WEATHERWAX: GAMETOGENESIS IN ZEA MAys 87 
half starchy and one half sweet. Webber (25, pp. 34-37) found 
an explanation for such irregularities.in the assumption that the 
sperm might sometimes fail to fuse with the polar nuclei, and thus 
be left to divide independently, or might fuse with only one of 
them, the other dividing independently. In either case there 
would be two, possibly genetically different, sources of endosperm 
formation, which might account for the two colors of aleurone. 
It has since been found that aleurone mosaics are amenable to 
Mendelian principles, but for the divided endosperms East and 
Hayes (7, pp. 34-35) consider Correns’s explanation well-founded. 
This assumption requires that either the sperm or one of the 
polar nuclei, both of which have been considered gametic in nature, 
develop without uniting with another gamete, which would be a 
very unusual behavior. If such an explanation of the occurrence 
is necessary, it seems more reasonable to suppose that the sperm 
united with the polar nucleus coming from the micropylar end of 
the embryo sac, since it is the sister of the egg and necessarily 
much like a gamete, and that the antipodal one of the polar nuclei, 
with its possibly irregular chromatin organization, divided inde- 
pendently, as its near relatives, the true antipodals, regularly do 
in maize and other grasses. Yet this would take from one polar 
nucleus much of its gamete-like nature and might seriously inter- 
fere with its ability to transmit hereditary characters in the ortho- 
dox fashion necessary for some of the other phases of our theory. 
In fact, any assumption of this kind will lead to the conclusion 
that something, elsewhere considered gametophyte, here contri- 
butes directly to endosperm formation and shows endosperm 
(xeniophyte) characteristics. : 
The responsibility for these rare occurrences of bilaterally 
differentiated seeds may be placed upon still another hypothesis 
which involves valid cytology and is not in conflict with Mendelism. 
The division of the primary endosperm nucleus may be heterotypic 
in a way. In fact, it is well known that the subsequent divisions 
often show fewer than 3x chromosomes; some chromosomes must, 
therefore, at some time, pass bodily to the poles of the spindle 
without themselves dividing, and, if one of these carries the gene 
for starchiness, reduction necessitates a sort of Mendelian segre- 
gation. 
