74 MITOSIS : THE VARIATION OF THE CHROMOSOMES 



5. It has usually been supposed that a simple linear arrangement 

 is the only kind that can be produced by structural change, since 

 that is the only kind found in the somatic chromosomes of mature 

 organisms. Probably it is in fact the only one compatible with 

 stable reproduction at meiosis and possibly also at mitosis. But in 

 the translocation of a piece of one chromosome on to the side of 

 another, lateral translocation, giving a branched structure, has been 

 seen in three plants, Tradescantia (D., 1929 c), Allium (Levan, 1932) 

 and Alstrcemeria (La Cour, unpub.). In all these cases the branch 

 seems to occur at the centromere. Lateral trabants have also been 

 produced by X-raying (Mather and Stone, 1933). In Drosophila 

 the same structure has been inferred at some distance from the 

 centromere (Bridges, 1923 ; Hamlett, 1926 ; Sturtevant, cf. 

 Dobzhansky, 1931 a), but Bridges (1935) and Muller (1935) have 

 been able to dismiss the possibility in particular cases. 



On the other hand ring chromosomes are now known in five 

 genera : Crepis (Navashin, 1930), Drosophila (L. V. Morgan, 1933), 

 Zea (McClintock, 1932) and Tulipa and Tradescantia (Upcott, 

 unpub.). Such rings may arise directly by interchange within a 

 chromosome with loss of the small end segments which will have no 

 centromere. Or they may arise by crossing-over between inverted 

 segments in a way that will be considered later. 



The behaviour of the ring chromosome is instructive in three 

 respects : (i) In Crepis it varies in degree of linear contraction, 

 sometimes appearing as a disc instead of a ring. Its structure 

 therefore interferes with spiralisation, perhaps through interfering 

 with the coiling of the chromatids which is characteristic of prophase 

 chromosomes, but which is necessarily absent in these ring chromo- 

 somes (D., 1935 h). (ii) In Zea a segment of the ring chromosome 

 is presumably lost during mitotic divisions, its broken ends joining 

 up again since it is replaced by a smaller ring. Sometimes also 

 the ring increases in size. These aberrations are perhaps due to 

 the third peculiarity of the ring, (iii) Two ring chromosomes are 

 sometimes found interlocked at metaphase in Crepis (Fig. 18) and 

 the same thing no doubt occurs in Zea. This is presumably respon- 

 sible for the irregular numbers in different cells both of these and 

 of the normal chromosomes, for such interlocking interferes with 



