146 HEREDITY IN RELATION TO EUGENICS 



not have affected offspring — but this is just what is alleged 

 commonly to happen. These cases, however, deserve care- 

 ful study. Frequently when both parents of the defective 

 child are normal one of them will belong to a fraternity 

 with the defect; occasionally, however, one must go back 

 to the second ancestral generation to find an affected rela- 

 tive. No eugenic instruction is, as yet possible. Corre- 

 spondence from affected persons, or their relatives, who will 

 volunteer to cooperate in studying the method of inherit- 

 ance of this trait is solicited. 



31. Cancer and Txjmor 



The question of *' inheritance of cancer" has been much 

 discussed and nothing but difference of opinion has r esulted. 

 This is largely due to the bad formulation of the problem. 

 In the first place, if, as seems probable, the stimulus to 

 cancer growth is an inoculable something — germ or fer- 

 ment — it does not follow that the consequence of .stimulus 

 is not determined by an inheritable factor. It is known 

 that certain strains or families of mice are uninoculable 

 while others will acquire cancer upon inoculation. The 

 question is, are there human strains that are easily and 

 others with difficulty inoculable? The whole question is 

 complicated by the fact that cancer is a disease of middle 

 or later life. Thus in the census for 1900 we find that the 

 heavy incidence of deaths from cancer occurs between 40. 

 and 80 years (84.4%). The detailed distribution is shown 

 in Table X. Here we see that the death rate of cancer 

 (as compared with deaths from all causes) reaches its high- 

 est point at between 50 and 60 years, but that absolutely 

 more deaths occur from that disease between 60 and 70 

 years. On account of this heavy mortality late in life many 

 who are inoculable never reveal the fact, owing to their 

 death before the cancer age. If cancer is communicable, 



