IN FUCHSTA GLOBOSA. 407 
parted company higher up than usual. The union may, it is true, 
be seen in some of our monstrous flowers too, but not so fre- 
quently as in Scaramouche. 
As for the modified insertion of the petals, Morren em- 
phatically points out that the petals have not been produced 
by the stamens, but have moved a longer or shorter way 
up the stamen. He calls this phenomenon * métaphérie" or 
* monstruosité par transport." This * métaphérie" may proceed 
so far that the petaloid appendage reaches the top ofthe con- 
nective and gives birth to a stamen, of which the anther-cells 
are placed on the edges of a small stalked leaf. “ Gliding ” is the 
term used by Morren to characterize this process, in which the 
petal may detach itself from the stamen at any height. But evi- 
dently Morren has not seen the petal disappear altogether. In 
the Searamouche flower represented in the drawing, the note- 
worthy fact of the floral parts being superposed to the sepals 
would constitute the greatest difficulty if it were a peculiarity of 
the whole sort. But the drawing of another flower shows that it 
is not one of the constant characteristies of Scaramouche. It is 
very probable that also in the flower described, the parts are in- 
serted at their ordinary place but have been forced aside by coa- 
lescence with the antisepalous stamens. 
The frequency of petals and antipetalous stamens growing to- 
gether renders it probable that the internal organization of the 
flower is such as to predetermine the irregularity. And, in fact, 
anatomical examination reveals the circumstance, that the petal 
and the superposed stamen are supplied by the same vascular 
bundle, which bifureates on the edge of the calyx-tube. This 
vascular bundle runs through the whole calyx-tube, and may be 
traced downward to the peduncle. The relation of the sepals to 
the antisepalous stamen is quite different ; both of them have a 
separate bundle, which may be followed up through the whole 
calyx-tube and the parietes of the ovary as far as the top of the 
peduncle. In consequence, twelve vascular bundles may be seen 
going up through ovary-wall and cal yx-tube—four supplying the 
sepals, four going to the antisepalous stamens, and four to the 
petals and the antipetalous stamens taken together. This, taken 
in connection with the monstrous adhesion of the antipetalous 
stamen to the petal as before described, proves that the petals 
and the outer row of stamens have been produced by the bifur- 
cation of one whorl, and that consequently the number of auto- 
nomous whorls is not five but four. 
