M. Beeton and K. Pearson 
77 
(ii) That the expectation of life is seriously modified by either the ages of 
death of relatives or their present ages*, and can be quantitatively 
determined so soon as the latter are known. 
(iii) The expectation of widowhood must be based not only on the present 
age of the husband, or on the age of relatives, but on the age at 
death of children. 
(d) Of biological interest are : 
(i) The possible correlation of a character or organ in a member of a family 
with his or her birth order. 
(ii) The estimate of the intensity of the selective death-rate in man. 
(e) Natural selection is manifestly at work in man, and is a factor in 50 to 80 
per cent, of the deaths which occur. 
We must place here a record of our great appreciation of the generous help we 
have received from Dr Alice Lee in the verification of the numerical work on the 
tables, and from Mr Karl Tressler in the pieparation of the diagrams and the 
recalculation of the averages on which they are based. 
* Thus that a man's father died at 80 is significant for his expectation of life, but it is also 
significant, if his father is alive at 75. 
