Divisions in Vicia Faba . 
847 
At the same time, perhaps we may say in consequence of this change, 
other meshes approximate so that a double thread is formed (Fig. 11). It 
shows an appearance similar to that of the split chromosomes when passing 
into the resting stage (cf. Figs. 5, 11), and may safely be identified with 
them. The spireme thus produced is from its formation a double structure. 
As development proceeds there becomes evident a definite relation 
between the nuclear thread and the nucleolus ; the spireme is arranged 
in coils and loops radiating somewhat irregularly from the nucleolus in 
a way that recalls the ‘ second contraction ’ of meiosis (Figs. 12, 13), and 
may perhaps be compared to the polarization observed by Farmer and 
Shove ( 3 ) at a similar stage in Tradescantia. The nucleolus becomes 
vacuolate and decreases in size, doubtless giving up material to the chro¬ 
matin thread, and as it disappears the spireme straightens (Fig. 14) and 
becomes coiled with more or less regularity either around or up and down 
the nuclear area (Fig. 15). The remains of the cross attachments have 
by this time disappeared, and the stainable substance has come to be pretty 
uniformly distributed along the thread, which is double throughout its length. 
At about this stage the segmentation of the spireme into chromosomes 
occurs (Figs. 16, 17); this takes place gradually, the spireme breaking 
across first at one point and then at another, and for some time the 
separated chromosomes may adhere one to another by fine threads (Fig. 18). 
Throughout the subsequent stages certain chromosomes are again and again 
observed, as in the telophase, to be made up of two, or in one or two cases 
more, distinct segments (Figs. 1, 19, 23), and when separation of the 
daughter chromosomes has taken place it is possible to recognize this 
in each member of a pair. 
While the development of the chromosomes is proceeding, spindle 
formation has taken place. A thickening of the cytoplasm around the 
nucleus occurs at an early stage (Fig. 16), and later, as the nuclear mem¬ 
brane breaks down, delicate fibres are visible (Fig. 18), running from end to 
end or from corner to corner of the cell. 
It would appear that the chromosomes take up their position on the 
spindle very soon after it is formed (Fig. 18) and remain for some time 
lying against it; we observed, at any rate, numerous stages such as that 
shown in Fig. 19, while Fig. 18 represents a stage comparatively rare. 
Throughout the development of the spireme the longitudinal fission of 
the thread is very clear, but after segmentation it becomes inconspicuous, 
and may sometimes seem to be completely obliterated (PI. LXIII, Fig. 27). 
Even at this stage, however, it can be identified in favourable preparations 
(PI. LXII, Fig. 18), and by the time that the chromosomes have passed 
on to the spindle it is once more evident. 
The double chromosome is attached to the spindle by one end, and at 
first lies more or less parallel to or across the spindle fibres (Fig. 19). 
3 K 2 
