1908] BROWN—PEPEROMIA 455 
pidula (JOHNSON ’07), where fourteen out of sixteen nuclei fuse to 
form the endosperm nucleus. 
Before applying the theory of PorscH it may be well to consider 
the theory itself in its relation to the ordinary gymnosperms and 
angiosperms. In Selaginella, Isoetes, and the gymnosperms, all of 
the first divisions of the megaspore are non-cellular, after which there 
is a number of cellular divisions in all species that form archegonia. 
It is in the cellular tissue that the archegonia are formed by cell 
divisions. The first two divisions of the ordinary angiosperm embryo 
sac are generally homologized with the free nuclear divisions of the 
gymnosperm prothallus, and the fact that in the derived sac of P. 
Sintensii, where cell walls are formed at the first two or megaspore- 
forming divisions, there is no sign of even a cell plate at the third or 
prothallial division, indicates that the character of the free cell 
divisions in the angiosperms is quite constant. It does not seem 
probable that nuclei formed in the angiosperm embryo sac by free 
nuclear division can be homologous with the nuclei in the cellular 
archegonia of the gymnosperms, for we would have to explain how 
the archegonia became shifted back from the cellular to the non- 
cellular phase of the prothallus. 
If the above idea is correct, the polar groups in the angiosperm 
embryo sac can have no such phylogenetic significance as is ascribed 
to them by Porscu, but all of the nuclei in the mature sac must be 
homologous or at least differentiated only in the last division. 
From what has been said, it seems possible that in some angio- 
sperms besides Peperomia, in which the embryo sac is not developed 
from one of a row of megaspores, the first four nuclei of the sac are 
megaspore nuclei. It must be remembered, however, that in most 
Cases we have no evidence for this, other than the presence of the 
reducing division in the embryo sac mother cell, and it is a mistake 
to suppose that the same structure may not come about in plants in 
more than one way. Besides this, it is hard to see how four embryo 
sacs can have become merged into one in the large number of cases 
in which a row of megaspores is not formed (CouLTER and CHAMBER- 
LAIN ’03) without disturbing the normal number and position of the 
nuclei, as has been done in Peperomia. 
Some workers have been inclined to regard the presence of the 
