The Nucleolus and Jlicrochromosomes in the Spermatogenesis etc. 185 
Figs. 15—17 illustrate the Separation of the gemini. This pro- 
cess seems to take place by the chromatic material sliding away 
aloüg one or more of the linin threads, and the latter then separate 
almost entirely from the original mass. These smaller masses then 
sub-divide again in the same way until the complete uumber of ge- 
mini is seen (Fig. 17). 
At this Stage, it is almost impossible to distinguish the nucleolus 
from the gemini, as these latter are crowded together. Fig. 17 was 
drawn from a very favourable cell, but even here, it would be difti- 
cult to decide which is the nucleolus. 
Even before the gemini have quite separated out (Fig. 16), the 
centrosomes, the central spindle and the asters, can be seen lying 
outside the nuclear membrane. Before this stage, I have not been 
able to make out any sign of the centrosomes in the cytoplasm. 
The archoplasm has long since disappeared (a, Figs. 10 — 12). 
When the chromosomes are moving up to the equatorial plane, 
it will be seen (Fig. 21) that the nucleolus remains outside the 
spindle, near one pole. It is generally nearer the pole than the 
equatorial plane, but it is never on a level with the latter (Fig. 18). 
In a large number of cells, 30 to 40^, it is possible to distin- 
guish on the equatorial plane, one of the gemini which is smaller 
than the rest, by at least a third (M Fig. 18). 
It will be as well to deal with the history of this and the other 
gemini now, and consider that of the nucleolus later on. 
This small chromosome, is undoubtedly bivalent like the others, 
and from its size, may be called the micro-chromosome. It is round 
like the others, only differing from them in size. Before this stage, 
it is not distinguishable from the rest of the gemini, nor do the uni- 
valent halves of which it is composed remain separate tili a late 
period, as described by Wilson (’05) in Anasa tristis, Archimerus 
calcarator, Ahjdus pilosulus and Chariesteriis. 
It also ditfers from those types in its dividing, not before, but 
after the others. However, its division is not long delayed. In 
some cells, it seems to divide almost at the same time as the other 
gemini. 
The other gemini assume three distinct forms. These separate 
forms, however, are not recognisable in the late prophase at the 
time the nuclear membrane is disappearing, or before it has dis- 
appeared. 
