84 SECTIONAL ADDRESSES. 
pre-eminent for his wide knowledge and his great reasoning powers, who 
mew the facts that had to be explained and gave us a theory that ex- 
plained them. The ‘ Origin of Species ’ still remains the one book essential 
for the student of evolution. 
Darwin has been criticised, because, we are told, he did not know that 
there were two sorts of variations—mutations, which are inherited, and 
fluctuations, which vary about a mean and are not inherited. But when 
you point out to a mutationist that the heredity of many fluctuating 
variations has been proved—parents above the mean, for example, giving 
ofispring above the mean—he tells you that that shows that the variation 
is not really fluctuating, but only apparently so, and that a large number 
of ‘factors’ must be involved. This is in effect a complete withdrawal, 
for it amounts to an admission that Darwin was right if he considered 
that these types of variation differed only in size and frequency. 
But there are other critics who admit that at any rate some fluctuations 
are inherited, but who say that the effect produced on a population by 
selection is limited ; elimination of certain types will change the average, 
but will produce nothing new. This criticism has also, as it seems to me, 
been disproved experimentally ; for example, by De Vries, who from two © 
plants of clover in which a few leaves were four-lobed produced by selec- 
tion a variety in which the number of lobes of the leaves varied from three 
to seven, fluctuating about a mean of five. Incidentally this experiment 
shows the relation between mutations and fluctuations. 
The criticism that many specific characters are non-adaptive merely 
amounts to this, that we do not know the meaning of many specific 
characters. And we are not likely to for a long time, for a prolonged study 
would be necessary to understand fully the meaning of the differences 
between any two species, to determine which characters were adaptive, 
which historical, which due to the environment, and which the expression 
of metabolic differences. 
But if these criticisms of the natural selection theory can be met 
it does not follow that it is a complete theory. It may be a sufficient — 
explanation of certain types of evolution, and one cannot wonder that those 
who have studied mimicry in insects are firmly convinced of its truth ; 
but the evolution of the Dodo, and of the blind fishes of subterranean 
waters, put rather a strain on the theory and almost demand the recogni- 
tion of the inheritance of the effects of use and disuse. 
And if this be admitted, if the adaptive responses of an organism to 
changed habits and changed conditions make it possible for subsequent — 
generations to respond with greater effect, then the part played by natural 
selection in evolution of this kind would be subsidiary, the selection of 
those individuals who responded earlier or better than their fellows. 
How well this idea fits in with that fundamental generalisation, the law of 
recapitulation, which states that ontogeny tends to repeat phylogeny, 
and that the more remote the ancestor the earlier it will be represented 
in the developmental history !_ This generalisation, based on embryological 
data, has since received strong support from paleontological evidence. 
No doubt all of you are aware that a flat-fish when first hatched is 
symmetrical and swims vertically, but that at an early age one eye migrates 
round the top of the head to the other side, and the little fish sinks to the 
bottom and henceforth lives with the eyed side uppermost. But perhaps” 
