92 DEVELOPMENT OF THE FLOWER. 
oriens, sulcis indistinctis tot quot stigmata, apice 5-divisus. 
Stigmata, 5, punctiformi-capitata, quodque vasorum fascicu- 
lis 3, Ovulum solidum, Raphe simplici. 
Carpella oblonga, rugosa, calyce patenti immutato cincta, 
armata stylo persistente, (divisionibus tantum lapsis) Peri- 
carpium subsucculentum. ' 
Semen anatropum conforme erectum, testa carnosiusculá 
embryonario sacculo membranaceo. Album. 0, Embryo vi- 
ride. Cotyledonze magne raphi parallela. Pl. 94.— Fig. 20. g. 
raphe, A. axis. 
Radicula brevis, conica, hilum spectans. 
I do not see any true perigynism in the genus, and it is 
well worthy of enquiry what are the limits of the application 
of this term; for in this case, and if I remember rightly in 
some others, it is merely applied because the base of the 
calyx is flat; true perigynism, may perhaps be limited to 
such insertions as are above the plane of the exsertion or 
origin of the calyx. Escholtzia is a manifest instance in 
point, and I know others proving that the torus may fre- 
quently be dilated before the exsertion of the parts consti- 
tuting the perianth. 
[Why should not the process of dilatation of a Composita 
take place in a simple flower ?] The style has 5 vascular 
fascicles, the ovaries are veinless, the 5 vascular bundles 
divide above, each branch of the style having 2, the third 
and longest vascular-looking line being the ventral suture, - 
the vessels supplying the ovule give off near their entrance 
into the ovule a branch which runs up the style. 
[The venation of the style is certainly Brunonian ? Conse- 
quently the five primary vessels correspond to the sinuses 
between the lobes of the ovary, an unheard of case.] 
The ovules have no distinction of integuments. 
This Brunonian venation at once led to the enquiry whe- 
ther the carpellary leaves and the styles might not really be 
bipartite basin usque, the styles and branches contracting an 
adhesion with each other. This supposition would however 
imply a dimidiation of one carpel, the number being five, but 
this I do not find to be the case. This instance I consider 
analogous to the venation of the corolla of Composite ; it at 
