IN FUCHSIA GLOBOSA. 411 
mentioned before (fig. 23) has two appendages, of which fig. 26 
gives a back view. In the latter two cases the appendages spring 
from the boundary between filament and anther, but in fig. 3 a 
filiform appendage is seen leaving the filament. This thread may 
be compared with an outgrowth described by Goebel *; according 
to whom it appears now and then on the inner side of the 
antisepalous stamens, and contains pollen-grains. It seems very 
probable that these thread-like outgrowths are sometimes addi- 
tional parts of the floral axis, as already explained in $ 1; but, 
on the other hand, there are cases in which it is quite open to us 
tolook upon them as the result of chorisis of the stamen which 
shows them. 
There is every reason to suppose that the stamens of Fuchsia, 
like those of so many other plants, have a tendency to petalody. 
Masters distinguishes the following cases :— 
1. The anther-cell becomes completely or partially petaloid, the 
filament remaining unaltered. 
2. The connective has grown out into a tubular petal. 
3. The whole stamen has been transformed into a cup-shaped 
petal. 
4. The filament is unchanged, the anther has the shape of a 
petaloid cup, from the middle of which spring two imperfect 
pollen-cavities, whereas the other pollen-cavities are petaloid. 
5. The filament is petaloid, and bears an anther-cell on each 
side. 
We can supplement these by the following cases, of which 
some, undoubtedly, correspond to those already mentioned. 
Suringar describes a flower in which three of the antisepalous 
stamens are replaced by three spathulate petals, and three of the 
outer whorl of stamens are normal, the fourth having become a 
petal. . 
In the flower described by Buchenau (Pl. LVIII. fig. 5), according 
to his explanation, one of the antisepalous stamens had developed to 
an organ that is partly sepaloid, partly petaloid. Whether the part 
thus interpreted is inserted at the right place, viz. opposite to a 
sepal, it is impossible to determine from the drawing. 
Sometimes we have ourselves found a stamen appearing petaloid 
on one side, normal on the other (P). LVII. fig. 27). As regards 
one of these cases, we are quite sure that the altered stamens 
were superposed to the petals. 
* In Pringsh. Jahrb, für Wiss. Bot. 1886, p. 247. 
