U4 KADI \I'I()N HI()L(»(;V 



lined lo wave lenfi;ths shortci- ilcm iMIOO A, the pliotodccoiupo.sition of 

 peroxide occurs throujilioiit the ultruviolet spcictrum. 



All these considerations plus the selective absorption of ultraviolet Wy 

 certain protoplasmic constituents indi<'ate that only a very minor portion 

 of the effects of direct ultra\iolet irradiation of bacterial cells by the wave 

 lengths usually employed can be ascribed to an indirect mechanism which 

 has as its first step the formation of hydrogen peroxide from water. 



A more conservative estimate of the significance of these results is that 

 treatment of nutrient broth or solutions of certain amino acids with 

 hydrogen peroxide or with ultraviolet of wave lengths shorter than 2000 A 

 can produce chemical mutagens. Such irradiation of so complex a sub- 

 strate as nutrient broth would be followed by a variety of chemical 

 changes including formation of hydrogen peroxide and organic peroxides. 

 Since both hydrogen peroxide and certain organic peroxides have been 

 shown to be mutagenic (Dickey et al., 1949; Wagner et al., 19o0; Demerec, 

 Bertani, and Flint, 1951), it is not surprising that broth irradiated under 

 such conditions is also mutagenic. 



MECHANISM OF RADIATION EFFECTS 



Various hypotheses have been proposed to explain the bactericidal 

 effects of radiations. In view of the recent evidence that a large propor- 

 tion of the effects of both ionizing and ultraviolet radiation are produced 

 by indirect mechanisms, many of the discussions are no longer appro- 

 priate, and no attempt will be made to reconsider them completely. 



Such discussions have generally been concerned with two central prob- 

 lems: (1) how best to account for the kinetics observed, and (2) the nature 

 of the damage and the mechanism by which it is produced. 



In regard to the former problem, several waiters (e.g.. Gates, 1929a) 

 have explained the occurrence of the survival curves observed on the basis 

 of variation in resistance of the individual cells comprising the bacterial 

 population. Although the occurrence of the sigmoidal survival curves 

 can l)e explained on this basis with not too violently skewed distributions, 

 the distribution recjuired to explain an exponential survival curve, namely, 

 an exponential distribution of resistances, is implausible. By contrast, 

 the first-order kinetics indicated by exponential survival curves are a 

 natural consequence of the target theory in its simplest form. Sigmoidal 

 survival curves are easily accounted foi' on the basis of the multihit 

 theory. Consequently, very few investigators still hold to the distri- 

 bution theory. 



In regard to the latter prol)lem, a number of explanations have been 

 proposed which are discussed by Lea et al. (1936). These may be broadly 

 grouped into two categories: (1) secondary poisoning of the cell by 

 chemical substances produced by the radiation and (2) decomposition of 

 molecules vital to the organism. Lea and cow^orkers reject the poison 



