5° BOTANICAL GAZETTE | JANUARY 
cell, while in many composites it is equally certain that the 
tapetum is derived from the wall layers. Warming’s account, 
reproduced by several text-books, refers the tapetum of angio- 
sperms to the wall layers. So far as concerns the spermat- 
ophytes, the only evidence offered as to the origin of the tap- 
etum has been the relation of its cells to those on each side 
of it. If, as appears to be true in this case, the tapetum may 
be derived either from the sporogenous cells or the wall cells; 
or perhaps partly from each, it would become clear that it is 
a physiological rather than a morphological layer. This accords 
with the conditions found in Ranunculus (17) and the sugges” 
tions made in that connection. 
When the microspore mother cells have become free by the 
breaking down of their cell walls, the tapetum sometimes divides, 
ie; tapetal 
IG. 19. Loculus with mother cells, the nuclei of which are in synapsis; tere 
cells dividing ; wall cells breaking down. X 1375. 
F 
x 1G. 20. Same as fig. 79, with no wall cells. Nuclei in early stages of division 
1375- 
forming groups of cells projecting into the mother cell Be 
(figs. AG 20, 27: 22). A large number of cases were observe 
in which cells of the tapetum were projecting deep into the 
cavity of the loculus, a very peculiar case being shown in fig. 23- 
These projecting cells evidently served to nourish the eer 
cells, as the latter were frequently found in close contact wit 
them, as shown in the figure last cited. _ 
