422 
AMERICAN JOURNAL OF BOTANY 
[Vol. 8 
Wx wx are concerned, just as the hypotheses of independent division of 
polar nuclei suggested by Webber (1900) do not fit the available facts where 
other than these aleurone and endosperm factors are concerned, was shown 
earlier in this paper. The hypothesis of vegetative segregation (East and 
Hayes, 191 1) is not sufhciently specific with respect to the mechanism of 
such supposed segregation to make it possible to apply crucial tests. More- 
over, several cases of somatic variations often referred to as cases of vegeta- 
tive segregation are quite as likely due to somatic mutation. There remains 
only the hypothesis of aberrant chromosome behavior (non-disjunction?) 
which is in accord with practically all the reported cases of aberrant endo- 
sperm development. It was shown earlier how that hypothesis fits the 
cases involving the linked genes C c (or / i) with Wx wx. That this 
hypothesis is not in disagreement with the cases where other endosperm and 
aleurone factors are concerned follows from the fact — determined by 
ordinary breeding tests — that these other factors are not genetically linked 
and that, therefore, they presumably have their loci in non-homologous 
chromosomes. Evidence of non-linkage for A a with Wx wx was presented 
by Bregger (1918), for R r with Wx wx by Kempton (1919), and for Su su 
with C c, R r, and Pr pr by Eyster (1921); and there is indirect evidence 
for Pr pr and Wx wx in Hutchison's data which show Wx wx to be linked 
with Sh sh and the latter to be independent of Pr pr. If none of these 
combinations of genes lies in the same chromosome, it is obvious that a 
non-disjunction of one chromosome could not affect more than one member 
of the combination, just as it is that both members of any combination 
lying in one chromosome must be affected by a single non-disjunction of 
that chromosome. 
While the writer feels that the genetic evidence in favor of the hypothesis 
of non-disjunction, or at least of some aberrant chromosome behavior giving 
a similar result, as the cause of most cases of the kind of aberrant endosperm 
here discussed is as convincing as such evidence can well be, it is realized 
that direct proof must come, if at all, from cytological studies. Whether 
it will ever be possible to detect non-disjunction cytologically in the endo- 
sperm of maize, granting that it occurs, cannot be said. The small size 
of maize chromosomes and their large number, 30 in the triploid nucleus, 
increase the difficulty of the undertaking. Moreover, the rarity of the 
phenomenon lessens the chance of a successful outcome. On this latter 
point, however, there is this to be said: non-disjunction is doubtless a 
much more common occurrence than are its visible manifestations. There 
is no reason to suppose, for instance, that, in material of the genotype 
c c C wx wx Wx, such as that recorded in table i , the chromosome carrying 
C and Wx is more often concerned in non-disjunction than are the other 
two homologous chromosomes each carrying c and wx. But a non-dis- 
junction involving either of the latter could result in no visible change in 
either the color of the aleurone or the texture of the endosperm. It may 
