2 I I 
Willis. — Age and Area. 
p. 175 of my book). Why should there be so many more relics, both in 
number and in proportion, upon the islands that are nearest to the main¬ 
land and least isolated ? The greater proportion of larger endemic genera 
upon the more outlying islands shows clearly that these genera were 
formed at an earlier date (on the whole), and thus, having had longer time 
in which to expand, have grown larger. 
Why has tropical America, a comparatively continuous area, 13 percent, 
of its endemic genera monotypic, while the much broken Old World tropics 
have only 6 per cent., and tropical Asia 36 per cent. ? The general distribu¬ 
tion of genera by sizes in different areas, more fully described in my book, 
offers very serious problems to the supporter of relicdom for most endemics. 
Prediction applied to the Flora of the British Islands. 
The predictions hitherto applied with such success have been mostly 
in reference to the flora of New Zealand and its surrounding islands, and as 
the criticism has been made that though prediction may succeed in such 
regions that have been comparatively little interfered with by man, it would 
be valueless in such regions as the British Islands, an attempt at prediction 
is here made for the latter. The attempt is to forecast the flora of the 
county of Dublin (a region of which we possess a good and recent flora ( 1 )) 
from what is known of the distribution of the plants of the adjacent island 
of Great Britain. For the latter there is no more recent general authority, 
so far as I know, than Watson’s ‘ Topographical Botany’, 2nd ed., 1883, and 
I have kept rigidly to this. I have not the time available to consult all the 
innumerable papers in which more recent work is enshrined, nor is it 
necessary for such work as the present, for, as will be seen below, the predic¬ 
tion is perhaps more successful than were any of those which I made in 
regard to the flora of New Zealand. 
It is generally admitted that Britain and Ireland were at one time 
united to one another and to the Continent, and received their floras thence. 
The nearest point to the Continent, and that a point not uncentral (in all pro¬ 
bability) as regards the land union, is the South Foreland in Kent. For the 
purpose of prediction let us assume that this was the centre of the main 
continental (northward-moving) invasion, and draw circles round it as 
a centre. If we take a circle that includes the county of Dublin, it will be 
seen that it also includes the counties of Wicklow and Wexford, and then 
crosses Scotland by the south of Ayrshire, through Lanark, Peebles, and 
Edinburgh, reaching the North Sea near Dunbar in Haddington. 
The prediction now is obvious. One will expect to find the flora of the 
south-east corner of Ireland made up, so far as this invasion is concerned, of 
those plants which reach Kent, and also reach Ayrshire or Haddington. 
But there are many plants in the British Islands that seem to have 
