12 BOTANICAL GAZETTE 



(JANUARY 



seems 



structure 



^ciri) m me resting stage of the nucleus, the cytoplasm shows a fine 

 fibrillar structure, almost homogeneously distributed. When the 



begins to enter into s}Tiapsis, long kinoplasmic 

 fibers appear from the rearrangmcnt of the cytoplasm surrounding 

 the nucleus, but far apart from the nuclear membrane. Between 

 these kinoplasmic fibers and the nuclear membrane there is present 

 a clear region free from any fine fibrillar structure, which was the 

 mam feature of the other parts of the cytoplasm before the appearance 

 of the kmoplasm {fig, 21). Some of the stages showing such a process 

 are represented in figs. 21-25. The nuclear membrane, which has 

 stamed distmctly, now becomes less and less affected by stains and 

 seems to be dissolved entirely, but for a short time the encroaching 

 and anastomosmg kinoplasmic fibers preser^-e the previous nuclear 

 outlme {figs. 25a, h, c). The kinoplasm, as a rule, surrounds the 

 nuclear membrane, but some of the threads run in somewhat radial 

 directions. No distinct polar formation was observed. 



\\Tien the kinoplasmic fibers push into the nuclear cavity, some 

 of them are seen to be attached to the chromosomes, while others pass 

 through, and a multipolar structure becomes noticeable (Jigs. 26a b); 

 from the fibers of the multipolar spindle there is worked out the bipolar 

 spindle, as mdicated in figs. 2ra, b. Shortening of the chromosomes 



time 



chromosomes 



ase 



different in aspect from those in the early 

 regularity in the position of the chromosomes, 



parts 



notice from successive stages, as figs. 26a, b and 27a, b tl 

 a tendency to move toward an equatorial region imtil 64 < 

 mosomes are arranged in the equatorial plate. 



The two parts of the bivalent chromosome in the equal 



throughout their entire length 

 are attached become directed toward 



m 



hile the other ends remain in 



The polar 



tonal Plate or a bttle earUer stage is favorable for counting the chro- 

 mosomes, which showed alwaj-5 64 or 66, each bivalent in nature 

 {fig. 29) . 



Sepakaiion of the daughiee chromosome 



u 



"'^. — ^N"ow the daug 



