270 JOURNAL OF SCIENCE, 
is said by Mr. Wallace to be connected with the antarctic 
plateau, but other geographers make a deep channel between 
Campbell Island and Macquarie Island, and another south of 
Macquarie Island. From what is known of the geology of the 
antarctic islands it appears that all are volcanic, except S. 
Georgia, which is part of an old slate mountain range. 
If we examine the faunas and floras of the islands along this 
track we find that Tristan d’Acunha, although three times as far 
from Fuegia as it is from the Cape of Good Hope, has its flora 
much more nearly allied to that of Fuegia than to that of Africa. 
Kerguelen Land also has its flora much more related to that of 
Fuegia than to that of the Auckland Islands, although the dis- 
tance is half as far again. This island has also 58 species of 
marine mollusca, of which 13 are found in South America, 6 or 
7 in New Zealand, and only 4 at the Cape of Good Hope ; and 
it has one endemic land shell—Helx Hookeri. Its fauna and 
flora must therefore have come from the west, and passed on by 
the east to New Zealand. We have already seen, in the early 
part of this address, that more land communication than at 
present exists is necessary to explain the migration of the 
antarctic fauna and flora ; and we have therefore in the antarctic 
plateau, stretching from near South America in an easterly 
direction to Victoria Land, and either connected with, or but 
slightly separated from, land that extended to 30° S. in the South 
Atlantic and Indian Oceans, the probable position of the con- 
tinent along which the migration took place, but which was 
always separated from New Zealand bya broad and deep 
channel south of Macquarie Island. 
There remains now only the question—What was the date of 
this migration into New Zealand? It is evident that it could 
not have taken place, asa whole, in the pliocene or later, because 
we have already seen that the floras of the outlying islands have 
only differentiated some 20 per cent. in species since the pliocene ; 
while the New Zealand antarctic flora; as I mentioned in my 
last address, has differentiated by about 65 per cent. in the 
species. Also it must, as a whole, have been before the eocene, 
as since then the differentiation of species has been at least 90 
per cent. The main immigration must, therefore, have taken 
place either in the miocene, when New Zealand was reduced to 
a number of islands,* or else part must have arrived in the plio- 
cene and part in the eocene, at both of which times New Zealand 
extended much further to the south. Let us try to see which 
of the two is the more probable. 
It would be a great mistake to suppose that our alpine 
flora is almost exclusively composed of plants of antarctic 
or north temperate origin. Of 189 species of alpine plants 
belonging to 64 genera, I find that 48 per cent. are of 
antarctic (including N. Temperate) origin, 37 per cent. are 
* There is evidence that an elevation occurred between the deposition of the 
Oamaru and Pareora Systems, but this elevation was slight, and New Zealand wa 
probably of no greater extent at that time than it is now. 
