other funiculi were destitute of these bodies, but were much
dilated and foliaceous in appearance.[129] In some flowers of
_Rhododendron_ I have observed a similar condition of the ovules, which,
moreover, in the primary flowers, were attached to the walls of the
carpels--parietal placentation.

In speaking of these as cases of intra-carpellary prolification, it is,
of course, impossible to overlook the fact that they differ in degree
only from those cases where the lengthened axis projects beyond the
cavity of the carpels; nevertheless they seem to demand special notice,
because in these particular plants the placenta or its prolongation
appears never to protrude beyond the carpels, or at least very rarely.
There are, however, numerous instances of such an extension of the
placenta and of prolification occurring among _Primulaceae_ in
conjunction with the more or less complete arrest of growth of the
carpels.[130] An instance of this kind has come under my own notice in a
monstrosity of the chinese primrose, in which the carpels were reduced
to a hardly discernible rim surrounding an umbel of five rays, each
terminated by a small normally constituted flower-bud.

The ovules of a prolified flower are either unaffected, or they occur in
a rudimentary form, or, lastly, they may be present in the guise of
small leaves.

Under the term prolification of the fruit two or three distinct kinds of
malformation appear to have been included. The term seems usually to be
applied to those cases where from the centre of one fruit a branch
bearing leaves, flowers, or another fruit, is seen to project, as
happens occasionally in pears. Now, in many instances, not only the
fruit, is repeated, but also the outer portions of the flower, which
wither and fall away as the adventitious fruit ripens; so that at length
the phenomenon of one fruit projecting from another is produced. It is
obvious that this form of prolification in no wise differs from ordinary
central prolification. Sometimes some of the whorls of the adventitious
flower are suppressed; thus, M. Duchartre describes some orange blossoms
as presenting alternating series of stamens and pistils one above
another, while the calyces and corollas belonging to each series of
stamens and pistils were entirely suppressed.[131] In other cases,
doubtless, the carpellary whorl is alone repeated, the other whorls of
the adventitious flower being completely absent.

Another condition, a