528 Reginald Ruggles Gates 
view by focussing for a considerable distance up and down tbe axis 
of tbe spindle. 
An anapkase is skown in tig. 5, tke fall 14 ckromosomes being 
in view at one end of tke spindle. In tkis case, as frequently, 
they kave not yet split for tke next mitosis. Fig. 6 is a slightlv 
later stage skowing several interesting features. Most of tke ckro- 
mosomes here sliow tkeir bivalent nature and in nearly all tkese tke 
split is apparently a transverse one. Whether this apparent direction 
of tke split is due to a ckange iu tke Orientation of tke ckromosomes 
on tkeir way to tke poles, or whetker tke split is actually longitu- 
dinal in some cases and transverse in otkers, cannot be determined 
at present. Tkere is certainly wide Variation in tke time at wkick 
tkis split appears. Whetker there is also Variation in tke direction 
of tke split must remain for tke present problematical. I tkink tkere 
is no doubt tkat tke split of tke chromosomes in fig. 6 is a preco- 
cious one anticipating tke line of Separation of the ckromosomes in 
the homotypic mitosis. Tke study of later stages shows tkat tke 
ckromosomes are bivalents from the time tkis split appears, and tke 
bivalents in tke telopkase of tke keterotypic mitosis are doubtless 
the same bodies as tke bivalents in tke propkase of tke homotypic. 
Fig. 7 is an early telopkase in wkick the chromosome bivalents are 
assuming various skapes in tke daughter nuclei. Tkree chromosomes 
are left bekind. It is not kuown whetker such mother cells will 
rnature pollen grains, but kere is exkibited the same tendency toward 
occasional irregularities during reduetion tkat has been observed in 
several of tke other mutants studied, such as are of frequent occur- 
rence also in hybrids. Two cases were also observed wkere one 
chromosome kad passed to tke wrong pole of tke keterotypic spindle, 
as I kave described elsewkere Gates 1908a) for some of tke otker 
forms. 
A telopkase is skown in fig. 8. Tke chromosome bivalents, 14 
in nurnber, are distributed in tke daughter nucleus, giving H’s, K’s, 
X’s, and otker forms wkich are ckaracteristic of tke keterotypic 
ckromosomes in other plants. Later tke ckromosomes may Stretch 
out and anastomose witk eack other, assuming irregulär skapes, be- 
coming less compact and taking a paler staiu. Tkis process kas 
already begun in fig. 8. Tke final result is a stage like fig. 9 in 
wkich tke chromosome boundaries are in most cases no longer dis- 
tinguiskable. It is probable tkat tkis stage of interkinesis is of vari- 
able length, and tke semi-resting condition of fig. 9 is probably in 
