1120 Thoday .— The Female Inflorescence and Ovules of 
between stage B and the mature ovule takes place in the middle region 
between the two integuments. 1 
The other changes in the development of the ovule consist in the 
differentiation of the elements of the vascular-bundle system and of the 
tissues of the outer integument, the complete closure of the micropylar 
tube and further withering of the tip, the disintegration of the pollen 
chamber and hardening of the apex of the nucellus. 2 A diagram of the 
mature ovule is given on p. 1113. 
Text-fig. 14 is drawn from part of a section through an abortive ovule in 
which these changes were almost complete. The ovule was about 1*5 mm. 
long, and a good deal of stretching of the region between the two integu¬ 
ments had already taken place. It is possible to distinguish the remains 
of the pollen chamber. 
D. Details of method of closmg micropyle . The method of closing the 
micropyle in G. africanum is in some respects different from that described 
recently in G. Gnemon by Miss Berridge; 3 in G. scandens I only saw very 
young stages and the mature stage with fully closed micropyle, but it 
would appear as if in that species the resemblance is closer. The closed 
micropylar tube of the mature ovule, of which the tissues are strongly ligni- 
fied, is shown in Fig. 12, PI. LXXXVI. The epidermal cells lining the canal 
appear to have elongated and interlocked with those opposite them, 
and thus to have obliterated the cavity. 
In G. africanum the actual ingrowth of separate epidermal cells as 
hairs or papillae plays only a small part in the closure of the micropyle. 
This is due mainly to the rapid division of the epidermis to form radial 
rows of cells, the pressure of which rapidly closes the micropyle. The 
innermost elements of the radial rows occasionally project inwards, and 
sometimes appear to have therefore got crushed together. As a rule, how¬ 
ever, this only happens above the region of the flange—that is to say, 
where the micropylar tube is clasped only by the outermost covering, and 
is not so tightly compressed as lower down. 
The down-growing flange is the result of rapid cell-division in the 
outer layer of the micropylar tube; it is fringed with short hairs. It adheres 
so closely to the outer edge of the outer integument that without close 
examination the tissues appear to be continuous. 
1 Earlier authors have noticed that the depth of attachment of the inner integument varies 
in Gnetum, but they appear to have regarded the difference as specific and not to have realized that 
they were dealing with different developmental stages. E. g. Karsten : Cohn’s Beitrage, 1893, p. 397. 
The stretching of the ovule and consequent change in position of the level at which the inner in¬ 
tegument becomes free is indicated in Coulter’s diagrams (Figs. 2 A, 3 A, 6 a), but not referred to in 
his text; Coulter, 1908 ; see also Strasburger, 1879, P* io1 - 
2 No sign of the development of a pedicel between the bases of the two outer coverings which 
is described by Coulter (1908, p. 381, Fig. 6 a) in G. Gnetnon was seen by Miss Berridge or by me. 
3 Berridge, 1911. This method of closing the micropyle is analogous to that described in some 
Conifers, e. g. Widdringtoitia , figured by Saxton, 1910, Fig. 2, PI. I. 
