186 Darling: Sex in dioecious plants 



body appears coming from the nucleolus (figs, 10-12); at first it 

 does not stain deeply with chromatin stains, but gradually it ac- 

 quires the same staining capacity as the spireme thread. This body 

 does not become freed for some time, at least not until the chromo- 

 somes derived from the thread are well formed, nor does it appear 

 at any time to be associated with the thread in any way. During 

 the first stages after synapsis this large chromatin body, which 

 may be called a chromosome, shows very clearly a splitting (fig. 

 13) ; this splitting is more or less visible until after this chromo- 

 some is freed. Occasionally, but not often, small chromatin 

 bodies are given off from the nucleolus at this stage, similar to 

 those given off before synapsis and these also function as those 

 which served to build up the spireme thread (fig. 14). They 

 certainly do not behave like the bivalent chromosome which re- 

 mains attached for so long a time to the nucleolus. 



While the eight chromosomes formed from the spireme are 

 becoming clearly defined, the one on the nucleolus gradually comes 

 farther out and in addition one or tw^o others make their appear- 

 ance in a similar manner (fig. 20). Each one of the chromosomes 

 thus formed from the nucleolus shows very clearly the bivalent 

 character ; as they become freed, however, they appear more and 

 more condensed so the splitting cannot be seen. There seems to 

 be some variation in the number of chromosomes which appear at 

 any one time upon the nucleolus, sometimes two or even four may 

 be seen ; but in every case no two seemed to be freed at the same 

 time nor do two ever arise in succession from the same place. In 

 all five chromosomes are thus derived. To illustrate the many 

 successive stages found in the same loculus of an anther, it may 

 be noted that figures 15, 16, 20, 21, 22, are drawings from the 

 same anther, while figures 17, 18, 19, were taken from another 

 single anther in the same stage of development. 



Soon after the fifth chromosome makes its appearance on the 

 nucleolus, the nuclear wall breaks down and the chromosomes 

 become arranged in the equatorial plate ; in the numerous nuclei 

 examined in this stage, hardly two polar views showed the same 

 arrangement of the chromosomes. There is also to be seen a 

 slight difference in size of the chromosomes but the difference is 

 not sufficient to enable one to identify any particular Individual 



