ably spreads
from the circumference to the centre,--that is, from the ray florets,
which so often include rudimentary organs, to those of the disc. I may
add, as bearing on this subject, that, with Asters, seeds taken from
the florets of the circumference have been found to yield the greatest
number of double flowers.[802] In these several cases we have a natural
tendency in certain parts to become rudimentary, and this under culture
spreads either to, or from, the axis of the plant. It deserves notice,
as showing how the same laws govern the changes which natural species
and artificial varieties undergo, that in a series of species in the
genus Carthamus, one of the Compositae, a tendency in the seeds to the
abortion of the pappus may be traced extending from the circumference
to the centre of the disc: thus, according to A. de Jussieu,[803] the
abortion is only partial in _Carthamus creticus_, but more extended in
_C. lanatus_; for in this species two or three alone of the central
seeds are furnished with a pappus, the surrounding seeds being either
quite naked or furnished with a few hairs; and lastly, in _C.
tinctorius_, even the central seeds are destitute of pappus, and the
abortion is complete.

With animals and plants under domestication, when an organ disappears,
leaving only a rudiment, the loss has generally been sudden, as with
hornless and tailless breeds; and such cases may be ranked as inherited
monstrosities. But in some few cases the loss has been gradual, and
{317} has been partly effected by selection, as with the rudimentary
combs and wattles of certain fowls. We have also seen that the wings of
some domesticated birds have been slightly reduced by disuse, and the
great reduction of the wings in certain silk-moths, with mere rudiments
left, has probably been aided by disuse.

With species in a state of nature, rudimentary organs are so extremely
common that scarcely one can be named which is wholly free from a
blemish of this nature. Such organs are generally variable, as several
naturalists have observed; for, being useless, they are not regulated
by natural selection, and they are more or less liable to reversion.
The same rule certainly holds good with parts which have become
rudimentary under domestication. We do not know through what steps
under nature rudim