170 TRANSACTIONS OF THE [may 11, 
Throughout all these stages the centrosomes are clearly trace- 
able and at no time disappear. During the formation of the 
segmentation nucleus the asters as a whole become far less dis- 
tinct, though they are seen to be composed of numerous ex- 
tremely long rays that converge directly to the centrosomes. 
Even at this stage fibres can be observed proceeding inward 
toward the nucleus. (Fig. 11.) 
THE CLEAVAGE-FIGURE. 
With the completion of the segmentation nucleus the asters 
again become distinct, the rays steadily increase in number 
and become more densely crowded as the development of the 
asters progresses. A small centrosphere now begins to differ- 
entiate about each centrosome. The spindle fibers from each 
aster are now more marked, throwing the nuclear membrane 
into folds at each pole. At no time, however, do they show the 
slightest evidence of entering the centrosphere, but in all cases 
end centrally at the periphery of the centrosphere. The nucleus 
now elongates, becomes spindle-shaped, its membrane fades 
away at the poles before the advancing spindle fibers, while 
laterally it is visible for some time, forming the lateral boundary 
of the spindle, its two halves converging toward the poles ex- 
actly like astral rays. (Fig. 12.) 
It persists thus for some time, in some cases until the process 
of chromosome formation is considerably advanced, but sooner 
or later it disappears. In the stage figured, each centrosphere 
contains but one centrosome occupying a central position. 
Before the equatorial plate is established, however, each centro- 
some becomes double and moves into a slightly eccentric posi- 
tion, nearer the outer periphery of the centrosphere. In favor- 
able preparations each pair of centrosomes is seen to lie in the 
center of a body or area, distinguished from the surrounding 
centrosphere by staining slightly deeper with Iron Hematox- 
ylin. This body, however, is in most cases extremely faint and 
can be made out only with difficulty. Up to this point the 
spindle has been steadily elongating, a phenomenon which ap- 
pears to be a continuation of the process of divergence of the 
centrosomes, subsequent to the division which gave rise to the 
sperm amphiaster. (Compare Figs. 8-12.) 
The fully formed mitotic figure (Fig. 13) is closely similar in 
appearance to that of Toxopneustes, as described by Wilson (11). 
In Thalassema, as in Toxopneustes, each aster contains an im- 
mense lightly staining centrosphere, sharply distinct from the 
dense outer crown of astral rays. In the former, however, the 
distance between the outer peripheries of the centrospheres is. 
