INTRA-HAPLOID PAIRING 



279 



1932), Triticum monococcum (Kihara and Katayama, 1932), Phar- 

 hitis (Katayama, 1935), and Zea (polymitotic pollen, Beadle, 1931, 

 1933)- That this pairing is by chiasmata resulting from homologous 

 association is indicated by the symmetrical lateral chiasmata in 

 (Enothera and the inversion-chiasmata giving chromatid bridges in 

 Triticum (Katayama, 1935), structures both characteristic of the 

 pairing of homologous segments intercalated in non-homologous 

 chromosomes. Quadruple chiasmata have been found in diploid 

 PcBonia (Dark, 1936), and evidence of reduplication within par- 

 ticular chromosomes in Tradescantia (D., 1929 c). Particular frag- 

 ments arise in trisomies in Solanum (Lesley and Lesley, 1929), 

 Matthiola (Frost, 1927) and Nicotiana (R. E. Clausen, 1931). 

 " Mutations " occur with specially high frequency in the progeny 

 of haploid (Enothera (Davis and Kulkarni, 1930 ; Stomps, 1931). 



The dyscentric crossing-over in Secale (Lamm, 1936), probably 

 depends on intra-haploid pairing. It appeared in diploid plants 

 produced by inbreeding and was associated with reduced precocity 

 and some failure of pairing. It seems probable that the legitimate 

 pairing was irregularly restricted on the time-limit principle, and 

 segments deprived of their legitimate partner by its premature 

 division have paired with reduplicated homologous segments else- 

 where, and some of these have made dyscentric partners. The high 

 frequency of intra-haploid crossing-over makes it clear that the 

 common notion of a diploid derived from non-reduction in a haploid 

 being necessarily homozygous is a delusion. Furthermore, a con- 

 dition of intra-haploid reduplication is shown by these observations 

 to be widespread if not universal. This conclusion is supported, as 

 we have seen, by salivary gland structures in Drosophila. Its 

 bearing on the theory of gene differentiation and haploidy will be 

 considered later. Its effect on the ordinary mechanism may be 

 summarised by saying that every organism with intercalary 

 reduplications can, by crossing-over between them, give rise to new 

 chromosome types and to new relatively unbalanced genotypes. 

 Furthermore, if chromosomes containing these reduplications are 

 unpaired, as in a triploid or trisomic, such changes will occur 

 especially frequently. That is why crossing-over between relatively 

 inverted segments seems to be commoner in triploids than in 



