194 BOTANICAL GAZETTE [SEPTEMBER 
larity with which the ro-rz segregation takes place, indicates that 
it is not merely a matter of chance, but that some mechanism, perhaps 
connected with the spindle, determines this regularity. A study of 
later generations of this hybrid should throw much light on the ques- 
tion whether the chromosomes of Oenothera are really unlike. 
Morcan (28) has recently shown that in certain phylloxerans, 
in which a generation of winged individuals produces (partheno- 
genetically) sexual males and females, the eggs are of two kinds, the 
female-producing eggs being large and the male-producing eggs 
small. Further, the females thus produced have the same number 
of chromosomes as the parthenogenetic females, while the males 
have two chromosomes less. Thus, in the formation of the polar 
body of the male egg, two extra chromosomes are extruded, so that 
the somatic cells of the male contain two less chromosomes than those 
of the female. Evidently there is here some sex-determining factor 
which antedates the chromosomal differences with the two kinds of 
eggs, and yet the chromosomes are the instruments of this factor, 
for the extrusion of the two chromosomes always precedes the develop- 
ment of the male individual. Studies of this sort will doubtless g!v° 
us clearer notions regarding the respective réles of chromosomes 
and cytoplasm in heredity. : 
In a former paper (11) I suggested that if the chromosomes of 
Oenothera are unlike in their hereditary capacities, then the occa- 
sional irregularities I have described in the chromosome distribu- 
tions on the heterotypic spindle would furnish a possible basis for the 
appearance of a series of types, such as the mutants of O. Lamarckian4. 
I have since studied the reduction phenomena in O. biennis an 
other types, an account of which will be published later. O. biennt’, 
in at least some of its geographical races, appears to be stable und ef 
ordinary conditions, though MacDouGat (24a) has obtained atypl@ 
forms by injections into its ovaries. The loose and frequently 
unpaired arrangement of the chromosomes in the central region. of 
the heterotypic spindle is evidently a delicately adjusted condition 
which can easily be thrown out of balance. Yet in some way te 
balance is ordinarily maintained and is only occasionally deranged 
in such a way that a different chromosome distribution results. It = 
not impossible that the condition of mutation in O. Lamarckian4 
