INTERACTION BETWEEN THE AUTOSOMES OF DROSOPHILA MELANOGASTER AS MEASURED BY VIABILITY AND RATE OF DEVELOPMENT
Females of an inbred al c: se ss: eyR mutant strain were crossed with males from a "wild" strain that had been inbred (brother × sister) for more than 60 generations. The F1 males were mated with mutant strain females. The backcross offspring would be expected to appear in eight genotypes with equal frequencies were there no differences in viability. The marker genes permitted the scoring of each fly as to whether it was heterozygous or homozygous for each of the mutant autosomes.The genes used as markers probably had a distinctly negative effect upon both viability and rate of development, but the second pair of autosomes in combination with the other mutant autosomes increased viability, when homozygous, to a greater extent than did a heterozygous mutant-wild type pair. Apparently the possible negative effects of the markers al and c were more than offset by other genes on this autosome which had positive effects. The negative effects of the homozygous mutant third and fourth chromosomes were severe when in combination.The interactions of the positive second chromosome and the negative third and fourth chromosomes in the eight genotypes were of some geometric order. It was found that the addition of a "positive" or "negative" autosome to any genotype caused a change in viability in the direction of the added autosome, but the amount of change is at present unpredictable and depends upon the particular combination to which the autosome was added.The relation between the different autosomes and the rate of development was quite different from the relation between autosomes and viability. The second chromosome (marked by al c), which gave the only positive contribution to viability, retarded development more than either the third (se ss) or the fourth (eyR). The effects of the three autosomes on rate of development were not strictly additive.It is concluded that there is interaction of the genes for quantitative characters and this interaction is geometric in nature. The significance of the results in their relation to some theories of the inheritance of quantitative characters is discussed.