(EXU'L'HEiiA MUT. LATxV AND MUT. SEMILATA. 
543 
111 fig. 28 one hoinotypic spindle of another inotber-cell 
is represented, showing 8 chromosomes in each anaphase 
group. In the mother-cell represented by fig. 29, all the 
chromosomes in the cell could be observed, showing very 
clearly 7 chromosomes in each of the anaphase groups on the 
upper spindle and 8 in each group on the lower spindle. 
This is the condition which should normally occur in all cases, 
giving two pollen grains with 8 and two with 7 chromo- 
somes. In fig. 30 a and h the full number of chromosomes is 
present, though they vary greatl}^ in size and some are 
degenerating. The left-hand spindle contains 6 chromosomes 
at one pole, 4 at the other, and 4 more degenerating in the 
median region of the spindle, showing that this spindle 
received 7 of the heterotypic chromosomes. The other 
homotypic spindle, which was cut at one end, contains 8 
chromosomes at or near either pole. 
The mother-cell represented in fig. 31a and h is in a late 
stage of the homotypic anaphase, the spindle on the left 
showing two clear groups of 8 chromosomes. On the 
right, one group of 7 is clearly countable; the other group 
has been cut. This group also contains 7 chromosomes, 5 
together and 2 displaced by the knife. One of the latter 
(marked c) has also been cut, the greater portion of it lying 
in section a, while a small part is in section h. 
It should be pointed out that only normal mitotic figures 
without degenerating chromosomes were used for making the 
chromosome counts, and cases such as those just described, 
>yhere difference of interpretation regarding the valency of 
particular chromosomes was occasionally possible, were not 
used at all as evidence in determining the chromosome- 
number for the plant. 
In addition to the irregularities and chromosome degenera- 
tions already described, lata rubricalyx shows various 
conditions of degeneration of the pollen mother-cells them- 
selves. Sometimes the whole anther fails to undergo proper 
development. In these cases (figs. 32, 33, 34) the tapetum 
wholly fails to differentiate and, probably as a result, the 
