PROBLEMS OF MEASUREMENT OF MUTATION RATES 43 



Cotterman: Did you find three kinds of cells that you could classify 

 as Ai, A2, and O? 



Atwood: In these individuals, no. We didn't use any agglutinin ex- 

 cept human anti-A, which will agglutinate both the Ao and the Ai. 



Cotterman: You would say that almost all the cells are Ai or A2 and 

 only a small fraction are O? 



Atwood: Yes, that's right. 



Cotterman: But you can't say what fraction of the cells among the 

 A's are phenotype Ai or Ao? You didn't test them? 



Atwood: No, we didn't have enough blood. 



Well, now, finally, I have this discouraging news. Among these 

 families with both parents AB, we had two B homozygotes, and so 

 were able to compare homozygotes and heterozygotes with respect to 

 frequency of non-B. 



The first instance here in Figure 12 again demonstrates absence of 

 age effect. This experiment is technically not so good as the other 

 because you are still removing some cells at the end of the experiment, 

 but it is approximately completed. The parents have levels of about 

 1.5 X 10"^ of non-B cells, and one child has about 10"^ of non-B cells. 

 He is the AB child. The B homozygote has 2 X 10"^. We have never 

 found an AB that had that few non-B cells. Even so, the observed fre- 

 quency is much too high to represent the product of the frequencies of 

 losses of the individual B alleles. The expected frequency is around 10"^ 

 on the assumption of independent behavior of the alleles. 



Lederberg: What is the product of the non-0 there? 



Atwood: We didn't have an opportunity to continue these experi- 

 ments with Ulex and carrier. 



Lederberg: What was the value with Ulex in the previous blood? 



Atwood: In the previous (AB) blood, the Ulex following anti-A and 

 anti-B removed all but about 10"^. 



Lederberg: Another factor of 10, then? 



Atwood: Yes, or nearly so. 



In Figure 13 the parents are both AiB. Their non-B frequencies are 

 different; one has about 10^ and the other about 2 X 10~^. One of the 

 children has about 10^ and the other about 2 X 10"^. You might think, 

 therefore, that the non-B frequency is a part of the phenotype of the 

 given B allele, as though this one got the B allele from this parent, 

 and this one got the B allele from that parent. I surmise that this is 

 just chance, however. 



The non-A frequencies were also done on these; they are the same 

 for all four of the individuals, 10"^. 



