20 MUTATIONS 



well as A2 and, to a lesser extent, some others. It is an extract of Ulex 

 seeds. Then, we have human anti-A. In the experiments I am about to 

 show human anti-B was one reagent, lima bean anti-A another, and 

 finally the Ulex anti-0. 



These experiments were designed to test the independence of the 

 losses of A and B in an AB individual. We thought that these might 

 be independent losses, because we were able to isolate non-A cells in 

 sufficient quantity from AB individuals to show that, as a population, 

 those non-A cells were phenotypically B. But we didn't know the exact 

 proportion of B to non-B in the isolated populations. The experiments 

 now are supposed to show what are the proportions of non-A, non-B, 

 and actually of non-0 in the same AB individual. 



Neel: Of non-0? 



Atwood: Yes, there are some non-0 cells; that is, they are Ulex- 

 negative, at least. 



The expectation was that in an AB individual, the product of the 

 non-A by the non-B frequencies should give the frequency of 0. This 

 is the expectation if the cells are formed by independent mutation of 

 the A and B alleles. 



Auerbach: That would be a means for ruling out somatic crossing 

 over as a mechanism, then? 



Atwood: Somatic crossing over would not affect that expectation in 

 the right direction at all, because what happens with somatic crossing 

 over is that the AB cells are changed into homozygous B and homo- 

 zygous A cells, which contribute to the non-A and to the non-B fre- 

 quencies, respectively, and so we get a higher value for their product 

 without having formed any O cells. The AB, A A and BB subpopula- 

 tions could create heterozygotes by mutation which again, through 

 somatic recombination, could produce homozygous A's and O's, and so 

 on. As you will see, I have not yet been able to explain the quantitative 

 results by any such scheme. 



Auerbach: What I meant was, if the whole effect was due to somatic 

 crossing over, you shouldn't get any cells. 



Atwood: Oh, yes, then the population would move toward an even 

 mixture of A and B. 



Lederberg: If this individual were homozygous for the Bombay 

 factor, would you have segregation for O's and A's and B's? 



Atwood: Well, I'll give you the result and we can talk about the 

 Bombay factor later. 



These experiments are rather difficult to do, which is why I have 

 so few. This experiment (Fig. 4) was done twice, once using the anti-A 



