MORTALITY IN PIKE-PERCH EGGS IN HATCHERIES, 7 
and it becomes a question whether these should be regarded as cells 
or not. On the other hand, at the 29-hour stage, abnormal eggs 
are often segmented into such small cells in such a way that, exter- 
nally, it is very difficult to tell them from normal eggs, although in- 
ternally they may be shown to be very irregular in behavior. The 
intermediate stages are therefore the best material for this phase of 
the investigation. 
Those cytological features in the development of normal eggs 
which bear on the work in hand are as follows: Up to the 16-cell 
stage cell walls are sometimes partially or completely absent, but 
asters and spindles are normal in size and occupy the same position 
that they would if there were a distinct separation into cells. Fol- 
lowing the 16-cell stage, the cleavage, which becomes externally com- 
plete at least in the surface layer of cells, gives evidence of this fact 
internally by the presence of very distinct cell walls. Mitosis is at 
first synchronous in all the cells, but this regularity is soon lost, so 
that certain cells of an egg may be in the resting condition while 
neighboring cells may be undergoing mitotic division. 
As already indicated, cytological examination of uncleaved eggs at 
4 to 8 hours showed the majority to be normal (fig. 11, p. 5). The 
few exceptions were found to have anomalous mitotic figures, and 
their number was increased in the 8-hour stage. At 29 hours every 
uncleaved egg showed anomalous internal features. The exceptional 
8-hour eggs often show a very large monaster (fig. 17, p. 5). Other 
eggs may show several cytasters and an occasional spindle (figs. 12 
to 16, p. 5). At 29 hours no such large monasters are found in eggs 
of this type or in those called abnormal and generally there is only 
an increase of cytasters in the former. 
The abnormal eggs often present a curious mixture of spindles 
and asters of varying sizes, drawn-out nuclei, chromosomal] irregu- 
Jarities, and partially formed cell walls (figs. 12 to 17, 21, and 22). 
Frequently an egg is found in which a part has undergone regular 
cleavage while the rest is filled with cytasters and shows no indica- 
tion of cell walls (fig. 15). As it was expected that such irregu- 
larities would be reflected in the distribution of the chromosomes in 
division, evidence of such chromosomal abnormalities was sought. 
But, as in other teleosts, the chromosomes are usually so clumped that 
an exact analysis of them is very difficult. In at least one case, how- 
ever, the metaphase plate in one cell showed close to 30 chromosomes 
(which seems to be the diploid number as obtained from counts in 
normal eggs, fig. 20), while the adjoining cell contained only about 
15 (figs. 18 and 19). This might be explained as a case of partial 
fertilization, in which the sperm has instigated a division of the 
egg nucleus and later has fused with one of the nuclei resulting from 
‘ this first division of the egg nucleus. The fusion nucleus would then 
be diploid and the purely maternal nucleus haploid. 
Irregularities in cleavage and mitotic figures practically identical 
with those here described have been obtained experimentally by a 
number of investigators. It will be noted that in every experiment 
of this nature the effect is to induce development with one of the 
2Reighard (1890a) mentions excrescences as occurring in correlation with the flow of 
protoplasm in the formation of the protoplasmic cap. Since the number of eggs showing 
the excrescences mentioned above increases Jong after the formation of the cap, in our 
case, the phenomenon described by Reighard is probably not related to it. 
