90 KANSAS UNIVERSITY SCIENCE BULLETIN. 



When the nucleus begins division the chromatin masses ar- 

 range themselves in a spireme (figs. 2, 3) and the linin reticu- 

 lum can no longer be seen. Perhaps the network has draw^n 

 tighter, the threads stretching out to form a framework for the 

 spireme. This stage has been called "close mother skein" 

 (Schaffner) . In figure 2 the nucleoli are still present, but they 

 are now without the surrounding clear spaces. 



The spireme soon becomes looped with the heads pointing 

 toward the poles. I have not observed any connection between 

 this fact and the fact that the chromosomes are J- and U- 

 shaped. The segmentation of the spireme does not seem to 

 occur regularly near the equator of the nucleus, as would be 

 necessary to make the heads of the chromosome loops identical 

 with those of the spireme loops. 



At about this stage the nuclear membrane disappears, as do 

 the nucleoli. There is, of course, variation about the time of 

 disappearance, for the membrane is still present in figure 4, 

 where the granules have split to form a double spireme. This 

 split is very clearly shown in figures 5 and 6. The spireme 

 granules here have segmented to form the chromosomes. Im- 

 mediately after segmentation the chromosome granules con- 

 dense and the segments shorten, thus giving the apparently 

 solid character to the chromosomes. In figure 6 the two lower 

 segments have lagged behind the others in condensation. 



At this stage or earlier the milage of the spindle should ap- 

 pear, for in the next stage observed the chromosomes were 

 arranged on the bipolar spindle in the metaphase. The origin of 

 the spindle was not observed, however, nor were any traces of 

 a multipolar spindle seen. This may be due to any one of the 

 three causes : the duration of the multipolar spindle may be 

 short ; the spindle substance at this stage may not take the stain 

 readily; or the bipolar spindle may have been formed in the 

 manner usual in the cases of higher plants, but with the omis- 

 sion of the multipolar stage. Of these three hypotheses for the 

 absence of the multipolar stage, the third seems, perhaps, the 

 best from the evidence in hand. The probability of the hy- 

 pothesis decreases when we consider that the central spindle 

 fibers stain more deeply than the traction fibers. No centro- 

 some or centrosphere was seen in these cells, although in one 

 or two cases a massing of protoplasmic granules at the poles 

 of the cells might have been taken on a superficial examination 

 for a centrosphere (figs. 7, 11). It was not difficult to de- 



