16 . ON THE STRUCTURE OF CRUCIFEROUS FLOWERS. 
CONCLUSION. 
If we recapitulate what has preceded, we shall arrive at the fol- 
lowing conclusion. 
The floral type of Cruciferæ is quaternary. The ca/yx is com- 
posed of 4 leaflets, the corolla of 4 petals, the receptacle has 4 
staminiferous glands, the andræceum 4 stamens, the gynwceum 4 
pistils, and the fruit 4 carpidia. 
. These verticils alternate regularly. Two stamens in the habitual 
state of the flower have been transformed into two pair by multi- 
plication (dédoublement), and two pistils have disappeared by abor- 
tion : hence the andreceum has two component parts more than it 
should have ; the gynceceum two less. 
'The four staminiferous glands are more or less irregular or in- 
complete, and are found above, below, or by the side of the fila- 
ments. "Their volume has caused a change in the position of two 
stamens and of two calycinal leaves, which makes the andreeceum 
and the calyx appear biverticillate.* 
* Since the above was written and prepared for the press, our attention has been 
called to a note of Mr. Brown, appended to his observations on Loronia acuminata 
Pl. Jav. 2, p. 106, in which he shows that each carpidium in a compound and unilo- 
cular ovarium has necessarily two stigmata (we have called this a bicephalous stigma,) 
and that the lobes, or as he has named them, stigmata of the same carpidium are 
usually confluent. 
“This rule,” he adds, “ admits of exceptions, as in Parnassia, in many Crucifere, 
. and in Papaveracee : in all vire cases the stigmata as well as placentæ of the ad- 
joining carpels are confluent." 
From this passage we are persuaded that Mr. Brown is of the same opinion with our- 
selves, and had the occasion allowed him to develop his ideas on the phyllidium or 
ovarian leaf of this order, they would have been found not very different from those we 
have attempted to explain above. "The portion, however, of our Memoir which treats of 
the gynæceum is not the less necessary ; for others have not interpreted the ideas of 
this profound observer in the same manner. 
Mr. Griffith (Trans. of Linn. Soc. vol. xix., 1845, p. 328) after citing the abiti 
passage, seems to suppose that in some genera at least the normal fruit of Cruciferz 
is composed of four carpidia, two anterior and posterior “subsequently much the — s 
smallest," whose stigmata are confluent, forming therefore what we have termed the — 
apparent stigma, and two lateral, distinct themselves as to their valves, but having 
their stigmata confluent with, and lost in, the apparent stigmata. * 
————— —— — Ht appears 
