CYTOLOGICAL STUDIES OF CENTRIFUGED EGGS 651 
the blastodisc at one pole) and since the egg orients on the ma- 
chine, the abnormahties that so frequently appear must be as- 
cribed to some other cause than separation of materials. This 
other cause I have discovered — polyspermy. If eggs are first 
fertilized, and then revolved they produce normal embr3^os even, 
when the eggs are left in the machine until the cleavage furrows 
have appeared. But if the eggs are first centrifuged, and then fer- 
tilized, polyspermy follows in a majority of the eggs. Only those 
produce embryos, as explained, in which one sperm nucleus com- 
bines with the egg to give a normal bipolar karyokinetic figure. 
The following diagrams show some of the more common types 
of embryos. They are not peculiar to centrifuged eggs, but may 
be found in polyspermic eggs that have arisen from any other 
source. 
In fig. J a young nearly normal embryo is shown, but at one side 
a separate blastodisc is present, which has not, however, interfered 
with the formation of the embryo up to this stage. In fig. K 
the embryo, although elongated, has failed to draw together in the 
middle line. In fig. L the anterior end only of the embryo has 
developed. The medullary plate is split open widely behind — 
only a thin layer of cells covers the V-shaped split. In fig. M 
the embryo has hardly differentiated at all, but appears as an 
enlargement at one point on the germ-ring. Such an embryo 
furnisher a step towards the condition when the germ-ring alone 
is present.^ In fig. N the embryonic material has the form of a 
horse-shoe owing most likely to the paternal portion of the blas- 
toderm lying at the posterior end of the embryonic region. In 
consequence the blastoderm fails to concentrate at this side to 
produce the primordium of the embryo. In fig. O the beginning of 
the embryo is shown, but one side is very incomplete. If an em- 
bryo were to develop later from such a beginning it would have 
only the head and one side of the body present and would proba- 
bly resemble the embryo shown in the next figure. This embryo, 
fig. P, is strictly a half-embryo except in the head region that is 
more nearly whole. On the defective side there is a mass of unor- 
See Morgan, The Formation of the Fish Embryo, Jour. Alorph. 10, 1895. 
