286 Willis . — The Floras of the Outlying Islands of 
found only in limited areas in the South Island. The South American 
wides are no commoner, in fact slightly rarer (less distribution area) in New 
Zealand than the Australian wides, and it may also be noticed that those 
South American wides that also reach Australia are very widespread indeed 
in New Zealand. 
There seems therefore some reason to suppose that the affinity between 
New Zealand and South America is really due to the fact that at one time 
there was land connexion, complete or nearly complete, between them, 
though a few species, especially those with northern range in New Zealand, 
may owe their presence to water carriage. The equatorial current from 
the west coast of South America passes by way of the Kermadec and 
Chatham Islands. 
As to the position and direction of the land bridge, it is very difficult 
to come to any definite conclusion in the present state of our knowledge. 
In general we may perhaps imagine that it went by way of the Antarctic 
continent, perhaps all the way to South Georgia, or perhaps by way of 
Juan Fernandez to Chili, though this seems unlikely. But a land connexion 
seems to me to be indicated by the facts of distribution. 
As so many Australian and Tasmanian forms are included in the plants 
of the southern invasion, it would seem probable that this land connexion 
extended also to Australia and Tasmania. 
Whether the South American and the Australian plants arrived by the 
same route from the south it is difficult to say. The general facts seem to 
me to indicate — it is impossible to put it into words without a vast amount 
of detail — that it is quite possible, if not probable, that the land joining 
New Zealand to the Antarctic continent was more or less broken up by 
water areas. It is, for example, noticeable that a number of species which 
occur in Kerguelen, the Crozets, Marion Island, &c., are found only in 
Macquarie, though of course this is quite possibly due to water carriage. 
The next question is, Did this connexion with Antarctica pass through 
any of the groups of islands that we are now considering? As even the 
Aucklands only possess 39 out of 127 wides of the southern invasion, and 
Campbell, the Antipodes, and Macquarie progressively less, it may be 
inferred that it did not pass directly through any of them, and that it was 
probably nearest to the Aucklands. 
The average rarity in New Zealand of the Auckland wides as a whole 
is 2*7, while that of the wides of the southern invasion is 3-1. This is little 
to go upon, but at least it tends to show that on the whole the arrivals of 
wides in the Aucklands were fairly early. 
There are a considerable number of South American genera which 
occur in New Zealand but not in the Aucklands, another fact which goes to 
show that the southern invasion, or some of the invasions, if there were 
more than one, did not pass directly through these islands. 
