GENERATION OF HYBRIDITY 421 



there is a high cell mortality even amongst the 3 per cent, of seeds 

 that germinate. Evidently many secondary changes take place 

 during development through the breakage of dicentric chromatids 

 at anaphase {cf. Cartledge and Blakeslee, 1934, on Datura). Peto 

 (1935) and Schwarnikow and Navashin (1934) found that the rate 

 might be greatly increased under special conditions of temperature 

 and moisture. 



In pollen grains of a triploid Tradescantia numerous structural 

 changes, including interchange between chromatids, have been 

 found by Upcott (unpubHshed) . These give rise to mitotic pseudo- 

 chiasmata between non-homologous chromosomes, and since the 

 interchanges are sometimes asymmetrical, to anaphase bridges. 

 Similar structures have been induced by Peto (1935) in Hordeum 

 vulgar e roots by heat treatment. Whether the chiasmata described 

 by Kaufmann (1934) at mitosis in Drosophila are related to the 

 somatic crossing-over that Stern (1934) and Friesen (1936) have 

 found here, we cannot say. 



Structural changes occurring within the organism occur 

 independently in homologous chromosomes. They therefore develop 

 a condition of hybridity within the organism in no way different 

 from that produced by crossing different individuals. Structural 

 hybridity therefore depends on the structural differentiation of 

 homologues, whether they are derived from the same parental 

 chromosomes, or from different chromosomes in the same parent, 

 or again from widely separated parents. We may therefore say 

 that the sticky-chromosome plant of Zea and the resting seeds of 

 Crepis are generating hybridity within themselves, while in most 

 organisms change is so slow that a high degree of hybridity arises 

 only when different organisms are crossed or in permanent hybrids 

 where, as we saw, hybridity is preserved with inbreeding or even 

 self -fertilisation from generation to generation. 



The internal generation of hybridity, by genetic changes within 

 organisms in experiment, raises the question of how far such changes 

 occur naturally. If they occur their results should appear in old 

 clones of cultivated plants. The extreme hybridity of such clones 

 has been shown in many species [cf. Crane and Lawrence, 1934). It 

 has been attributed to crossing. But the occurrence of large 



