Abstract
The variability of recombination frequencies between alleles of the ftr cistron is assessed and shown to be normally in the range of ± 25% though occasionally replicates differed by a factor of two and in rare cases even more extremely. It is shown that the variation results from biological differences between separate attempts at the same cross and cannot be ascribed to sampling error or any other controllable cause. In the light of this variability the marker effect alleles already recognized in this cistron are re-examined with a new series of crosses. It is confirmed that marker effect may act to enhance or reduce the recombination frequency but it is shown that enhancement can be expressed in two entirely different ways. There is one type of site which causes a uniform doubling of the recombination frequency in all crosses no matter where the second site is placed in the map. The second type of marker effect enhancement site causes a very great increase in recombination frequency in crosses against relatively closely linked mutants, but its effect diminishes as the distance between the marker effect site and the second site is increased.