106 BOTANICAL GAZETTE [FEBRUARY 
though not proportionally so great (plate fig. 9). Once this 
increase in size has been effected in the mother cells and their walls, 
the cell contents appear to round up and shrink away from the 
walls more or less; just how much I am unable to say owing to the 
difficulty of determining whether part or all of the shrinkage may 
not be due to the difficulties of securing satisfactory fixation. The 
walls themselves have a tendency to persist until after the young 
spores have been formed, but apparently they change their chemical 
condition and become more or less mucilaginous in character. 
Sometimes no walls at all are visible, but the spaces among the 
mother cells are filled with a thin sort of mucilage. The greater the 
amount of this mucilage present, the poorer the fixation. 
Owing to the difficulties already mentioned, I have but little 
trustworthy information concerning the course of events during the 
reduction divisions. Stages in the formation of the presynaptic 
spireme, synapsis, and diakinesis were observed. At the metaphase 
of the heterotypic division there are 8 bivalent chromosomes, but 
the material did not permit one to follow the method of their 
formation. The spindle fibers are attached to their apices (plate 
fig. 11), and they are drawn apart as short stubby masses and 
collect at the pole in a close mass. They appear at the spindle of 
the homeotypic division as much longer rods curved into more OF 
less U-shaped chromosomes. No walls are formed, apparently, 
until after the spore nuclei have passed into the resting condition, 
when a system of fibers is present between the nuclei, on which the 
plasmatic membranes separating the young spores arise (plate 
figs. 13 and 14). 
The young spores now form walls around themselves, entirely 
within the old mother cell wall if it is still present. The spores 
then begin a slow enlargement and thickening of the walls. At 
first the growth consists merely in enlargement of the wall without 
any apparent increase of cytoplasm or nucleus. There is either on¢ 
very large vacuole almost entirely filling the spore or many sma ler 
ones separated by only the most tenuous cytoplasmic walls. 
this time no inclusions of starch or other food materials are visible. 
After the spore wall has reached its mature size, but not final 
thickness, the cytoplasm becomes more abundant and the nucleus 
enlarges (plate figs. 16 and 17). In the large nucleus there is a larg¢ 
eae i ee Ne 
tes 
Bra) eae tees 
# 
Be sph: tk Seine F ene ea 
se ot eae 
Aji Seda ON Oe 
Woe ro ee aay ee es 
