290 



C. LEVINTHAL 



frequency between the three pairs of genetic markers which would have been 

 obtained if the three two-factor crosses which are possible with the markers 

 a, b, and c had been carried out. The fourth parameter measures the prob- 

 abihty that a particle which has had a recombination between two of the 

 markers also has a recombination between the other two. It is the first three 

 of these parameters, namely, the three distances, which allow one to deter- 

 mine the order of the markers along the genetic map (see Fig. 3) and the 



hfe 



ht4 



ht7 



ht, 



hfs htio 



l.oM I 2 7. 



38.3 



Fig. 3. A linkage map of the phage T2 with most of the known markers indicated; 

 rl, rl, rl3, m, and h from Hershey and Rotman (1948); MQ through htl3 from Baylor et al. 

 (1957); the c region contains at least three different mutable sites (Brenner, 1957); ii 

 (McFall and Stent, 1958) is a mutation which suppresses the action of h^, an extended 

 host-range mutant studied by Baylor et al; Si is a star-producing mutant which produces 

 slow growiih under some plating conditions (Symonds, 1958); Tg (McFall and Stent, 

 1958) is an apparent "deletion" situated between the distantly linked rl and rl; in the 

 A and B cistrons of the rll region, approximately 2,500 independently arising mutants 

 have been mapped (Benzer, 1957). 



additivity of distances along this map, while the fourth parameter is a 

 measure of the interference between two adjacent recombinations. This latter 

 point will be discussed further below. 



IV. Multiple Matings in Individual Crosses 



If a cross is made in which three different parental mutants are used to 

 infect the same bacterium, progeny particles are observed (Hershey and 

 Chase, 1951) in which all three mutations are combined in the same particle. 

 One can then conclude that if the elementary mating act involves the com- 

 bination of only two particles, repeated matings must take place in the cell, 

 and the recombinants formed in one event must be able to enter into other 

 mating events. This conclusion is supported by the fact (Doermann and 

 Hill, 1953; Visconti and Delbriick, 1953) that in a cross with one parent in 

 great excess over the other more recombinants can be observed than minority 

 parental types. If each particle could only mate once, one would expect to 

 find, at most, an equal number of recombinants and minority parents. In 

 addition to the fact that recombinants can engage in further mating, it is 

 also clear that they are able to multiply, since the frequency of recombinants 

 from a cross increases as the burst size increases with delayed time of lysis 

 (Hershey and Chase, 1951; Levinthal and Visconti, 1953; Doermann, 1953). 



