80 DEVELOPMENT OF THE FLOWER. 
vascular fascicles end just where the stigmatic surfaces pass 
on to the dorsum of the style, or rather pass beyond the 
midrib in a confluent state ; the style is at the base anteriorly 
very gibbous, and the space between the two lowermost car- 
pella is considerable; the 2 uppermost being in apposition. 
'The venation of the divisions of the ovary is, one outer 
or dorsal vein, and about 2 others on either side, confluent 
with the dorsal one, which passes almost entirely round the 
division, but does not as it ought to do (?) run again into 
the fascicles supplying the style. 
ous base of the style of the anterior carpellum is 
worthy of remark, as proving that even filiform growths may 
have a tendency to division. 
As it has thus an independent venation of its own, it may 
be assumed to represent a pinnate leaf, with a single pair of 
leaflets, and an exserted petiole. 
This hypothesis however is merely arbitrary, merely refer- - 
ring to the resemblance in configuration between the two, 
equally assumable as a palmate leaf.  Labiatze prove clear- 
ly, that great inequality may exist between the stigmatic 
surfaces of the two carpella, without any corresponding in- 
equality of function; but this is as it should be, for what- 
ever inequality may exist between the stigmata is com- 
pensated by the equal, free communication all the carpella 
have with the stigmatie canal Of this Viola is a marked 
instance, as well as Orchidez. 
In a Carduaceous genus, the stigmata might be described 
as linear subulate scabrous, commencing from a hispid ring 
at the bifurcation of the style. Pl. 30.— Fig. 3, A. 1, 1, stig- 
matic surface. 
This would not be the true structure, the real stigmatic 
surface is linear sub.involute, occupying all the inner face 
of the styles, and produced beyond its apex into a mucro. 
Both, form a channel leading directly into the stigmatic 
canal, having 2 vessels, terminating just below the scabrous 
end of dorsum. 
In a Lactucaceous genus, formed on the same plan, the 
limits of the stigmatie surfaces are obscure by its being 
