958 Davis.—A Comparison of the Reduction Divisions of 
The chromosomes become somewhat irregular in form in their passage 
to the poles of the spindle during anaphase (PI. LXXIII, Fig. 75) and after 
the organization of the daughter nuclei that are to enter the pollen-grains 
(Figs. 76-78). The fourteen chromosomes may at first be easily counted in 
favourable preparations of the young pollen nuclei (Fig. 77), but as the nuclei 
enlarge the chromosomes become attenuated, and by anastomosis soon 
develop a chromatic network in which their outlines become lost. Deeply 
staining regions in the reticulum may at certain stages be recognized as 
chromatic bodies undoubtedly representing the chromosomes, but they 
cannot be positively identified in the nuclei of older pollen-grains (Fig. 79). 
Thus it will be noted that the events of the homotypic mitosis in gigas 
are in agreement with those of Lamarckiana and the other forms of Oenothera 
that have been studied. 
The size of the cells and nuclei of gigas as compared with that of 
Lamarckiana. It is important to compare the size of the cells and nuclei 
of gigas and Lamarckiana with reference to Boveri’s well known conclusions 
from experimental studies on the eggs and larvae of sea-urchins. He was 
able to institute comparisons between larvae developed parthenogenetically 
(x chromosomes), normal larvae (2 # chromosomes), and giant larvae 
( 4 x chromosomes) obtained by shaking eggs and thus arresting the process 
of mitosis after the division of the chromosomes, all of which become 
included in the same nucleus through the formation of a monaster. Among 
other conclusions, Boveri (’ 05 ) formulated a law that the size of the cells in 
these sea-urchin larvae is a function of the chromatin content, and that the 
cell volume is proportional to the number of the chromosomes. 
Gates (’ 09 ) has brought together from a variety of tissues an interesting 
series of comparative measurements between the cells of Lamarckiana and 
gigas , which indicate that the cell volume throughout the gigas plant is 
uniformly greater than in Lamarckiana. However, the ratios between the 
cell volumes of the two plants in the homologous tissues studied by Gates 
vary greatly, ranging from a ratio of approximately 1: 1 -5 to a ratio of 1 : 3-8, 
instead of being in a uniform ratio of 1: 2 as might be expected from Boveri’s 
conclusions. The discrepancies between these results are perhaps difficult 
to understand, but it should be noted that the cells were measured from 
sections cut in paraffin (in which some distortion is inevitable), and also that 
no allowance was made for the size of the vacuoles, which in plant cells are 
large and variable, except as a rule in embryonic tissues. Thus it is 
extremely difficult to institute close comparisons between the size of the 
nuclei and plasma masses of most plant cells. 
It is clear from Gates’s results, nevertheless, that the cells of gigas are 
probably larger throughout the entire plant than those of Lamarckiana , 
and the greater size of many of the plant’s organs (leaves, flower parts, 
pollen, &c.) is quite certainly correlated with this fact. An examination 
