170 Blackburn and Harrison . — The Status of the British 
and blacker than the remainder. First the bivalents separate in their 
anaphase, actually reaching the poles ere their companions divide (see 
PI. X, Fig. 36). 
When the latter do so we have an anaphase totally distinct and much 
less in accord with type. With extraordinary fidelity the events recorded 
for R. Sabini are reproduced. Nor does the resemblance end here; just as 
related for that plant, the attraction (or repulsion ?) urging the chromosomes 
to the poles seems to lose its strength long before the whole of the split 
univalents reach their goal. Some of the failures wander and degenerate, 
whilst others join company in little groups of three and four, to appear 
later in interkinesis, as micronuclei, comparable with those recorded long 
ago for the same stage in Hemerocallis fulva by Strasburger and Juel. 
That the majority do manage to reach the seven separated bivalents in the 
‘ polar cap’ is demonstrated in a convincing fashion in the homotype division, 
for there, once again, the plates, ideal in their simplicity, show numbers 
varying from sixteen to twenty-five. Once more the seven whole chromo- 
somes, derived from the former associated pairs, travel to the plate first 
and arrange themselves centrally, with somewhat less precision, to be. 
enclosed later by the chromosomes representing the split halves of the 
heterotype phase. An equatorial division then takes place for the first 
set, when their halves travel rapidly towards the poles (see PI. X, Fig. 39). 
Although delayed in the end the others divide, but rarely, if ever, pass 
more than two-thirds of the way up the spindle fibres. As the outcome : 
of this premature halting the major daughter nuclei contain almost 
uniformly seven chromosomes, that number having been determined in 
scores of cases. Nor is there any great difficulty in making out the same 
number, when the chromosomes display themselves with much less sharp- ! 
ness of outline, in the major nuclei of the final octad. 
To go back to the fate of the delayed number ; with some regularity, 
much greater than in the analogous case of R. Sabini y they tend to crowd 
together, thus entailing, with the reconstruction of the daughter nuclei, the 
development of eight major groups. From this we learn that usually, as 
already stated, an octad rather than a tetrad is generated ; still, however, 
we must bear in mind that the term, although allowable in the circumstances, 
is not strictly correct, when due cognizance is taken of the number of 
micronuclei likewise included (see PI. X, Fig. 41). Be that as it may we 
have, arising from the pollen mother-cell and lying somewhat carefully 
spaced within the octad, eight nuclei, markedly different from the rest. 
These, acting independently, take unto themselves separate portions of 
cytoplasm, secrete the normal exine, and finally yield pollen grains. The 
fate of the micronuclei varies. If entrapped within the sphere of influence 
of one of their larger comrades they are included within the microspore 
to which it gives rise. On the other hand, one, two, three, or even more 
