PROBLEMS OF MEASUREMENT OF MUTATION RATES 3 



certain number of sites, a change in any one of which gives the pheno- 

 typic change being screened, then the forward and back mutation rates 

 should differ by a factor representing the number of sites. 



Of course, the fact that not all the sites have the same rate will 

 make this only an approximation, but it is probably a very good one. 



Benzer: Are you trying to give an explanation of why human rates 

 are higher? 



Atwood: Yes. One reason is that, usually, we can surmise that they 

 represent loss mutations or forward mutations; in other words, one 

 would be, in your system, looking at the rate only from r* to r, rather 

 than any of the reverse rates. What is the difference between the over- 

 all mutation rate r* to r and the average of the various r's back to r"", 

 for example? 



Benzer: There is no average reversal rate. 



Atwood: The reversion rates vary all the way from nondetectible to 

 about several per cent per growth cycle? 



Benzer: It is a range of 10^. 



Atwood: How many possible r's are there? 



Benzer: At least 500. 



Atwood: All right; that's the factor I'm looking for. 



Auerbach: Would it be possible to compare forward and reverse 

 mutation rates? The latter is much lower in Drosophila. 



Atwood: Well, I was able to find only one example of forward and 

 back mutation rate measured with respect to nutritional requirement 

 in bacteria, and that is the h — <=±h+ histidine system in which 

 Lieb(20) was able to get an estimate of the rates in both directions. 

 The difference was about a hundredfold. 



Goodgal: You're only considering one locus in one gene. Seymour 

 points out that the reversion rates vary by a factor of lO*'. 



Atwood: Does anyone wish to maintain that these factors are mean- 

 ingless in the sense that there shouldn't be any difference? 



Lederberg: I'll accept your challenge. You get a very wide disper- 

 sion of rates in both cases, but they overlap so much that it's hard to 

 know what particular conclusion you should be able to draw. 



Zamenhof: The unstable genes are not as infrequent as one might 

 think, and there is no evidence that the instability is caused by a mu- 

 tator gene in every case. A frequency of 10 ■* or 10"- is very often found, 

 if one uses the proper techniques (44) . 



Demerec: May I raise a question? How frequent are hot spots in 

 phages? Our experience with Salmonella indicates that if they occur 

 at all in this bacterium their frequency is considerably lower than that 



