194 
Willis. — Age and Area. 
pp. 441-4, and especially 7 , 12 ). Papers based upon Age and Area were 
read by Mr. G. Udny Yule and myself, the title of mine being ‘On the 
Inadequacy of the Theory of Natural Selection as an Explanation of the 
Facts of Geographical Distribution and Evolution We had read papers 
upon somewhat similar lines at a meeting of the Linnean Society in 
February (abstract, 13 ), and the change of view-point of the biologists since 
that meeting was clearly marked. Many of the speakers were concerned 
less with defending Natural Selection than with attacking Age and Area, 
which was variously described, then or in subsequent newspaper articles, as 
‘worth less than nothing’, ‘utter balderdash’, ‘a few well-known facts 
distorted and 4 a' theory whose career has already ended with more to 
this effect. 
As usual the objections were mostly due to a misunderstanding of the 
hypothesis. In spite of the fact that throughout I have made clear that it 
must not be applied to less than ten species at once, and those allied, and 
that these must only be compared with other groups allied to the first, 
persistent and somewhat pathetic attempts are made to apply it to 
individual cases, or to unrelated forms. As an instance of the kind of 
argument employed, I may quote from the second review of my book in ‘ The 
Times ’ the sentence (which falls into both errors) ‘ Is the barn-owl, almost 
world-wide, more ancient than the three-toed ostrich, found only in South 
America?’. In my book, upon p. 62, I have given an illustration of the 
way in which, though only dealing with fives instead of tens (as should 
always be the case), Age and Area is inapplicable to three out of five cases, 
although all species are following the law with complete exactness, a thing 
that is somewhat improbable in real life. In chapter ix I have devoted 
much attention to this difficulty, and I have repeatedly pointed out, there 
and elsewhere, that inapplicability to single cases does not in the very 
slightest degree invalidate the theory. 
The single-species difficulty seems a stumbling-block to a very great 
number of people, and it is needful once more to emphasize the fact that 
one cannot reduce statistics to individual cases—the very name statistics 
indicates this. To take an illustrative case from ordinary life: no amount 
of study of individual members of the Goddard family would tell us much 
about the origin of the family, and it might be very difficult to make out 
anything, so far as I know, from history (corresponding to palaeobotany). 
We might ultimately know every detail about their anatomy and physiology, 
and even about their genetics, but none of these would be likely to give 
a clue. But supposing that we add up the numbers of Goddards in every 
county, in the more slowly-moving section of the population, the farmers, 
we obtain in this way statistics which show that they form 0*55 per cent, 
of the farmers in Berkshire, and no more than 0*31 in the adjacent counties, 
whilst they almost disappear beyond Dorset, Derby, and Norfolk. Con- 
