1912] CHAMBERLAIN—CERATOZAMIA 15 
A strong reason for believing that fertilization is sometimes 
effected by a ventral canal nucleus is found in a paper by VAN 
TIEGHEM (13) published in 1870. He secured four seedlings, the 
result of fertilization of the ovules of Ceratozamia longifolia by ‘the 
pollen of C. mexicana, which had been preserved for three years.” 
VAN TIEGHEM speaks of these seedlings as hybrids, but I do not 
believe the pollen of Ceratozamia will live for three years. Pollen 
of C. mexicana, shed April 22, 1909, in cultures started on that 
date and also a week later, germinated immediately, but in cultures 
made a month later from the same collection of pollen, the grains 
simply became spherical, but would not germinate. In January 
1911, I pollinated two cones of Zamia Ottonis with some of the same 
pollen, at about the same time pollinating another cone of Z. 
Ottonis with pollen of Z. floridana. I have not yet examined the 
cones, except to note that they are in fine condition, preferring to 
wait for the later embryo and seedling stages, if there should be 
any. At the time of this pollination I again made cultures of the 
old pollen of Ceratozamia, but not a single pollen grain germinated, 
and recently I repeated the attempt, but no germination occurred. 
The old pollen is doubtless dead, and VAN TIEGHEM’s seedlings were 
parthenogenetic or were the result of fertilization by a ventral 
canal nucleus. I might mention here that I have preparations of 
Encephalartos from a greenhouse specimen where there had been 
no possibility of pollination, in which the ventral canal nucleus 
has become greatly enlarged and has moved toward the egg nucleus. 
I should not be surprised to find occasional seedlings from cycads 
in greenhouses where there has been no pollination. 
The archegonial chamber is conspicuous before the pollen tubes 
are half way through the nucleus, and during the early stages 
in its development it contains a fluid, doubtless secreted by the 
gametophyte, for the megaspore membrane is torn loose from the 
bottom of the chamber. At the time of fertilization the chamber, 
although moist, does not contain liquid. 
The megaspore membrane is thin, only 2.5-3 # in thickness. 
It has about the same structure as in Dioon edule (g), a compara- 
tively homogenous inner layer beset with an outer layer of irregular 
club-shaped bodies. These bodies, which in some gymnosperms 
