218 PROBLEMS OF HUMAN EVOLUTION 



Natural Selection. In order to gauge it with approximate 

 correctness, we must find out the lowest age at which marriage 

 commonly takes place and the percentage of the population who 

 die before they attain this age. For all those who leave no off- 

 spring we may, for our present purpose, look upon as having 

 been eliminated ; they are of no account in the physical evolu- 

 tion of the race. Now in 1896, of the bachelors who married 

 in England less than 6 per cent, were under 21. Women 

 marry younger, but of the spinsters who married, less than 

 10 per cent, were minors. Turning now to the deaths 1 of the 

 same year I find that 47-9 per cent, of the males and 44-5 per 

 cent, of the females were under 2 1. 2 In the male sex we may 

 hold that quite 47 per cent, die before the marriage age since 

 the number of men who marry under 21 is so small that it is 

 practically a negligible quantity. Of the spinsters who became 

 wives in the same year 1 8*8 per cent, were minors, a far larger 

 proportion. But if only 1 8*8 per cent, of the spinster brides are 

 under 21, obviously quite an insignificant percentage of the 

 whole number of females born marry before that age. More- 

 over, of this insignificant percentage a large proportion live to be 

 much older and consequently are not to be accounted in the 

 number of those who die under 21. Now if we take the year 

 1896 as typical (as we fairly may) these amount, as I have said, 

 to no less than 44*5 per cent, of the females born. Nearly all 

 this high percentage, therefore, we may reckon among the 

 eliminated. 



But it is possible to quote figures that will show the facts in a 

 still more striking light. The mean age at which bachelors 

 married in 1896 was 26*59 5 * n t ^ ie case or " spinsters it was 25*08. 

 For brides and bridegrooms together, therefore, it was 25-83. 

 Now the percentage of the total population who die under 25, 

 if we may judge by the returns for 1896, is 48*2. These figures 



1 See the Registrar General's Report for 1896, p. 1 18. For ages of marriage see 

 pp. 13, 14. 



2 To find this number has required some calculation as the Registrar General's returns 

 do not give the deaths in each age-year but in each group of five years. To find the 

 number of those who died in their zist year, I have taken one-fifth of those who died 

 from 20-25. The death-rate at this age is low and the possibility of error is small. 



