72 MUTATIONS 



Motulsky: With man, you can almost generalize that every genetic 

 disease that has been investigated biochemically during the last few- 

 years has turned out to be heterogeneous. 



Auerbach: But what is the factor? It won't be more than 2 or 3. 



Neel: Professor Dobzhansky, will you summarize now? 



Summary of Discussion 



Dobzhansky: Last night, when our Chairman asked me to undertake 

 the function of summarizing, I reluctantly agreed. I certainly did 

 not know what I was getting into. I do not undertake the impossible, 

 but shall try to comment on some of the concluding sentences of 

 Atwood's presentation. The way we look at mutations, the way we 

 go about studying them, depends on the purposes of the study. I think 

 that the discussion this afternoon was revolving around this point. 



In dealing with a difficult material like the human species, and 

 being interested in the role of mutation in the maintenance of the 

 genetic load, we adopt a different definition of mutation and go 

 differently about studying mutations than when we want to find the 

 causes of mutation. 



It would be rather unfortunate to go back to the state of things 

 in 1905 and define mutation as any and all heritable changes. 

 As far as I can see, this is, how'ever, the definition which we have 

 to adopt for quite a number of purposes, especially in dealing 

 with man. Obviously, we must try to analyze what sort of heritable 

 changes we are observing in order to distinguish the chromosomal 

 changes, the cytoplasmic changes, and the changes which, for want of a 

 better explanation, we ascribe to qualitative alterations of the gene 

 structure. 



I think that our Chairman tried, unsuccessfully, in the last fifteen 

 minutes to start a discussion of the differences between major muta- 

 tions which produce clearly visible, clearly classifiable phenotypic 

 effects, and the so-called polygenic mutations. About the latter we 

 know very little, and, in many organisms, nothing at all. 



An analogy has been suggested between the study of mutation and 

 the study of such an obvious phenomenon as death. A statistician or a 

 statistical epidemiologist will try to ascertain how many people died, 

 of let us say tuberculosis, during a given time interval in a given 

 territory, to discover whether the mortality is periodic or shows no 

 regular periodicity, whether the increases and the decreases in the 

 mortality can be correlated with some climatic, economic, sociological, 

 and other factors, etc. But sooner or later we shall want to know what 



