THE NATURE OF THE GENE 307 



performed also the necessary tests to exclude other explanations. 

 He found, further, a similar result for the gene hairy in the third 

 chromosome — in fact, both effects simultaneously, when the 

 translocation involved breaks near both the hairy and the 

 cubitus interruptus loci. In a crossover experiment, Dubinin 

 even claims to have introduced the hairy locus itself into the 

 broken partner chromosome with the result that this replaced 

 hairy gene also showed the dominance effect. 



Similar results were obtained by Dobzhansky and Dobzhansky 

 (1933) for the gene bobbed and by Dobzhansky and Sturtevant 

 (1932) for the genes yellow, kurz, rudimentary, and forked. But 

 in these cases duplications were involved, i.e., chromosome 

 fragments containing the Wild-type alleles in addition to the two 

 recessive genes in the normal chromosomes (one + being domi- 

 nant over two recessives). The increased dominance of the 

 recessives in these cases might have different reasons, usually 

 expressed in terms of disturbance of genie balance. But the 

 remarkable fact is that the dominance-shifting effect is produced 

 only when the duplicate fragment is broken off near the locus 

 of the gene showing the dominance effect. This is certainly 

 in favor of the assumption that here, also, the position effect 

 of the cubitus interruptus type is involved. 



4. There are cases that belong to more than one of these 

 categories; i.e., allelomorphic mutations are found as a conse- 

 quence of translocations, inversions, duplications, and deficiencies. 

 But the decisive point is that, again, the effective locus is near or 

 identical with the point of breakage. A complication within this 

 group of cases is that these effects behave frequently like unstable 

 genes and therefore appear as variegation, mottling. All these 

 cases are phenotypically dominant eye colors, studied first in 

 detail by Muller (1930) and later especially by Glass (1933, 1934) 

 and Schultz and Dobzhansky (1934). Without going into the 

 complicated and not yet completely analyzed genetical details 

 (now checked by the salivary-chromosome method), it may be 

 stated that in the best known set of Plum alleles a break has 

 always occurred near the brown locus in the second chromosome 

 to which all these mutations are allelomorphic and another one 

 near the light locus. According to Schultz (1934), in these 

 cases a transfer of chromosome material to the inert region 

 (chromocenter in the salivary chromosomes) is always involved, 



