Tllli GROWTH OF GENES 



ductioii ot the broken end gene is interfered with as much as its 

 action, and by sister reunion or non-division one continuous 

 chromatid loop is formed at metaphase and a bridge at anaphase 

 instead of two sister chromatids. 



Complex Genes 



On these general grounds we may approach the question of 

 certain special lines of development that genes may take. One of 

 the clearest is shown by the self-incompatibility allelomorphs in 

 plants. These are of outstanding importance for our purpose, because 

 on no view can they be regarded as produced by degradation from 

 an ancestral type : each is unique and all are fit. When, on the other 

 hand, these natural genes are damaged by X-rays they might be 

 made to mutate to give degradation products revealing their 

 structure and organization. 



The degradation of an 5 allelomorph has been brought about 

 by Lewis in Oenothera organensis. The mutant derived by irradiation 

 from Sq was unable to produce an incompatibility reaction by itself 

 in the pollen: it would grow down styles carrying 5g. But on the 

 diploid female side, when either homozygous or combined in the 

 style with S^ or any other allelomorph, it gave the 5g reaction: it 

 inhibited S^ pollen. Thus the mutation evidently depended on the 

 knocking out of enough of the power to produce incompatibility 

 substance to prevent the gene being effective in the haploid pollen, 

 but not enough to prevent it being effective in the diploid style. At 

 the same time it has not impaired power to stamp its owoi specific 

 shape or quality on that substance when once produced. Evidently 

 the S gene has two parts. The larger one, the carrier or primer, 

 makes the key blanks; the smaller, the specifying group, gives them 

 their specific shape, their unique character. 



The same principle is expressed in certain special reactions of the 

 natural S allelomorphs in the wild Petunia violacea. S^ pollen will 

 grow in S2S0 styles more readily than in S^S^- Evidently So is a 

 less efficient producer of the carrier than 5i, but an equally efficient 

 stamp for an excess of materials produced by S^. And S^ is doing 

 two different things which can be separated by its interaction 

 with ^2. 



11^ 



