NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. 145 
These threads from their very first appearance are moored to the membrane—being 
a continuation of the same cytoplasmic substance—and that association continues 
throughout all of these stages. As the membrane recedes, the threads are drawn in 
with it, and so it comes about that as the membrane becomes closely applied to the 
surface of the chromosomes, each of these bodies is furnished with its own system of 
fibrils, which to all intents and purposes are attached, as shown in fig. 9. This seems to 
me a much more rational method of attachment than the commonly accepted view that 
the threads push into the nuclear area and attach themselves individually to the 
chromosomes with their free ends. It also accounts for the more or less equal 
distribution of the fibrils among the chromosomes; for these threads would continue 
their fairly uniform occurrence over the osmotic surfaces, that is, over the membranes 
enveloping each chromosome, and as these latter are approximately the same size and 
shape, the amount of fibrils furnished to each chromosome would be about equal. 
It is obvious that as osomotic diffusion progresses and the nuclear vacuole becomes 
smaller and smaller, there will follow not only a corresponding shifting of the lines of 
tension as expressed in the kinoplasmic threads, but also an acceleration of the changes 
oceasioned by such a shifting. So that the interval between the stages represented in 
figs. 6 and 9 would be a very brief one, and such critical periods would not be very 
frequently observed. It is also obvious that, as the karyolymph becomes gradually 
exhausted by the continued osmotic diffusion, it eventually becomes no longer visible 
as a clear nuclear sap. This transition is illustrated in figs. 7 and 8. As the condition 
shown in fig. 8 is approached there naturally follows a further readjustment of the lines 
of tension. In the earlier stages, as shown in figs. 3, 4, 5, and 6, these lines of tension 
radiate out from the nuclear membrane with the nucleus itself as a centre; but when we 
reach the stage represented in figs. 8 and 9, this condition no longer prevails. The lines 
of tension have readjusted themselves to meet the new condition. Hach bivalent 
chromosome becomes the centre of a system of fibrils, but on account of the crowded 
position of the chromosomes (fig. 8) at this time, the radiations are not so regularly 
disposed as in the earlier stages. Now, I do not for a moment believe that these 
radiating threads—which merely express lines of tension—move individually or 
collectively through the cytoplasm. The apparent change in their position is due to 
the relaxing of the tensions along certain lines and establishment of new tensions 
along others. In other words, while the lines of tension do shift, the actual 
threads of cytoplasm do not. ‘These latter withdraw or reappear, according to the 
shifting of the position of the osmotic surfaces, that is, the membrane enveloping 
each chromosome. 
Another point of great interest in this connection, and revealed in the stage shown 
in fig. 8, is the distribution of the fibrils over the surface of the chromosomes. Here it 
will be seen that the chromosomes are longer than broad, and the line of division 
between the two daughter chromosomes is parallel with the long axis. If the chromo- 
| somes were perfectly spherical, one would expect a uniform and symmetrical radiation 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 7). 23 
