806 
for some time about 180° from the oil globule giving the impression 
that the head is growing forward. The facts are however different. 
Delaying Fertilization. 
A few other observations may be briefly described here, although 
they do not bear directly on the main problem before us. If the eggs 
of Ctenolabrus be allowed to remain a short time in sea water that 
has been boiled, before adding spermatozoa, a curious modification 
results. As small an interval as ten minutes is sufficient and the 
same result followed if two or three hours elapsed although fewer 
eggs would then develop. Assuming ten minutes between stripping 
the female and adding the sperm to have passed, it would be seen 
that when the blastodisc began to thicken in the region of the micro- 
pyle that the protoplasm accumulated around several centers instead 
of a single one. About one hour later the protoplasmic dise broke 
up simultaneously into from three to ten distinct cells with a nucleus 
in the center of each. The delay in fertilization seems to have caused 
either a rapid division of the segmentation nucleus or else many sper- 
matozoa have entered. The boiling has nothing to do with. the result 
for normal (check) egg placed in such water and immediately fertilized 
develop normally. I have not examined the material as yet with suffi- 
cient care to state which of these views is correct but I incline strongly 
to the second assumption. If a greater time passes between stripping 
and fertilization the blastodisc breaks up into these characteristic cells 
but usually into a larger number the longer the interval. 
In both cases described the cells soon pass into a resting stage 
accompanied by partial refusion as in normal eggs and there follows 
a new division of nucleus and cells. A flat blastoderm results which 
grows half way over the yolk. From such lots of eggs a few normal 
embryos are formed but I can not state that these have actually come 
from such polyspermic(?) eggs. The larger number of eggs die with- 
out forming embryos. Further details I hope to add later. If it prove 
true that we have here a case of polyspermy it is remarkable that 
we find so perfect a coordination of these male nuclei, so that the 
blastoderm after passing through alternate periods of rest and activity 
grows half over the yolk-sphere as in a normal embryo. 
In a few cases after shaking the egg at the two-cell-state, one 
of the two blastomeres has been killed and its sister-cell continued 
to develop. In such cases the single cell produced a small circular 
blastoderm that grew over the dead cell but I could not succeed in 
producing embryos from such eggs. 
