1 6 THE CAVENDISH LECTURE 



Registrar- GcfteraPs Life Tables. 



Now I do not assert that the rising infantile death-rate is the 

 ^:ause of the falling child death-rate. All that I ask you to note is 

 that three out of these important tables indicate that when the infan- 

 tile death-rate is high the child death-rate is low. No such general 

 law as that stated by Dr. Newsholme really holds, even when 

 no correction has been made for differential environment. 



But, because I state that the infantile death-rate is selective, 

 and assert that it by no means follows that a low infantile death- 

 rate will compensate racially for a falling birth-rate, why should I 

 be described as a Herod, and those who hold the same views as 

 supporters of the " better-dead " doctrine ? I feel sure that many 

 of you who have, by your skill, helped into the world the cripple, 

 or the child of deformed or diseased parents, must have said to 

 yourselves, when you found it viable, better it had not been born. 

 Many of you, I take it, hold with me the "better-not-born" doctrine, 

 but the recognition of the fact that the infantile death-rate is 

 selective cannot of itself justify the charge that we wish the 

 weaklings killed off. 



Nevertheless, medical science has to face the fact that the 

 upward progress of man in the past has been largely controlled by 

 a stringent Darwinian selection. We shall gain nothing for 

 racial efficiency by neglecting that central fact of human develop- 

 ment. 



Now if there be — and I, for one, think that two independent 

 lines of inquiry demonstrate that there is — a fairly stringent 

 selection of the weaker individuals by the mortality of infancy and 

 childhood, what will happen, if by increased medical skill and by 

 increased state support and private charity, we enable the weaklings 

 to survive and to propagate their kind ? Why, undoubtedly we 

 shall have a weaker race. Remember that " general health " is 

 undoubtedly inherited (see Figs. 3-5). Now couple this partial 

 suspension of the stringency of natural selection with a reduced 



