1 98 Sargant. — Formation of the Sexual N tic lei 
this is the effect of unconscious comparison between the 
breadth of the ribbon and the size of the nucleus. Wherever 
the full breadth of the twisted ribbon can be seen, it turns out 
to be about the same in both. Sometimes the nucleus has 
a single nucleolus at this stage (Fig. 4), but commonly there 
are two or more. The differentiation of the chromatin-granules 
on the ribbon is perfectly clear even in thick sections from 
alcohol-material (Fig. 4) ; thin tangential sections from nuclei 
fixed in Hermann’s solution show it very beautifully, as in the 
preparation from which Fig. 4 a is drawn. This preparation 
is stained with Flemming’s triple stain (I, p. 474 Appendix), 
in which the characteristic colour of chromatin is a dull 
purple, and that of nucleolus and cytoplasm orange-red. The 
matrix of the spirem-ribbon stains red under this treatment, 
while the granules which border it are dark. Other double 
stains bring out the same distinction : the ribbon itself stains 
like the nucleolus and cytoplasm, while the dots take the 
colour of chromatin. 
We have now traced the history of the pollen-mother-cell 
nucleus through the first period of its development. That 
period is occupied by the formation of the spirem-ribbon. 
We have seen that it is differentiated while the nucleus is in 
the contracted condition known as synapsis, and it seems 
probable that the linin-matrix of the ribbon is formed in great 
part from the substance of the half-dissolved nucleolus. The 
double row of chromatin-granules can be first demonstrated 
with certainty during the condition of synapsis, but there is 
reason to think that it was formed by the fission of the 
pre-existing single row before the final contraction took place. 
In all these points, as also in minor details, the development 
of the pollen-mother-cell nucleus agrees exactly with that of 
the embryo-sac nucleus. 
The nucleus of the pollen-mother-cell remains for a short 
time in the spirem-condition. When development proceeds 
again, it still shows the closest agreement with that of the 
embryo-sac nucleus. As before, however, the smaller size of 
the former nuclei renders detail less clear. On the other 
