THE MUTATED GENE 137 



in Fig. 26. These are incidentally the facts to which we alluded 

 in the last chapter as proofs of Goldschmidt's interpretation of 

 the exaggeration phenomenon. 



A number of comparable facts have become known since. 

 Muller, League, and Offermann (1931) produced series of gene 

 quantities for such mutant genes of Drosophila as scute, eosin, 

 apricot by adding small fragments of the chromosome in question 

 containing the respective mutant locus, which had been secured 

 by irradiation. For most of the loci the results parallel com- 

 pletely those of Stern and lead therefore to the same conclusions. 



It is in these same experiments on dosage changes that Muller 

 et al. found the less simple behavior, as reported earlier, for one- 

 dose relations in the case of deficiencies, in which two doses of the 

 mutant gene have more effect away from normal than one. The 

 probable explanation was given on page 135 for Mohr's example 

 of abnormal abdomen. Muller (1932) mentions another some- 

 what different case. Homozygous hairy wing is twice as hairy 

 as heterozygous. Adding the normal allele in a duplication does 

 not change the situation. On the other hand, a duplication 

 containing the Hairy gene added to a normal individual makes it 

 hairy. One might express this as dominance of Hairy wing over 

 one or two doses of the normal allele (Muller calls this a neo- 

 morphic mutation. We are speaking here, with Muller, of 

 genes and alleles, though probably all the cases thus far allotted 

 to this group are major changes in chromosome architecture, like 

 translocations). It is too early to consider these facts in simple 

 terms of dosage, but there is another case, which Muller puts 

 in the same category with his neomorphs and which is most 

 thoroughly known and therefore more suitable for an analysis. 

 This is the case of Bar eye in Drosophila. 



Bar eye was originally described as a semidominant mutant 

 reducing the number of facets in the eye. Subsequently, a 

 number of what seemed to be mutations arose at the same locus, 

 partly more, partly less, reducing the number of facets. The 

 one that is important for our present discussion is Zeleny's 

 Ultrabar, reducing the number of facets to about 25 under stand- 

 ard conditions (Normal 780, Bar 68). The next important step 

 was Sturtevant's (1925) proof (see also L. V. Morgan, 1931) that 

 Ultrabar is Double-Bar, two Bar "genes" being located in the 

 same chromosome as a consequence of unequal crossing over 



