(338) 
chalaza and at the micropyle. In the latter place it has 
practically filled the micropyle by a proliferation of cells on 
its inner surface. At about the time of fertilization the outer 
epidermis of the outer integument begins the tangential divi- 
sions which, as Harz*® has described for other Cucurbitaceae, 
give rise to the characteristic histological features which are 
well known in the ripe seeds of this family. Subsequently 
to this period, growth in the ovule is principally in the outer 
integument, the nucellus enlarging mainly by the expansion 
of cells already formed. 
When the first periblem layer in the very young ovule has 
assumed some degree of regularity, a cell situated imme- 
diately below the epidermis at the apex of the young nucellus 
increases in size and becomes clearly differentiated from its 
neighbors (fig. 9). The first division of this cell is tangen- 
tial and a single tapetal cell is cut off. The inner daughter- 
cell, the megaspore-mother-cell, remains undivided and enters 
upon a period of growth. The tapetal cell in the meantime 
divides by walls both perpendicular and tangential to the epi- 
dermis, so that the spore-mother-cell soon becomes covered 
by a row of cells to the number of ten or twelve. At this 
time divisions are going on in the cells to the right and left of 
this row so that several rows of cells are formed and the ap- 
pearance is such as is shown in figure 2. 
Both first and second divisions of the spore-mother-cell 
were observed. Four spores are formed, but only the lower- 
most one is functional. The four megaspores were some- 
times observed soon after the division when they were all in 
good condition and again when the three uppermost ones 
were beginning to disintegrate (fg. 76). In figure 76 the 
two upper cells appear a little more advanced in the process 
of disintegration than does the third cell. Usually light 
walls are formed between the potential megaspores, but 
sometimes none are formed in the first division. It might be 
inquired as to whether this condition does not represent the 
second division of the megaspore rather than the second di- 
vision of the megaspore-mother-cell. It may be answered 
