912 PROF. F. W. OLIVER ON THE STRUCTURE 
The reason for this severance of the beak and collar from the nucellus is to be sought 
probably in the fact that the nucellus underwent contraction on the death of the seed. 
The beak being trapped in the micropyle, with which it closely engages (Pl. 41. fig. 14), 
was unable to follow, consequently rupture occurred immediately below its insertion. "The 
remarkable similarity to whieh attention has been drawn, between the detached collars of 
different specimens, may be interpreted as an indication of identity in the area over which 
the induration of the epidermal cells in this region spread. The whole phenomenon 
Text-fig. 1. 
Partial restoration of Stephanospermum akenioides, Brongn. 
Showing the relations of nucellus and integument, the nucellar beak, and intact tracheal mantle.—sel.t., sclerotesta ; 
end.t., endotesta; m., micropyle; g., gap between nucellus and testa; n.e., nucellar epidermis; n.b., nucellar 
beak (indurated) ; p.c., pollen-chamber ; sh., shoulder of pollen-chamber ; t.d., tracheal dise; t.m., tracheal 
mantle; ¢.m.’, position in which schizotracheal gap developed on death of seed; m.w., macrospore-wall ; 
a., archegonium. The spongy tissue of the trough is not shown. 
seems to accentuate the fact that the epidermis at the apex of the nucellus was peculiar 
and coherent, recalling the apical patch of a Marattiaceous sporangium or Cycadean 
pollen-sae. 
| In this connection there is one peeuliarity of preservation that requires a word of 
explanation. Reference has been made to the apparently structureless continuation of 
the collar (shown in Pl. 41. figs. 14, 3, and Pl. 43, fig. 29) which is occasionally to be 
traced to the base of the seed. Actually this continuation is not a portion of the collar at 
all—that is to say, the structureless residue does not indicate that a continuous stripping 
