— A -——Á — ——————mÁÓGC T HE 
bid 
ON THE SEXUAL ORGANS OF THE CYCADACER, 13 
cannot be regarded as actual membranes existing as such originally. 
The outer one is intimately blended with the woody portion of the 
coat, and seems to form part of it ; the inner is nothing more than the 
compressed remains of the nucleus, mentioned above, combined with 
what has been termed the epithelium of the nucleus. As soon as the 
embryo-sac is for the second time filled with cells for the formation, 
properly so called, of the albumen, and its cavity has attained in con- 
sequence a considerable enlargement, the tissues of the nucleus are 
pushed out and compressed in all directions, especially laterally, and 
transformed into a kind of membrane. This compression, in most 
of the species, is least towards the base; and in many of them, such as 
C. Rumphii and C. spherica, a thick brown layer remains, on which the 
broad base of the albumen rests. 
In Macrozamia, Dion, Encephalartos, and many species of Zamia, 
on the contrary, this layer is entirely converted even at the base of 
fruit into a sort of membrane (Plate XCI. fig. 15 and 17 c).* The de- 
gree of conversion is very variable in the same genus, and even in the 
same species, especially when the fruit has not been fertilized. In C. 
angulata, for example (Plate XCI. fig. 14 c), the layer is completely 
wasted by compression ; in C. revoluta the enlargement which the cavity 
undergoes to make room for the endosperm takes place unequally, so 
that the tissues of the nucleus may be more or less preserved or effaced 
at the base. Modifications of the entire fruit result from this, and the 
fruit becomes ovoid, elliptical, or obovoid (Plate XCI. fig. 2-6). Gene- 
rally speaking, this membrane (the remains of the nucleus which in 
its earliest stage is intimately united with the internal layer of the 
coat, but which gradually separates as this layer becomes woody) is so 
pressed by the dilation of the endosperm against the inner layer and 
the vascular network, that it can only be detached by maceration and 
boiling. In its earliest state, and in a living condition, it is often yel- 
lowish in colour. Later on, if still existing in sufficient quantity, it 
is brown when dry, and exhibits between the parenchymatous cells 
others of an elongated form. 
When the nuclear tissue has been removed in ripe seeds, the vas- 
Gottsche (l. c. p. 384) states that in Encephalartos a thin white membrane 
covers the vascular layer. Possibly there may have been in this case some of 
the cells of the first endosperm in addition. M Ne 
They suggest the spicular cells which Hooker has found in certain tissues 
of Welwitschia. 
