140 Nature of the Genetic Material 



is more normal, in agreement with nos. 7 and 8. Nevertheless, in no. 

 14 the increase of the ci eflFect goes again in the opposite direction. 



Our question is whether in view of these (at first sight) rather 

 strange facts (which Stern explains by general ideas concerning the 

 action of the gene, to which we shall return), the Dubinin effect is 

 a special kind of position effect. We saw that this assumption is 

 partly caused by semantics: instead of saying that a break near the 

 normal ci locus makes it act like the recessive mutant, it is described 

 as the break making the opposite ci act as a dominant. On the other 

 hand, there is a real diflFerence from ordinary position effects, which 

 is clearly visible in the details of action just tabulated. 



The confusing results are, however, understandable if we realize 

 the special feature of the case and forget about misleading termi- 

 nology like "the position effect stretches to the gene." The decisive 

 fact, as already emphasized above, is that in the Dubinin effect we 

 have the following arrangement in the translocation of the right arm 

 of the fourth chromosome, say to the X-chromosome: X euchrom. — 

 break — 4 heterochrom. — ci locus. 



We discussed the fact previously that a position effect based 

 upon a break in chromocentric heterochromatin is weaker and there- 

 fore, under proper threshold conditions, results in the mottling effect, 

 which is rather sensitive to temperature, extra heterochromatin, and 

 so on. We therefore expect a weak position effect for the break R ( + ) 

 in the rearrangement above. In the present case, the absence of a 

 piece of vein, variegation or lack of it cannot be distinguished. 

 Hence we expect for R( + )/+ a plus-minus normal condition and, 

 with sufficient weakening of the effect, the same for R(4-)/R( + ). 

 This is actually found (nos. 6 and 8) in the foregoing enumeration. 

 However, one rearrangement, R(2)+ acts, when homozygous, like 

 a standard position effect (deficient venation), showing an expected 

 range of seriation for the weakening of the position effect from a 

 complete one to almost nil. The next test is the break opposite a 

 deficiency, R( + )/— . According to no. 6 and the fact of the exaggera- 

 tion effect in ci/— , we expect a little less normality. No. 7 does not 

 show this. Obviously, the shfft is still below the threshold for the 

 ci phenotype. But no. 13 shows it in action, when a little weaker 

 allele is involved. Now we come to the most interesting combinations: 

 ci/-|- is normal, but with some isoalleles and in low temperature may 

 be slightly abnormal. What is expected for R(ci)/+? If we use the 

 definition of the position effect that a locus near the break is made 

 to act as if mutated, the already mutated locus should not be in- 



