Polynesian Flora— COPELAND 
295 
derived from the comparatively rich flora of 
Malaya. Today, it is more generally recognized 
that the better explanation is that the two floras 
have a common origin of their common ele¬ 
ments, and New Guinea is looked to for the 
common source. But again, taking a lesson 
from the ferns, and bearing in mind the diffi¬ 
culty of direct eastward migration, it seems 
more reasonable to look for a common outside 
source than to suppose that one area was 
colonized from the other. To the extent that 
this common flora is of ultimate austral origin, 
it would have had to back-track to the Solomons 
and New Hebrides before entering Polynesia, 
and to do this against the winds and the oceanic 
currents; I cannot believe that on any major 
scale this ever happened. So, of present land 
areas, I regard New Caledonia and the New 
Hebrides as the roughly approximate region 
from which Fiji and thence Polynesia received 
the bulk of their vegetation. 
Some such history as I propose has been re¬ 
jected by some very respectable authorities. 
Thus Diels, in the Setchell Festschrift (1936: 
191), treating of groups of plants of distinctive¬ 
ly austral occurrence, wrote that "it is impossible 
to infer that they originated in the south.” 
This is evidently not true, because I do so infer, 
and so have Hooker, and Christ (1910: 248), 
and Skottsberg, and still others. 
More disquieting to me, and not so summarily 
to be dismissed, is Merrill’s recent (1945) 
statement: "It has also been suggested that 
scattered throughout this vast region are cer¬ 
tain types that were apparently derived from 
ancient Antarctica. But this idea is purely 
theoretical and is one that can scarcely be 
proved.” As I am expounding this theoretical 
and scarcely provable idea, I must of course 
concede that direct proof is impossible. But 
other proof can carry conviction. A frog in the 
milk pail is not direct proof that the milk has 
been diluted. None of us has seen any plant 
emigrate from Antarctica, and no plants are 
now so emigrating. In the absence of direct 
evidence, the indirect evidence by present dis¬ 
tribution, coupled with the fact, unknown to 
Hooker and to Christ, that Antarctica was at a 
definite and not too remote past time a fit place 
for plants, is the nearest possible approach to 
direct proof that they did migrate thence, into 
and across temperate lands. 
Although migration from Antarctica has long 
ceased (the last warm era there was Miocene, 
say twenty million years ago), I suppose that 
it goes on now along the old paths where the 
climate now permits. Because Merrill has pro¬ 
vided dependable figures on the distribution of 
species, I will use them to show how this mi¬ 
gration seems to have occurred more recently 
in two families, Elaeocarpaceae and Orchidaceae. 
The pertinent figures on the occurrence of 
species of Elaeocarpus are: 
New Guinea . 114 
Philippines . 49 
Borneo . 40 
Java . 18 
New Caledonia . 30 
Fiji . 12 
Samoa . 6 
Rarotonga . 1 
Hawaii . 1 
Merrill concludes, in harmony with general 
practice, that New Guinea is the focus of dis¬ 
tribution. As to all lands farther west, I read 
his evidence in the same way. As to the rapid 
decrease in number of species from group to 
group in Polynesia, his figures agree with the 
evidence of the ferns in showing the great diffi¬ 
culty of eastward migration in this region. But 
his figure of 30 species in New Caledonia is 
very significant to me. New Caledonia has one- 
fortieth of the area of New Guinea, but has 
more than one-fourth as many species of 
Elaeocarpus. The winds and the currents are 
always from New Caledonia to New Guinea, 
and I find it hard to doubt that this is the direc¬ 
tion of migration. The genus is of minor im¬ 
portance in Polynesia, but what species there 
are seem most probably to have come from the 
west or southwest, not from the northwest, not 
from New Guinea. 
