438 FRANK R. LILLIE 
at the point of the sperm nucleus farthest from the cone. In 
other words, the position of origin of the sperm aster is a function of 
polarity of the sperm nucleus, and it is this which explains its invari- 
able origin, so far as has been recorded in the literature, in the 
position of the middle piece of the spermatozo6n; and the theory 
that a centrosome introduced by the spermatozo6n is necessary 
for such formation is therefore shown to be incorrect. 
The results so far show that sperm fragments, even of very 
minute size, may enter the egg in conjunction with the cone, rotate 
in the normal manner and produce an aster in the egg-cytoplasm. 
The question now arises, what is the ultimate fate of such frag- 
ments? Is their fertilizing power adequate to produce segmen- 
tation of the egg? 
5. The later history of the partial sperm nucler 
Partial sperm nuclei separate from the cone and penetrate 
towards the center of the egg like normal ones (figs. 38 and 39). 
The sperm aster divides and forms an amphiaster in the stage of 
the anaphase of the second maturation division characterized by 
inequality of the two poles as in the normal. But apparently the 
size of the centrosomes and the extent of the astral radiations 
are directly proportional to the mass of the sperm nucleus con- 
cerned. This is brought out very well in fig. 41, which is a recon- 
struction from three successive sections showing two sperm nuclei 
of unequal size in the same egg. The egg in question had been 
centrifuged forty-four minutes after insemination and was pre- 
served forty-seven minutes later in the stage of the telophase of 
the second maturation division. It will be observed that the 
larger nucleus (left) is accompanied by a larger centrosome and 
aster than the smaller one (right), and it should be stated that the 
aster in each case is the larger one of an amphiaster. Here, where 
direct comparison between nuclei of unequal size and their accom- 
panying asters within the same egg is possible, the proportional 
size of asters to nuclei is obvious. I do not mean of course to 
assert that the proportions are mathematically accurate for this 
would be impossible to determine. 
