914 Gates.—Pollen Formation in Oenothera gig as. 
are therefore (i) the expansion of the nucleus, with frequently the partial 
disappearance of the nuclear membrane for a time, and (2) the rearrange¬ 
ment, without growth, of the structural content of the nucleus, from the 
reticulum to the spireme condition. This rearrangement has not begun in 
PL LXVII, Figs. 2-4, is partly completed in Fig. 5, and is quite completed 
in Figs. 6 and 7. It is not uncommon for threads to remain attached to the 
nuclear membrane, as in Figs. 4 and 5. PL I, Figs. 12 and 13 of a former paper 
(Gates, ’ 08 ) show the same thing, and are also instructive (when compared 
with PL I, Figs. 6-11) in showing that expansion of the nucleus, rather than 
contraction of its contents, is taking place. In ordinary nuclear growth, 
such as occurs between successive mitoses, the reticulum also grows so that 
it continues to fill the nuclear cavity. But in the growth of the synaptic 
nucleus, the reticulum does not continue to grow, at least not to a corre¬ 
sponding extent, and probably not at all, and the result is that the reticulum 
soon occupies only a small part of the cavity of the synaptic nucleus. 
Regarding the presynaptic nucleus, it will be seen from Figs. 1-4 that 
darker-staining portions of the threadwork frequently occur, especially near 
the periphery of the nucleus, but they show no features of regularity in 
size, shape, or number. 
One feature which has been of much value in the study, particularly 
of the synaptic stages in O. gigas , is the occurrence of slightly different 
stages or conditions of synapsis in different anthers of the same flower, 
while all the mother-cells of each anther are in about the same condition. 
•4 
Thus in one flower the mother-cells of one anther were presynaptic, except 
a few which showed the beginning of the synaptic nuclear expansion. 
In another anther of this flower the mothei'Tcells were all presynaptic and 
with numerous dark bodies lining the nuclear membrane as in Fig. 1, while 
other anthers showed the same condition except that the dark bodies were 
wholly absent. This probably indicated in this case a different physio¬ 
logical condition of the pollen mother-cells in these anthers rather than 
two successive stages during which these bodies made their appearance. 
Figs. 6 and 7 show the typical synaptic condition, in which a rather 
dense aggregation of threads, usually with darker areas or bodies in its 
meshes, appears on one side of the nucleus. The threadwork appears to 
be more or less continuous and frequently shows a typical ‘ chromomeric ’ 
condition such as figured by Mottier for Lilium (Mottier, ’ 09 , Figs. 4 and 5). 
The threads are extremely delicate. This chromomeric condition of the 
threads will be referred to again later. The only difference between Figs. 6 
and 7 is in the condition of the nucleolus. In the anther from which Fig. 6 
was taken, the nucleolus was, in all the mother-cells, spherical and free 
in the nuclear cavity. Fig. 7 was drawn from a different flower, and in 
this anther the nucleoli were always flattened against the nuclear mem¬ 
brane as in the figure. 
