in Li Hum Mart agon : II. Spermatogenesis . 201 
very conspicuous. As the nucleus emerges from the con- 
tracted state and the nucleoli are reconstituted, this feature 
disappears. The filaments, which looked ragged when first 
formed — an appearance easily understood when we remember 
the lumpy distribution of the linin in the spirem-ribbon during 
longitudinal fission (Figs. 4 a and 5 a ) — become of uniform 
thickness during this second synapsis (Fig. 8), and before the 
end of that period two rows of dots can be made out in each 
filament. The young chromosome now consists of two lengths 
of ribbon twisted round each other 1 . Each length bears 
two rows of dots (Fig. 8#). These four rows of dots are 
the product of two successive longitudinal fissions. The 
ribbons continue to grow in breadth (Fig. 9 a), while the 
nuclear membrane reappears and the nucleolus resumes its 
definite outline (Fig. 9). No further change in the structure 
of the chromosome takes place for some time. It continues 
to grow shorter and broader, and the twist of its two segments 
on each other becomes closer. 
The chromosomes which will take part in the first karyo- 
kinesis of the pollen-mother-cell nucleus are now nearly 
mature. They lie loose within the nuclear membrane in 
company with the nucleoli (Fig. 10). Each consists of two 
segments which are already perfectly distinct, and which 
will be separated from each other in the course of the coming 
karyokinesis. The chromosomes of a vegetative nucleus 
show no fission until it is mature and has taken its place 
on the spindle (I, p. 454 and Fig. 5). But in the pollen- 
mother-cell nucleus we have seen that the fission of chromatin- 
granules took place within the immature spirem-ribbon 
many days before karyokinesis. The fission thus begun was 
completed by the splitting of the spirem-ribbon before it 
1 In very deeply stained preparations from material fixed in Flemming’s solution, 
it is seen that some of the young chromosomes are connected with each other 
by irregular fibrils. These disappear at a later stage. In the same preparations 
similar fibrils usually connect these chromosomes with the nuclear membrane. 
I believe the fibrils to be cytoplasmic, and to proceed from the adherence of 
certain chromosomes to the fibrillar envelope during the absence of the nuclear 
membrane. 
