196 BOTANICAL GAZETTE [MARCH 
higher magnification (fig. 11) it was seen that plerome, dermatogen, 
and calyptrogen all arose from a common group of meristematic 
cells, and that these regions were not yet clearly defined. The 
periblem was more easily picked out by the larger size of its cells 
and their relatively less cytoplasmic contents. At this stage of 
the embryo it was evident that the cotyledons were to be net- 
veined, as might be expected. The rapid growth of the embryo 
continued until at maturity it approached 1 dm. in length. Fig. 
12 is a diagram showing the size of the seed and the position of the 
embryo within it. 
The seed 
The tremendous rate of growth of the seed made it seem worth 
while to look somewhat into the method of food supply. In 
the development of the ovule there was a very early differen- 
tiation of tissue in the chalaza. On March 25, when the embryo 
sac was in the two-nucleate stage, there was seen extending across 
the chalaza, from the base of the inner integument, a rather narrow 
layer of cells, which, having more densely granular cytoplasmic 
content, took a deeper stain than the other cells of the region 
(fig. 14). Immediately outside of this layer of cells, which for 
convenience we will speak of as the nutritive layer, there was a 
layer of cells elongated transversely to the axis of the ovule, which 
was continuous with the vascular bundle of the funiculus (figs. 13 
and 14). 
The cells of the nutritive layer divided repeatedly, so that by 
the time the development of the embryo sac was completed there 
was quite a mass of them showing very clearly, even under low 
magnification, on account of their deeper stain. At about the time 
that the embryo sac was ready for fertilization, the appearance of 
these cells changed, and under low magnification they could now 
be picked out by their lighter stain. Higher magnification showed 
them to have almost no granular cytoplasmic content (fig. 15). 
Very shortly after this it was seen that there was being deposited 
in the cells some reserve material, probably tannin, which did not 
stain with the iron-haemotoxylin combination. This deposition 
began at. the periphery of the cell and proceeded toward the 
