47 2 Willis . — Plant Invasions of New Zealand . 
which had a greater distribution abroad than merely to South India, I 
deduced the conclusion that in a given country the area of distribution of 
a species (working always with groups of at least ten allied species) depended 
upon its age in that country. The correctness of this deduction was then 
confirmed by observations upon the floras of New Zealand and its surround- 
ing islands, which gave exactly parallel results, and also by observations on 
the floras of Hawaii, Jamaica, &c. 
In my first paper on the New Zealand flora ( 9 ), I assumed the correct- 
ness of the hypothesis (which for convenience I have termed that of 1 age and 
area ’), and by means of the diagram here reproduced I predicted the result 
of the entrance of a single species (w) at the centre of New Zealand, which 
spread at a uniform rate (represented by the triangle), and which sub- 
sequently gave rise to local endemic species (E i to E io) in a casual way. 
The length of New Zealand, supposed i,ooo miles, being divided into ten 
zones of equal width, and the number of endemics found in each zone being 
counted (each endemic of course being supposed to spread uniformly like 
the parent species, as indicated by the similar triangles), the result shows 
a curve rising and falling to and from a maximum (sometimes two) which is 
somewhere near the point of entry of the original species. In the present 
case, for example, the numbers of endemics in each consecutive zone are — * 
o 3 5 8 9 1 8 7 3 2 2 
It will be well to make clear at this point that if the entry of the 
original species be at one end, instead of in the middle of New Zealand, the 
curve will in general show its maximum at or near that end. The maximum, 
1 The thicker type indicates the maximum in this and the following table. 
