9i3 
Gates.—Pollen Formation in Oenothera gigas . 
completely, leaving the nucleus in open communication with the cytoplasm. 
This is certainly true in fixed preparations, and is probably also true in the 
living cell. Fig. 3 illustrates a not uncommon condition at this time, in which 
there appears to have been a sudden forcible expansion or rupture of the 
nucleus on one side, following perhaps such a condition as shown in Fig. 2. 
This gives a clear area in the cell containing nothing but karyolymph or 
cell-sap. The main body of the nucleus is in such a case unaltered, and 
the amount of increase in the karyolymph can be estimated from the size 
of the pouch-like lateral enlargement of the nucleus. That this lateral 
expansion is relatively sudden is shown by the fact that the nuclear 
reticulum has not begun to undergo the changes which will transform it 
into the typical spireme condition of synapsis. Surrounding the clear area 
in such cases as Fig. 3, there may be a definite membrane, or it may be 
quite indistinct and indefinite. 
Another condition very commonly found in mother-cells before the 
completion of synapsis is that shown by Fig. 4. Here the nuclear 
membrane evidently first began to enlarge, as in Fig. 2, on the left side 
of the nucleus. Then the membrane was ruptured, probably by the 
increasing pressure from within, which was causing its expansion. The 
result is an indefinite area on the right side of the original nucleus, which 
contains much karyolymph and little cytoplasm, but with no definite 
limiting membrane. The expansion of the nucleus in Fig. 5 probably 
took place in much the same way. There is thus a period during which, 
on one side or portion of the nucleus, the membrane partly or completely 
disappears. Later a definite membrane is again formed. It thus appears 
that the nuclear expansion, which is one of the main features of the period 
known as synapsis, may be gradual or relatively sudden, and that there is 
very commonly at this time a disappearance of the nuclear membrane on 
one side of the nucleus, leaving the latter and its contents in open connexion 
with the cytoplasm. This condition is brought about either by a rupture 
or by the solution of the nuclear membrane on one side of the nucleus, and 
there follows a period during which it is impossible to define accurately the 
limits of the nucleus. A portion of the membrane seems always, however, 
to remain intact. A little later the nuclear membrane is once more complete 
and perfect, the nucleus spherical, and we have the typical synapsis condition, 
as represented with a somewhat lower magnification in Figs. 6 and 7. 
During the period when the limits of the nucleus are so poorly defined, 
the cytoplasm also shows a f ragged ’ condition rather than a well-defined 
reticulum. This will be seen in comparing Figs. 3 and 5 with Figs. 1, 6 , 
and 7. It might be concluded that these differences were due to differences 
in the treatment. To obviate the possibility of this conclusion, Figs. 1-7 
were all drawn from the same preparation. 
The two most characteristic features of this first period of synapsis 
