ORIENTATION 527 



of unorientated bivalents and of those of different bivalents. It 

 indicates that the structure of the spindle which enhances repulsions 

 in one direction diminishes them in another. Hence the length- 

 wise orientation of the contained water which is associated with the 

 differential lengthwise repulsion must be responsible for this 

 differential repulsion. But it is significant in two other respects. 

 It enables us to understand why the chromosomes are bunched 

 together in the short interval between diakinesis and the metaphase 

 orientation (" pro-metaphase," Lawrence, 1931). Evidently the 

 invasion of the nucleus by the spindle reduces repulsions until 

 orientation has occurred. The second respect is more important. 

 The reduction of repulsions does not imply a reduction in specific 

 attractions, since the two are physically unrelated. The equili- 

 brium position between repulsions and secondary attractions will 

 therefore be influenced more strongly by the attractions crosswise 

 in the spindle than at any time, such as diakinesis, where there 

 is no spindle. Hence secondary pairing of chromosomes is found 

 at metaphase in organisms which do not show it at prophase (cf. 

 Matsuura, 1935). 



Orientation must consist in the centromeres lying in such a 

 direction as to exercise maximal repulsion on one another. This 

 repulsion is shown by the pairs of centromeres only in meiotic 

 chromosomes, for where mitotic chromosomes are united by pseudo- 

 chiasmata as a result of X-ray breakage their centromeres show 

 no relative orientation (White, 1935). 



As we saw, owing to their body-repulsions, the bodies of long 

 chromosomes whose centromeres are lying on the periphery of the 

 spindle themselves lie outside it in the cytoplasm. Further, where 

 the parts of the chromosomes between the centromeres are under 

 very little tension, the pairs of centromeres lie equidistant, in two 

 parallel planes (Fig. 152, Plate XV). This represents the simplest 

 equilibrium position between the repulsions of the two poles and 

 the two centromeres. When, however, the parts of the chromo- 

 somes between the centromeres are very short they are drawn out 

 under tension. The equilibrium position is modified and the pair 

 of centromeres lie a little closer together. This happens necessarily 

 whenever a chiasma is formed in a very short arm of a chromosome 



