ACTIVE FROM INACTIVATED BACTERIOPHAGE 119 



tive particle, once formed, should be favored in multiplication. Without 

 completely disproving it, our results make this explanation very unlikely. 

 The fact that the probability of reactivation increases greatly by increas- 

 ing (for example, from 8 to 20) the multiplicity of infection with heavily 

 irradiated particles would require a very high number of exchanges. To obtain 

 high yields of active phage in these cases we should assume, moreover, that 

 the exchanges take place early, and that the active particles, once formed, 

 multiply with little or no interference from the inactive particles in large excess. 

 This seems contradicted by the fact already mentioned that active particles 

 can actually undergo interchanges with inactive ones. Altogether, there is 

 strong evidence that inactive units have less chance than active units of enter- 

 ing the final particles, even when the active units derive from inactive particles. 



In search for a mechanism that could selectively bring together the active 

 units, the senior author (Luria 1947) suggested the hypothesis of independent 

 reproduction of individual units to form a "gene pool," from which the new 

 active particles could be derived. Inactive units were considered to be those 

 that cannot reproduce and that have, therefore, little chance of incorporation 

 into the final particles. The tendency to reduction in yield may be due to oc- 

 casional incorporation of some of the original inactive units. No hypothesis 

 is made as to how the units reproduce or reassemble. The last step is the most 

 difficult to visualize, and we incline to the belief that the original particles 

 may play a role in it, possibly by supplying a framework for reassembly. This 

 is suggested by the limited efficiency of collaboration among large groups of 

 particles, which indicates a certain tendency of the units to remain together 

 with their original companions. 



One may ask whether, inside bacteria in which reactivation does not take 

 place, the active units present in the infecting particles reproduce or not. 

 Cohen (1948) states that no desoxyribose nucleic acid is synthesized in bac- 

 teria infected with particles of T2 exposed to doses of ultraviolet light much 

 higher than those employed in our study. Preliminary cytological evidence, 

 collected with the collaboration of Dr. C. F. Robinow in our laboratory, 

 indicates that in infected bacteria, in which reactivation does not take place, 

 there is no accumulation of stainable material supposedly representing des- 

 oxyribose nucleotides. If this evidence is confirmed and found to apply to the 

 conditions of our experiments, it might then suggest, either that there is no 

 reproduction of active units when they are not all present (which might al- 

 together invalidate the hypothesis of independent reproduction), or that at 

 least part of the reproduction of the active units may take place without 

 increase in desoxyribose nucleotides. 



The hypothesis of a "gene pool," although by no means the only possible 

 one,^ fits all results of reactivation. Its validity may soon be amenable to 



^ Another possibility, suggested by Dr. A. H. Sturtevant, would be a process of zipperwise 

 replication of the various units of a phage particle. When in this process an inactive unit was 

 reached, replication could only continue if the partial replica came in contact with another phage 

 particle in which that unit was active. The process would then continue by addition of replicas 

 of the active units of the second phage particle. If repeated several times, such a mechanism 

 would provide for selective recombination of all active units. 



273 



