traits are essential to survival, and therefore evolution is more rapid as regards them. Many traits essential in a wild state for survival, in a cultivated or domesticated state are not essential, and therefore as regards them there is retrogression. In a few generations "sports" appear; but these sports, if I am right, must generally be examples of retrogression, not of evolution (retrogression in traits which were essential to the wild ancestors, but are no longer essential to the cultivated or domesticated descendants, or far-reaching retrogression in traits rapidly acquired under artificial selection); must generally be due to reversion to the ancestral form, not to an advance beyond it.
Again, atavism is not the only cause of retrogression. Evolution may be a cause of apparent retrogression, as in the case of certain insects, which, living as they do in storm-swept islands, are exposed when flying to the danger of being carried to sea. In them natural selection, reversed selection as it is called in such cases, has co-operated with atavism to deprive them of the power of flight. Therefore, if this theory of retrogression be correct, it may afford us a not unimportant insight into the past life-history of species, and enable us to decide what retrogressive changes are due to atavism and what to natural selection. For instance, man as we know has descended from a hairy ancestor: if his present partially hairless condition is due to atavism, then since he has reverted in this respect to a remote hairless ancestry, his embryo will not be hairy, nor will he in cases of atavism be hairy. We know that this is not the case; therefore his present hairlessness is due mainly to natural selection (reversed selection), possibly to that form of natural selection known as sexual selection.
This theory has not, so far as I am aware, been propounded before, and it is opposed to another theory
