20 DEVELOPMENT OF THE FLOWER 
lines of septa, it becomes a question whether such instances 
as Composite or Graminez are reducible to a parietal pla- 
centation or to its obverse, a central free placentation. 
[—Go through parietal placentation. ] 
For excellent remarks on the anomalous situations of 
ovula, see Lindl. Introd. p. 204, para. 1. Turpin’s instance of 
Ornithogalum is doubtful, unless he found the ovules on the 
veins; for I believe that in every vascular part, the ovules 
and buds will have direct communication with the vessels. 
In a Liliacea from Pushut, much like Royle's Fritillaria 
Thomsoniana, the connection of ovuliform buds with the 
leaves is most marked. _ 
As the midrib of leaves is branched in various situations, 
although always least so towards its apex, it may be asked 
cannot a style of a simple ovary be branched? For, it is 
at once evident, that the style does not consist merely of the 
midrib; if it did, we should require instances of the midrib 
being branched after its production beyond the lamina. 
But it consists of more or less of the lamina, often perhaps 
of the mere margin. And if we take into consideration what 
would be the nature of an ovarium formed from the convolu- 
tion of a divided leaf, such as a palmate leaf, we can hardly 
deny the possibility of the occurrence of several apparent 
styles belonging to a simple ovary. 
It may be urged against this with every apparent truth, that 
such a structure is incompatible with a carpellary leaf, such 
leaf being a leaf in its earliest simplest form, and I need not 
add that all divisions in a leaf are comparatively late deve- , 
lopments. By such an assumption I do not wish to attempt 
to explain such instances as Statice, Composite, and Grami- 
new, because absolutely the existence of several styles to a 
simple ovary will require an amount of unilaterality of the 
styles, with respect to the ventral suture. 
Additional proof would be afforded by the examination of 
the vessels, for that vessel which corresponded to the midrib 
would undoubtedly be the largest in a simple ovarium, and 
in the most extreme case imaginable the lateral ones would 
converge towards it at some point or another. 
