328 Willis . — The Distribution of the Plants of 
The Kermadec Islands, it will be noticed, are divided from New 
Zealand by a greater depth than 1,000 fathoms at one part of the inter- 
vening sea. The ridge on which they stand leads to the Tongas and Fiji. 
Now the distribution of their plants in New Zealand does not exactly agree 
with that of the plants of the other islands, which are not divided from New 
Zealand by so great a depth of water. It agrees perfectly with my age 
and area hypothesis, but it shows several very special and interesting 
features, which tend to indicate that the period when the Kermadecs were 
united to New Zealand was not quite coincident with that of the union 
with the other islands. To work out the matter in detail requires geological 
aid, but my figures come out with such astonishing simplicity and accuracy 
that it is becoming increasingly clear that evidence based upon ‘age and 
area’ cannot be altogether neglected, even in dealing with geological 
problems. 
We shall deal principally with the species common to these islands and 
New Zealand, or New Zealand and Australia, leaving out those only found 
elsewhere in New Zealand and South America, which again present a very 
special problem, indicating that the union with South America was perhaps 
not exactly synchronous with that with Australia. 
No hypothesis as yet put forward, whether Natural Selection or any 
other, with the exception of ‘ age and area ’, will enable us to make any 
prediction with regard to the distribution in New Zealand of the plants of 
the outlying islands, as to whether they are or are not widespread there ; 
but age and area permits us to do this. The prophecy is obvious, and the 
fact that it is completely borne out by the actual state of the case has made 
it worth while to write this little paper. 
If age and area be the general rule, then it is evident from the con- 
figuration of what we may, for the purposes of this paper, term the New 
Zealand archipelago, that the earliest arrivals in New Zealand would be the 
most likely to reach the islands, whilst the later ones would not do so. 
The three chief groups of islands, which bear enough plants to make 
argument from their floras by age and area fairly safe, are the Kermadecs, 
Chathams, and Aucklands. Examination of the little map showing the 
soundings will show that they lie at more or less the same distances 
from the narrow strip of less than 1,000 fathoms which runs down from 
Australia, and which, in the absence of any evidence to the contrary, one 
must look upon as probably the centre of the line of immigration. It is 
safe to prophesy, however, that those plants which reach the islands will 
have been the first to reach New Zealand, and should therefore be more 
widespread there than those that do not reach them. 
Let us now examine the Australian wides of New Zealand which also 
reach the islands, and classify them according to their range in New 
Zealand, as was done in Table VIII of the previous paper on that country, 
