ation, it appeared that the
petaloid expansion to which the dilated filament was apparently
attached, was equally a part of the stamens; in other words, that the
filament was provided with four petal-like wings, two on each side
[Symbol: 00 topped by (turned cw 90 deg., an o above and another ( turned cw
above that]. This disposition was well seen in the anther, half of which
was, in some cases, petaloid like the filament; in fact, the inner wing
of the latter was directly continuous with the petal-like expansion from
the anther. A section through the latter showed, going from within
outwards, the cut edges of two perfect polliniferous lobes in the
centre; and on either side the petaloid wing representing the remaining
anther-lobe; outside these were the edges of the remaining wings, one on
each side. (See p. 290, fig. 155.)

=Enation from the carpels.=--The only instances of this that need be
referred to are such cases as those in which spur-like projections,
horns, tubercles, or winged expansions, are formed from the surface of
the ovary during the course of its development. The extraordinary
cornute oranges described and figured by Ferrari, Gallesio, and other
writers on the genus _Citrus_, may be mentioned under this head. A
similar formation occurs in the fruit of some species of _Solanum_. (See
p. 316.)

FOOTNOTES:

[518] It is desirable in this place to allude to a singular case of
fissiparous division of a leaf of _Prunus Laurocerasus_ described by
Prof. Alexander Dickson ('Seemann's Journ. Botany,' vol. v, 1867, p.
323), and which did not come under the writer's notice till after the
sheet relating to fission, p. 61, had been sent to press. Dr. Dickson
thus speaks of this abnormal leaf:--"The petiole (unchanged) supported
two laminae, placed back to back, and united by their midribs (_i.e._ not
separated) to within about an inch from their extremities, which were
perfectly free from each other. These laminae stood vertically, their
edges being directed towards and away from the axis; and as they were
placed back to back, the shining surfaces, corresponding in structure to
the normal upper leaf-surface, were directed laterally outwards. In the
axil of this abnormal leaf were two axillary buds. The existence of two
leaf-apices and two axillary buds shows that this was not due to an
accidental exuberance of development, but to fissiparous division,
which, had it been complete, would have resulted in the replace