14 JOURNAL OF SCIENCE. 
This is a mistake which could not be made with the Australian 
genera. 
The Kermadec Islands occupy a very important. position for 
furnishing evidence of migrations into New Zealand from the 
north, but unfortunately very little is known of their flora. 
What is known shows a remarkable affinity to the flora of New 
Zealand. Of the 21 species of flowering plants collected by 
Dr. Macgillivray, only three (14 per cent.) are endemic; 17 are 
found in New Zealand (one of which is supposed to have been 
introduced into both places), and the other—WMetrosideros poly- 
morpha—inhabits Polynesia and New Caledonia. From this we 
must infer that at a comparatively late period New Zealani ex- 
tended further to the north-east than at present, for if it had not 
done so the Kermadec plants would have been far more differen- 
tiated from those of New Zealand than they are. At the same 
time as but few sub-tropical species are common to New Zea- 
land and Australia, this land could not have extended far to the 
north-west ; but we may perhaps refer to this period the intro- 
duction of several of those tropical species, such as Avicennia 
officinalis and Sicyos angnilaris, which are also found in Aus- 
tralia. 
It would thus appear that there have been three migrations 
of plants from the north into New Zealand. Two of very 
ancient date; the third comparatively recent, and comparatively 
unimportant. The supposition that New Zealand was at one 
time connected with a South Pacific continent, from which plants 
spread into South America, and into New Guinea ; and that, at 
a subsequent period, Eastern Australia was attached to New 
Guinea, and received from thence fragments of this Polynesian 
flora, together with plants of the Indian Archipelago, will ex- 
plain, I think, why some Polynesian and South American genera 
are found in New Zealand but not in Australia, and why some 
occur in Australia but not in New Zealand. 
Passing on now toa consideration of our. fauna, we find it 
composed of the same elements that we recognised in the flora, 
viz.—(1) Australian ; (2) Polynesian; (3) S. American; (4) 
Antarctic; and (5) North Temperate. The South American 
element seems to be the weakest, but until the distribution of our 
insects, land-mollusca, and land-worms is better known we 
cannot speak with any confidence on this point. One of our two 
bats was formerly thought to belong to an American family, 
but this has been shown to be a mistake, and it now seems that 
both are of Old World extraction. This removes a difficulty, 
for bats are certainly not more ancient a group than birds, and 
it would have been very puzzling if their distribution had coin- 
cided with that of the frogs instead of with that of the birds. 
Our birds show only three elements :—(1) An Antarctic, 
which comprises the Penguins, the Petrels, three out of five gulls, 
and four out of nine cormorants ; (2) a Polynesian, consisting of 
the Paroquets, Aplonis, and the long-tailed Cuckoo; and (3) 
an Australian, which includes all the rest, except a few which 
