CISSAMPELOS. 311 
duced to one sepal and one petal which is from its situation 
most likely a transformed stamen. "There can be no doubt 
that both these answer to parts of the female perianth. The 
outer scale being obviously a sepal, the inner a part of the 
cup-shaped body, which from the alternation of the stamens 
with the sepals cannot be petals, but most probably results 
from a modification of the stamina, This is too perhaps the 
most natural way; but the occasional occurrence in the 
order of several carpella suggests an idea as to the composi- 
tion of the flowers of Cissampelos. In this case we shall 
have 4 sepals, and 4 scales, and 4 ovaries. 
„The greatest arguments against this are the invariable 
greater development of the 2 posticous ovaria and the sepa- 
ration of the pedicel, which last seems to me nearly con- 
clusive, the plurality of stigmata. 
Those in its favour are the invariable quaternary aggrega- 
tion of the flowers, and the situation of the scales, which 
Seem never to be developed next the axis. 
All these fine speculations are knocked on the head by the 
fact that the ventral suture is not next the axis. The first 
View is therefore the correct one. 
Ovaria pilosa postica gibba, convexava sutura ventrali an- 
tica Sita, l-locularia, biovulata an semper ovula collateralia 
subantitropa, foramen superum, stylus cylindraceus, brevis, 
stigmata subtria, simplicia. 
Sepalum unicum, anticum externumve pilosum, late obo- 
m. 
Squamula subreniformis, huic opposita, carnosa evascu- 
After fecundation the ovarium suddenly enlarges and al- 
ters ita Shape, becoming as it were curved on itself, the 
Càuse seems to be the elongation of the placenta, the 
Style and pedicel are at length brought close together. 
ltis hardened somewhat by the time the embryo has 
"PPeared, its sac which previously existed soon disappearing. 
