128 BOTANICAL GAZETTE [FEBRUARY 
might take the curved form shown in fig. 60. The origin of the 
two bodies is of importance in determining the homology of the 
centrosome and blepharoplast and will be discussed later. 
Starting with the stage in which the centrosome-like bodies are 
on opposite sides of the nucleus, the nuclear division takes place 
in the following manner. At first the bodies are at a little distance 
from the nuclear membrane, then the nucleus elongates so that the 
membrane closely approaches the bodies, becoming somewhat 
pointed. At the same time one observes that there is a collection 
of kinoplasm at the poles of the nucleus and extending along the 
nuclear membrane for some distance. At this time the bodies at 
the poles do not show radiations in any direction, but are very distinct 
(jig. 54). The spindle is formed from the kinoplasm which has been 
described, and when formed consists of a few thick fibres which 
converge at the poles, so that the centrosome-like bodies occupy the 
position of true centrosomes. About the time when the spindle 
develops, the chromosomes are formed from the central mass of the 
nucleus and become arranged in the nuclear plate. They are closely 
crowded together in this stage, and not so easily counted as when 
they have moved to the poles. The photograph (jigs. 75, 76) shows 
the dense mass formed by the chromosomes when arranged in the 
nuclear plate. It was impossible to determine how the division takes 
place in the chromosomes as they move to the poles. The changes 
take place so rapidly that stages are rarely found in which the chromo- 
somes are on their way to the poles. 
The centrosome-like bodies disappear during the division, but 
it is difficult to say at just what point. Fig. 56, a cell taken from 
an antheridium in which only one or two more divisions will take 
place, shows the centrosome-like bodies quite distinctly when the 
chromosomes are almost at the poles, but by the time the chromosomes 
are at the poles and before the daughter nuclei are formed, the bodies 
disappear (jig. 57). 
These bodies are best seen in preparations which have been over- 
stained and washed out. In some cases my preparations were 
stained deeply enough to show the spindle and chromosomes well, 
but only an cccasional spindle showed the bodies at the poles. When 
these slides were over-stained and carefully washed out, the bodies 
were brought out very distinctly in all cases. 
