18 ETHEL M. ELDERTON 



points of doubt and difficulty ; for this reason we have not yet modified the general 

 average of the First Series by including these higher results. But it is conceivable 

 that we may have to raise the general measure of resemblance of cousins from "28 to 

 •33, when other large series already observed have been tabulated and reduced. 



7. One further point may be finally touched upon, namely the inheritance of 

 disease. We cannot in the least hope here for accurate numerical estimates, but the 

 data of our first series may suffice to show that cousins are of value even from the 

 standpoint of medical diagnosis. The difficulties of accurate determination are as 

 follows : 



((() While on the schedules the record of brothers and sisters, of children, of parents 

 and of grandparents is fairly complete, that of cousins must necessarily be defective. It 

 is quite possible- — nay not infrequent— to have more than 50 first cousins. And while 

 one or two recorders actually were patient enough to enter details of a cousinship 

 as large or even larger than this, the bulk of recorders contented themselves with 

 entering a much more limited number, 10 to 12, and thus we have the first limitation; 

 our cousins, as the recorders themselves state, are a selection. It is probable, also, that 

 the selection has been made more frequently of living than of dead cousins, and more 

 frequently of accessible than possibly inaccessible cousins ; thus the individuals 

 suffering from phthisis or insanity, or having died from these diseases, may without 

 direct intention to deceive have been more frequently omitted than in the case of 

 relatives all of whom were included. 



(b) The cousins in our family record schedules are those of the subject. In order 

 to get full ancestral information a young adult has been very often taken as the subject 

 and the cousins belong accordingly to the third generation, and are themselves often 

 young adults. It follows accordingly that their medical history is in many cases 

 incomplete. They have not passed wholly through the danger zone in the case of 

 either tuberculosis or insanity. 



In the case of tuberculosis, we have for instance among males only 206 tuberculous 

 out of 2990 individuals, and among females only 205 out of 3242, whereas 10 p.c. 

 would probably be affected if we had the full record. 



In our records for example there are in the case of women 130 cases of individuals 

 classed as cousins with some form of brain disease or mental defect*. These 

 130 individuals have 6 insane and 124 sane cousins. If in the remainder of their 

 lives 4 persons out of those 124 sane cousins were to suffer from some form of brain 

 attack, then our table would be as follows : 



* "Insanity" for the purpose of this investigation has been taken to include the neuroses: confirmed 

 alcoholism and marked hysteria. These were not included by Heron in using Pearson's Family Records 

 (Eugenics Laboratory Pali/iratiiut.i, n. p. 33). Its use here approaches "want of mental balance." 



