296 C. LEVINTHAL 



mean number of recombinants of a given type produced by a single cell was 

 considerably greater than unity. This finding seems to be at variance with 

 the idea of a very large loss in the sampling and maturation of the mating 

 particles. 



Bresch (1955), in a similar experiment with the phage Tl, has been able 

 to supply what seems to be a definitive answer to this question. He used 

 three genetic markers and examined each single burst for all the possible 

 recombinants. Since the number of mating events per bacterium in Tl is 

 very much smaller than that in T2, it is possible to test the correlation under 

 conditions where many of the single bursts produce no recombinant particles 

 of a particular type. A small correlation was observed between reciprocal 

 recombinants, but the same small correlation was also observed between 

 pairs of recombinants which should have no relationship to each 

 other on the basis of crossing over. Thus, the results suggest that in some 

 bacteria conditions are such that the probability of mating is higher than 

 usual, and the frequency of all recombinants is therefore increased. There is 

 no special correlation between the two recombinants that would have been 

 expected had they arisen in a single crossing-over event. The data for Tl 

 showed even more clearly than those for T2 the difficulty of explaining the 

 lack of an observed correlation by a random loss of all phage particles before 

 they were able to produce plaques on indicator bacteria. If a single particle 

 of a given recombinant tjrpe appeared on a petri dish, there was a high prob- 

 abihty of another particle of the same type being found in the same burst, 

 whereas the probability of finding a particle of the reciprocal type was con- 

 siderably lower. Were it the random loss of particles which accounted for the 

 loss of an observable correlation, then one would expect to lose the self- 

 correlation as well as the reciprocal correlation. Thus it seems likely that the 

 mating event which forms recombinants in phage is not identical to the 

 process of crossing over which occurs in higher organisms. 



B. Heterozygotes 



The second phenomenon, which suggests a difference between the recom- 

 bination event in phage and that in the more classic systems, was reported 

 by Hershey and Chase (1951). They showed that a bacterium mixedly infected 

 with phage particles differing by one genetic marker would produce about 

 two particles in a hundred which behave in a very unusual manner. These 

 particles, called partial heterozygotes, were found in the same frequency for 

 each of the six genetic markers tested; they were characterized by the fact 

 that they produced two different kinds of offspring. The only way these 

 abnormal particles can be detected easily is by means of the mottled plaques 

 they produce when the two kinds of offspring differ with respect to the r 

 character. This mottling, which is due to the presence of both the r phage 



