[page] 107

The Difficulties of Natural Selection.

I find, on looking again at Mr. Bennett's article, that I have misrepresented him on one point, for which I beg to apologise. On his supposition, that the first twenty possible steps on the road to mimicry are absolutely useless, his argument will have some weight. This supposition, however, is entirely unsupported by facts. Very large variations of colour are exceedingly common in butterflies; and when such variations are in the right direction, they must in some cases be useful. I believe myself that far less then fifty, or even twenty, steps of variation would in some cases produce very good mimicry.

Alfred R. Wallace