76 Evolution of Intdligence 



erly classified, age at death may be used to deter- 

 mine relative longevity as distinguished from 

 actual longevity. 



It is not possible to explain here all of the 

 steps taken to eliminate errors in an investiga- 

 tion which is bristling with potential errors, but 

 a few points will be mentioned. The material 

 taken consisted of large families of brothers and 

 sisters who lived into the reproductive age be- 

 yond twenty-five. The average family consisted 

 of a little more than six persons, and the total 

 number of persons was 1,105. These persons 

 were made into a composite family in which there 

 was a direct comparison between the age of the 

 father when the child was bom, and the age of 

 the child at its death. All known cases of vio- 

 lent deaths were eliminated, and care was taken 

 to eliminate as much as possible all families of a 

 distinctly pathological character. Also to get a 

 line on variations in health of parents. The final 

 result may be set forth as follows : 



As long as parents retain their health and 

 strength, the older they are when their children 

 are bom, the greater will be the natural longevity 

 of those children. The figures showed that each 



