W. C. Worsdell. 
58 
is my opinion that in the case of Cycads and Cephalolaxus the 
well-characterised and contrasted parts of the seed-envelope 
represent a congenital fusion between the two integuments which, 
in the remote ancestors of the plants, were quite free. In the same 
way, congenital fusion between nucellus and integuments along the 
greater part of the ovular axis, exists in all Gymnosperms. 
In the ovule of the ancient ancestors of these plants, as seen 
in the case of Stephanos per mum, the nucellus existed free and 
independent of the integument. Oliver regards the modern 
structure, e.g., that of Cycads and Torreya , as having been derived 
from the older one by means of a “ new intercalation ” of tissue at 
the base of the ovule, which elevated the “ real ovule,” as he terms 
it, so that in the modern ovule nucellus and integument are free 
only at the apex of the organ. This view, while highly ingenious, 
is, to my mind, a mistaken and artificial one. In the first 
place, there is no evidence for supposing that any elongation of 
the ovule has ever taken place. The seed of Stepha>iosperinuin 
is approximately of the same length as that of most species 
of modern Cycads. Oliver admits that the “ newly intercalated” tissue 
consists of integument and nucellus; if this be so, then he is surely 
illogical in the position assumed; for these two organs must, in that 
case, be congenitally fused together, since he starts out with the very 
proposition that the primitive condition from which the ovule is 
derived is that in which nucellus and integument are free; it 
therefore follows that these organs in the region of the “ new 
intercalation ” must also have formerly been free, for it is to my 
mind an absolute impossibility for a new tissue thus to suddenly 
drop upon the scene from nowhere, having arisen, that is to say, 
out of no pre-existing organ or organs. Again, if the “intercalated” 
region consists of a congenital fusion of nucellus and integument, 
as our author admits, why could not that region have been formed 
by simple congenital fusion of those two organs, for the greater 
part of their length in the ancestral ovule itself, which is the view 
1 hold ? The case before us is perfectly analogous to that of the 
formation of a gamopetalous corolla, which I regard as having 
originated from the congenital fusion of primitively free petals, as 
the facts of dialysis prove; it is not only the ontogenetically free 
tips which represent the real corolla, but the tubular portion as 
well, and there is here no “new intercalation”! In the case of 
Torreya it would indeed be “an evolutionary freak ” for the chalaza 
to become elevated into the position to which our author assigns 
