.4. D. HERSHEY AND MARTHA CHASE 



whether we were really measuring this proportion. Two possible sources of error, 

 namely, accidental overlaps of two plaques, and clumps containing two or more 

 phage particles, were excluded by the following experiment. Plates showing not 

 more than 50 plaques (as opposed to about 100 in other experiments) were pre- 

 pared from a population of phage from the cross hrl X wild-type, of which 90 per 

 cent had first been neutralized by antiserum. The antiserum treatment should 

 have eliminated any clumps of phage, and the small number of plaques per plate 

 should have eliminated overlaps. Eighty-three mottled plaques were sampled 

 from these plates, of which five proved to contain both parental types of phage. 

 This is the same proportion foimd in 170 plaques examined in other experiments. 

 The estimate of six per cent shown in Table 2 is therefore correct. 



The data of Table 2, together with the estimated total frequency (2%) of r-r+ 

 heterozygotes, measure the frequencies of three classes of heterozygotes among 

 the progeny of crosses between h and r markers. These frequencies are, for the 

 cross hr7 X wild-type, 0.12 per cent hr-wild; 0.94 per cent h-hr; and 0.94 per cent 

 r-wild; expressed in round numbers. The corresponding frequencies for the cross 

 hrl 3 X wild-type are 1.48, 0.26, and 0.26 per cent, respectively. What other 

 heterozygotes might we expect to find among these progeny? 



Two possible classes remam to be looked for; namely, the classes segregating 

 into the pairs hr, r; and h, wild; coming from particles heterozygous for h but not 

 for r. If the distribution of heterozygotes is symmetrical, that is, if the total 

 frequency of heterozygosis for h is two per cent, and if the two undetected classes 

 are of equal size, their individual frequencies would be 0.94 per cent for the cross 

 hr7 X wild-type, and 0.26 per cent for the cross hrlS X wild-type. 



One of the midetected classes can be efficiently measured by sampling clear r+ 

 plaques from platings of the progeny of the cross on mixed indicator, and retest- 



TABLE 3. THE FREQUENCY OF h-wild HETEROZYGOTES PRODUCED IN 

 CROSSES BETWEEN hr AND WILD-TYPE 



ing to determine how many of the samples contain mixtures of h and wild-type 

 phage, and how many contam h only. The proportion of clear r+ plaques that 

 yield mixtures, multiplied by the proportion of clear r+ plaque-formers among the 

 progeny of the cross, gives the frequency of h-wild heterozygotes in the popu- 

 lation. 



The results of this measurement for two crosses are compared with the expecta- 

 tion for symmetrical distributions of heterozygotes in Table 3. The findings are 

 similar to those already described for r heterozygotes, namely: 



186 



