s in prolified flowers; they
assume, for instance, a leaf-like or petal-like condition, or take on
them more or less of a carpellary form, or they may be entirely absent;
but none of these changes seem to be at all necessarily connected with
the proliferous state of the flower. Of more interest is the alteration
in the position of these organs which sometimes necessarily accrues from
the elongation of the axis and the disjunction of the calyx; thus, in
proliferous roses the stamens become strictly hypogynous, instead of
remaining perigynous. In _Umbelliferae_ the epigynous condition is
changed for the perigynous, &c.

The condition of the pistillary organs in prolified flowers has already
been alluded to. Hitherto those instances have been considered in which
either the carpels were absent, or the new bud proceeded from between
the carpels. There is also an interesting class of cases where the
prolification is strictly intra-carpellary; the axis is so slightly
prolonged that it does not protrude beyond the carpels, does not
separate them in any way, but is wholly enclosed within their cavity.
Doubtless, in many cases, this is merely a less perfect development of
that change in which the axis protrudes beyond the carpels. This
intra-carpellary prolification occurs most frequently in plants having a
free central placenta, though it is not confined to them, as it is
recorded among _Boragineae_. A remarkable instance of this is described
by Mr. H. C. Watson in the first volume of Henfrey's 'Botanical
Gazette,' p. 88. In this specimen a raceme of small flowers was included
within the enlarged pericarp of a species of _Anchusa_. But the most
curious instances of this form of prolification are, no doubt, those
which are met with among _Primulaceae_ and other orders with free central
placentation.

Duchartre, in his memoir on the organogeny of plants with a free central
placenta, in the 'Ann. des Sc. Nat.,' 3 ser., 1844, p. 290, among other
similar instances, mentions two flowers of _Cortusa Matthioli_, wherein
the placenta was ovuliferous at the base; but the upper portion, instead
of simply elongating itself into a sterile cone, had produced a little
flower with its parts slightly different from those of the normal
flowers. M. Alph. de Candolle has likewise described somewhat similar
deviations, and one in particular in _Primula Auricula_, where the
elongated placenta gave off long and dilated funiculi bearing ovules,
while