1916] NOTHNAGEL— REDUCTION DIVISIONS 459 



this being accompanied by further growth of fibers. These fibers 

 gradually appear within the nuclear space even before the mem- 

 brane is entirely gone (fig. 24). When this activity first begins, 

 the chromosomes move slowly toward the center, so that by the 

 time of the stage shown in fig. 25, where the membrane has entirely 

 disappeared, the bivalents are tightly massed in the center, form- 

 ing a third contraction. Step by step the kinoplasmic fibers encroach 

 upon the chromosomes (figs. 25, 26), until they come in contact 

 with them (fig. 27), this being immediately followed by the loosening 

 up of the aggregation (figs. 28, 29). By this time there is a strong 

 multipolar complex, which is apparent as early as the disappearance 

 of the membrane. The number of poles may be many and irregu- 

 larly arranged (figs. 28-31), and as the fibers are rearranged into 

 sharper points, the chromosomes become more and more disso- 

 ciated (figs. 28-31) and scattered upon the fibers. Those bivalents 

 which formerly consisted of the two end to end somatic chro- 

 mosomes and formed a loop (fig. 22) have separated transversely, 

 making 16 chromosomes or 8 pairs (figs. 29, 30, 30.V). Fastened 

 to each member of a pair, approximately at the middle, is a group 

 of kinoplasmic fibers (fig. 30*) extending to one of the several 

 poles. As the multipolar spindle changes to the bipolar shape, 

 the chromosomes that formerly were scattered irregularly upon 

 the fibers (figs. 29-32) gradually arrange themselves upon the 

 spindle's equator (figs. 32-35). During this latter period the 

 chromosomes shorten considerably, having reached their largest 

 size during multipolar phase, as well as slowly untwisting. 



The forms that the pairs may assume now are various; some 

 remaining slightly twisted, others form a U, while still others may 

 become linked with one another, forming an X, or the ends of tin- 

 pair may remain in contact, forming a ring. No form was found 

 to be conspicuously dominant dig. 31). Polar views of the spindle 

 first show the chromosomes not to be peripherally placed (fig. 32), 

 although before metaphase is completed they take that posi- 

 tion (fig. 33). At this time the chromosome count again can be 

 taken without difficulty. 



The individual fibers that are attached to each chromosome 

 may become so closely associated at the point of union with them 



