1306 BIOLOGICAL EFFECTS OF RADIATION 



At anaphase separation the two pairs of strands may be distributed 

 to the daughter nuclei in either of two ways: 



(a) One nucleus 1+2 (normal); the other 1^ + 2^ (interchange). 

 (6) One nucleus 1+2^; the other 1^ + 2 (both duplicate-deficient). 



If assortment is at random, the two types of distribution should occur 

 equally freely. 



If the two segments with spindle attachments unite with each other 

 and likewise the terminal segments, there will be two normal chromo- 

 somes, one with double spindle attachment and one with none. The 

 two latter will be lost within a few divisions. The two normal chromo- 

 somes may go to the same pole or to opposite poles. The resulting 

 daughter nuclei should eventually be either one nucleus normal, the other 

 deficient for both chromosomes, or one nucleus deficient for chromosome 

 1, the other for chromosome 2. 



In the case of alterations due to irradiation of pollen, there is little, 

 if any, elimination before fertilization. Most of the deficient chromo- 

 somes except those with two spindle attachments or none are probably 

 capable of surviving in the zygote. But very few will be capable of 

 surviving the haploid gametophyte generation, unless the deficient 

 region is very small. Interchanges and inversions will survive with 

 varying degrees of pollen and egg sterility. As in general the X-rays 

 have been applied to the sperm nucleus after the last haploid mitosis, 

 the two strands present must become separated at the first cleavage 

 of the embryo. This implies that where alterations are induced by 

 the irradiation the embryo is mosaic. Owing to the type of growth, the 

 large majority of plants with induced alterations will have all of the 

 observed tissue of the same constitution. But mosaic plants should be 

 rather frequent. This checks with the observations. The mosaic endo- 

 sperm seeds reported by Stadler (48), where part of the endosperm is 

 deficient for C while the rest is deficient for Su, are strong support for 

 this interpretation. 



Inversions and interchanges would not be detected in the plant 

 itself, except by pollen or egg sterility or by linkage relations in the 

 progeny. Duplications would probably not be detected by any technique 

 used in maize. Deficiencies involving certain known genes are readily 

 detected. So far as visible gene characters are concerned, most of the 

 internal deficiencies and all of the duplicate-deficient classes both from 

 intra- and interchromosomal interchanges would be classified as terminal 

 deficiencies. Even cytological characteristics and behavior are not to 

 be relied upon except where long terminal sections are involved or con- 

 spicuous knobs or other markers are present. The prevalence of non- 

 homologous pairing in unmatched short sections makes it impo.ssible to 

 rely upon synapsis alone for the detection of small terminal 

 fragments. 



