636 Kershaw. — Structure and Development of the 
The cells forming this ring are long and narrow and obliquely placed, 
making an acute angle with the micropylar tube. The position of these cells 
shows them to be part of the inner fleshy layer of the integument, but they 
are very closely related to the stone layer. In some parts they are actually 
in contact with the sclerized cells of the stone layer, and in other parts the 
separating parenchymatous cells themselves show signs of sclerification 
beginning. It seems probable that in an older ovule the whole of this 
basal region of the micropyle would be one sclerized mass of cells- 
There seems to be no exact parallel to this condition among living 
Cycads or fossil seeds. The upper part ot the micropylar tube is now 
closed, and can scarcely be seen in a transverse section. Lower down 
the tube becomes slightly wider, and the long pointed beak of the 
nucellus extends up into it. The separating line between inner flesh and 
stone layer is difficult to make out in the parts where the latter is not 
sclerized, also there is no obvious line of division between nucellus and 
inner flesh. The cells of both consist of rather elongated parenchymatous 
cells with scattered tannin sacs. The inner series of vascular bundles lie in 
this region and follows an imaginary line dividing nucellus and integument, 
and obtained by continuing the slit between their free portions towards the 
base. In many of the ovules the bundles terminate just below the division 
into free nucellus and integument. In several of those examined, however, 
the bundles of this inner series were found to extend for a short distance into 
the free part of the nucellus (Fig. 12, PI. LXI). From the earliest stages 
of the ovule where the bundles were beginning to develop it could be seen 
that they were not integumental, as has been recently generally assumed 
for Cycads. From the later stages examined it seems that one would be 
justified in stating that in the case of Bowenia there are strong reasons for 
assuming the inner vascular supply to be nucellar. A transverse section of 
a bundle of the inner series in the lower part of the ovule is seen in Fig. 13, 
PI. LXI. The bundles differed much from those of the outer series. The 
protoxylem is mesarch, and the greatest development of wood is in the 
centripetal direction with only a few tracheides of centrifugal direction. 
The phloem is small in amount and difficult to identify. The possible 
elements seem to be arranged all round the xylem. As compared with the 
outer collateral bundles, the inner ones have an almost concentric appear- 
ance. As the bundles branch and pass upward through the integument 
they gradually lose their mesarch appearance. The protoxylem is no longer 
recognizable, and the bundle consists of a flattened mass of short thick 
tracheides without phloem. These bundles are very numerous and close 
together in the upper part of the ovule (Text-fig. 6). 
The free area of the nucellus is a comparatively small part 1-4 mm. in 
length which caps the top of the prothallus. It is limited by a regular 
epidermis with cuticularized outer walls and terminates in a long beak-like 
