166 REV. PROF. G. HENSLOW ON THE 
stigma were remarkably long, being suggestive ofan approach to an 
anemophilous condition. Around these papille the pollen-tubes 
clung and twisted just like tendrils (1) Figure 2 represents a 
portion of a staminal whorl, the filaments have a contraction at 
the base, while the receptacle has a papillose surface. 
The cord which supplies a carpel divides at first into two 
branches, one being dorsal, the other marginal or placental. The 
latter supplies the basal ovule with a branch and then passes over 
the ovary and curls backwards, finally joining the dorsal cord 
near the summit (3). 
IV. Erantuis nvEMALIs.—The peduncle contains a single 
eircle of large and small cords surrounding a central lacuna (1). 
After supplying the involucre the ring re-forms and then the 
cords increase till they form a complete pentangular prism (2), 
which sends off five branches to the sepals ; the method of sup- 
plying the stamens is the same as in other cases. On arriving 
at the summit of the receptacle the remaining cords are at first 
circular (3); they then become grouped into triplets (4), each 
group being finally enclosed in a quadrangular mass of tissue, 
these masses, which vary in number, being all united by their 
bases (5). Each of them now becomes rounded and isolated, 
containing three cords (6). The trachez are central in position. 
One cord separates from the other two (7) and becomes dorsal 
(8); the others are placentary (9). 
II. PAPAVERACEÆ.—v. Papaver Rraas.—The pedicel 
contains a circle of large and small cords (1). These increase 
until a complete circle is formed (2,3). A large number of cords 
are then given off, which branch in a dendritic manner towards 
the circumference in order to supply the two sepals (3) and the 
four petals (4, p). The somewhat disarranged circle now closes 
up again and, becoming symmetrical, proceeds to supply the 
stamens, just as in Ranunculacee (5, st.). Just below the pistil, 
sections show from five to ten large cords which stand in front of, 
and are destined to su pply, the placentiferous plate-like processes 
(6). Intermediate with these cords there are at first no tracheids 
at all, but a branching meristematic process can be seen iuserted 
in the cellular tissue (7). This will be differentiated into inde- 
pendent groups of tracheids(7 a), finall y branching in all directions, 
especially towards the ovarian cavity (8, Zr) and up the wall of 
the ovary. Sections a little higher up than that represented in 
