3^4 
Gates and Rees . — A Cytological Study of 
tapetal cells remain in the binucleate condition. This is particularly true 
of cells near the end of a loculus, but one not infrequently finds the tapetal 
cells quadrinucleate on one side of a loculus and binucleate on the other 
(Fig. 4). Binucleate cells are nearly always shorter and broader than the 
quadrinucleate, and in the latter the nuclei are usually, though not invariably, 
in a single row. As Figs. 1-4 and 73-8° show, there is a great variety 
of sizes and shapes of cells. Figs. 79 and 85 show two extremes. The 
long narrow cells are usually towards the centre of the loculus. In these 
binucleate cells particularly the appearance of the nuclei is extraordinarily 
like that of some of the synaptic stages in the pollen mother-cell nuclei. 
Figs. 75-81 show the various appearances of the tapetal cells at about 
the time of diakinesis. It will be seen that the cells are binucleate or 
quadrinucleate, with different arrangements of the nuclei and various 
appearances of the nuclear content. The latter may be in the form of j 
a coarse reticulum with karyosomes at the nodes (Fig. 75), or a finer j 
network with sharply marked chromatin bodies which appear to be more 
independent of the network (Figs. 77, 78). These bodies are not chromo- 
somes ; they are usually rather angular in shape, and are variable in size and 
number, though often about twelve or thirteen are present. Sometimes 
they have the appearance of being longitudinally split, and they give the 
nuclei a hyperchromatic appearance. Winge ( 1917 ) has observed and 
figured (Figs. 5 and 6, p. 1 50) chromatic bodies having a very similar 
appearance in the cytoplasm of the young spores of Entorrhiza Raunkiaeriana , 
a member of the Plasmodiophoraceae, at a time when the nucleus is small, 
amoeboid in shape, and non-chromatic. 
Another type of tapetal cell exists in Lactuca (Fig. 79), small in size, 
with small nuclei, having a fine network and only one to four large angular 
chromatin bodies. 
It is difficult to say what proportion of the tapetal cells remain 
binucleate, but probably it is a considerable number. At a later stage of i 
development the nuclei of the tapetal cells frequently fuse with one another, 
sometimes forming one large nucleus. Or some of the nuclei may break 
down (Figs. 81, 85). Finally, at the stage of pollen tetrads the nuclear ; 
membranes become very faint (Figs. 83, 86) and disappear, the nuclei more- 
or less completely disintegrating and forming irregular or globular fragments 
in the cytoplasm. At the same time or a little later the cell-walls, which 
have become very thin, disappear, and the cytoplasm, containing nuclei in 
various stages of disintegration, flows in amongst the pollen grains and forms 
a tapetal plasmodium (Figs. 72, 84). 
A tapetal plasmodium was formerly supposed to be characteristic of 
sporogenesis in Pteridophytes, but not of the Angiosperms. Juel ( 1915 ), 
from a comparative study of the later stages of tapetal development in 
Angiosperms, finds two extreme types, (1) in which the cells form a true 
