END-TO-END PAIRING 519 



It remains to be said that the property of terminal affinity can 

 be supposed to develop only after the beginning of diplotene; for, 

 were the ends already attractive at this stage, terminal chiasmata 

 would arise at every chromosome end immediately and as a matter 

 of course. Actually, it is improbable that any chiasmata are terminal 

 at the moment of origin, for the reasons given above. 



Terminal affinity is exerted as a rule only when one lateral 

 association is displaced by another in terminalisation. But if 

 terminal affinity develops after diplotene the possibility arises of 

 its being a secondary cause of pairing at diakinesis or metaphase. 

 Exceptional cases of this have been found in the sex chromosomes 

 of LygcBus (Wilson, 1905), the microchromosomes of Anasa (Wilson, 

 1905), and Alydus (Renter, 1930). These chromosomes do not as a 

 rule pair at the prophase of meiosis. In Alydus this is apparently 

 because they have already divided at the pachytene stage when the 

 paired chromosomes have not yet divided. The microchromosomes 

 pair momentarily at metaphase and separate at anaphase. The 

 sex chromosomes divide equationally at the first division and their 

 daughter chromatids pair at the second prophase. In the anomalous 

 meiosis of Llaveia a pair of autosomes behaves in the same way 

 (Hughes-Schrader, 1931). In either case the pairing is end-to-end, 

 and not side by side. It indicates a special capacity of attraction 

 in the ends of the chromosomes such as that which is assumed to 

 maintain the terminal chiasma. Probably the sex chromosomes 

 which pair in this way are homologous for too short a length to 

 permit the formation of a chiasma if they were to pair at pachytene, 

 but retain minute homologous segments that are sufficient to exert 

 a specific terminal affinity. 



Conclusion. Terminalisation has now enabled us to distin- 

 guish three forces acting within the resting nucleus. The two 

 repulsions are non-specific. One is generalised and probably due 

 to a surface charge evenly distributed over alj the parts of the 

 chromosomes. The other is localised and is probably due to a 

 specially high charge, concentrated on the centromere and develop- 

 ing after chiasma formation. The different balance between these 

 two forces of body repulsion and centromere repulsion is responsible 

 for the forms of bivalents with different degrees of terminahsation. 



