ON THE STRUCTURE OF CRUCIFEROUS FLOWERS. 5 
curious monstrosity became known, inquired whether such might 
not be the real primitive type of the order, and whether in the 
usual state of these plants there might not exist a constant abor- 
tion of the whole of the two lateral flowers, excepting one stamen. 
This explanation is inadmissible, if not absurd, and has been suc- 
cessfully combated by M. Lestiboudois. In the Teratology of 
our own species, it might as safely be asserted, when a six-fingered 
child is produced, that three embryonary ova had met together, 
and that two of the foetuses, save one finger of each, had disappeared 
by abortion. 
De Candolle, himself, has shown in his Memoir on Cruciferæ, 
that each pair of geminate stamens has really only the value of a 
single organ, and consequently that the andrœceum in Cruciferæ 
may, like the corolla and calyx, be reduced to the quaternary type. 
The filaments in this order are usually thin, and widened by 
compression, like ribands: those of the longer stamens occupy, 
therefore, much more space than a regular alternation requires. 
Their bases extend right and left, at times so far as even to place 
themselves in front of the margins of the petals. It is this, pro- 
bably, which led several botanists, (as we have seen,) to imagine 
that the longer stamens were opposed to the elements: of the 
corolla. If, however, we consider the two to be in reality but 
one, we shall find that their point of separation, which represents — : 
the middle of the primitive organ, is opposite to no part of the 
corolla, but invariably alternate with it. This is still more appa- 
rent in the flowers of Sterigma tomentosum, and Anchonium Bil- 
— lardieri, in which these stamens remain undivided below, and 
the common filament is in strict alternation with the petals. 
. One of us, long since, adopted this opinion, in his Essay on the 
m reduplieation of Organs,* a work in which he called the attention 
of botanists to the numerical increase of organs, and showed its 
importance in organography, teratology, and taxonomy. 2t 
De Candolle had clearly indicated this phenomenon in his - 
Memoir cited above, since he compares the double stamens to the 
_ * Essai sur les dédoublements, Montpel. 1826. in 4to.—Elem. de Térat. végét. 
a 1841, p. 337. 
