O NATURE AND NURTURE 



been called as witnesses because they represented opposite 

 views or because they had a popular reputation as 

 authorities. Should we expect to learn the truth about 

 the influence of parental alcohohsm on the offspring by 

 taking the evidence of a brewer, a pubHcan, the Secretary 

 of the United Kingdom AUiance, and a fashionable 

 medical consultant who has had to deal with extreme 

 instances in a special social class ? Is it not rather 

 a case where we must use the careful methods of exact 

 science ? Hundreds of homes must be visited, hundreds 

 of parents, alcohoUc and non-alcohoHc, must be observed 

 and reported oq ; more than hundreds of children must 

 be measured and weighed, their physique and their 

 mentality must be recorded by medical officer and teacher, 

 and then, what ? Shall we know the answer to our 

 problem ? No, certainly not. We shall have the data 

 from which an answer may be extracted, when and when 

 only we use all the caution and refinement possible to 

 modern statistical methods. To emphasize what I mean 

 by this caution and refinement let me point out some of 

 the difficulties involved in this very problem of alcoholism. 

 Let us suppose that we have found out that the child 

 death-rate of alcoholic mothers is higher than that of 

 sober mothers — have we any right to infer that alcoholism 

 leads to many child-deaths ? It seems straightforward 

 enough to make the inference, but yet note how terribty 

 complex these problems are ! We find that alcoholism 

 of the mother is even more intimately associated with 

 her employment than it is with the death-rate of the 

 child. Is it possible, therefore, that employment of 

 mothers, and not alcohol, is the source of this increased 

 death-rate ? No, we cannot stop there. The employ- 

 ment of women is associated with an ' unhealthy ' trade 



