MUTAGENESIS 135 



coli are much more sensitive to delay of mitosis by UV than ah'eady es- 

 tablished prototrophs. Similarly, Kaudewitz (37) found in E. coli that 

 immediately after the induction of lysine reversions by UV the mutant 

 cells are exceedingly sensitive to very small changes in temperature. 



Quite recently, Luzzati and his collaborators (49) reported an 

 interesting case of this kind. They scored "gene conversions" or what- 

 ever you wish to call it in yeast cells that were heterozygous for two 

 alleles of a gene causing a double requirement for adenine and histidine. 

 In these strains, wild-type segregants arose through some kind of 

 nonreciprocal recombination. The occurrence of such recombinants 

 could be suppressed by histidine in the medium. Yet, in reconstruction 

 experiments histidine did not discriminate against prototrophs. 



In all these cases, an incipient mutation is affected by conditions 

 that do not affect the established mutant. Any differences in the de- 

 gree to which this happens when different mutagens are used and 

 different mutations are scored would result in observable mutagen 

 specificity. 



Even in the ordinary reconstruction experiments, in which one tests 

 for possible selection for or against established mutants, specificity may 

 arise in response to plating conditions. You have to remember that 

 plating density on the mutation plates is very high and one may 

 obtain what is now usually called a Grigg effect, that is, a suppression 

 of the prototrophs by the residual growth of the vast number of 

 auxotrophs. Several years ago, Stevens and Mylroie (68) found 

 interesting differences between strains of Neurospora in this respect. 

 On plates with minimal medium, a large inoculum of leucineless spores 

 inhibited the growth of a small admixture of prototrophs; an equally 

 large inoculum of inositolless spores did not do so, presumably be- 

 cause there was less residual growth. In a doubly requiring leucineless 

 inositolless strain, one might then expect suppression of inositol- 

 revertants by residual growth of leucineless spores on minimal medium, 

 but not suppression of leucine-revertants by residual growth of 

 inositolless spores. Hollaender and his co-workers (92) found that 

 after ultraviolet radiation there is a Grigg effect for mutations to 

 purine independence, but not for mutations to streptomycin independ- 

 ence. 



There remains one possible source of mutagen specificity that, to 

 my mind, is the most difficult to detect. This is that the mutagenic effect 

 of a treatment is measured by the ratio of mutants to survivors and 

 that the numerator and denominator in this fraction are obtained under 

 completely different conditions. I have already discussed this when I 



