Six of eight inbred lines, segregating for the same B complex alleles, permitted a comparison of genetic backgrounds on resistance to Bryan high titre Rous sarcoma virus (RSV) of subgroups A, B and C. Tests were based both on wing web challenge (WWC), using 682 five-week-old chicks, and on the chorioallantoic membranes (CAMs) of 606 twelve-day-old embryos (CAM test). Absence of tumour development on WWC and of pock formation on CAMs indicated cellular resistance to RSV (primary mechanism of genetic resistance). The differential response of tumours (regressive or progressive tumour growth in WWC) measured the secondary mechanism of genetic resistance.

The results verified previously reported studies that the tv loci for cellular resistance map outside the B histocompatibility complex, and that variation in RSV-induced tumour regression was mostly but not entirely linked to the B complex. Although cellular resistance and tumour regression are not wholly independent events, the data generally support the dual mechanism hypothesis of resistance.

The ratios of variance between lines with the same B serotype, σ2L, to that between serotypes with common inbred line background, σ2B, demonstrated that σ5L were 3.2 and 17.3-fold larger than σ2B, respectively, to the CAM and WWC tests of cellular resistance (tv genes). In contrast, σ2B was 5-fold larger than σ2L for tumour regression (rs genes).

The results further indicate that segregation at both the tv and rs loci is greater than expected within these inbred lines in view of their high coefficients of inbreeding. Theoretical expectations show that for eight inbred lines, each with an inbreeding coefficient of 80% assuming initial gene frequencies of .5, the probability of complete gene fixation at the same tv locus is .43 and the probability that only one line is segregating is .38.

Finally, consideration is given to the heritability of cellular resistance and tumour regression when viewed as quantitative traits in a selection programme. As expected, the heritable fraction of variance is substantial even when the frequency of an unfavourable dominant allele, as tv, is low.