66 
the anterior end. This view does not seem probable to the writer, 
since no chromatin granules were seen escaping into the cytoplasm, 
and no stages that could be recognized as steps in the formation of 
a nucleus in this way were observed. Furthermore, the ‘ waist-line’ 
of the fully developed egg is not accounted for by this theory. 
The most probable view, in spite of its uniqueness, is that ex- 
pressed in the preceding pages and illustrated in Figs. 15, 16, 17, 
and 18. According to this hypothesis pairs of oöcytes fuse end to 
end. The proximal oöcyte is the older and contains the keimbahn- 
chromatin: the distal one possesses a spindle which breaks down and 
becomes transformed into a resting nucleus just as is supposed to 
happen in the case of the ‘uterine’ spindle of the Turbellarians. 
There are numerous cases of cell fusion in both Protozoa and Meta- 
zoa, and germ cells and somatic cells. For example, Protozoa engulf 
other cells; the fully grown ova of Hydra consist of several germ 
cells fused together; and leucocytes may fuse with one another. In 
all such cases the nucleus of one cell persists whereas those of the 
other cells disintegrate and disappear. Among certain leucocytes of 
Axolotl, however, WALKER (’07) has described a sort of fusion which 
results in the transference of the chromatin from one cell to another 
without the disintegration of the migrating chromatin. In plants also 
Gates (11) has shown that chromatin may migrate from one pollen 
mother-cell of Oenothera gigas into a neighboring mother-cell where 
it remains visible for some time before becoming incorporated with 
the surrounding cytoplasm. Many more cases of cellular fusion might 
be mentioned, but in no instance, so far as I am aware, has the 
union of two well developed odcytes to form one egg been reported. 
It is true that in Copidosoma the chromatin in one (the proximal) 
oöcyte (the keimbahn-chromatin) finally disintegrates and disappears 
in the cytoplasm, and thus the condition here may be compared with 
that in the cases mentioned above, buth the stage of fusion in Copi- 
dosoma is extremely late in the growth period and the chromatin 
material remains visible for a remarkably long interval of the germ- 
cell cycle. 
Summary of Observations. 
(1) The ‘nucleolo’ which lies near the posterior end of the eggs 
of certain parasitic Hymenoptera and which serves as a keimbahn- 
determinant during early embryonic development, does not arise from 
the nucleolus of the germinal vesicle as claimed by SıLvestkı (’06, 
