IN FUCHSIA GLOBOSA. 401 
the sepals, the petals, and the stamens. In a double flower, how- 
ever, some additional bundles may appear, which can be traced a 
long way downward. But since in this direction, especially in 
the circumference of the ovary, they gradually become very 
indistinct, I could not detect any connection with the bundles 
normally occurring. Although I readily admit—from the per- 
sonal observation of full-grown flowers—that chorisis of the 
petals causes doubling, yet I think there is another cause of aug- 
mentation, viz., the intercalation of parts connected with bundles 
which have their origin in the lower part of the flower *. 
§ 2. The Variability of the Petals and their Relation to the 
Antipetalous Stamens. 
One of the commonest occurrences in Fuchsia is the tendency 
to vary its petals; even flowers looking quite normal at first 
sight, show in many cases slight traces of alteration. The modi- 
fieations which deserve most attention, because they are most 
frequent, are of three sorts. The first appears as a cup-like 
formation, the second is due to an enation from the inner side of 
the midrib, the third by the growing together of the petal with 
the antipetalous stamen. In the last case the petal displays a 
tendency to grow smaller, and even to disappear altogether. 
Staminody of the petals seems to be very rare; the changes 
that are usually classed as such are nearly always of the second 
or the third kind. Only once, besides other alterations,we have 
observed something resembling an anther, in a petal (PI. LVIII. 
fig. 4, an). That, however, actual staminody of petals may occur, 
is to be inferred from a statement by Suringar, who mentions 
a flower of which two petals had changed into stamens +. The 
same authority describes foliaceous petals, where he mentions a 
corolla consisting of four spathulate petals, whose upper side is 
slightly tinged with green, while the margins of three of them 
are toothed. Whether petals can ever take the shape of small 
tubercles, as Buchenau admits (Pl. LVIII. fig. 5), seems very 
doubtful; our objections to this view will be found on p. 414. 
We shall now treat the principal modifications in detail. 
* In the calyx-tube of single flowers, now and then some small groups of 
minute cells may be seen; they are not unlikely the first degrees of develop- 
ment of vascular tubes persisting in this stage. Are such flowers tbe offspring 
of double flowers ? 
t The two other petals are described to have changed halfway into stamens ; 
it is probable, however, that they have grown together with the stamens, as we 
shall by-and-by show is of frequent occurrence. 
