Oenothera Lamarckiana and O. gigas. 959 
of the figures accompanying this paper will illustrate the main principle. 
They were drawn for their cytological features alone, and without reference 
to comparative size, and many of the figures of nuclei do not give the 
outlines of the optical sections. Nevertheless, a comparison of the figures 
of gigas with corresponding figures of Lamarckiana will show almost 
uniformly the fact that the cells and nuclei in the pollen mother-cells of 
gigas are much larger than those of Lamarckiana. The point is brought 
out especially clearly in comparisons between Figs. 4 and 51, 7 and 53, 27 
and 66, 32 and 66, 33 and 72, 40 and 78. 
The origin and significance of the double chromosome number in gigas. 
As might be expected, we have no direct evidence of the manner in which 
gigas obtained its double quota of chromosomes at the time of its origin 
from Lamarckiana , and since the type arises at such exceedingly rare 
intervals, there is little hope of obtaining such evidence unless experimental 
means are devised for very greatly increasing its production. There is only 
one observation on the Oenotheras that can have any possible bearing on 
the problem, and this is the statement of Geerts (’09, p. 52) that he found 
an example of the heterotypic mitosis of Lamarckiana showing about 
twenty-eight chromosomes in place of the seven pairs (i. e. fourteen sporo- 
phytic chromosomes) normally present. This interesting preparation is not 
figured and there are no comments upon its possible bearing on the present 
problem. 
If material of Lamarckiana should be found presenting a fair pro¬ 
portion of abnormalities such as that described above, it might be discovered 
that pollen-grains and embryo sacs are occasionally formed with nuclei 
containing fourteen chromosomes derived from twenty-eight chromosomes 
appearing during the heterotypic mitosis. Chance mating of germ-cells 
with this double number of chromosomes would give rise to individuals 
agreeing with gigas as to the chromosome count. The presence of twenty- 
eight chromosomes during a heterotypic mitosis of Lamarckiana might 
come about from a somewhat earlier appearance of that premature division 
of the chromosomes which normally takes place in Lamarckiana as early as 
the anaphase of this mitosis. Thus the pushing forward of this premature 
fission of the chromosomes from the anaphase to the metaphase of the 
heterotypic mitosis would result in the distribution of fourteen chromosomes 
to each pole of the spindle. Another fission introduced before the metaphase 
of the homotypic mitosis would make possible a group of four nuclei at the 
end of the reduction divisions, each with fourteen chromosomes. Such 
conceivable irregularities, however unlikely they may appear at first thought, 
should at least be borne in mind by those who study the Oenotheras. 
Gates (’09, p. 544) has brought forward several suggestions as to the 
origin of the double number of chromosomes in gigas , not including the 
possibility given above. Of these, the one that seems most probable to 
3 R 2 
