130 Nature of the Genetic Material 



mutation to position effect. We saw that apparent return mutations 

 are suspected of being the result of a rearrangement with dominant 

 position effect and that there is a general suspicion regarding the 

 occurrence of genuine return mutation (I 3 B a). In position effect 

 the equivalent of a return mutation would be the complete restoration 

 of a rearrangement to normal. There is clearly very little chance 

 that the same two or four breaks would reoccur unless easily break- 

 able points are involved. It is not surprising that only two such 

 events have been described, by Griineberg (1937) and Hinton (1950). 

 However, there is also the chance that an apparent reversal of a 

 position effect is due to the occurrence of a new break near the one 

 causing the first position effect and acting in the opposite direction. 

 Actually, a number of counteracting position effects have been 

 described, meaning that one position effect disappears if another is 

 superimposed. Such cases have been found for Curly, Glazed, Bar 

 (all of them belonging to the dominant effects just discussed) by 

 Griffen (1941), Griffen and Stone (1938), and others. It should be 

 realized that this is just the opposite of the action that simulates 

 return mutation of a recessive point mutant. There the rearrange- 

 ment causes a normal dominant effect which covers the recessive 

 mutant effect. In the present case, a dominant mutant effect is re- 

 turned to normal by a superimposed position effect. This shows 

 that reversal of the effect is a problem which offers the same diffi- 

 culties and pitfalls for position effect that it does for so-called point 

 mutants. 



Furthermore, if a break near a given locus produces the position 

 effect, we should expect to find frequently two position effects, one 

 for each euchromatic break. Such cases seem to be very rare. One 

 reason might be the small probability that two breaks are each 

 located near loci with discernible phenotypic effect. Another reason 

 is that, in view of the greater breakability of heterochromatin, there 

 is a smaller chance of two euchromatic breaks. Sometimes an indirect 

 demonstration may be possible. Thus I could show the following 

 relation ( Goldschmidt, 1945a ) : the recessive mutant bran at or near 

 the arc locus in Drosophila produces broad wings. The mutant 

 pointed ( svr^"' ) near the tip of the first chromosome makes for pointed 

 wings. The combination bran/bran svr^"* has short, soft, and blistered 

 wings. A small inversion in the first chromosome In( l)y''''^'' has one 

 break near the yellow locus and therefore a yellow phenotype by 

 position effect. The other break is in the silver region, but no second 

 position effect is visible, perhaps because yellow and the paleness of 



