172 BOTANICAL GAZETTE [MARCH 
could not be determined, since there was no opportunity to observe 
living material. 
The wall first formed about the spore is the exine. Very early, 
before any thickening of this wall has taken place, the color changes 
from purple to bright red, and the spores increase in size so that the 
tapetal protoplasm is now.close against them. As the exine thickens, 
a difference in staining becomes apparent, the inner part being red, 
- the outer yellow, but with no clear line of demarkation between them 
(fig. 2, ex, r, andy). No chemical tests could be applied to determine 
the nature of the substances taking these colors, but from what THOM- 
SON (16) says of the colors of megaspore membranes of gymnosperms 
when stained with safranin, and from what he says of the course of 
development in these membranes, it seems probable that the red indi- 
cated the presence of suberin, and the yellow of pectin. 
By this time the intine (fig. 2, 7) has been laid down just within 
the exine. In the sporangia showing the stage just preceding, the 
spores were so collapsed that it was impossible to determine whether 
or not the intine was formed before the exine had differentiated into 
red and yellow regions. 4 
At a stage shown in fig. 2 there are here and there in the cytoplasm 
groups of reddish granules (fig. 2, g). These are first recognized as 
very small, deeply staining bodies at the intersections of the proto- 
plasmic network. They increase in size and the protoplasm, either 
by the breaking down or drawing together of some of the connecting 
strands, assumes the appearance of a network of coarser mesh, in 
which, both at the intersections and along the strands, are the now 
-_redly staining granules. The nuclei (fig. 2, 2) show the beginning of 
degeneration in the irregular clumping of the chromatin, and later 
in the disappearance of the membrane, and it is very likely that the 
nuclear substance contributes largely to the formation of the granules, 
which take on a deeper and deeper stain as the nuclei become unrecog- 
nizable. 
In the older sporangia the granules are larger and larger, of fairly 
regular shape, as if from the rounding up of viscid matter, but of vary 
ing size (fig. 3). Some of them are close together, and their position 
suggests that the larger ones may have come from the running together 
of two or more of the smaller ones. Those in proximity to the spore 
