62 A. D. HERSHEY AND RAQUEL ROTMAN 



slightly exaggerated owing to the fact that the 54 tubes contain on the average 

 mixed yields of virus from about 1.4 bacteria per tube. 



THE EFFECT OF A SHORT PERIOD OF ADSORPTION AND LOW MULTIPLICITY 

 OF INFECTION ON THE DISTRIBUTION OF RECOMBINANTS 



The following experiment shows that when the multiplicity of infection in 

 the cross hXr7 is reduced from five to about one of each viral type per bac- 

 terium, the distribution of recombinants among single mixed bursts is little if 

 any altered. 



Six crosses were made in the usual way, except that the period allowed for 

 adsorption was reduced to one minute, without reducing the total input of 

 virus. The amount of virus adsorbed was too small to be measured, but the 

 multiplicity of infection can be estimated from the data given below. 



The single burst cultures collected from the six crosses are sufficiently simi- 

 lar to be considered together. The mean number of bacteria per tube for the 

 six sets, determined by colony counts from the growth tubes immediately 

 before adding virus, is 0.23. The mean number of infected bacteria per tube 

 determined by plaque counts of samples taken before lysis is 0.20. The mean 

 number of bursts per tube calculated from the proportion, 141 out of 720, of 

 tubes containing virus is 0.22. The 141 tubes therefore contained about 157 

 infected bacteria. 



From the distribution of viral types among the tubes, namely, 69 containing 

 h only, 22 containing r only, 50 containing both, and 579 containing neither, 

 one finds the probable distribution with respect to bacteria to be 80 infected 

 with h only, 28 with r only, 49 both, and 46 neither. The multiplicity of in- 

 fection is therefore about 1.0 with respect to h, and 0.48 with respect to r. One 

 can estimate further that about seven of the tubes contained one or more 

 mixed bursts plus one or more unmixed; and that about three contained both 

 h and r bursts without any mixed bursts. Also, among the mixedly infected 

 bacteria, 45 percent were infected with one particle only of each viral type. 



In making the above computations we have neglected the probability that 

 the h mutant suppresses the growth of r in some bacteria adsorbing both 

 types of virus. The apparent inequality of infection is probably due in some 

 part to this effect. However, in other experiments with low multiplicity of in- 

 fection with h and r7 designed to check this point, the split into h, r, and mixed 

 yielders was nearly equal. It seems likely, therefore, that in the experiments 

 reported here the two viral types were unequally adsorbed for unknown 

 reasons. 



The tubes containing only one viral type may be dismissed by saying that 

 their average content of virus did not differ significantly from that of the 

 mixed yields, and that the h and r yields per bacterium were the same. There 

 was one exceptional burst containing only h r + and h r phenotypes. The char- 

 acteristics of the remaining 49 cultures containing h and r virus are summarized 

 in table 8. The data show no unusual features excepting the small burst size, 

 which is a direct effect of the low multiplicity of infection, and the somewhat 



256 



