Beer, Development of tlie pollen grain and anther of some Onagraceae. 303 
fragmentation. AYben tlie special-mother-cells bave been establi- 
slied mitotic divisions are rarely met witb whilst fragmenting 
nuclei occur on eveiy side. Most of tlie tapetal nuclei now 
eontain a single large nucleolus and a very deeply staining 
nuclear wall, besides tliis only a very little finely granulär 
cbromatic material can be seen lying near or upon tlie nnclear 
wall (see n Fig. 44). 
In antliers in wliicli tlie first pollen-wall is jnst making its 
appearance. I liave severgl tinies seen tbe tapetal nuclei in tbe 
propliases of mitosis butlbave never, at tbis period. succeeded 
in finding tlie later stages of division and I believe tbat mitosis 
is no longer completed by tbe nuclei. Xuclei wliicli liave every 
appearance of undergoing fragmentation are, liowever, very 
abundant botb at tbis and at later stages of development 
(Fig. 32. 33. 34. 40. 42) Strasburger 1 ) and later writers, in 
describing tbe tapetum of other plants, bave found mitosis to 
be tbe onlv mode of nuclear division and tbev bebeve tbe con- 
kJ 
stricted nuclei wbicb occur in tbe cells to represent fusion and 
not fragmentation of tbe nuclei. 
In Oenothera it is impossible to imagine tbat karyokinesis 
can be tbe only mode of nuclear multipbcation. 
In tbe first place mitotic divisions are never very frequent 
and it is difficult to account for tbe presence of six or seven 
nuclei in a young tapetal cell tbrougli tben* agency alone, 
Moreover, mitotic figures cannot be found in tbe tapetum 
of Oenothera after tbe appearance of tbe first pollen wall so tbat 
if tbis is tbe onlv mode of division and tbe constricted nuclei. 
c • 
wbicb are common botb at tbis and at subsequent stages, really 
represent fusions it is impossible to see wbence tbe constant 
supply of nuclei comes for tbese repeated fusions and wbicb 
leaves tbe older tapetal cell witli two or tbree nuclei to tbe last. 
Tbe way in wbicb tbese constricted nuclei often bang togetber 
by a narrow neck also favours tbe view tbat tbey are separating 
from one anotlier and are not uniting (see especially Fig. 34). 
Tbe great disparity wbicb after exists in tbe sizes of tbe 
nuclei of a cell is also wbat one would expect witb direct ratber 
tlian witb indirect division (compare sizes of tbe two nuclei in 
Fig. 37). 
For all of tbese reasons I consider tbat most of tbese con¬ 
stricted nuclei represent a fragmentation and not a fusion of nuclei. 
Every constricted nucleus does not. bowever. necessarily 
imply nuclear multipbcation. 
. Tbere is no doubt tbat tbe tapetal nuclei alter tbeir sbape 
and often beeome very irregulär in outline witbout tbis leading 
to a division of tbe nucleus or representing a fusion (see Fig. 36. 37). 
Tbese cbanges in sbape are evidently signs of tbe occurrence of 
an active metabolism in tbe cell and may be compared to tbe 
similar pbenomena wbicb bave been described in tbe secreting 
cells of may an im als. 
kJ 
fi Strasburger, E. ..TeibmgsVorgang d. Zellkerne etc.(Arcli. f. Mikro. 
Anat. Bd. 21. 1832. pp. 574—575.) 
