3 i8 THE TREND OF THE RACE 



would be a regular average increase of age according to the order 

 of their birth. Let us consider families of four children the 

 fathers marrying at the ages of 20, 25, 30, and 35. Suppose these 

 fathers, by virtue of differences in inherited vitality, live to the 

 ages of 40, 45, 50, and 55 years, respectively. Suppose also that 

 at intervals of five years each father has a son who lived to be 

 several years older than himself. We may represent the ages of 

 the four fathers A, B, C, and D at the time of the birth of their 

 sons in the upper horizontal column and the ages of the sons 

 begotten at these respective ages immediately below. 



Averages of sons 40 42 44 46 48 50 52 



In the cases of these four families thus arbitrarily chosen the 

 sons in each family have a diminished duration of life as the age 

 of their fathers increases, but their average ages give an entirely 

 misleading indication of the relation of parental age to longevity 

 of offspring. In our table the older fathers produce the older sons, 

 but the influence of age per se is to reduce the son's expectation of 

 life. Of course, the supposition we have made is very artificial 

 and arbitrary, but it will make it clear, I think, that the data 

 which Redfield presents do not necessarily prove his case, or 

 obviate the objection which he admits might plausibly be urged 

 against his conclusions. The arbitrary assumption may be not 

 far from the truth, however, since stocks which marry early, 

 such as unskilled laborers, do not have as great longevity as 

 stocks which, like the professional classes, marry late in life. 1 



The chief thesis of Redfield's book on The Control of Heredity 



1 And it must not be forgotten that the decline in the general rate of mortality 

 tends to give the later born members of a family a greater expectation of life. 



