494 CELL MECHANICS 



(Ch. IV). In such cases it does not seem impossible that pairing may 

 sometimes be continued by a false association of divided parts of 

 chromosomes, due to torsion and not to attraction, due in fact to 

 torsion overcoming repulsion in short intercalary unpaired seg- 

 ments. 



The evidence shows that none of these kinds of torsion pairing 

 results in association that will permit of crossing-over. Many cases 

 are inconclusive. Thus where there is non-homologous torsion pair- 

 ing it is impossible to prove the complete absence of homologous 

 attraction-pairing (Ch. VIII). Pairing and crossing-over within 

 haploid sets, for example, may always be due to the duplications 

 that are known to occur within them (Ch. XI). On the other hand, 

 non-homologous parts pair at pachytene in a Zea-Euchlcena deriva- 

 tive (Beadle, 1932), and the same pairing is characteristic of the 

 differential segments of some sex chromosomes (Ch. IX). But in no 

 case do these segments form chiasmata and in no case does the 

 progeny reveal crossing-over. 



It follows from all these considerations that the analysis of the 

 conditions of zygotene pairing is more complex than it was formerly 

 believed to be. From time to time observers allege that they have 

 discovered an unexpected peculiarity of the chromosomes in regard 

 to their pairing (e.g., Huskins and Smith, 1934 ; Lorbeer, 1934 ; 

 Fig. 86). In order to arrive at such conclusions in the first place and 

 a fortiori in order to assess their value for the general theory of 

 chromosome mechanics it is necessary to see and understand the 

 succession of changes before and after the supposed abnormality, a 

 precaution that its discoverers have usually omitted to take. 



The underlying principle in zygotene and pachytene association 

 is therefore incontestable : an attraction between all homologous 

 parts of threads in pairs. When we turn to the later stages of the 

 prophase of meiosis we find that the same principle continues to 

 apply in a system where the units have a different value. Instead 

 of the single threads being chromosomes they are the chromatids 

 produced by the division of the chromosomes at the end of the 

 pachytene stage. Association between chromosomes continues by 

 virtue of the changes of partner at chiasmata amongst the pairs of 

 associated chromatids. This is shown by the failure of pairing 



