132 BOTANICAL GAZETTE [AUGUST 
still had a multicellular archesporium, as the more primitive dicoty- 
ledons do yet. 
Megaspores.—There are usually four megaspore nuclei formed. 
However, in a few instances, only one daughter cell divided. In 
many cases the wall separating these megaspore nuclei at the second 
division was seen to be forming (figs. 14, 16, 19) but always it had 
disappeared at maturity (figs. 20, 23). In a few cases there was no 
indication of a wall forming (jigs. 15, 17,20). This omission of all cell 
formation in the second division is the condition found in Cypripedium 
(14). The fact that Calopogon usually has the four megaspore nuclei 
was rather unexpected in this highly specialized group of the mono- 
cotyledons. For the tendency is not only to a row of three, due to the 
failure of one of the daughter nuclei to divide, but to the condition 
known as the mother cell functioning as a megaspore (COULTER and 
CHAMBERLAIN 7, p. 80), which is common in monocotyledons. 
SCHNIEWIND-THIEs (16) has shown that in Lilium the heterotypic 
division takes place within the embryo sac, only ephemeral walls. 
being formed in either of the divisions from the mother cell to the 
megaspore. In Cypripedium it was found (14) that no wall appeared 
at the second division, and that two megaspore nuclei are used in 
forming the embryo sac. CouLTEr has shown (6) that this may be 
a very important point in connection with embryo sacs of more than 
eight nuclei. For if more than one megaspore nucleus enters into 
the organization of the sac, it is evident that more than eight nuclei in 
the sac would be the result of the usual five divisions from mother cell 
to egg found in angiosperms. So that while Cypripedium with its 
four-nucleate sac seems more reduced than Lilium, it really has just 
the same number of divisions from mother cell to egg (14). 
BRowN (4) objects to calling these nuclei in Cypripedium meg 
spore nuclei, yet he proceeds to do it in Peperomia, although there 
is only the additional evidence of ephemeral walls. JOHNSON (11) 
has also shown these four nuclei in the tetrad position in Peperomia. 
The origin of the nuclei is identical in both Cypripedium and Peper 
mia, with the usual megaspore story except in the matter of wall—4 
mother cell passing from synapsis through the heterotypic and homo 
typic divisions. In Cypripedium a permanent wall appears at the 
heterotypic division, but none at all at the homotypic; in Peperomia 
a te eee ae 
A: Sa ee ee 
Ne 
er WAT Re ee ee 
~The s 
