MANNER OF PRODUCTION OF MUTATIONS 543 



nonmutagenic light (short visible and long ultraviolet) on the effective- 

 ness of ultraviolet. Several years after certain similar observations had 

 been made along these lines by Hollaender (personal communication) on 

 the survival of fungi and by Whitaker (1942) on that of Fucus spores, it 

 was noticed independently by Kelner (1949, 1950, 1952), Dulbecco (1949, 

 1950), and Novick and Szilard (1949) that Streptomyces, bacteriophage, 

 and E. coli, respectively, after having been irradiated with ultraviolet, 

 have their survival frequency considerably raised by a posttreatment 

 with longer ultraviolet or visible light, a phenomenon referred to by them 

 as "reactivation" because of its effect on the activity (survival) of the 

 organisms. Although in connection with mutation (to which it was later 

 found to apply) it might more appropriately have been termed "deactiva- 

 tion," there would be confusion in introducing so opposite an expression 

 at this time, and the phenomenon will accordingly be referred to, without 

 commitment concerning its nature, merely as "repair." Of course, this 

 term is not intended to indicate that it is the final end effect which 

 becomes repaired but that some intermediate step of the reaction chain is 

 interfered with. 



It was shown by Kelner and by Novick and Szilard that a given dose of 

 such light, in largely time-independent fashion, has the effect of prevent- 

 ing or repairing a constant proportion of a given part p (the repairable 

 part) of the potential ultraviolet damage to life. The remainder of the 

 damage q {= 1 — p) is a fixed irreversible fraction. Thus, regardless of 

 the dose of damaging ultraviolet used, a given dose of the longer-wave 

 light used as a posttreatment causes the damaging ultraviolet to act as if 

 there had been the same factorial reduction of its own dose, that is, the 

 ratio of the effective dose of damaging ultraviolet to the actual dose 

 is the same for all doses used, provided the dose of reparative light is held 

 constant. For increasing doses of the reparative light, the damage sinks 

 as though the repairable fraction of it (p) had been subjected to an 

 exponential decrease in effect. That is, if a reparative dose d reduced the 

 repairable part of the damage to a fraction dp, a reparative dose 2d would 

 reduce it to d'^p, and, as the reparative dose continued to increase, the 

 product 5"p, representing the remaining repairable damage, would 

 approach zero, so that the observed damage would approach a stable 

 value q (the part which was not repairable). In the work on E. coli the 

 repairable fraction p was about two-thirds, i.e., at least one-third of the 

 effect remained, regardless of how large a reparative dose was applied. 



This distinctive manner of change of the amount of effect with the dose 

 of reparative light shows that the original elements (activations) of the 

 reparative light act independently of one another in the prevention of the 

 damage. A further significant discovery regarding the nature of the 

 effect, made by Novick and Szilard (1949), was that the reparative process 

 is temperature dependent, proceeding more rapidly and being carried 



