AND AFFINITIES OF STEPHANOSPERMUM, 367 
the nucellar epidermis, and sharing, in common with these two layers (with which it is 
continuous), the characteristic blackened structureless contents. Above this layer is a 
considerable zone of thin-walled parenchyma ( p.), continuous in the peripheral direction 
with the hypodermal tissue of the nucellus. In the direction of the entering bundle this 
layer all but dies out, being intercepted by the longest of the palisade-cells of the sclero- 
testa in this region. Nevertheless it was probably continuous with the delicate one-layered 
parenchyma-sheath (or possibly phloem) which accompanied the tracheal strand into the 
seed. The layer of thin-walled parenchyma (p.) is very conspicuous in transverse sections 
of the seed cut in this plane, and occupies a large portion of such sections. 
The tracheal bundle which enters at the little basal papilla is slender, and runs 
unaltered till the thick palisade-layer is traversed. In optical section it is found to be 
four or five elements across, each tracheal element being very narrow (10 p) and having 
rather close scalariform markings. A protoxylem was not detected. Level with the 
internal limit of the palisade it broadens out funnelwise, ultimately expanding into 
the broad tracheal plate, so usual a feature in the seeds of this period (/.d.). The elements 
composing the plate and the funnel are broad and short somewhat irregular tracheides (see 
Pl. 43. fig. 21, ¢.d., and fig. 27), and they form a close plexus: 234 X404 are average 
dimensions. As the margins of the plate are approached the tracheides become longer 
and thinner and more regularly arranged, and as we draw level with the angle where the 
nucellus becomes free, the tracheal sheath has assumed the characters which, as regards 
form of elements, dimensions, and thickness, characterise it as a tracheal sheath or mantle 
enclosing the macrospore in its entirety. It is a continuous sheath without trace of 
local segregation into bundles. In the living seed the base of the macrospore rested on 
the tracheal plate, as indeed is often the case in the fossil. In the case rendered in the 
drawing (fig. 21), the spore-wall (m.w.) is withdrawn some little distance, and has carried 
with it residues of tissue. Not far above the level shown in fig. 21 the macrospore-wall 
is found in close contact with the remains of the nucellus (Pl. 41. fig. 7, a). Thus it 
appears that the delicate bundle expands into a thin but continuous sheath, which forms 
a complete investment to the macrospore; whilst the layer of parenchyma, which 
accompanies it into the seed after undergoing considerable augmentation below the 
tracheal plate, thins out and is continuous with the hypoderm of the nucellus. 
The chalazal papilla (Pl. 41. figs. 1 5 & 7) has a diameter of 120; if this represent the 
full width of its attachment, the seed was doubtless pendulous. But it is premature to 
draw a conclusion until decisive evidence is forthcoming as to the existence of a 
sarcotesta. 
The Endotesta.—For the existence of a layer interior to the sclerotesta adequate 
evidence exists. In Renault's figure of the chalaza * a delicate network of thin-walled 
cells is represented lying in the angle between the base of the nucellus and the lining of 
the sclerotesta, an observation confirmed by specimens that have passed through my 
hands. Unfortunately this network of cells is rarely preserved so as to make it possible 
to determine whether it was continuous with the tissues between which it lay. In the 
* Renault, ‘ Cours de bot. fossile, vol. iv. pl. 22. fig. 1. 
