed shrubby species, 
but none so arborescent as in some of the preceding more per- 
fectly isolated island groups. 
D. CoLoNIZING Composit OR INTRODUCED SPECIES. 
In Composite, as in other orders, there are a certain number 
of species which when carried through the agency of man, inten- 
tionally or unintentionally, into distant lands become there 
established, sometimes even to the exclusion of much of the 
native vegetation. We might, indeed, have expected in an order 
where there are apparently so many provisions for the dispersion 
of the seeds, and where we so frequently observe the greatest 
readiness in germination, that such colonization would be very 
extended. But when the list has to be made out, it is by no 
means a long one; such species appear to be fewer in proportion 
