EARLY STAGES OE SEGMENTATION OF ECHINUS HYBRIDS. 489 
Some inerelj become swollen and faintly stained in the 
centre; others develop a vesicle at one side or end while the rest 
of the chromosome appears normal ; others, again, split more 
or less completely, but a vesicle is formed by one or possibly 
both of the longitudinal halves. AVhether a chromosome of 
which one half forms a vesicle divides completely or remains 
undivided in anaphase, seems to depend to some extent on its 
position on the spindle at the time when the vesicle develops. 
Jjefoie the equatorial plate stage the chromosomes are scat- 
tered quite irregularly over the spindle, but already may 
show traces of the longitudinal split (h‘g. 9, a, h). If at this 
sfage one half of a chromosome develops a vesicle, it prevents 
the whole chromosome from reaching the equatorial plate, 
with the result that both halves of the chromosome are 
included in the anaphase group belonging to the pole of the 
spindle to which the chromosome happened to be nearer in 
the pi-ophase. This is especially well seen in the case of one 
chromosome wh.ich is conspicuously longer than the rest. 
Two such long chromosomes ai-e found in both esculentus 
and acutus ; the hybrid also has two, one doubtless derived 
from each parent. In the pure species these long chromo- 
somes complete their division rather later than the smaller 
ones, making bracket-shaj)ed figures, just before the 
complete separation of the halves. In the hybrids it ma}' 
sometimes be seen that one of them is in this stage in the middle 
])lane of the spindle, while the second, lialf of it bearing a 
vesicle, lies between the equator and one of the poles (fig. \8h). 
At the close of the prophase the normal chromosomes 
arrange themselves in an equatorial plate, and are at this 
stage rather long and bent. Some of them are already 
completely transformed into vesicles; to some, vesicles are 
attached, and probably others have already^ formed and 
thrown off a vesicle, d'he vesicles may be included in the 
equatorial })late, or remain scattered on the spindle; some 
are nearly always left in the equatorial plane just outside the 
spindle, and take no further part in the division. The 
chromosomes now divide along the split of which traces can 
