LINKAGE 



of lethality of a semilethal ' factor differs 

 widely under diverse environmental conditions 

 and in different stocks. This has been shown 

 in Habrobracon for a number of visible mutant 

 types of low viability. The mutant genes are 

 here serailethals which have a recognizable ef- 

 fect on the surviving individuals. Factors or- 

 dinarily designated as semilethals presumably 

 differ from these only because of failure to 

 find a conveniently visible trait difference 

 (Schaeffer, 1945). In Habrobracon, with its 

 haploid males, stocks very quickly become pure 

 for modifying factors by lethal selection. It 

 seems likely that a balance of factors should 

 thus be readily attained tending toward greater 

 viability. Relatively disharmonic combinations 

 may be expected to enhance by their cumulative 

 effect the influence of a semilethal among the 

 progeny of hybrid females. A larger proportion 

 of the La/Lb daughters than of the La/Lb sisters 

 of the hybrid mother might then be inviable. 

 In five tests involving reciprocal crosses of 

 two orange-eyed stocks and in one test of one 

 of these stocks having 'the loosely sex-linked 

 gene stubby with a third stock, lethal effects 

 v/ere present associated with the dominant allele 

 to stubby. The lethal effect is regarded as 

 complementary. It is probable that this lethal 

 factor lies to the right of stubby (Schaeffer, 

 1945). Another lethal factor (Helsel, 1942) 

 showing 25 per cent recombination with white 

 and pellucid has been detected. 



In the case of linkage between a and b, 

 AB/(AB + aB) might be expected to equal ab/(Ab 

 + ab) were it not for one or more of the three 

 disturbing factors--somatic overlapping, non- 



115 



