RADIOBIOLOGY OF BACTERIOPHAGE 365 



entire population may be thought of as a large number of subclasses within 

 any one of which the given phage marker has a different sensitivity than in 

 any other subclass. The basic philosophy of this idea is that for the most 

 part a marker is knocked out not by a hit on it but by a hit near it; a pre- 

 diction of this idea is that the survival curve for a marker should have a 

 slope which decreases continuously with increasing dose of radiation. 



Doermann and Chase (personal communication) have obtained survival 

 curves for single markers up to doses of about 800 phage-lethal hits. Such a 

 curve is shown in Fig. 6. The legend to the figure describes the technique 

 by which the data were collected; here we wish only to indicate the essence 

 of the method and point out several features of the survival curve. The 

 technique measures the probability that an irradiated phage particle carry- 

 ing a given marker will contribute the marker to at least one progeny 

 particle in a cross-reactivation experiment. Any damage which reduces this 

 probability in any way will contribute to the knockout of the marker, 

 whether this damage is genie or nongenic. Two features of the curve argue 

 for a very small upper limit to the amount of nongenic damage: (1) The 

 final slope of the curve is less than 1 % as steep as the curve for the survival 

 of infectivity; (2) the initial slope of the curve is almost the same as the 

 initial slope of the curve determined under conditions which measure only 

 genie damage (Fig. 5). 



The progressively decreasing slope of the survival curve argues strongly 

 for the recombinational nature of the marker reactivation. The significance 

 of the apparently linear portion of the marker survival curve at high dose is 

 quite in doubt. It has been variously interpreted as a measure of the mini- 

 mum distance within which the two rescuing genetic exchanges can occur, 

 a measure of the length of an ultraviolet -induced lesion (Barricelli, personal 

 communication), or a measure of the sensitivity of a target upon whose 

 survival the opportimity for cross reactivation of any part of the genome 

 depends (i.e., a measure of the amount of nongenic damage). 



Two modifications of this experiment lend support to the interpretation 

 we have put upon it. In the first, wild-type phage is irradiated, and the 

 reactivating phage is mutant for two closely linked markers. A reactivation 

 is scored whenever at least one progeny particle contains both wild-type 

 alleles. According to our interpretation, the low-dose survival curve in such 

 an experiment should be approximately the same as that found when a single 

 mutant phage is used as the reactivating parent. At high dose, however the 

 survival curve should be steeper since any hit occurring between the pair of 

 wild-type alleles would be expected to make reactivation of both markers 

 into one midamaged vegetative phage essentially impossible. Furthermore, 

 the slope should increase with increasing distance between the genetic 

 markers employed. In the second experiment, the irradiated phage is mutant 



