also to be borne out by what happens in
the _Pomaceae_. In some cases in this sub-order, the calyx becomes
detached from the carpels, so that the latter organs become more or less
"superior," and distinct one from the other. This happens constantly in
the double-flowered thorn, _Crataegus Oxyacantha_, in some blossoms of
which the hollowed end of the peduncle still invests the base of the
carpels, leaving the upper portions detached. In apples flowers are
occasionally met with of greater size than usual and on longer stalks,
so that the whole looks more like a rose than an apple blossom. In these
cases it will usually be found that the calyx consists of distinct
sepals, without a trace of the ordinary swelling beneath the flower. The
petals are often more numerous than usual; the stamens variously
changed, and the carpels sometimes absent; at other times, as in the
instance figured in the adjacent woodcuts, figs. 36, 37, consisting of
separate, superior ovaries, sometimes destitute of ovules, or, at other
times, having two of these bodies.[85]

[Illustration: FIG. 35.--Proliferous Rose. Showing an absence of the
usual dilatation of the flower-stalk, and other changes.]

This condition accords precisely with the account of the development of
the flowers in _Pomaceae_ as given by Payer, Caspary, and others, so that
the flowers above described would owe their deficiency of the swollen
receptacle to an arrest of development. M. Germain de Saint Pierre,
among other malformations of the rose, presented to the Botanical
Society of France in 1854[86] two specimens which are of special
interest as relating to this contested point. In the one, the swollen
portion beneath the flower was surmounted by five perfect leaves, as,
indeed, is not infrequent in such malformations; here, then, the calyx
could have had little or no share in the production of the swelling in
question. In the other, the swollen portion was actually above the
insertion of the sepals here represented by five perfect leaves.

[Illustration: FIG. 36.--Section through Apple blossom, showing
detachment of calyx from ovaries, absence of dilated flower-stalk, &c.]

[Illustration: FIG. 37.--Calyx detached from carpels in Apple.]

On the other hand, M. Planchon's specimen of the Quince before alluded
to, not to mention other instances, tends to show that the bases of the
sepals do sometimes enter into the composition of the pome. And, indeed,
in many of these case