50 Genetical and Gytological Studies on Pisum 



segregations also occur on the female side we may expect to find trisomies 

 occurring occasionally in the progeny of semi-sterile plants. Hakansson, 

 found siinilar cases in which three chromosomes of the ring go to one pole 

 and one to the other, and suggested the possible occurrence of trisomies." 



Discussion. 

 The presence of a ring, on the principle that only homologous parts 

 of chromosomes pair, indicates that each of the four chromosomes is 

 homologous at one end with part of one chromosome and at the other 

 with part of another. Instead of two pairs of homologous chromosomes 

 we have to consider four pairs of homologous segments, and, giving 

 letters to the segments, the four chromosomes can be represented as 

 AB, BC, CD, DA respectively. There is no evidence that any segment 

 is reduphcated, since pairing is always between the same two parts. The 

 appearance of such a ring in a hybrid between two non-ring-forming 

 lines indicates that in one Kne the segments concerned in the ring were 

 arranged AB . CD AB . CD, whereas the other hnes had them arranged 

 AD . BC AD . BC in the diploid condition. On crossing, the ring 



AB CD 

 BC DA 



is constituted. Such a different arrangement of segments in the two lines 

 could have been caused by segmental interchange between non-homo- 

 logous chromosomes in one of them. 



When adjacent chromosomes go to the same pole, gametes of the type 

 AB . BC and CD . DA or BC . CD and DA . AB are formed. In every case 

 there is a redupUcation of one segment and absence of another, and the 

 assumption is made that such gametes are non-viable. When opposite 

 chromosomes go to the same pole only two types of gametes, AB.CD 

 and BC . DA are formed, and as these have each segment present in their 

 nucleus, they are viable. The semi-sterility associated with ring formation 

 is therefore due to non-disjunction in the ring. The proportion of dis- 

 junction to non-disjunction observed agrees with the proportion of 

 viable to non- viable gametes, being approximately 1:1. Hakansson has 

 made counts in which the number of zigzag arrangements is 268 and of 

 the other arrangements 271. 



The hypothesis of segmental interchange was put forward by Belling, 

 to explain certain cytological configurations in trisomic Daturas, and 

 also to explain the occurrence of semi-sterility in Stizolobium (Belling, 

 1925). Semi-sterility has been found to be associated with ring formation 



