The Influence of Genetic Constitution 



at 2 1 years of age or later, but that there was a far lower sib-sib 

 correlation between those dying as minors. Haldane (1949) has 

 pointed out that this is the type of correlation which would be 

 expected where the heterozygote is fitter in the Darwinian sense 

 than either homozygote: insofar as natural selection operates to 

 elimate homozygosis, not to promote it, such fitness must imply 

 a higher correlation between sibs in an equilibrium population 

 than between parent and child. In any case, the degree of 

 parent-child correlation observed by Beeton and Pearson is only 

 a quarter that between parental and filial statures in comparable 

 studies. 



Dublin and his colleagues (1949) have summarized most of 

 the historic studies on the inheritance of longevity in man. They 

 conclude that the popular idea of inheritance as a factor in 

 longevity is probably correct, that the evidence from actuarial 

 studies is heavily vitiated by all kinds of environmental influ- 

 ences, and that the order of advantage to the sons of long-lived 

 fathers is small compared with the secular increase in life-span 

 during recent generations. The difference in life expectation at 

 25 years between those with better and poorer parental lon- 

 gevity records is between 2 and 4 years — this compares with a 

 gain of 6-7 years in the general expectation of life at 25 years in 

 the U.S.A. between 1900 and 1946. Tt may be well, as has been 

 suggested, to seek advantages in longevity by being careful in 

 the choice of one's grandparents, but the method is not very 

 practicable. It is simpler and more effective to adapt the 

 environment more closely to man, (Dublin, 1949, p. 118). 



It does not follow from these considerations that longer life 

 cannot be obtained in a given population by selective breeding, 

 and in mice this has, in fact, been done (Strong and Smith, 

 1936). There may well be single-gene characters where the 

 homozygote is significantly longer-lived. 'Vigour', on the other 

 hand, which is a correlate of both longevity and fertility, and 

 hence of Darwinian fitness, is likely in most cases to be an 

 expression of heterozygosis, and one would not expect to be 

 able necessarily to produce abnormally long-lived animals by 

 inbreeding long-lived parents. 



Agricultural genetics, like natural selection, has for the most 

 part attempted to increase lifetime production averages by 



123 



