182 BOTANICAL GAZETTE {SEPTEMBER 
results directly from the doubling of the chromosome number; at 
least the double number of chromosomes and the larger size of cells 
occur simultaneously. Whether the latter isa necessary consequence 
upon the increase in cell size and a resulting change in cell relations, 
or is due to an independent factor, is uncertain. It should be said 
that characters, such as leaf shape, which might be accounted for 
by a change in the relative dimensions of the cells (though I have not 
as yet made measurements of the leaf cells to determine this), are 
extremely variable, and it seems not unlikely that this extreme varia- 
bility may result from variation in cell dimensions consequent UpuD 
the readjustment to the double chromosome number. : 
DESCRIPTION 
The cytological account will begin with the telophase of the 
heterotypic mitosis in the pollen mother cell. The ten figures in 
plate XII are from the plant having 20 chromosomes, and were drawn 
on a smaller scale than those of the other two plates. At this stage 
the chromosomes can be counted with perfect accuracy. A large 
number of counts of this telophase show that 10 chromosomes enter 
each daughter nucleus. In a number of cases 10 were counted at each 
end of the same spindle. In several instances 11 were found in one — 
of the daughter groups, and in a few cases it was possible to show that 
the corresponding daughter group contained only 9 chromosomes. 
In this plant, then, there are 20 chromosomes which segregate into 
two groups of ro each in the reduction division, one chromosome 
occasionally going to the wrong group. Counts of somatic cells, 
made long ago in tissues of the anther, also showed that 20 chromo- 
somes were present. 
_ The material from which the figures in plate XII were drawn was 
subjected to an exceptionally high temperature in the process of | 
imbedding, and in a few cases this has apparently affected the shape 
of the viscous chromosomes. Figs. 1-5 are early telophases before 
the formation of a nuclear membrane. In fig. 1 the chromosomes 
nearly all show their bivalent nature, and in most of them the two 
halves are dumb-bell shaped. This clubbing of the chromosome? 
at the ends is a common phenomenon both at this time and in the 
telophase ofthe homotypic and somatic mitoses. I have already 
