FAUNA AND FLORA OF NEW ZEALAND. 13 
Statistical results like these are always open to the objection 
that the data on which they rest are incomplete and more or 
less erroneous (for example Corzarza and Gunnera may belong 
to the Antarctic element, and Drosera to the South American). 
They also assume that the rate of variation is equable, which of 
course cannot be strictly accurate. But this method of investiga- 
tion has been used with great success in geology, and it can, I 
think, be trusted here for establishing the two following con- 
clusions :—First, that the northern immigration, taken as a whole, 
was anterior to the southern immigration, also taken as a whole ; 
and second, that the immigration of the sub-tropical South 
American genera belongs to the first period, and not to the last. 
The first conclusion is similar to that of Mr. Wallace, but arrived 
at in a different way. The second is opposed to Mr. Wallace’s 
idea that the South American plants passed through New Zea- 
land and Antarctic lands during a warm miocene period, which 
is also opposed by the fact that a number of Australian genera. 
are found in South America, but not in New Zealand. The 
fact that very few of our South American genera are absent 
from Australia, while a large number of our Australian genera 
are absent from South America, makes it probable that there 
have been at least two migrations into New Zealand from the 
north, and that the South American element belongs to the 
first of these only. This is borne out by the distribution of some 
the groups. The best example perhaps is the Orchids, of which 
18 genera occur in New Zealand. Of these two are endemic, 
and the other 16 are all found in Australia. Two occur also in 
New Caledonia, three in Polynesia, four in the Indian Archipelago, 
and three in India, while one consists of a single species widely 
spread over Asia and Australia. None of them are found in 
South America. The path of the Orchids into New Zealand, by 
the Indian Archipelago and New Caledonia, is thus plainly 
mapped out, and as none have passed into South America the 
migration probably took place after the South Pacific continent 
had disappeared. The number of New Zealand species of this 
order is 38, of which 32—or 84 per cent.—are endemic, so that 
the immigration must have been an early one. Other examples 
are found in Prttosporee, Rutace@, and Santalacee. Examples 
of the earlier South American migration are seen in the M/onz- 
miacee and Chloranthacee, while examples of the Antarctic 
migration are the Caryophyllee, the Geraniacee and the Rutacee. 
It may be objected that the percentage of endemic species is 
greater in the Australian than in the South American element, 
and therefore that the first must be the older. But the objec- 
tion is not fatal, because, in the first place, we must remember 
that the American genera would continue to live in Polynesia, 
and would migrate into New Zealand again with the Australian 
forms, thus making the percentage nearly the same in each case ; 
and in the second place, one or two genera may be included in 
the South American element which ate really Antarctic, and 
this would at once bring down the percentage of endemic species, 
