PREFERENTIAL SEGREGATION IN MAIZE 79 



non-random orientation on the metaphase II spindles. The monocentric, non- 

 crossover chromatids are free to pass to the two terminal poles, while the two 

 centromeres from the dicentric chromatid are directed to the two inner poles. 

 Consequently, at anaphase II the terminal poles each receive a non-crossover 

 chromatid. Since the egg nucleus arises from the innermost terminal pole it 

 would contain a non-crossover chromatid with a full set of genes. The cor- 

 rectness of this ingenious hypothesis was established by Darlington and La 

 Cour (1941) in Lilium and Tulipa and by Carson (1946) in Sciara. 



It is possible that a somewhat similar mechanism is operating in Zea to 

 produce preferential segregation. In maize, as in Drosophila, the two 

 spindles of the second meiotic division of megasporogenesis are arranged in 

 a linear order. The basal megaspore of the linear set of four develops into 

 the female gametophyte, the remaining three aborting. We know that in 

 plants heterozygous for knobbed and knobless chromosomes, one arm of 

 some of the disjoining dyads at anaphase I possess precociously-acting 

 chromosomal fibers not present in the homologous arm. There is reason to 

 believe that the knobbed arms form precocious neo-centromeres while knob- 

 less arms do not. Owing to the rapidity with which neo-centric regions pass 

 poleward at anaphase I, those chromatids with neo-centromeres reach the 

 pole in advance of knobless arms lacking neo-centromeres. In a dyad con- 

 sisting of one knobbed and one knobless chromatid, the knobbed chromatid 

 would come to lie closer to the pole, while the knobless one would face the 

 spindle plate. 



In order to account for preferential segregation, it is necessary to assume 

 that this orientation persists until the second metaphase, and that it results 

 in the knobbed chromatids facing the two terminal poles while the two knob- 

 less ones would be oriented toward the two inner poles. On such a mechanism, 

 preferential segregation would occur only when a crossover takes place be- 

 tween the knob and the true centromere in a heterozygous bivalent. The 

 extent of preferential segregation would be a direct function of the amount of 

 crossing over in the knob-centromere region. 



Such an explanation can only be considered as a working hypothesis. It 

 can be critically tested, however, and such experiments are being conducted 

 by Jean Werner Morgan, who also participated in the studies reported here. 

 They include varying the crossover distance between knob and centromere 

 by translocation and inversion, testing for preferential segregation of hetero- 

 morphic chromosomes other than chromosome 10 in plants homozygous for 

 abnormal 10, determining neo-centric activity in chromatids with knobs in 

 both the long and short arm, etc. I prefer not to mention her incomplete 

 findings at this time, since to do so would detract from continued interest in 

 her work. 



The phenomenon of preferential segregation is by no means confined to 

 maize. Sturtevant (1936) found a non-random segregation of three chromo- 



