Morphology of Cycadean Sporangia. 429 
by the cutting off of tabular cells from the outer layer or. less 
commonly, by the direct conversion of the outer cells into 
cells of the tapetum. Fig. 13 shows the tapetal layer at this 
stage, and demonstrates, both by the arrangement of the cell- 
walls and by its position with regard to the double layer of 
cells of the wall, that it has arisen from sporogenous cells. 
Usually the tapetum is a single layer, but it may become at 
places two or even three cells thick, by the deeper cells of the 
sporogenous mass assuming the characters of tapetal cells. 
Its cells at first resemble those of the sporogenous tissue so 
closely that care is necessary to distinguish them, but changes 
soon occur in the latter which render their distinction an easy 
matter, and, by comparison with earlier stages, place their 
mode of origin beyond doubt. The sporogenous cells are 
now the mother-cells of the pollen-grains, and their walls 
undergo a change which causes them to stain deeply with 
Bismarck-brown, while the walls of the tabular tapetal cells 
remain thin and are more difficult to distinguish than before. 
Comparison of a large number of sporangia shows that the 
mode of origin of the tapetal layer described above is the 
ordinary one in Stangeria ; but I am not prepared to state 
that this layer is never derived in part from cells adjoining 
the sporogenous tissue, although no sporangium in which this 
was clearly the case has been found. 
Immediately outside the tapetum the double layer of cells 
can in most cases be recognized. Its cells are usually flattened 
and contain elongated nuclei which appear homogeneous and 
highly refractive. Between these cells and the epidermis, the 
cell-walls of which are becoming thickened, the remaining 
layers of the wall are present. Their nuclei as a rule do not 
take on the same highly refractive appearance as those of the 
double layer of cells. 
Some of the spore-mother-cells do not develop further, but 
undergo a change closely resembling that described by Treub 1 
in Zamia , both nucleus and cytoplasm becoming highly 
1 Recherch.es sur les Cycadees : Ann. Jard. Buitenzorg, Vol. ii, p. 36, PI. Ill, 
1, 4> 5- 
G g 2 
