92 BOTANICAL GAZETTE [avcust 
cut longitudinally it is uncertain to which group the chromosome ~ 
lying near the middle belongs. F ig. 18 shows that the chromosomes 
do not always pass simultaneously to the poles, and it is possible that 
the distribution is not always equal. TI have frequently been able to _ 
count only seven chromosomes in a group, but such evidence is uncer 
tain, since there is always the possibility that one has been removed 
by the razor in making the section. In case the number exceeds. | 
eight the difficulties are fully as great, since there is always the possi: 
bility of a tetrad being broken apart. It is true that in such an 
example the size of the bodies is some check, but still there is great : 
uncertainty. Also the nucleolus, which as has been stated stains 
as the chromosomes, is to be reckoned with, if the count is made ata 
stage when that body is present. However, I believe that while the — 
number of the chromosomes is normally eight, occasionally a variation” 
from this number will be met, due no doubt to an unequal distribution — 
during division. : 
The number of chromosomes in the sporophyte is undoubtedly : 
sixteen, though I have not made an actual count. Figs. 29 and 30 — 
Tepresent the two parts into which a single cell of the seta has been — 
cut. It will be observed that the spirem is just segmenting into the — 
elongated chromosomes; two nucleoli are still visible (fig. 29). The 
count cannot be made with absolute certainty, but the number 5 — 
approximately sixteen. Fig. 31 shows one section of an early telo- 
phase from a cell of the seta. There are seven and eight chromo 
somes at the respective poles. The other section of the same cell 
shows about the same number of chromosomes in each group, P 
the masses are too confused to admit of an accurate count. I have 
observed figures in dividing spermatogenous cells, and here 
the number of chromosomes is without doubt sixteen. 
It seems desirable to point out that my final conclusions agre 
in all essentials with my preliminary paper of 1903, and are in conflict 
with FARMER’s views in the fundamental feature of his accoult 
taneous distribution of the chromatin 
the four daughter nuclei through a quadripolar spindle. It is per : 
fectly clear from my studies that the chromosomes in Pallavi ims 3 
Lyellit are distributed by two successive mitoses, each with well: 
defined bipolar spindles, and that the chromosomes are organized 
