MATURATION OF OVUM AND DEVELOPMENT OF ALLOPORA, 589 
When the germinal vesicie reaches the distal pole of the 
ovum, two polar vesicles are successively thrown out. I am 
unable to find any evidence that this is accompanied by 
karyokinesis. There seems to be a concentration of the 
chromatin meshwork in the part of the germinal vesicle that 
is protruded, but that is all. I can find no stars or loops at 
any time in the history of the germinal vesicle of Allopora. 
After the discharge of the polar bodies the germinal vesicle 
becomes flattened to a hemispherical shape on the surface of 
the ovum, and from the large number of ova that can be 
found of that shape in that position I infer that it remains so 
for a considerable time (fig. 11). Itis while the germinal vesicle 
is in that position that I believe fertilisation takes place. I 
have not been fortunate enough to find any steps in the process 
of fertilisation, nor to discover the precise way in which the 
spermatozoa reach the ovum. It should be noted here, how- 
ever, that as the ampulla grows in size it breaks into various 
canals of the superficial meshwork of the coenosteum, and con- 
sequently the ovum is brought into close contact at various 
points with the superficial canal systems. If therefore we 
are justified in supposing that a number of spermatozoa are 
driven by ciliary currents into the canal systems by way 
of the mouths of the gastrozoids, there is little difficulty 
in understanding the way in which the spermatozoa reach 
the ova. 
In the next stages the large nucleus which must now be 
supposed to be formed of the conjugated germinal vesicle and 
sperm nucleus retreats from the surface of the ovum. At 
first it has a rough resemblance to the germinal vesicle in the 
stage just before it reaches the periphery, but a careful exami- 
nation is sufficient to avoid confounding the two stages (fig. 
12). When the oosperm nucleus retreats, it is regularly hemi- 
spherical on the proximal side but irregular and provided 
with four or five delicate amoeboid processes on the distal side. 
The membrana limitans of the oosperm nucleus becomes very 
indistinct, especially on the proximal side, and in the later 
stages entirely disappears. By these characters it is possible, 
