4 8 
Thoday, — Anatomy of the Ovule and Seed in 
of the tube, where the wall has not increased so much in thickness, are 
mostly drawn out with the broken base of the tube. Thus the lower end of 
the tube is now found projecting as a solid, hard, lignified beak from the top 
of the middle covering. 
Other characteristics of G. Gnemon are the angular outline of the 
middle covering, and the freedom of the two outer coverings from each 
other except at the apex of the seed. 
In the seeds of G. africanum the micropylar tube is also closed by 
the ingrowth and subdivision of the epidermal cells lining the tube, and 
there is also a small flange which projects over the apex of the middle 
covering, but is not, unless at a later stage than I have seen, fused on to the 
outer covering. The micropylar tube has no papillate region, and is 
nowhere torn across. The middle covering in this seed is of highly complex 
structure. It is free from the outer covering. 
In G. scandens and G . funicidare , which closely resemble one another, 
the two outer coverings are fused. The flange projecting over the tip of the 
middle covering is very slightly developed, consisting in G. funicidare 
of a few peculiarly large lignified hairs only. The tube in G. scandens 
is closed by ingrowth and lignification of the epidermal cells, and the 
concurrent growth in thickness of its whole wall. In G. funicidare the wall 
of the tube does not increase in thickness ; consequently there is no pressure 
to assist in the closing of the lumen, and the ingrowing epidermal cells are 
looser and more irregular, like the papillate region in G. Gnemon. In 
G. scandens the middle covering, and in G. funicidare the outer and inner 
also, contain many strongly lignified fibres. 
These four species thus show very different degrees of efficiency in the 
closure of the micropyle. In G . funicidare and G. scandens there is no 
gap between the two outer coverings of the ovule to fill, for these are fused 
in the mature seed, so there is no need of an elaborate stopper, and the 
small flange of hairs growing out of the micropylar tube and adhering 
to the tip of the middle covering is sufficient. In G. africanum and 
G. Gnemon the outermost covering is free from the middle, but in the' 
former species the chink between the two is not completely closed, the 
stopper being well developed and projecting over the middle covering, but 
not being adherent to the outermost covering. In G. Gnemon the closing 
arrangement is complete, and the adherence of the stopper to the outermost 
covering has resulted in the breaking of the tube and the separation of the 
stopper from the micropylar beak. 
Comparison zvith the Bennettitales . 
From time to time comparisons have been made between the Gnetales 
and the Bennettitales; recently both Miss Berridge and I have drawn 
