1909] PACE—GAMETOPHYTES OF CALOPOGON 133 
ephemeral walls appear at both divisions. This gives apparently 
four megaspore nuclei forming the sac in Peperomia, while only two 
enter into its organization in Cypripedium. It might be better to 
select another name for these nuclei, but it would always involve the 
explanation of their similarity to megaspore nuclei, with apparently 
no gain in clearness. For our notion of spores, extending back 
through pteridophytes and bryophytes, is that when a mother cell 
divides there are four spores produced, the spore being the cell formed 
by the heterotypic and homotypic divisions. The matter of walls 
would not seem to be so important as the behavior of the nuclear mate- 
vials; for all sorts of variation in the kinds and shapes of these are 
found, these differences being related apparently to the environment 
of the spore. The egg is of course a descendant of only one mega- 
spore, however many may enter into the construction of the sac. And 
it seems to involve less change from what have seemed to be the essen- 
tials to regard these as megaspore nuclei than to conceive of a mother 
cell functioning directly as a megaspore and yet proceeding to the 
very same heterotypic and homotypic divisions. This is more evident 
still when Calopogon is taken into consideration, as it seems to present 
a stage that might be considered intermediate between Cypripedium 
and the usual one. For here there are certainly two megaspore nuclei 
in the sac, one finally disintegrating, although some evidence of its 
division is present in two cases examined (figs. 24, 25) 
This lack of walls is not very rare. Lxoyp (12) reports them as 
usually absent in Rubiaceae, although the tetrad is formed. Here 
only one megaspore nucleus functions. The same condition is 
reported by CANNON (5) in Avena, and by SMITH (17) in Eichhornia. 
WIEGAND (19) finds the second divisions unaccompanied by walls 
in Potamogeton foliosus, and HOLFERTY (10) finds the wall omitted 
in the division of the micropylar daughter cell in Potamogeton natans. 
BILLINGs (2) reports one case of megaspores without separating walls 
in Tillandsia usneoides. And the many cases where the mother cell 
is said to function as a megaspore seem to represent the same con- 
dition, except that the non-functioning megaspores not only do not 
disintegrate but even divide and thus contribute to the contents of 
the embryo sac. . 
In Calopogon only one megaspore nucleus is used in forming 
