16 ON THE SEXUAL ORGANS OF THE CYCADACEÆ. 
tissue, a true endosperm. lt is now no longer possible to dis- 
tinguish the true wall of the embryo-sac; the space which it oc- 
cupies is bounded by the dense, smooth, and shining surface of the 
dilated nuclear tissue, to which perhaps the débris of the original 
amniotic membrane adhere.* It is this which has previously led me 
into error, in regarding the embryo-sac as a free cavity in the albumen 
which I regarded as derived from the nucleus, so that I could not 
recognize the morphological meaning of the true nucleus, although I 
had observed figured and described the different stages of development. 
(Ann. des Se. Nat. l.c. p. 199; Monogr. plate i. fig. R, s.) In nu- 
merous unfertilized seeds the endosperm occurred just as in those that 
were fertilized; to my great astonishment, however, I have observed 
several isolated cases where it was wanting, though the cavity for its 
reception existed. 
I know nothing of the changes which take place in the upper part 
of the embryo-sac at the first appearance of the second endospermie 
formation, nor of the way in which the corpuscles of Brown originate. 
I only know the period at which the corpuscles already exist both in 
the unfertilized ovules and in the ripe seeds containing an embryo. The 
vault or upper part of the embryo-sac is very persistent, and becomes 
a soft pulpy, often yellowish membrane, to which adheres above, the in- 
ternal tissue of the conus nuclei, below, the tops of the corpuscles. 
Plate XCI. fig. 12 6, where the corpuscles do not yet exist; Plate 
XCII. fig. 9 a, the remains of the cone with the amniotic membrane ad- 
herent, below which are the corpuscles; fig. 10, the part removed with 
the corpuscles, which are attached to it; fig. 1, the embryo-sac with 
the cone removed and viewed from above, with the six areole or places 
where the interior canals of the cone terminate, and to which are 
attached on the opposite sides by their opercular rosettes the tops of 
the corpuscles ;+ fig. 2, tops of the corpuscles situated at this level; 
fig. 8, corpuscles whose tops exhibit regularly arranged fragments of 
tissuef (opercular rosettes (?) or shreds torn from the part where ad- 
* Hooker has observed the same thing in * Welwitschia, Z. c. p. 3 
t “ Juniore state membrana tenuis n fere gelatinosa saccos ob it 
et eorum — adheeret, — obscuris vel areolis parumper elevatis ex 
ins quæ eum saccorum subjacentiu ge correspondent,” etc. ^re 
des Se. Peur i e. 
ragmen enta regulari s probabiliter canalium conductorum coni 
acral pi eet oman Bie attached to plates.—W. T. D. 
CP oC Oc OPUS ee 
