- 
72 ON THE SEXUAL ORGANS OF THE CYCADACES. 
mistaken when I regarded (in the memoir quoted) the cavity of the 
embryo-sac, which soon loses its membrane, as formed by absorption 
and dilatation in the nucleus. The same error had already been com- 
mitted by others in the case of the Conifere, and the point has only 
been cleared up by the researches of Pineau. 
he free conical summit of the nucleus, in which I formerly looked 
for the embryo-sac, is situated above it, and rests upon its membrane. 
I considered as belonging to the nucleus a special vascular expansion 
formed of bundles which, after penetrating the ovules, rise above the 
external bundles, perforate the woody layer of the coat (producing the 
holes in its base; see Plate XCI. fig. 16), and distribute themselves, by 
ramifying and anastomosing, on the interior surface of the coat. They 
terminate above just at the point where the nucleus becomes free; it is 
consequently blended with the coat for two-thirds of its height. I 
had noticed this internal vascular system in all the Cycadacee, but it 
escaped my notice at first that it exists previous to fertilization. It 
has since been also made out in the Conifere.* Guided by analogy, 
I considered myself justified in terming it an expansion of the cha- 
laza.t In the ripe fruits it appears much more distinctly ; and when 
the remains of the nucleus which cover it are reduced to a thin mem- 
brane, as in Macrozamia and in a Cycad from New Holland, it is seen 
through it, and produces reticulated impressions on the surface of the 
endosperm. (Plate XCI. fig. 13 and 14, fig. 15 and 17; Plate XCII. 
fig. 11 
As these ovules perforate the coat, and are situated between it and 
the enlarged part of the nucleus, it seems that they cannot be regarded 
as belonging to the coat. Heinzel (Diss. de Macrozamia) states that 
the vascular network is included between two membranes; but this 
view does not seem altogether accurate, since these cellular layers 
* The analogue of these vascular bundles may be "pes at the base of the 
nucleus in W'elwitschia (Hooker * On Welwitschia,' p. 33, tab. 9, fig. 11 and iz 
Trans. Linn. Soc. vol. xxiv.). "They become afterwards iia scis ah (..¢. 
+A nn. des Se. Nat. iii. p. kein A vascular network which eus to be 
of the same nature has been rved more recently in some Huphorbiacee. A. 
Gris. has studied it cael in m Ricin us. He also adopts the pun de it of 
exponi ^ E chalaza, a pus am astonished that its resemblance to what 
ean ades escaped his e: Just as in it, the nu- 
deni is nile n the són, MR. the endosperm in its enlargement reduces by 
compressio nto the state of a spongy NM. (Ann. des Sc. Nat., sér. xv. 
p-7; pl. ii. fig. 3.) 
