ACTIVE FROM INACTIVATED BACTERIOPHAGE 113 



increasing multiplicities is less evident for phages TZ than for T6, and still 

 less for T4, diminishing in the same order as the calculated number of units. 



To summarize, the results show that the probability of production of active 

 phage from inactive particles depends on the dose of radiation and on the 

 multiplicity of infection in a way similar to the one predicted by the simple 

 theory, which assumes lethal mutations in discrete transferable units of equal 

 radiation sensitivity and a hundred percent efficient recombination of active 

 units to reconstitute active particles. All deviations can be accounted for by a 

 limitation in the efficiency of recombination when the active units derive from 

 more than two inactive particles. 



Several explanations may be offered for this limitation; some of them have 

 been tested experimentally. Lysis from without — failure to liberate phage due 

 to excessive multiplicity of infection (Delbruck 1940) — was found not to 

 take place for multiplicities of the order of those for which the deviations from 

 theory occur. Limitations in the number of particles of a given phage that can 

 participate in phage growth were looked for and found (see Dulbecco 1949a), 

 but their magnitude cannot account for the differences between w and y. 



The process of reactivation must involve complex mechanisms of transfer 

 of genetic material among phage particles. Whatever these mechanisms, it 

 is reasonable to expect that they will work less efficiently as the number of 

 phage particles increases. Limitations may conceivably be caused by steric 

 reasons — shape of the particles, position in the bacterium — or by physiological 

 reasons — limited number of units of some catalyst, competition for substrates. 



Estimation of the number of units 



We have compared our results with the theoretical curves for y calculated 

 for different values of n, the unknown number of transferable units per particle. 

 The curves that best fit the results for phages TZ and TZr are those for « = 25; 

 for T6 and T6r, n = 30; for T4 and T4r, w=15. For phage T5, no accurate 

 estimate of n was obtained, but n appears to be lower than for T4. 



If the interpretation of the results based on the simple theory is justified, 

 we must consider the values thus obtained for n as minimum estimates of the 

 number of radiation-sensitive, transferable units per phage particle. The 

 estimates are minima because of the assumption of equal sensitivity of all 

 units. Should there be units more sensitive than others, they would be hit more 

 often, and in order to obtain the correct probability of reactivation we should 

 assume more units of the less sensitive type. For example, if one unit were 

 twice as sensitive as the average of the others, one locus of the average sensi- 

 tivity should be added to our estimate in order to distribute the probability 

 of inactivation over all units in such a way that the reactivation probability 

 remains the same. 



In our preliminary report (Luria 1947) we calculated n in a different man- 

 ner, by assuming that for low multiplicities and low doses, where the probabil- 

 ity of reactivation appeared to be approximately constant as a function of x, we 

 could consider all multiple-infected bacteria as double-infected. This corre- 



286 



