268 
Willis . — The Floras of the Outlying Islands of 
keeping to the larger floras and papers it must not be forgotten that I gain 
the very considerable advantage of dealing with a flora worked up by one 
or by a few persons, and consequently one which is fairly uniformly treated 
in its conception of specific forms, &c. 
The second map-diagram of my last paper, reproduced here, represents 
an imaginary distribution of the plants of the northern and southern invasions 
of New Zealand already dealt with, and supposed, for the sake of simplicity, 
to centre in Auckland (city) and Dunedin. A very brief consideration of 
it shows that while on the whole the plants of Stewart Island will be older 
in New Zealand itself than the general average of the plants of New 
Zealand, those of the outlying islands will be on the whole at least as old, 
and in most cases probably 
older, as they would have to 
start very early from New Zea- 
land to reach the islands be- 
fore they were cut off. The 
Kermadecs and Aucklands, 
however, having been to some 
extent in the track of invasions, 
may of course contain species 
which arrived there too late to 
reach New Zealand at all, or 
only just reached it at the last 
minute before it was cut off. 
The Chathams, on the other 
hand, must clearly have re- 
ceived all, or nearly all, their 
flora by way of New Zealand ; 
no invasion can have passed 
New Zealand and inlands to show imaginary distribu- . xr , c 
tion from Auckland and Dunedin. that way. We shall start from 
this conception of the great 
relative age of the floras of the outlying islands as a fundamental fact, 
and study the floras of these islands by the method of prediction and 
verification, just as was done in the case of Stewart. 
One will expect to find that there is what one may perhaps term 
a selection list (the Scottish term ‘ short leet ’ exactly expresses it) of the 
(in New Zealand) older families, from which Stewart as the nearest island 
will select practically all, while the more outlying islands will have fewer 
of them. But the families of these islands should be all included in the 
Stewart list, unless they be the later arrivals of the northern or southern 
invasions, or families which have their maximum development in the centre 
of New Zealand, from which it is less distance to the Chathams than to 
Stewart. 
