PROBLEMS OF MEASUREMENT OF MUTATION RATES 59 



which are single events, such as deamination of cytosine in viral RNA, 

 and there are mutations which are composed of at least two steps, 

 because the DNA cannot reproduce with the changes which one has 

 inflicted by mutagenic agents, such as mustards. This means that the 

 final fault does not occur at the moment of reaction with mustard, 

 but this final fault, substitution of the wrong base, occurs later on at 

 the moment of cell division, or DNA replication. Thus, there must be 

 two steps for such a mutational process (43). 



Auerhach: It may be so from what you call a mutation in the 

 chemical sense, but from what I observe, I must draw the conclusion 

 that mustard gas in spermatozoa, which presumably have a double 

 helix of DNA, produces mutations, a high percentage of mutations, 

 which affect the whole offspring. These cannot have occurred at the 

 next division. 



Atwood: Maybe the complementary strand was lethal. 



Auerhach: Yes, this is a possibility. But there is, in fact, no evidence 

 for believing that mustard gas-induced mutations must occur at cell 

 division. There is a theoretical model under which this should occur, 

 but if one talks of mutations, one should really talk of observable 

 biological facts, and there are so far no facts that support this model. 



Lederberg: You never observe a mutation, either. 



Neel: Then we have arrived at the position that we don't really 

 know what we're measuring when we measure mutation rates. We 

 can't compare from one species to the next, but if we could compare, 

 the denominator that would pull all rates closer together would be cell 

 division. 



Auerhach: There are mutations without cell division. 



Zamenhof: That's what I just said. 



Novick: The events which lead to mutations — 



Auerhach: No, spontaneous mutations. 



Magni: Why don't we separate the spontaneous and induced muta- 

 tion rates? In the induced, we know what we are measuring, i.e., the 

 effect of a given treatment on the frequency of mutants. For the 

 spontaneous mutations we don't know exactly which is the frequency 

 of the primary event. 



Zamenhof: But spontaneous mutations are induced. 



Magni: Well, that has to be proved. 



Novick: That has been proved. 



Atwood: I think that we ought to discuss actual evidence on this, 

 of which there is some. Aaron, you might tell about the chlorampheni- 

 col-inhibited cells and how spontaneous mutation is occurring in 

 them. 



