Let us take a single illustration. Many of the species of beetles on
oceanic islands have very rudimentary wings, or none at all, and yet
their nearest relatives are winged forms on some neighboring
continent. Mr. Darwin would explain the origin of these evidently
distinct wingless species as follows: They are descended from winged
ancestors blown or otherwise transported thither from the
neighboring continent. But beetles are slow and clumsy fliers, and
on these wind-swept islands those which flew most would be blown out
to sea and drowned. Those which flew the least, and these would
include the individuals with more poorly developed wings, would
survive. There would thus be a survival in every generation of a
larger proportion of those having the poorest wings, and destruction
of those whose wings were strong, or whose habits most active. We
have here a natural selection which must in time produce a species
with rudimentary or aborted wings, just as surely as a human
breeder, by artificial selection can produce such an animal as a pug
or a poodle. These, like sin, are a human device; nature should not
be held responsible for them.

But you may urge that the variation which would take place in a
single generation would be, as a rule, too slight to be of any
practical value to the animal, and could not be fostered by natural
selection until greatly enhanced by some other means. Let us think a
moment. If ten ordinary men run in a foot-race, the two foremost may
lead by several feet. But if the number of runners be continually
increased the finish will be ever closer until finally but an atom
more wind or muscle or pluck would make all the difference between
winning and losing the prize.

Similarly the million or more young of any species of insect in a