7 8 Oliver . — ( 9 ^ Phy so stoma elegans , Williamson , 
the seed. There is no indication of any general sclerosis, nor is there 
any localized sclerotic layer such as characterized many of the seeds of 
the same period. The study of the histological character of the seed-wall 
gives the impression that, without being succulent, the seed was far from 
being nut-like. In this respect Physostoma was rather exceptional ; perhaps 
the comparative softness of the tissues is to be correlated with the remark- 
able enveloping layer of hairs to which reference has been made. 
Another feature relating to the testa merits passing comment, though it 
is mainly a question of preservation. In the case of the Lagenostomas 
(L. Lomaxii and Z. ovoides ) it is very rare to find the vascular strands in 
situ ; they break away with a connecting sheath of tissue from the testa 
(the ‘ bundle ring ’) and this ring is generally to be found lying contracted 
some distance away from the outer part of the testa. 1 In Physostoma such 
a separation has not been met with, and the bundles, or in their absence — 
which is not infrequent — the lacunae marking their position are always . 
found in situ (PI. V, Figs. 2, 5, and 6, vb.). The possible explanation of 
this peculiarity will be discussed later on. 
The central body of the seed (nucellus) conforms to the Cycadean type, 
and is largely occupied by the long embryo-sac or megaspore. The free 
apex is, as usual, modified to form a pollen-chamber which is very large in 
relation to the size of the seed. The internal cavity of the nucellus or 
megaspore-chamber attains a length of over four millims., and is sharply 
delimited by a uniform layer which is generally preserved as a black, opaque, 
structureless border, continuous from end to end. A characteristic feature 
in the form of the embryo-sac is the conical papilla at the apex (PI. VI, 
Fig. 18, mg.pl) which projects or bulges for a distance of f millim. into the 
pollen-chamber, like the incurved bottom of a wine-bottle. 
The pollen-chamber itself is a bell-shaped crevice embracing the pro- 
jecting apex of the embryo-sac ‘ as if it were a soft bladder half full of water, 
allowed to rest on one of the old-fashioned soda-water bottles.’ 2 
This marked overlapping of the pollen-chamber and megaspore regions 
of the nucellus is a very unusual feature in the seeds that have so far come 
to light. It recalls on a grand scale the tent-pole mechanism of Ginkgo , and 
will require further consideration. 
The orifice of the pollen-chamber is to be found at the summit, where 
it takes the form of a circular opening situated on a low papilla (PI. VI, 
Fig. 18, o. p. cl). The diagrammatic longitudinal and transverse sections, 
given on pp. 76-77, will serve to epitomize the main relations in the organiza- 
tion of Physostoma. 
1 Oliver and Scott, loc. cit., PI. VII, Fig. 2, i.s. and PI. X, Fig. 31, i. s. and b. r. 
2 Williamson, eighth Memoir, Phil. Trans., 1877, p. 242. 
