86 BOTANICAL GAZETTE [FEBRUARY 
My observations on the later stages of tapetal development agree 
in the main with those of BEER, and so will not be taken up here in 
detail. In some cases appearances suggest simultaneous division 
of a nucleus into several smaller nuclei, the latter often forming 
closely aggregated masses, as figured by Beer. At a still later stage, 
when the persistent pollen grains have begun to assume their adult 
shape, fusion of these masses of nuclei into one or two apparently 
occurs, and the tapetal cells return to a binucleate or even uninucleate 
condition. Both nuclei and cytoplasm now, however, show a very 
different appearance from the early uninucleate and _ binucleate 
condition, the nuclei being at this time much larger and having 
extremely little chromatic content. The cytoplasm, which is loosely 
reticular rather than granular at this time, frequently contains more 
or less indefinite and irregular chromatic masses. BEER describes 
the origin of these masses and compares them with the chromidial 
structures found by MevEs (19) in the tapetal cells of Nymphaea 
alba, and with similar bodies found by GoLtpscumipt (8) in the cyto- 
plasm of certain animal tissues. These tapetal cells of the genus 
Oenothera offer excellent material for a further study of amitosis. 
It is interesting to note that the structure of the tapetal cells in the 
mutant O. lata is apparently identical with that of the well-known 
members of the genus, such as O. biennis. During these later nuclear 
divisions of the tapetal cells in O. Jata, radial elongation of the cells 
may take place, but rarely enough to nearly fill the cavity of the 
loculus left by the degeneration (in such cases) of the mother cells. 
Transverse divisions of the tapetal cells are comparatively rare. 
In the great majority of cases the tapetal cells begin to break down 
before disintegration in the mother cells sets in. As already stated, 
this disintegration may appear as early as the resting stage of the 
mother cells, where scattered cells begin to lose their turgor and 
degenerate. In later stages of development a great variety of condi- 
tions of disintegration of this tissue may be found. The tapetum 
may break down irregularly in different parts of a given loculus, oF 
only on one side, or in patches. Different loculi of the same anther 
and different anthers of the same flower show wide variations in this 
matter. 
Pout (22) states, followed by DeVries (/.c. 12292), that in O- 
