122 MUTATIONS 



Benzer: You have to test each one. 



Atwood: It is possible, but it would require doing it without screen- 

 ing, which is then almost impossible. 



Auerbach: In fact, if that is the explanation, short deletions in one 

 of the redundant segments would not even be picked up, because it 

 would still be — 



Atwood: Well, it has the part that overlaps, the nonredundant part, 

 then. 



Benzer: I'm afraid you would have to stretch the duplication idea 

 to having duplications within duplications within duplications, because 

 I can give you a set of big deletions all the same, a smaller set all 

 the same included in the first one, and a third one included in the 

 second. 



Demerec: Why not, particularly if the duplicated segments are 

 heterochromatic? There is good evidence in Drosophila work that 

 heterochromatic segments are scattered throughout the chromosomes 

 (33). One suggestion is that it may be an effect on polysaccharide, and 

 the other is just an accidental formation of the loop without any 

 duplication, and so on. You have uniformity there; they are all just 

 the same, and a large number of them indicate or suggest duplication. 



Goldstein: Some of us don't understand the connection between the 

 data you presented and the logic that leads you to talk about duplica- 

 tion. Could you explain that? 



Demerec: The evidence in support of duplication has not been 

 mentioned, and I am glad that you have raised the question because 

 we have studied it in some detail. The results are described in the last 

 Annual Report of the Department of Genetics (16), and I shall sum- 

 marize them briefly. All the long deletions of cysC are similar in length. 

 One end of each falls within a section occupying the left third of the 

 locus, and the other end is either at the extreme right of the locus or 

 beyond its right-hand boundary. It is possible to analyze the left-hand 

 section because genetic markers are available to either side of it. 

 The frequency of recombination between these markers indicates that 

 the section is of considerable length, comprising almost one-quarter of 

 the locus. Although more than a hundred single-site mutations have 

 been identified in cysC, none of them is located within that section — • 

 which thus appears to be genetically silent, as would be expected of 

 a duplication whether it consisted of genetically functional or of 

 genetically inert material. 



Stern: If this is a duplication, I would say it is at a selective dis- 

 advantage from the point of view of producing deletions constantly. 



