1196 BIOLOGICAL EFFECTS OF RADIATION 



istic properties (Bridges, Mohr) : they produce an exaggeration of the 

 effects of the genes induced, they behave as the most extreme known 

 allelomorphs of these genes, and they frequently include the loci of more 

 than one neighboring gene. In cases in which these criteria could have 

 been applied to mutations associated with breakages (Schultz and 

 Dobzhansky, 116) the results were consistently negative. The case of 

 baroid (see above) seems to be directly contradictory to the deficiency 

 explanation. According to Sturtevant (130, 132) there is no wild-type 

 allelomorph of Bar. A mutation from wild-type to an allelomorph of 

 Bar is, hence, an addition rather than a loss. 



The second explanation is, likewise, purely formal. Gene mutations 

 may take place simultaneously with breakages. Oliver (90) examined 

 the statistical consequences of this explanation and came to the conclusion 

 that mutation at loci closely adjacent to breakages takes place more fre- 

 quently than it might be expected on a chance basis. One is therefore, 

 forced to assume that the occurrence of breakages increases the proba- 

 bility of mutations taking place in the neighboring loci. In such a form 

 the hypothesis becomes a mere restatement of the facts. If, as seems 

 probable in certain Plum allelomorphs, a definite chromosome rearrange- 

 ment is always correlated with a definite mutation (Schultz and Dobzhan- 

 sky, 116), this explanation begs the question. 



The third explanation assumes that the functioning of a gene depends 

 on its structure as well as on the structure of its neighbors. The adjacent 

 genes in the chromosomes may not be totally independent of each 

 other. The genes A, B, and C may produce an effect which we call 

 "normal" when they are arranged in the order ABC, and a different 

 effect in case the order becomes CAB. Such a rearrangement deprives 

 each of these genes of the neighbors with which they are usually asso- 

 ciated, and gives them new neighbors which are normally located in a 

 different part of the same chromosome, or even in a different chromosome. 

 Either the rupture of the normal intergenic connections, or the establish- 

 ing of the new ones, or both, may alter the functioning of the genes. 

 Any chromosome rearrangement obviously involves such changes in the 

 surroundings of the genes located at the breakage points. Mutations 

 at the loci of breakages are, then, due to "position effect." 



Position effect is not a new principle invoked specially to explain 

 the association of mutations with breakages. Sturtevant (130, 132) 

 has firmly established the existence of a position effect in the case of 

 Bar. Two Bar genes located in the same chromosome (double Bar) are 

 more effective than the same two Bar genes located in different chromo- 

 somes. Neither the deficiency explanation nor the mutation hypothesis 

 is apj)licable to the Bar case of Sturtevant. Sturtevant clearly foresaw 

 the possibility of finding further instances of position effect by studying 

 chromosomal rearrangements, predicting in a sense the association of 



