Darling: Sex in dioecious plants 187 



with certainty. During the interval from the collapse of the 

 nuclear wall to the arrangement of the chromosomes in meta- 

 phase, the chromosomes have become more condensed (figs. 22, 

 23). At this metaphase stage one or more bodies are found in 

 the cytoplasm, which are undoubtedly remains of the nucleolus, 

 and which finally disappear. It is very evident that these are not 

 chromosomes located at some distance from the others, for in sub- 



w 



J 



sequent stages where counts can be made only thirteen chromo- 

 somes exist ; and in side view these bodies have no connection 

 with the spindle. 



Particular attention was given to the first nuclear division to 

 determine if there is any unequal distribution of the number or in 

 the size of the chromosomes analogous to what has been found in 

 the spermatogenesis of insects ; but all observations indicate that 

 each chromosome splits into equal halves and each half moves to 

 opposite poles of the spindle. Polar views of the nuclei in ana- 

 phase stages of the first division (figs. 24 a and d), and in meta- 

 phase of the second division (fig. 33), both show the same number, 

 thirteen ; numerous counts were made of these stages so there can 

 be no doubt about the equal distribution. Side views of the spindle 

 during the metaphase and anaphase stages also show, as nearly as 

 could be determined, that each chromosome splits into equal 

 halves ; the only irregularity noticed was that some of the chromo- 

 somes in some nuclei divide before others, but in late anaphase all 

 seem to be in about the same plane. 



As the chromosomes are gathered at the poles, they are so 

 massed that their individuality cannot be determined ; very soon, 

 how^ever, the nuclear membrane is formed and as it increases in 

 size chromatin masses become readily distinguishable (fig. 25). 

 In each daughter nucleus there is found one large chromatin mass 

 and a few smaller ones varying somewhat in size. Many observa- 

 tions seem to indicate that these smaller chromatin masses pass 

 into one large mass in one of the daughter nuclei, while in the 

 other daughter nucleus they unite so as to form two chromatin 

 masses of unequal size (figs. 25, 26, 28). As a consequence, the 

 tw^o daughter nuclei present a different appearance at this stage in 

 their reconstruction by one containing one large and one small 



r 



chromatin-staining mass, w^hile the other contains but one large 



