Data for the Problem of Evolution in Man. 

 Expectation of Life. 



Here there is an increased expectation at birth for males, but very 

 small increases between the first and last periods at 20 and 25. For 

 females there is an immense increase at birth and sensible increase in 

 the other cases. Possibly a good deal of this may be due to more 

 exact returns for the ages of women being now obtainable. 



If we take the earliest Table of the Probabilities of Life, that 

 deduced by J. P., F.R.S., for the bills of mortality in London for 1728 

 to 1757, and printed in the work cited on column 5, we find the number 

 of deaths of 1000 persons born given for each year of life, male and 

 female being combined. According to the Eegistrar-General's returns 

 for 1881-90 of 1000 persons born, 728 survive to 20 and 709 to 25, 

 but from J. P.'s table only 485 survive to 20 and 448 to 25. This 

 tremendous mortality of infancy and youth was probably largely a 

 selective death-rate. We find accordingly the expectation of life at 

 birth to be only 25 -5-9 years ; at 20, however, to be 49'56 years ; and at 

 25 to be 51-30 years.* These results are for London, not for 

 England in general, but making all possible allowances for the- 

 difference between city and country, they suggest a most stringent 

 selection. An increased expectation of life at birth of anything 

 like 25 years in less than two centuries could not be achieved even 

 at the American rate of two years per generation. Nor is it possible 

 that the whole of the increase in the Registrar-General's returns for 

 expectation of life at birth for the periods 1838-54 and 1881-90 an- 

 increase of somewhere about four to five years could be due to repro- 

 ductive selection, unless we suppose the correlation between age at 

 death of minors and of their parents to be considerably greater than 

 0-1682. On the other hand, if we confined our attention to adults of 20 1 

 to 25 of both sexes, we have, roughly, an increase of about a year in 

 the expectation of life, and this result with nine months per generation 

 could easily have been reached within the two-generation period in* 



* We do not know how J. P.'s table was deduced, but we got the above 

 results by averaging the years lived by those surviving at any age out of the 1000 

 born. 



