658 Balls .— The Mechanism of Nuclear Division . 
should have been; in this case the spindle fibre which ends in it remains 
at the side of the spindle, often slack and bent (Fig. 8) instead of being 
drawn taut. 
The polar dots at the ends of the fibres have been noticed by other 
observers, but I can find no mention of their cross-connexion by the threads. 
When separation of the chromosomes takes place in anaphase, the 
halves retain their continuity by means of fibres ; these inter-chromosome 
fibres appear to be drawn out from the achromatic matrix of the chromo¬ 
some. The original fibres between chromosome and dot become shorter 
without any noticeable thickening. 
In telophase of this first division the cytoplasm invades the space 
between the two daughter nuclei, and the inter-chromosome fibres seem to 
be dissolved in it. There is, however, no disintegration of the thread-rings 
themselves ; it seems quite plain, from the position of their loops in late 
telophase, and in early prophase of the second division (Fig. 9), that each 
of the two rings retracts into the daughter nucleus at its own pole, viz. at 
the pole where it is attached to spindle fibres. In this way the continuity 
of the thread-ring is never broken ; the chromosomes are retracted into it, 
and it proceeds to divide again. 
The second division follows the first without any noticeable delay. It 
is even less easy to observe, the spindle being smaller and barrel-shaped and 
leaving hardly any clear zone, but the threads are again visible, and*the 
fibres terminate on them in dots as before. 
The thread-ring divides longitudinally again. This division is visible 
in the earliest second division prophase (Fig. 10). Indeed, there is some 
probability (though this is not certain, on account of the complex looping) 
that it is visible in metaphase of the first division, or even earlier ; when the 
two spireme halves are separating in first prophase there is occasionally 
a slight indication of a second division of each half. This would produce 
a fourfold spireme thread, in places, from the very beginning of the reduc¬ 
tion division ; it should be noted that this quartering is not effected in the 
same way as the quartering of the bivalent chromosome. The super¬ 
position of the loops of the threads in early prophase of the second division 
is very marked. 
The mode of isolation of the chromosomes from the rings has not been 
clearly seen in second division. If the splitting of the rings precedes this 
division, then each chromosome is drawn up to the pole by at least two 
fibres, and the rings have simply to separate. Otherwise, the chromosome 
is presumably retracted into the ring, and the process of splitting repeated 
as in the first division. 
In the beginning of the process of reconstitution of the four micro¬ 
spore nuclei it can be seen that the chromosomes are connected by threads 
and fibres, with the ‘ black dots’ at the junction of these latter. It would 
