210 NORMAL STRUCTURE OF CRUCIFEROUS FLOWERS. 
the same line as the four long stamens in Cheiranthus, they were in the 
direction of the two short ones in Jéeris. If we 
then assume the longitudinal cells suppressed in 
Cheiranthus, and the transverse ones developed, 
and the longitudinal cells of Zéeris developed, 
and the transverse suppressed by supplying the 
missing cells of Cheiranthus from Iberis, we get 
a 4-celled pod (Fig. 3),—without doubt the 
normal condition of the fruit of Crucifere, as in 
Tetrapoma pyriforme, Seem. Botany Herald, t. 2. 
Somewhat in proof of this the two missing cells 
of Cheiranthus are indicated by two small gland- 
like bodies at the base of the pod, and between 
the two fully-developed cells (a, Fig. 1, and x, Fig. 4). Now that the 
4 cells of the pod are arrived at, indications of the normal number of 
stamens is to be ascertained; of course the six stamens are altogether 
out of the question. 
The usual number of stamens found in Crucifere is six—two pairs 
with long, and two single stamens with short filaments, the shorter 
ones further removed from the pistil, or making part of a second whorl. 
But if a large number of flowers are examined, it will be observed that 
it is not at all unusual, or very rare, for the short stamens to occur in 
pairs also (D,D, Fig. 3), or with one short stamen on one side, and a 
pair of short ones on the other. TI therefore assume provisionally that 
the number may be put down as eight, two pairs of long and two pairs 
of short in two whorls, The only difference this makes as to whether 
the short stamens are assumed to grow singly or in pairs is, that the 
number will range between 14 and 16, according to whether the two 
extra stamens are absent or present, but I believe the ultimate number 
to be 16. This agrees, in the most remarkable manner, with the Kast 
Indian Megacarpea polyandra, Benth. in Kew Journ. vii. t. 7 and 8.* 
* Mr. Bentham, after r giving the history of this remarkable plant, its discover 
by Strachey and Winterbottom in t the Himalayas, its ibeliiita by ss srias Madden 
to Glasnevin, and its distribution of fresh specimens by Dr. Moore to various bota- 
nists, continues :—“ Both Dr. Hooker and myself took the aly 4 of pee 
ing a considerable number of buds in various stages of dev velopment, as well as 
oe veggies but again failed in detecting any daa or puma. in se 
' ment, even when the number of stamens, twelve or 
Multiple of that of the petals or sepals. Dr. Lin ndley indeed believed "m had cual 
traces of an arrangement in two distinct series, each double in number to that of 
