THE MITOTIC CYCLE 547 



repulsion, accompanied at mitosis by the division of the centro- 

 meres into daughter bodies repelhng one another like those of biva- 

 lents at meiosis and therefore leading to their anaphase separation. 



6. The separation of the chromosomes at anaphase is assisted 

 by the change of shape which the part of the spindle between them 

 undergoes owing to its lying between and being determined by, two 

 plates instead of two poles and one plate. In this third stage it 

 is a centromere spindle. 



7. The formation of the spindle and the successful orientation 

 and division of the chromosomes therefore depends on a balance 

 in the strength and timing of the cycles of division and repulsion 

 of the centrosomes and centromeres. This balance is determined 

 in different ways in different types of mitosis. In higher organisms 

 the most important condition is in the centromeres lying inside 

 the nucleus while the centrosomes lie outside in a substrate with a 

 presumably different cycle of changes in hydrogen-ion concentration. 

 Most changes in the balance give unregulated results, such as 

 polymitosis and non-pairing at meiosis. The one regulated change 

 is that produced by the degree of precocity of the external cycle 

 which has turned mitosis into meiosis. 



4. ULTRA-MECHANICS 



(a) Introduction. The mechanical hypotheses that have been dis- 

 cussed so far provide the clue to certain problems of chromosome 

 movement, especially to those concerned with visible movements. 

 In doing so they should also present us with a new means of approach- 

 ing other problems of chromosome movement, those concerned with 

 movements that are inferred from their later consequences. And 

 these invisible movements in their turn should tell us something 

 about the physico-chemical conditions underlying the mechanical 

 properties of the chromosomes and other cell-components. 



The first of these problems that we have to consider are the 

 problems of chromosome breakage. They are concerned with two 

 not unrelated aspects of breakage, viz., crossing-over and structural 

 change. These properties, as we shall see, being universal and 

 fundamental, are presumably inherent in the chromosome thread. 



(h) Crossing-over. Many of those who have put forward hypo- 



'^- 18—2 



