202 
Willis .— Age and Area . 
Age and Area—‘has not yet proceeded long enough to spread them all 
over the island.’ There is no suggestion here of competition, of adaptation, 
or of any of the other outstanding features of the Darwinian theory. 
The author of this paragraph explained that he imagined all the fishes 
to arrive at the same time, or in other words that they were all at the same 
time at the line of junction with the Continent. This of course means that 
though in Britain they might be distributed ‘ at varying rates according to 
circumstancesthis was not so upon the mainland, where they all simul¬ 
taneously reached the same line. Or if this interpretation be rejected, then 
we must suppose that those fishes that first reached the line of division 
must have waited there until all had arrived, so as to effect a kind of 
Norman conquest of the English rivers. And this although England was 
attached to the Continent for a very long period. 
The only reasonable explanation of the distribution of the freshwater 
fishes, as of the plants, is that they came over so soon as they reached 
sufficiently far down the Seine and the Rhine, &c., each one spreading at 
its own average rate, when long periods are considered, and that the dis¬ 
persal in Britain has been mainly determined by the time factor, so that, in 
the words of Mr. Tate Regan, ‘ Ireland has only ten of the twenty-two 
species. ... In Britain there is a considerable diminution in the number of 
species towards the north ... in the Northern Highlands, as in the out¬ 
lying islands, there seem to be no indigenous true freshwater fishes . . . 
a similar decrease . . . takes place from east to west. . . . Ouite a number 
. . . are absent from Wales west of the Severn system. . . .’ 
With regard to the question whether areas give a true hollow curve, 
the matter seems to me one of comparatively small importance. It is really 
a matter of the size of the units of area employed. Taking fairly large 
units, as I have done, there can be no question that the greatest number of 
species are found, in any genus of reasonable size, upon the smallest areas. 
But if one try to divide these units into smaller ones, one finds only a very 
few genera with a sufficient number of species to enable such a thing to be 
done. It is obviously unfair to divide the smallest area into yet smaller 
without at the same time doing the same thing to the larger areas, and 
there are so few in these larger areas that one may often find many of the 
smaller units without any species in them at all. Suppose, for the sake of 
example, that we divide the flora of New Zealand by smaller areas than at 
present (cf. 11 , pp. 64, 77, &c.). To take the first large genus, Ranunculus , 
one finds the endemic species to have the following ranges: 
83°’ 670, 580, 570, 540, 460, 420, 340, 340, 320, 310, 280, 260, 220, 180, 
170, 90, 60, 60, 20, 20, 20, 10, 10, 10, to, 10 (miles of the length of New 
Zealand). 
It is clear that the smallest unit, ten miles of longitudinal range in 
