280 THE SECOND-CHROMOSOME GROUP 



Only those which resulted from crossing-over between star and the 

 ocus for dachs would fail to carry the lethal. If these two loci had 

 been closer together, then fewer such cross-overs would occur and 

 selection could be correspondingly relaxed. In the case of a pure 

 breeding stock of ''beaded," Muller found that there was an autosomal 

 lethal in the not-beaded third chromosome, and very close indeed to 

 the locus of the beaded allelomorph, so that no selection at all was 

 needed. This principle, first used consciously in carrying on the 

 stock of dachs-lethal, was called by IMuller "balanced lethals" as 

 worked out by him in the analysis of the beaded stock. Muller has 

 shown that this principle has wide application, and may solve some 

 of the knotty problems of the genetics of Oenothera, such as pure- 

 breeding heterozygotes, the continual production of rare forms called 

 mutants (which by this principle are due to crossing-over rather than 

 to a fresh occurrence of the mutative process), and also the appearance 

 of t\Adn hybrids from certain crosses. 



It was quickly recognized that the dachs-deficiency explanation was 

 alternative to that of a simple recessive autosomal lethal occurring 

 in a locus close to that of dachs, the recessive dachs gene being present 

 and unchanged, but prevented from giving rise to the dachs character, 

 because all (or nearly all) of the homozygous dachs flies were also 

 homozygous lethal, and hence never appeared as adults. All of the 

 three parallels to forked-deficiency were equally explainable on the 

 linked lethal view. A possible method of distinguishing between the 

 two conditions was offered by the appearance or non-appearance of 

 dachs flies upon inbreeding. If the phenomena were due to dachs- 

 deficiency, then, of course, no dachs could ever appear, since the lethal 

 effect involved the dachs locus itself. But if two separate and dis- 

 tinct loci were involved — dachs and a neighboring lethal locus — then 

 by crossing-over between them dachs should reappear. For this reason 

 a most careful count was kept of the early stock cultures, which were 

 run by the method of inbreeding. For four generations this was 

 continued (table 126) and not a single dachs fly appeared among the 

 fhes. Besides the pair cultures recorded in table 126 (which were 

 necessary in order to avoid all danger of losing the stock by crossing- 

 over between star and the lethal), many other mass-cultures were 

 raised for the purpose of giving full opportunity for dachs to reappear. 

 These were not counted, since the composition of the parents was of 

 two sorts and the ratios correspondingly confused. Approximately 

 5,000 flies were examined, however, without finding any dachs. 



The appearance of a dachs fly would have established the linked 

 lethal view; but the non-appearance of such flies did not prove the 

 deficienc}^ view, but only that if a linked lethal were present its locus 

 was extraordinarily close to that of dachs. Such an appearance of a 

 dachs fly would be parallel to the appearance of certain Oenothera 

 "mutants," according to the application made by Muller. 



