82 Oliver . — On the Structure , Development , 
posterior aspect), so that the anterior lamella towers above 
it. A longitudinal median section shows the exact relations 
of the lamellae (Fig. 22). The surfaces, corresponding to the 
inner applied faces of an unmodified stigma, are covered with 
stigmatic hairs — the posterior, reduced lobe (p-l.) having its 
papillose surface continued some little distance down its 
posterior face (see Fig. 22). The two vascular bundles, 
running respectively dorsally and ventrally in the style, die 
out at the base of the stigmatic lobes. 
Passing on to the ovary itself, I have been able to show 
that this is in reality two-celled ; but from the fact of one of 
these cells — the anterior — being quite rudimentary, it escaped 
notice in the original diagnosis made from imperfect material. 
The placentation is axile, and the two ovules are inserted 
high up in the free part of the ovary. These are shown 
in situ in Fig. 18; here we have represented the upper part 
of the ovary after the removal of the right side. Both are 
pendulous, and apparently anatropous, with superior (and 
exterior) micropyle. They are attached, right and left of 
the median line, to the top of the partition separating the 
reduced and fully-developed loculi. The upper ( ov d) at- 
tached on the right side of the median line is sessile, but 
the lower one (ov. 2 ) is suspended by a longish funicle (fun.). 
The point of insertion of the funicle is slightly below that 
of the sessile ovule (ov. 1 ), and to the left of the median 
line. In Figs. 19 and 20 are given longitudinal sections 
through the insertion of the ovules ov . 1 and ov . 2 respect- 
ively; Fig. 19 being slightly to the right and Fig. 20 to the 
left of the absolutely median section. 
The rudimentary loculus (red. /.) is on the anterior side 
of the partition, and there is projecting into it, near the top, 
a small cushion of tissue (r Figs. 18 and 21), representing 
perhaps a rudimentary ovule. 
especially well seen in a median anteroposterior section of a cleistogamic flower 
(Fig. 21), in which the stigma is sessile on the ovary. This would point to the fact 
that entirely different factors have been at work in causing the reduction in the two 
cases, and that we have not to deal simply with the gradual atrophy of one or 
other carpel, as might at first be supposed. 
