CO-ORIENTATION 



529 



and we must suppose it is the initiation of division, the mode of 

 which will be considered later. 



If, then, the centromeres of a bivalent at metaphase are in the 

 same condition as those of a meiotic chromosome at anaphase why, 

 it may be asked, do they linger in equilibrium on either side of the 

 plate instead of passing at once to the poles ? This question can be 

 answered from the arrangements of multiple configurations on the 

 plate. It is a matter of indifference for their movements on the 

 spindle whether such configurations are the result of polyploidy 



■^•O 



V^ J 



3 i V al c/\ t i 



T'r L y a / en ts 



Fig. 155. — Equilibrium positions of centromeres of bivalents and 

 trivalents with different chiasma positions in relation to the 

 two poles. (D., 1935) 



or structural hybridity. They arrange themselves according to 

 five different systems : (i) linear, (ii) convergent, (iii) parallel, 

 (iv) discordant, (v) indifferent (Fig. 153). 



The linear arrangement is usual where the centromeres are close 

 together as in small chromosomes or with chiasmata very close to 

 the centromere. Again we see that there is a limit to the distance 

 apart at which orientated centromeres can lie, and this distance is 

 less than the length of the spindle. 



The convergent system is the characteristic " disjunctional " 

 arrangement in rings of numerous chromosomes, but it is also the 



