AND AFFINITIES OF STEPHANOSPERMUM. 371 
The Nucellar Beak. 
The pollen-chamber is surmounted by a long canal which engages distally with the 
micropyle. At its base of insertion on the broader part of the nucellus this canal 
widens out into an inverted funnel, which overlies the pollen-chamber; but in thin 
median sections this funnel always presents the appearance of being detached from the 
wall of the pollen-chamber (Pl. 41. fig. 14, co., and Pl. 43. fig. 29, co.). The tubular 
part of this structure is evidently perforated (fig. 29), whilst in fig. 3 a local swelling 
is shown ( p.g.), due to a contained pollen-grain which has failed to reach the pollen- 
chamber. At the base of the tube the section falls just outside the cavity; higher up, 
the lumen is laid open (Pl. 43. fig. 29, from just above J to beyond m.), finally the surface 
of the tube comes into the section showing two long rectilinear cells with carbonised 
contents (Pl. 41. fig. 1 4, from the same preparation as fig. 29). 
The funnel or collar is due to the sudden spreading of the sides of the tube above the 
pollen-chamber. It is one cell thick, and ends suddenly and always in the same way 
(Pl. 43. fig. 29, x x, also Pl. 41. figs. 2, co., & 3, and Pl. 42. fig. 15 *). This was evidently the 
case in the specimens described by Brongniart}, though his figures are not highly 
magnified. The cells of the collar appear thick-walled and their contents strongly 
earbonised. Preservation ceases abruptly at a distance of about *4 mm. from the insertion 
of the tube and apparently in the middle of a cell (see fig. 29, x x). From this spot the 
collar is continued in the form of a structureless residue traceable down the flanks of the 
pollen-chamber for varying distances, and sometimes to the base of the seed where 
the space between the nucellus and sclerotesta dies out. "This smear-like continuation 
of the collar is shown in Pl. 41. figs. 1 a (end./.) and3. In another specimen (not figured) 
traces of cellular structure are apparent in the residue. 
The roof of the pollen-chamber immediately below the collar has all the appearance of 
a tissue without its delimiting layer (Pl. 42. fig. 14); i.e., the collar is in all probability the 
missing epidermis which has become detached from the apex of the pollen-chamber. If 
this be the right interpretion, it follows that the true apex of the nucellus is beak-like 
and engages with the micropyle, whilst the tubular cavity of the beak is the entrance to 
the pollen-chamber. This view is corroborated by the specimen represented in Pl. 41. 
fig. 6. This section is as nearly as possible median. The distal portion of the beak had 
undergone maceration prior to mineralisation, but traces of several pollen-grains which 
it contained are present in the micropyle (but not shown in the photograph). Of the 
funnel or collar, three fragments are present in the section and are distinctly visible 
just above the pollen-chamber. Of these fragments the two on the right are detached 
portions of cells of the same character as co. in fig. 14. The third fragment, shown 
asa tooth at the apex, is in situ; it has not been drawn away from the subjacent 
tissues. 
Figure 3, from another specimen, lends itself to interpretation in the same way. 
Here the epidermal layer of the pollen-chamber (e., e.) is continued up to the level from 
which the collar (co.) bas been drawn away. 
* Figure 15 should be examined with a hand-lens. + * Graines fossiles silicifi¢es,’ pl. 16. figs. 4, 5, 7, & 9. 
9H2 
