30 BOTANICAL GAZETTE [JANUARY 
stages, sometimes, when the plants were afterward placed under the 
best of growing conditions, only part of the microspores survived the 
change. The surviving ones would then respond with vigor, becom- 
ing sometimes three times the usual size for the normal microspores 
(figs. 26, 28). It was noticed also that when a sporocarp, whose 
oldest sporangia were just forming tetrads, was subjected to a spray 
of cold water for several days, the megaspores would abort, and the 
microspores, which at this time would be just coming into the mother 
cell stage, would not undergo the reduction division while under the 
spray, but would continue to grow until they became much larger 
than the ordinary microspores, assuming at the same time many 
unsymmetrical and unusual shapes (jigs. 15-22). When these were 
placed later in good growing conditions, both reduction divisions 
occurred without the formation of more than mere traces of walls 
(fig. 14, c), but the entire wall of the mother cell became much heavier. 
Repeated attempts at germination under the most favorable conditions 
have so far failed to demonstrate what these spores might become 
ultimately. The failure to germinate seems to find its explanation, 
not in any want of vitality of the cell, but rather in the rigidity and 
continuity of the surrounding wall. As is well known, the heavy 
walls of both megaspores and microspores are interrupted at the 
papilla, which has only a light membranous covering, due to the pres- 
ence of radiating protoplasmic strands (figured in Sacus 22 for Pilu- 
laria globulifera) that hold the four spores in close proximity until the 
wall is well developed over the free surfaces. The tapetal layer 
always incloses the tetrad, but forms a perinium only where it 
comes in contact with the cell membrane of the functioning spore. — 
It is prevented from touching this wall at the papilla by the other 
three members of the tetrad (figs. 6-10), yet the perinium will not form 
around an aborting spore. The microspores separate earlier than the 
megaspores, hence the papilla is less evident in them than in the 
megaspores where the abortive members of the tetrad may not be cast 
off at all, often remaining until germination has begun. When the 
tetrad divisions are delayed, there is not a corresponding delay in 
wall formation, which goes forward at such a rate that when the 
divisions occur the wall is too strong to disrupt, though in the second 
division evanescent -cell plates are sometimes evident (fig. 14, ©)- 
