PSE U DO-CHI A SMA TA 433 



time of division in relation to the chromosomes. The distinction 

 between acentric fragments and functional monocentric chromo- 

 somes first became clear from the results of irradiation. They 

 showed that the centromere was a self-perpetuating body and 

 incapable of arising de novo or from other chromomeres. The 

 inference had already been drawn both directly and on 

 mechanical grounds that in mitosis the chromomeres divided 

 during the resting stage and the centromeres divided at metaphase, 

 while both these processes were relatively delayed owing to the 

 precocity of meiosis (D., 1932). These conclusions are now borne 

 out by both the genetical and cytologic al evidence of irradiated 

 material. Chromatid breaks and chromosome breaks are found 

 at the first mitosis after treatment of the resting nucleus at different 

 stages. The evidence of the diplochromosomes is of especial 

 significance in relation to the centromere, for it shows two of its 

 properties ; first, that its division is dependent on the action of the 

 spindle or more properly of the centrosomes and is independent of 

 the division of the chromomeres ; secondly, that its division is 

 later than that of the chromomeres. Finally, we see that mitotic 

 " bivalents " united by pseudo-chiasmata which result from 

 chromatid interchange do not orientate themselves on the spindle, 

 and do not separate Hke meiotic bivalents owing to their centromeres 

 being already in process of division and already having, therefore, an 

 internal property such as polarity, which the centromeres of chromo- 

 somes at meiosis do not acquire until a later stage. These observa- 

 tions have an important bearing on the external mechanics of the 

 chromosomes which will be considered later. 



