Relation between Nucleolus and Chromosomes. 49 
change of nuclear and cytoplasmic contents (figs. 53, 54). The puckering 
and subsequent dissolution of the nuclear wall extend progressively over the 
entire circumference until after an interval of about 30 minutes, or by the 
time the first polar spindle is fully formed (though still tangential to sur- 
face of egg), the nuclear wall has completely disappeared (figs. 52, 55, 62). 
Mathews claims to have been able to trace the origin of the centrosomes 
from a small granule (dividing into two before passing out of the germinal 
vesicle) within the nuclear membrane, and figures, the centrosomes passing 
through the ruptured wall. JI am unable, as I described above, to identify 
any chromatic bodies (or such as were differentiated by the stains employed ) 
with a possible centrosome. I am unable to trace any of the several bodies 
seen in many nuclei in passage through the nuclear wall. Nor is it pos- 
sible to identify any of the many small dark-staining bodies outside of 
the nuclear wall with centrosomes. If there really be such, they are indis- 
tinguishable from the yolk-granules abundantly scattered through the cyto- 
plasm. It is only where an archoplasmic sphere appears about such bodies 
that they become recognizable as centrosomes. As such I have seen them 
arise in the narrow strip of cytoplasm between the nucleus and periphery 
of egg while the nuclear wall was still, as far as the microscope revealed, 
wholly intact (fig. 50). This is conclusive proof, I think, against their 
intranuclear origin, though it still remains possible that the centrosomes 
may arise from the outer layer of the nuclear wall. I believe it more 
probable, however, that here also, as described by Griffin for Thallasema 
and by other investigators for various forms, the centrosomes arise in the 
cytoplasm. Hartmann (1902), without making any mention as to their 
origin in Asterias glacialis, shows illustrations where two fully formed 
centrosomes with surrounding centrospheres and rays are present, identical 
with my own in Asterias forbesi. 
Bryce (1903) thinks that in Echinus esculentus the centrosomes arise 
in the mass of cytoplasm which projects into the nucleus at the time of the 
rupture of the nuclear wall, thus also indicating cytoplasmic origin for 
these structures. Their origin in this location may be due to the fact that 
“the wall of the vesicle is always very close to the surface of the egg, 
leaving no room for such a formation, and the aster seems to form within 
the process” (p. 188). The astral and spindle fibers, however, he believes 
to be spun from the nuclear reticulum. In Asterias forbesti there is no such 
difinite mass of cytoplasm projecting into the nucleus, nor are there any facts 
to support the view that the asters and spindle fibers arise from the nuclear 
reticulum. 
The centrosome in Asterias forbesii is at first a single structure of the 
form of a central granule (staining black with iron hematoxylin) within a 
surrounding sphere (centrosphere) of orange-staining (with orange G) 
archoplasm. There is unmistakable evidence that this single body divides 
