170 BOTANICAL GAZETTE [SEPTEMBER 
Scarcely any trace of jacket cells can be seen in the group of arche- — 
gonia of fig. 16, and they are never more than very feebly organized. 4 
In this respect the condition in Cephalotaxus as described by COKER 4 
(9) may be compared, where jacket cells are sometimes replaced by 4 
ordinary prothallial cells; and in Torreya where CouLTER and LAND a 
(II) note the absence of jacket cells until after fertilization. Their = 
absence in this genus and in Widdringtonia is probably to be corre- 4 
lated with the small size of the archegonia. 7 
If the nuclear phenomena in these partially abortive archegonia — 
are to be taken as representing normal conditions, the central nucleus 4 
divides (fig. 16, sixth archegonium from bottom) to form the egg and 
ventral nuclei (see also the top archegonium). ‘The ventral nucleus a 
seems usually to disappear completely, leaving only a centrally placed 4 
egg nucleus, as shown in four of the archegonia in the figure. q 
The lower of the two proembryos referred to above (fig. 16) closely 4 
resembles the proembryo of Sequoia (Lawson 20), but the upper a 
shows that more cells are formed than in that genus. It is noticeable 4 
that the proembryo practically fills the archegonium in each case. : 3 
The only conifers previously described in which this is the case an q 
Torreya (CoULTER and LAND II) and Sequoia (LAWSON 20). This 
fact is likely in all three cases to be correlated with the small size of 
the archegonium, and is probably of no phylogenetic importance. 
Only one of the several cells of the proembryo forms the suspensor 4 
and one forms the embryo. The others must disintegrate rapidly, ce 
as in the next stage they are no longer recognizable (jig. 17). _ 
Figs. 16, 17, and 21 show stages in the development of the multi: 
nucleate endosperm mentioned above. Although in early stage 
numerous uninucleate cells may be seen, as well as a large numbet 
of binucleate cells, yet the actual origin of the binucleate condition 
has only been indicated by two karyokinetic figures and the remains 
of a spindle between the two nuclei of one cell. Sometimes tO 
nuclei come to lie almost in contact in a cell and have thus often 
suggested that direct division of the nucleus has occu red, but 4 
careful search has failed to confirm this suggestion. It seems Pe 
fectly clear, therefore, that the binucleate (and in some cells — 
nucleate) condition arises by karyokinetic division of the origin# a 
single nucleus. : 
