IN FUCHSIA GLOBOSA. 403 
But as it is, this view cannot be reasonably held. We again draw 
attention to the fact that in the same way the additional parts may 
thicken their midrib, and consequently become equally capable 
of producing a more or less incomplete stamen (fig. 2). 
As to the variations described under the heading a and 5, it will 
have become clear that either or both may affect the same petal *. 
(c) Adhesion of the petal to the antipetalous stamen and its 
attenuation.—The process of enatiou, in its results just referred 
to, may easily give rise to a confusion with staminody, but 
the danger becomes much greater in the case of the variation 
which we are about to describe. A well-marked instance of 
mimie staminody of this sort came under my observation some 
years ago, and was described by me as an example of genuine 
staminody f. A closer examination, however, of a great number 
of flowers revealed the real cause, viz. adhesion of a petal to a 
stamen. 
Generally speaking the process may be supposed to take 
place as follows :—starting from the fact that in a normal flower 
a stamen is anteposed to a petal, the first modification is a taper- 
ing of the basis of the petal so as to form a claw. 
In this stage the petal either remains quite independent of 
the stamen, or the elaw grows together with the filament. In 
a subsequent stage, the claw and the filament bave quite coa- 
lesced into one single part, bearing an anther and a reduced 
lamina at top. Let this petaloid appendage be supposed to grow 
gradually smaller (for instance, as shown in Pl. LVII. fig. 15) and 
finally even to disappear, then the ultimate result of this variation 
must be a stamen inserted at the spot as a rule occupied by both a 
petal and a stamen. The different stages of coalescence of petal 
and stamen in conjunction with a gradual falling off of the 
former part are illustrated by figs. 15, 16, 17, 18,and 19. For the 
sake of clearness we have added to the figures floral formule, 
which briefly express the degree of adhesion. 
s, indicates that petal and stamen are wholly separate, no 
matter whether or not the petal has a claw, whether it has the 
ordinary shape, or has become cup-shaped. (Pl. LVIL figs. 16 &17.) 
* Since in double flowers of Fuchsia additional stamens occasionally appear 
between the additional petals, it is not improbable that the former are pro- 
duced by the latter. 
* Nature, 1885, May 21. 
LINN. JOURN.—BOTANY, VOL. XXV. 2r 
