nuous_ degeneration culminating in the
total or almost total disappearance of an organ. Romanes and others have
pointed out this weakness in his theory, but he does not notice it, and
goes on calmly throughout the essay to _assume_ that mere panmixia must
cause progressive degeneration to an unlimited extent; whereas all it
can do is to effect a reduction to the average of the total population
on which selection has been previously worked. He says "individuals with
weak eyes would not be eliminated," but omits to notice that individuals
with strong eyes would also "not be eliminated," and as there is no
reason alleged why variations in _all directions_ should not occur as
before, the free intercrossing would tend to keep up a mean condition
only a little below that which was kept up by selection. It is clear
that some form of selection must always co-operate in degeneration, such
as economy of growth, which he hardly notices except as a possible but
not a necessary factor, or actual injuriousness. It appears to me that
what is wanted is to take a number of typical cases, and in each of them
show how Natural Selection comes in to carry on the degeneration begun
by panmixia. Weismann's treatment of the subject is merely begging the
question.--Yours faithfully,

A.R. WALLACE.

* * * * *

TO PROF. POULTON

_Parkstone, Dorset. August 29, 1892._

My dear Mr. Poulton,--As to panmixia you have quite misunderstood my
position. By the "mean condition," I do not mean the "mean" during the
whole course of development of the organ, as you seem to take it. That
would indeed be absurd. I do mean the "mean" of the whole series of
individual variations now occurring, during a period sufficient to
contain all or almost all the variations to which the species is _now_
subject. Take, for instance, such a case as the wings of the swallow, on
the full development of which the life of the bird depends. Many
individuals no doubt perish for lack of wing-power, due to deficiency in
size or form of wing, or in the muscles which move it. The extreme
limits of variation would be seen probably if we examined every swallow
that had reached maturity during the last century. The average of all
those would perhaps be 5 or 10 per cent. below the average of those that
survive to become the parents of the next generation in any year; and
what I maintain is, that panmixia alone could not reduce a swallow's
wings below