514 CELL MECHANICS 



(i) Closed loops repel one another more than open arms : for 

 distal chiasmata move towards the ends even though no other 

 movement is detectable. And chiasmata in Tiilipa move away 

 from the centromere when it is in a closed loop but not when it is an 

 open arm [i.e., when only single chiasmata are present). 



(ii) Loops or arms, including the centromere, repel one another 

 more than others : for the proximal loop is always the largest, 

 where considerable movement takes place. But if the centromere 

 is in an open arm and on the other side of the chiasma is a closed 

 loop the extra repulsion between the centromere may not give 

 equilibrium until the centromeres are drawn closer by the chiasma 

 moving nearer them. 



From these two series of observations, it may be inferred that 

 two kinds of repulsion occur between the chromosomes (paired 

 chromatids) : First, a general repulsion between all parts of the 

 chromosomes such as might be determined by a surface electrical 

 charge ; this would be stronger between loops than between free 

 arms, because their parts would be held closer together. Secondly, 

 a localised repulsion between the centromeres of the chromosomesi 

 The first assumption is justified by the staining properties of the 

 chromosomes, by the fact that unpaired chromosomes never touch 

 one another in the nucleus, and by the known behaviour of ampho- 

 lytes in solution. The second assumption is justified by the observed 

 special state of tension found at diakinesis and metaphase at all 

 terminal chiasmata and between the centromere and the first inter- 

 stitial chiasma wherever these are close together, as they regularly 

 are in small chromosomes and in large chromosomes where the 

 chiasmata are localised (Fig. 145) . 



These two assumptions explain the movements of chiasmata so 

 far as they are yet known (D. and Dark, 1932). In organisms with 

 the least degree of terminalisation, such as Lilium, Hyacinthus, 

 Fritillaria and Vicia, the generalised repulsion alone is effective and 

 no fusion of chiasmata seems to occur. On the contrary, an equili- 

 brium position is reached in which all the chiasmata in each arm 

 are equidistant (D. 1933 on Agapanthus). In organisms with 

 slightly more terminalisation such as Stenobothrus and Tulipa the 

 centromeres are seen to repel one another by their loops becoming 



