THE LINEAR DIFFERENTIATION OF THE CHROMOSOMES III 



Oenothera are often associated in a ring at metaphase, but this was 

 originally interpreted on the basis of the theory of telosynapsis (p. 1x3), 

 the assumption being that the postulated continuous spireme of pro- 

 phase was for some unknown reason not broken up into separate 

 chromosomes in these forms. Cleland^ was the first to show that the 

 phenomenon is regular and that rings (sometimes replaced by chains) 

 of definite even numbers of chromosomes are characteristic of the 

 different races and hybrids. The ring formation can be explained as a 

 consequence of segmental interchange similar to that described by 

 Belling in Datura.^ As an example, Oe. Lamarckiana forms a ring of 



Fig. 53. Chromosome Rings in 



Oenothera. — The figure shows 

 the side view of meiotic meta- 

 phase in Oenothera. The chro- 

 mosomes are held together in 

 two rings by terminal chiasmata, 

 alternate chromosomes going to 

 the same pole (1, 3, 5, 7 down- 

 wards and 2, 4, 6, 8 upwards in 

 the ring on the right; 9, 11, 13 

 downwards and 10, 12, 14 up- 

 wards in the ring on the left). 

 In this figure the regularity of 

 the arrangement is disturbed by 

 one of the rare cross-overs 

 between chromosomes belonging 

 to different complexes (1 and 6). 

 This will give rise to a mutant. 

 (From Darlington.) 



twelve and one pair at meiosis, and the chromosomes in the ring of 

 twelve become associated terminally and are arranged so that alternate 

 chromosomes regularly pass to the different poles. The species con- 

 tains two complexes, velans and gaudens, and we may suppose that the 

 genes composing these complexes lie in the middle regions of the 

 chromosomes, while the end segments are the mutually translocated 

 parts which pair. The constancy of the complexes depends on the fact 

 that the genes of one complex lie in alternate chromosomes, and on the 

 regularity of the arrangement of the ring on the spindle, while the 

 true-breeding of the heterozygote depends on recessive or gametic 

 lethal effects of the complexes. 



The formation of de Vriesian mutants can occur in two ways. 



(i) by a segmental interchange between the two complexes which 

 ^ Cleland 1922, 1931. ^ Darlington 19296, 1931c. 



