242 REPRODUCTION, PUBERTY AND LONGEVITY. 



abstracted from the table all of those persons living more than 

 eighty years and those living less than sixty years. Those living 

 more than eighty years were distributed according to birth-ranks 

 and were compared with the normal distribution of persons living 

 to this age as determined from all of these persons. From this 

 comparison it was found that when the fathers were less than 

 twenty-five years of age only forty per cent of the normal number 

 lived to be eighty years old. When the fathers were over fifty 

 years of age, one hundred and forty per cent of the normal number 

 lived to this age. In other words, the chances of the adult son of 

 a fifty-year-old father living to eighty years of age are three and 

 one-half times as good as are the chances of the son of a twenty- 

 five-year-old father. When the mark was set at sixty years of life 

 instead of eighty years, it was found that the chances of long life in 

 favor of the son of the older father were 1.6 to i. 



BIRTH-RANKS AND EXPECTANCY. 



The results of this tabulation were then put into the following 

 table of expectancy : 



Table of Expectancy by Birth-Ranks and Ages. 

 Expectancy of life at age of 



Birth-Ranks. 25 30 40 



Less than 30 39.53 35.88 29.22 



3<> to 39 40.33 36.23 29.62 



40 to 49 42.44 38.52 30.81 



50 and over 45.27 41.45 31.95 



The expectancy of this table is somewhat higher than that of 

 insurance tables, and this difference arises from two causes. First, 

 I eliminated all of those persons known to have been killed by acci- 



