200 
Willis . — The Relative Age of Endemic 
or did not once occupy a greater area. The view adopted by Mr. Ridley 
apparently is that they are the relics of a once extensive flora which is being 
driven in. 
But if 233 of the Ceylon endemics are VR relics of this flora, why are 
222 wides also VR ? Are they also relics ? And why has Ceylon 222 such 
relics, while New Zealand has only 21 (with one-fifth the flora of wides), of 
which at least half have no right to be included in the list ( 7 , p. 452) ? Why 
do the endemics choose mountain-tops to such an extent? Ceylon has 108 
of them confined to summits or to small areas in the high mountains (12). 
Why has New Zealand proportionately fewer, and why in New Zealand do 
they comparatively rarely occupy one summit only, but more often a small 
range ? 
The usual theory of the supporters of Natural Selection is that they have 
gone up the mountains as the last refuges from the invading flora of the low 
country. But in such small and uniform countries as South-western 
Ceylon and New Zealand it is hardly possible to suppose that there was a 
separate Eugenia or Celmisia at every few miles. And why did they climb 
right to the summit ? It suggests an unnecessary degree of alarm about the 
coming competition. Further, it would suggest that endemics are not so 
incapable of adaptation to new conditions that they need fear the competi- 
tion at all. If they can undergo the great adaptive changes necessary 
to reach a summit of 3,000 to 10,000 feet, they must have a very fair capacity 
for modification, and should be able to hold their own. 
Why are the wides which are VR in Ceylon, 222 in number, not found 
confined to mountain-tops ? Are they not dying out, and, if not, why not, 
when they are as rare as the endemics which are supposed to be doing so ? 
It suggests that they did not care to waste time in modification to suit high 
altitudes, when they were to be killed out in any case. 
But the great difficulty of all, perhaps, for the supporter of Natural 
Selection is to explain why the dying out of the endemics (assuming that 
they are doing so) is purely mechanical. Why does every family and genus 
behave in the same way, whether it does or does not contain wides, and 
whether it be species in an endemic or in a widely distributed genus ? 
Natural Selection could not cause a mechanical dying out unless it meant 
that the arrival of the first few widely distributed species (i. e. assuming that 
they are the younger and arrived later) was the signal for the dying out of 
the whole of the old flora. Why should a genus, as we have just seen, die 
out sooner than a species ? Why should a genus die out in the regularly 
graduated way shown in the map of Doona (6, p. 14), or in the Tables IV, 
V, and VI of my last paper ( 7 )? Why should the endemics be most 
numerous where there are most wides (see below, crucial case No. 1) ? 
We shall now give two crucial cases (already given at Newcastle), 
which speak strongly for age against youth, and a third, still more con- 
