that many
intermediate stages could be traced between the ordinary ovule and the
ovary (fig. 147, 1, _a_, 2, _b_). Some of these carpels, thus derived
from the ovules, themselves bore secondary ovules on a marginal
placenta, as shown in the sections at _c_, _d_, _e_. Could such a change
occur in the animal kingdom, there would be the unfertilised ovum
converted into an ovary, and this again bearing Graafian vesicles! In
Mr. Berkeley's carnation the change was not so great, seeing that the
nucleus of the ovule was not developed, and sufficient evidence has been
above given as to the foliar nature of the primine, while for a leaf to
be folded up so as to form a carpel is an ordinary occurrence.

It is worthy of remark that in these foliaceous ovules there is never
more than one coat, the secondine and other integuments do not make
their appearance in these cases, and that very generally the change in
question accompanies a similar foliaceous condition in the carpel, the
margins of which are more or less disunited.

Prof. A. Braun remarks that up to this date no such change has been
observed in the ovules of Monocotyledons.

=Changes in the nucleus of the ovule.=--The preceding remarks have had
reference especially to the ovular coats, but it is desirable also to
allude to certain points connected with the nucleus. Very frequently,
when the coat of the ovule is phylloid, as before described, the nucleus
is altogether wanting, though sometimes it is present as a small
cellular papilla; very rarely is it to be found in its perfect state.
Occasionally the nucleus is present in the guise of a small elongated
branch. Wigand cites ovular buds in every stage of progress into a
branch, sometimes even bearing indications of anthers. Wydler has
observed a similar occurrence in ovules of _Alliaria officinalis_, and
Schimper has described and figured specimens of _Nigella damascena_ in
which the outer coats of the ovule were but little changed, while the
nucleus was replaced by a leafy shoot. On one of the leaves of this
latter was found an imperfect ovule--an ovule on an ovule!

Fig. 148 shows a floret of a species of _Gaillardia_, in which the ovule
was replaced by a leafy shoot which had made its way through a chink in
the ovary. In this specimen, however, there was no evidence to show
whether the shoot in question was a perverted development of the
nucleus, or whether it was wholly independent of the ovule.

[Illustration: FI