A critical test of a current theory of genetic recombination in bacteriophage.

IGH Negative Interference: In an analysis of closely linked markers of bacteriophage T4, CHASE and DOERMANN (1958) found a phenomenon which they called “high negative interference” (HNI) : the frequency of genetic recombinants whose formation requires multiple exchanges is much greater than would be predicted if exchanges in neighboring regions were statistically independent. This is not a marginal effect. For three-factor biparental crosses in which the outermost markers are less than two map units apart, the fraction of double recombinants exceeds the random expectation by a factor of 14-31. For fourfactor crosses the factor is still greater. EDGAR and STEINBERG (1958) concluded that excess double recombinants due to HNI were produced in a single mating event. This conclusion was based principally upon the results of a series of biparental, three-factor crosses in which the relative frequency of the two parents was systematically varied. Using the type of reasoning common in chemical kinetics, it was demonstrated that the excess recombinants derived their three selected markers from only two parents. It is also possible to design triparental, thre-factor crosses so that the selected double recombinant must derive its three markers from three parents. In this case, there is little, if any, negative interference (STEINBERG and EDGAR 1961). Possible explanations for HNI: Since HNI is such a striking phenomenon, we will center our discussion of recombination over short distances around the question: how can we account for HNI? We can conceive of three general directions in which an answer can be sought. The first possibility might be called The Trivial Hypothesis. By this we mean an explanation which has little to do with the mechanism of recombination per se. As examples, we might mention technical artifacts-that the experiments do not really measure what we think they measure-and various types of statistical inhomogeneities. Inhomogeneity of “mating experience” can satisfactorily account for the small degree of negative interference found for crosses involving loosely linked markers. It is easy to imagine more insidious types of inhomogeneities which would be important only when relatively rare events are being con-