206 
Allen . — Nuclear Division in the 
particular order is followed in this matter, such as a breaking of the thread 
first into two parts, each of these into two or three, and so on, I have been 
unable to determine. 
The massing of the thread toward the centre becomes more and more 
marked, until the condition shown in Figs. 28 and 29 is reached. The 
close and apparently tangled mass at the centre includes a considerable 
part of the spirem, together with one or more rounded nucleolar masses. 
Another nucleole, or nucleoles, may frequently be seen in the peripheral 
region of the nucleus. It is while the nuclear contents are in this condition 
that segmentation, for the most part, occurs, although, as has been said, 
the first breaks may appear before the tendency toward massing has 
become so marked. In Fig. 28, showing a comparatively early stage of 
the massed condition, few free ends of the thread are to be seen, and these 
are all at or near the periphery of the nucleus ; for the most part, the 
portions of the thread in the peripheral region consist of loops which 
originate in and return to the central aggregation. At this stage also 
two pairs of the free ends commonly lie close together (as at a and a\ 
b and b\ Fig. 28). As time goes on, fewer of the continuous loops are 
visible, and correspondingly more free ends. Fig. 29 shows one of the 
later stages, at which a considerable number of free ends are visible, all 
in the peripheral region. The general effect of a section of such a nucleus 
is that of strands radiating like the spokes of a wheel from the central mass, 
each spoke consisting of two separate threads more or less twisted about 
each other. The exact number of free ends at this stage is very difficult 
to determine with accuracy; but in favourable cases enough may be counted 
to make sure that nearly all at least of the twenty-four pairs which would 
be present at the time of complete segmentation are located in the 
peripheral region. It follows from the facts described that the central 
massing of the spirem is due to a rearrangement of the twisted, un- 
segmented double thread in such a way as to form twelve loops, continuous 
with each other by means of the centrally massed strands, and with the 
bend which produces each loop located in the peripheral region of the 
nucleus ; and that transverse segmentation occurs in the peripheral portion 
of the loops, so that each of the chromosomes so formed has its ends 
at the periphery and its central portion in the central region of the 
nucleus. 
Miss Sargant (’ 96 , ’ 97 ) has observed a similar stage of aggregation in 
both the pollen mother-cells and the embryo-sac mother-cells of Lilium 
Martagon. She describes this ‘second synapsis’ as occurring just after 
the completion of segmentation, and finds in it the phenomena charac- 
teristic of the earlier stage to which the term ‘ synapsis ’ is more generally 
applied — namely, a clustering of the chromosomes about an amorphous 
nucleolar mass against one side of the nuclear membrane, and the apparent 
