Davenport and Weeks : Inheritance oe Epii^Epsy 29 



through immigration, through trauma, and through the chance 

 union of defective germ cells of normal persons a thin stream 

 would be maintained, but the State would have control of the sit- 

 uation and the expense would be ever diminishing instead, as now, 

 ever increasing. 



Summary 



1. The method of field-study of epileptic families combined 

 with the modern biological methods of analysis of hereditary 

 data constitute a vastly improved means of inquiry into inheri- 

 tance of epilepsy. 



2. Epilepsy and feeble-mindedness show a great similarity of 

 behavior in heredity supporting the hypothesis that each is due to 

 the absence of a protoplasmic factor that determines complete 

 nervous development. 



3. When both parents are either epileptic or feeble-minded all 

 their offspring are so likewise. 



4. The conditions named migraine, chorea, paralysis, and ex- 

 treme nervousness behave as though due to a simplex condition 

 of the protoplasmic factor that conditions complete nervous devel- 

 opment; i. e., persons belonging to these classes usually carry 

 some wholly defective germ cells. Such persons may be called 

 "tainted." 



5. When such a tainted individual is mated to a defective about 

 one half of the offspring are defective. 



6. When a simplex normal is mated with a defective about 

 half the offspring are normal ; the others defective or neurotic. 



7. When both parents are simplex in nervous development 

 and "tainted" about one quarter (actually 30 per cent.) are 

 defective. 



8. The proportion of tainted offspring is not noticeably higher 

 when both parents show the same nervous defect. 



9. Normal parents that have epileptic offspring usually show 

 gross nervous defect in their close relatives. 



ID. While we recognize that "epilepsy" is a complex, yet 

 there is a classical type numerically so preponderant that, in the 

 mass, " epilepsy " acts like a unit defect. 



11. Our data point to a poisoning in slight degree of germ 

 cells by alcohol, but the evidence is hardly crucial. 



12. There is evidence that in epileptic strains the proportion 



