474 llADIATION HIOLOGY 



Novick (personal communication) has checked the poison theory with 

 the followinji; experiment : A sample of E. coli B/r was iria<liat('(l with a. 

 dose /> ol' nil i:i\i()l('t and then pholoreactivated. Ailci I his opcrnt ion the 

 sur\'i\;il corresponded to a dose Ij\{L\ < D) ot ultraviolet ji;i\('ii in t he- 

 dark. According to the poison theory, there should ha\(! been no ditTer- 

 ence between these photoreactivated bacteria and bacteria which received 

 oi-i«i;iiuUly a dose />i of ultraviolet and were photoreactivated. If now a 

 new ultraviolet dose is given to the pliotoreactivated bacteria, their 

 survival curve should be ecjual to the original survival curve in absence of 

 photoreactivation, starting at the ultraviolet dose Li. This was actually 

 found. After a second stage of photoreactivation the survival of the 

 bacteria became equivalent to a dark ultraviolet dose L2; the twice- 

 photoreactivated sample was exposed to ultraviolet for the third time, 

 and again the survival cui\e was similar to the original one, starting 

 at a dose L-i. 



5-5. .\CTI()X SPECTRUM OF Tin<: FHOTOREACTIVATING LIGHT 



Preliminary results on the shape of the action spectrum for bacterial 

 photoreactivation have been published by Kelner (1950b), who has 

 described in E. coli a peak near 4000 A, and by Knowles and Taylor 

 (1950), who determined that the greatest effect occurs for wave lengths 

 between 3500 and 4500 A, with a maximum at 3600 A. A more extensive 

 determination of the action spectra of FJ. coli B/r and Streptomijces griseus 

 spores has been published by Kelner (1951). The relative efficiency of 

 lights of different wave lengths was determined by comparing the amount 

 of light energy re(iviired to produce a standard "degree of photoreacti- 

 vation "(the fraction of inactivated cells that have been photoreactivated) 

 after a standard ultraxiolet irradiation. For values of the degree of 

 photoreactivation near the standard \'alue, the degree of photoreacti- 

 vation is a linear function of the logarithm of the light energy, and inter- 

 polation of the data is possible. In the range of light intensities used, the 

 reciprocity law was found to hold for exposures varying between 1.6 and 

 75 min for S. griseus and between 5 and 52 min for E. coli. However, long 

 exposures should be relatively less effective in E. coli owing to the decay of 

 photoreactivability with the time of sojourn of this organism at 37°(' in 

 saline (Kelner, 1949c; Novick and Szilard, 1949); this decay might be 

 masked by saturation occurring at high light intensity. This may affect 

 somewhat the shape of the action spectrum. 



The spectral region examined consisted partly of lines of the mercury 

 spectrum isolated by glass filters and partly of bands separated out of the 

 continuous spectrum of an incandescent bulb by interference filters. The 

 spectral regions of the second type are not well defined because the inter- 

 ference filters do not cut off very sharply. This introduces some uncer- 

 tainty for the points so determined. 



