629 
Ovule of Bowenia spectahilis. 
and appear more densely granular than the surrounding nucellar cells. This 
tissue appears to act as a nutritive tissue or tapetum. All the evidence points 
to the tapetum being a modification of the sporogenous tissue as in Stangeria. 
In rather later stages of the ovule, crushed and disorganized cells of 
the tissue lie immediately around the embryo-sac (Text-fig. 5). The nutri- 
ment which they contained has evidently been absorbed by its encroach- 
ment. In this same ovule, which is of the size indicated in Text-fig. 3, the 
outermost cells of the sporogenous tissue show signs of active division. It 
is evident that for some time the increase by division of the outer sporogenous 
cells keeps pace with the destruction of the innermost cells by the embryo- 
sac, and the result is that the embryo-sac is surrounded until a much later 
stage by a layer of nutritive tissue. Cf. Text-fig. 5, and Fig. 5, PL LXI. 
The last of the series of normal, unpollinated ovules is much larger 
than the previous stages described, and shows marked advances in the 
structure and differentiation of the various parts. One of these ovules 
is seen in median longitudinal section in Text-fig. 4, which shows the 
general relation of parts. The free part of the nucellus is a large dome- 
shaped mass (Fig. 7, PI. LXI), surrounded by the integument, which is 
o*6 mm. thick, and shows signs of differentiation into the three characteristic 
layers. The micropylar tube is wide and contains mucilage, the function 
of which is probably to assist later in catching the pollen-grains. The 
embryo-sac is so large that it almost fills the lower part of the nucellus. 
The chief interest of an ovule of this age is that it shows the first 
indication of the formation of the pollen-chamber (Fig. 7, PI. LXI). Run- 
ning down the centre of the nucellus from the tip to a short distance above 
the level at which it fuses with the integument, a length of 1 mm., is a strand 
of cells. These are conspicuous since they are elongated in the vertical 
direction, while the cells immediately around them are slightly elongated in 
the horizontal direction and thus appear to radiate from the central strand. 
Also the surrounding cells have denser contents, larger nuclei, and appear in 
a more active state of growth than the central cells. Warming 1 figured 
cells of this nature in the nucellus of Zamia , and described them as a kind 
of conducting tissue. A similar tissue was also figured by Brongniart 2 in 
Cavdiocarpus augustodensis , also by Stopes 3 in Cycas Rumphii . Although 
there is at this stage no sign of breaking down of tissue, it is evident that 
this central strand indicates the area which will ultimately dissolve in the 
formation of the pollen-chamber. The embryo-sac has grown with the rest 
of the ovule, being i-8 mm. long by 1-2 mm. wide. It still appears essen- 
tially as in the last stage, a hollow sac with a thin lining of protoplasm con- 
taining nuclei, some of which are undergoing division, but with no sign of 
1 Warming t Rds. du Bull, de l’Acad. Roy. Dan. des Sciences et des Lettres, 1877. 
2 Brongniart : Recherches sur les graines silicifides. Paris, 1881. 
3 Stopes : Beitrage zur Kenntnis der Fortpflanzungsorgane der Cycadeen. Flora, vol. xciii, 1904. 
