612 CAROLINE M. HOLT 



in the nucleolus. At this period the nucleolus may take a pale 

 stain so that the chromatin can be clearly seen collected about 

 it (fig. 6) or, as the chromatin threads become more distinct, a 

 deeply staining mass may appear in the region of the nucleolus. 



The chromatin threads are gathered into an irregular mass 

 which becomes three-lobed and finally resolves itself into three 

 vesicles in which the granular chromatin threads are fairly dis- 

 tinct (figs. 7, 8, and 9). The vesicular appearance is then lost, 

 but the chromosomes remain in three distinct groups of very 

 long, folded, and somewhat twisted threads (figs. 10, 11, 12). 

 These threads immediately begin to straighten and condense 

 (figs. 11 to 16). This process of conden.sation is very interest- 

 ing. Not enough cells at exactly the same stage of condensation 

 could be found to make it possible to decide definitely, but com- 

 parison of the longest pairs of chromosomes in figures 13, 15, 

 16 and figure 14, and of the medium chromosomes in figures 13, 

 15, and 16 and of the short chromosomes (especially c) in figure 13 

 with the short one in figure 15, certainly do suggest very strongly 

 that condensation takes place in a very definite and orderly 

 manner, and that in all nuclei at the same stage of condensation, 

 we should find the same number of granules in corresponding 

 chromosomes. During this period the chromatin threads split. 

 Whether, in the resting stage, the multiples of a given chromosome 

 really become fused into a single thread, it is not possible to 

 say. But the fact that in individuals which show complexes of 

 thirty-six or forty-eight chromosomes in metaphase, early pro- 

 phases in adjoining cells show but three or six splitting threads, 

 with far too much chromatin for a like number of chromosomes, 

 suggests that a parasynaptic union of sister chromosomes takes 

 place in the telophase and also that the re-separation may come 

 at any time during prophase (figs. 10 to 16 and 19). 



There is nowhere any evidence of anything but a longitudinal 

 sphtting of the threads. Figures 10 and 11 appear to be cases 

 in which an early division has occurred, while figures 13 to 16 

 show division taking place in threads in which the chromatin 

 has become condensed to comparatively few granules. Since each 

 normal complex is made up of three pairs of chromosomes, and 



