WHAT IS TO BE DONE? 115 



children and ten normal. Both of these are in accord- 

 ance with Mendelian expectations. 



We further find that in the cases where one parent 

 was feeble-minded and the other undetermined, the 

 children were nearly all feeble-minded, from which we 

 might infer that the probabilities are great that the 

 unknown parent was also feeble-minded. 



We shall not go further into this matter in the present 

 paper, but leave the detailed study of this family from 

 the Mendelian standpoint for further consideration, 

 when we take up the large amount of data which we 

 have on three hundred other families. Enough is here 

 given to show the possibility that the Mendelian law 

 applies to human heredity. If it does, then the neces- 

 sity follows of our understanding the exact mental con- 

 dition of the ancestors of any person upon whom we 

 may propose to practice sterilization. 



From all of this the one caution follows. At best, 

 sterilization is not likely to be a final solution of this 

 problem. We may, and indeed I believe must, use it 

 as a help, as something that will contribute toward the 

 solution, until we can get segregation thoroughly es- 

 tablished. But in using it, we must realize that the 

 first necessity is the careful study of the whole subject, 

 to the end that we may know more both about the 



