368 PROF. F. W. OLIVER ON THE STRUCTURE 
majority of specimens this layer is represented by a structureless fold with occasional 
traces of cellular structure preserved in it. Thus in the median longitudinal section of 
the seed shown in Pl. 41. figs. 1 a & B the remains of what seems to be such an inner layer 
are represented by the light shade (end.¢.) between nucellus and sclerotesta. In the right- 
hand half of the seed the shading is continuous from the apex of the nucellus to the 
chalaza, and for a large part of the distance is in contact with the sclerotesta. On the 
left-hand side of the nucellus a similar condition obtains, though the continuity of the 
layer in question is less perfect. 
In PI. 41. fig. 3 a similar fold is seen in contact with the sclerotesta lying to the left of 
the nucellar beak at end.t. A portion of this layer (ec) has become separated from the 
main part by a longitudinal fissure and lies near the pollen-chamber wall; lower down, 
however, this detached part is continuous with the main fold (¢.e. with the downward 
prolongation of end.t.). Again, in Pl. 42. fig. 13, from a somewhat younger seed, the 
endotesta is well shown at exd.t.—here evidently displaced from its continuity with the 
sclerotesta. This preparation demonstrates, perhaps better than any other, that the 
endotesta was continuous with the sclerotesta. Its relations with the nucellus are obscure, 
but no positive evidence of its continuity with the nucellar surface has been found. The 
point is one of some interest, as upon it depend the relations of nucellus and testa, whether 
free as in Lepidocarpon or joined as in Cycas. On the whole, I incline to the former 
view, which is embodied in the attempted restoration given in text-fig. 1, p. 372. 
The Nucellus and Pollen-chamber. 
Apart from the pollen-chamber and adjacent regions, in which the main interest centres, 
there is little to be said about the nucellus. As in so many gymnospermous seeds of 
corresponding age, its whole proximal portion has become transformed into a mere pouch 
for the accommodation of the large solitary macrospore. We rarely find preserved 
(restricting ourselves to the prothallial region) more than (1) the epidermis of thick flat 
cells with blackened contents; (2) a narrow zone within the epidermis, the elements of 
which are flattened and rarely recognisable; (3) the tracheal mantle of delicate tracheides 
with close-wound spiral and scalariform thickenings. The inner limit of the tracheal 
zone abuts almost directly on the thick wall of the macrospore. The hypoderm is con- 
tinuous with the parenchyma at the chalaza, and in the best specimens it is evident that 
it consists of several layers of cells which have become flattened upon one another. The 
tracheides are of various length and have an average diameter of ‘V1 mm. They are not 
readily seen in the nucellar wall, except in the very thinnest sections. When this con- 
dition is fulfilled the tracheides are found to be present in extraordinary abundance ia all 
parts of the wall. In the aggregate they form a continuous mantle, two or three deep, 
from the chalaza to the pollen-chamber, and there is little doubt that they closely 
invested the top of the macrospore in the same way in the living seed (cf. text-fig. 1). 
Thus we find in this seed (and doubtless in some of the allied forms) a complete isolation 
of the inner parts of the nucellus, with the macrospore and prothallium therein contained, 
by a continuous sheet of tracheides. Two photographs in illustration of these tracheides 
