NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. Lod 
YucCA. 
A perfect series of stages was obtained in the microspore mother-cells of Yucca, 
which furnished convincing evidence that the nuclear membrane neither breaks down 
nor collapses during the period of spindle formation. In fig. 26 we have represented a 
median section of a mother-cell with the bivalent chromosomes just about completely 
organised. The nuclear vacuole has reached its maximum size and occupies a space 
nearly equal to the volume occupied by the cytoplasm. ‘That the nuclear vacuole is 
under high osmotic pressure seems evident from its spherical form and the distended 
condition of the enveloping membrane. The cytoplasm is finely and densely granular, 
but perfectly uniform in its reticulate structure. The peripheral cytoplasm is in 
close touch with the inner surface of the thick cell-wall. 
Fig. 27 represents a similar section at a somewhat later stage. It will at once be 
seen that the nuclear vacuole is now less than half the original volume shown in fig. 26. 
It will also be seen that this reduction is accompanied by a differentiation or rather a 
changed configuration of the cytoplasm. With the nuclear membrane as a base there 
appear numerous long radiating threads or fibrils which show every indication of having 
been drawn out from the cytoplasmic reticulum. Many of these threads are so fine that 
they can only be seen with difficulty, being more or less obscured by food granules. 
Others, again, are well defined and extend for some distance towards the periphery. In 
the stage shown in fig. 28 we find the nuclear vacuole has diminished still further and 
the kinoplasmic threads have increased in number. Fig. 29 represents a little later 
stage, where the nuclear vacuole has been reduced to such an extent that the chromo- 
somes have become crowded together by the enclosing nuclear membrane. The 
kinoplasmic threads are still more numerous and much more sharply defined. The lines 
of tension represented in the threads or fibrils seem to shift when the stage represented 
in fig. 30 has been reached. As indicated in this figure, the threads appear to arrange 
themselves in conical-shaped sheaves or groups, until, as shown in fig. 31, there is a 
distinct tendency to the multipolar arrangement. It should also be noted in this last 
figure that the nuclear membrane is now in close touch with the majority of the 
chromosomes. 
A comparative study of these serial stages (figs. 26 to 31) establishes a number of 
facts in regard to spindle development that are of vital importance. In the first 
place, it becomes obvious that there has been a gradual and progressive diminution in 
the amount of karyolymph. In the second place, there is no doubt whatever that the 
nuclear membrane persists throughout all of these stages. In the third place, these 
figures demonstrate quite clearly that the differentiation of the cytoplasmic reticulum 
into the kinoplasmic threads or fibrils progresses with the diminution in the volume of 
the nuclear vacuole. 
I have already stated my interpretation of these facts in the case of Disporwm and 
Gladiolus. The diminishing of the karyolymph I believe to be due to osmotic 
